THE WORLD BANK ECONOMIC REVIEW Volume 1 September 1987 Number 4 1175 5q9 A SYMPOSIUM ISSUE ON THE MULTILATERAL TRADE NEGOTIATIONS AND DEVELOPING-COUNTRY INTERESTS lD Introduction Jagdish N. Bhagwati, Anne 0. Krueger, and Richard H. Snape Sectors: New and Old El Trade in Services and the Multilateral Trade Negotiations Jagdish N. Bhagwati a Agriculture in the Uruguay Round: Interests of Developing Countries Alberto Valdes El GATT Law, Agricultural Trade, and Developing Countries: Lessons from Two Case Studies Surakiart Sathirathai and Ammar Siamwalla El Textiles and Clothing in a New Round of Trade Negotiations Vincent Cable Systemic Issues a Differential and More Favorable Treatment of Developing Countries and the International Trading System Martin Wolf Oi Balance of Payments and Related Issues in the Uruguay Round of Trade Negotiations Shailendra J. Anjaria ai GATT Safeguards and Voluntary Export Restraints: What Are the Interests of Developing Countries? Brian Hindley Li International Control of Subsidies and Countervailing Duties J. Michael Finger and Julio Nogues Li Export-Promoting Subsidies, Countervailing Threats, and the General Agreement on Tariffs and Trade Chong-Hyun Nam THE WORLD BANK ECONOMIC REVIEW EDITOR Richard H. Snape ASSISTANT EDITOR Clara L. Else EDITORIAL BOARD Assar Lindbeck, Institute for International Economics, Stockholm John Holsen T. N. Srinivasan, Yale University Sarath Rajapatirana Joseph Stiglitz, Princeton University Marcelo Selowsky Dennis N. de Tray Richard H. Snape The World Bank Economic Review is a professional journal for the dissemination of World Bank-sponsored research which informs policy analyses and choices. It is issued three times a year, with an occasional additional issue devoted to a special area of current research or policy interest. The Review is directed to an international readership among economists and social scientists in government, business, and international agencies, as well as in universities and development research institutions. The Review emphasizes policy relevance and operationial aspects of economics, rather than primarily theoretical and methodological issues. It is intended for readers familiar with economic theory and analysis but not necessarily proficient in advanced mathematical or econometric techniques. Articles will illustrate how professional research can shed light on policy choices without becoming essays in advocacy. Inconsistency with Bank policy will not be grounds for rejection of an article. Articles will be drawn primarily from work conducted by World Bank staff and consultants. Before being accepted for publication by the Editorial Board, all articles are reviewed by two referees who are not members of the Bank's staff; articles must also be recommended by two out of the three external members of the Editorial Board. The Review may on occasion publish articles on specified topics by non-Bank contributors. Any reader interested in preparing such an article is invited to submit a proposal of not more than two pages in length to the managing editor. The views and interpretations in articles published are those of the authors and do not necessarily represent the views and policies of the World Bank or of its executive directors or the countries they represent. When maps are used, the designations employed are solely for the convenience of the reader and do not imply the expression of any opinion whatsoever on the part of the World Bank or its affiliates concerning the legal status of any country, territory, city, or area, or concerning the delimitations of its boundaries or national affiliation. Commnents or brief notes responding to Review articles are welcome and will be considered for publication to the extent that space permits. The Editorial Office welcomes inquiries and requests to translate or reprint material from the Review. Please direct all editorial correspondence to the Managing Editor, The World Bank Economic Review, The World Bank, Washington, D.C. 20433, U.S.A. Subscriptions to The World Bank Economic Review for the first volumes will be made available without charge upon written request to the Publications Department. To remain on the mailing list, each year readers must reconfirm in writing their interest in renewing their subscription. See the last pages for subscription information and subscription request forms. Inquiries concerning subscriptions should be addressed to World Bank Publications, Washington, D.C. 20433, U.S.A. C 1987 by the International Bank for Reconstruction and Development / THE WORLD BANK All rights reserved Manufactured in the United States of America ISSN 0258-6770 Material in this journal is copyrighted. Requests for permission to reproduce articles should be addressed to the Director, Publications Department, The World Bank, Washington, D.C. 20433, U.S.A. The World Bank encourages dissemination of its work and will normally give permission promptly and, when the intended reproduction is for noncommercial purposes, without asking a fee. Permission is not required to make photocopies for classroom and similar immediate use. This journal is indexed regularly in Current Contenrts / Social &- Behavioral Sciences and the Social Sciences Citation Indexg. Important Renewal Notice You have been receiving The World Bank Economic Review on a complimentary basis since September 1986. If since that date you have not asked in writing to be on the subscription list, you must complete this card and return it to the World Bank. If you do not return this card, you will no longer receive the Review. Peel off the mailing label from the envelope in which you received the Review and place it below. Mark any address changes necessary. If your label is missing, fill in your name and complete address below. Detach card here and mail. ED Yes, continue sending me complimentary copies of The World Bank Eco- nomic Review. Name r r Title l l Organization name D Address i City and province or state Postal code Country I b Oraiainnm Affix Postage Here World Bank Publications 1818 H Street, N.W, Room J-1085 Washington, D.C. 20433, U.S.A. THE WORLD BANK ECONOMIC REVIEW Volume 1 September 1987 Number 4 THIS SYMPOSIUM ISSUE draws on papers presented at a conference on the role and interests of the developing countries in the multilateral trade negotiations held in Bangkok, Thailand from October 30 to November 1, 1986, under the auspices of the Thailand Development Research Institute and the World Bank. The organizer of the conference, Jagdish N. Bhagwati, is guest editor of this issue. THE MULTILATERAL TRADE NEGOTIATIONS AND DEVELOPING-COUNTRY INTERESTS Introduction 539 Jagdish N. Bhagwati, Anne 0. Krueger, and Richard H. Snape Sectors: New and Old Trade in Services and the Multilateral Trade Negotiations 549 Jagdish N. Bhagwati Agriculture in the Uruguay Round: Interests of Developing Countries 571 Alberto Valdes GATT Law, Agricultural Trade, and Developing Countries: Lessons from Two Case Studies 595 Surakiart Sathirathai and Ammar Siamwalla Textiles and Clothing in a New Round of Trade Negotiations 619 Vincent Cable Systemic Issues Differential and More Favorable Treatment of Developing Countries and the International Trading System 647 Martin Wolf Balance of Payments and Related Issues in the Uruguay Round of Trade Negotiations 669 Shailendra J. Anjaria GATT Safeguards and Voluntary Export Restraints: What Are the Interests of Developing Countries? 689 Brian Hindley International Control of Subsidies and Countervailing Duties 707 J. Michael Finger and Julio Nogues Export-Promoting Subsidies, Countervailing Threats, and the General Agreement on Tariffs and Trade 727 Chong-Hyun Nam Index to Volume 1 745 THE NEXT ISSUE of The World Bank Economic Review will be Number 1 of Volume 2; volumes will then be on a calendar-year basis. THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4: 539-547 Introduction Jagdish N. Bhagwati, Anne 0. Krueger, and Richard H. Snape This symposium issue of The World Bank Economic Review contains a selection of articles originally presented as papers at a conference on the role and interests of the developing countries in the multilateral trade negotiations held from October 30 to November 1, 1986, in Bangkok.' Proposed by Anne Krueger, then Vice President of Economics and Research, the World Bank, the conference was designed by Jagdish Bhagwati and held under the auspices of the World Bank and the Thailand Development Research Institute. A few words to intro- duce the issue and to situate its contents in the current policy context are in order. I. MARKET ACCESS AND PROTECTIONISM Trhe developing countries no longer need to be persuaded that access to for- eign markets is important for their development. The economic success of the outward-oriented economies of East Asia has been decisive in changing many minds on this subject.2 The real worry now is whether protectionism will not frustrate the converts as they seek the benefits of the export-promoting strategy. The postwar period opened with much skepticism as to the feasibility of an export-oriented growth strategy. This "first export pessimism" (Bhagwati, forth- coming) was based on an assessment of natural market forces: the markets were not thought to be deep enough to absorb the exports-which were then predom- inantly of primary products-of an expanding developing world.3 By contrast, the current "second export pessimism" is based not on the view that markets do not exist but rather on the gloomy view that once these markets are entered protectionism will close them. 1. In addition, one paper by Chong-Hyun Nam has been included because it complements well the analysis by Finger and Nogues of the issues raised by countervailing duty actions. 2. In addition, numerous studies have shown that the outward-oriented countries have consistently performed better than those countries following an import-substituting strategy (see Balassa 1986; Bhagwati, forthcoming; Krueger 1983). 3. Among the contributory factors in this pessimism was the worry that synthetics were replacing natural products and that technical progress was simultaneously reducing inputs per unit of output (see Nurkse 1959). Jagdish Bhagwati is a professor of economics and politics at Columbia University and currently a consultant to the World Bank. Anne 0. Krueger is a professor of economics at Duke University. Richard H. Snape is the editor of The World Bank Economic Review and of The World Bank Research Observer. Copyright © 1987 by the International Bank for Reconstruction and Development / THE WORLD BANK. 539 540 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 The second pessimism is thus man-made. It can therefore be eliminated by human action. Hence the new (as distinct from the old) pessimism requires, not adaptation to a situation that must be accepted as given, but rather a strategic response to alter the situation. II. MULTILATERAL TRADE NEGOTIATIONS: RATIONALE The Uruguay Round of Multilateral Trade Negotiations (MTN) is part of that strategy. First, there is the old "bicycle" theory. As protectionism intensifies in the congresses and parliaments, which respond to sectoral interests, the execu- tives find it profitable to engage in trade negotiations and talks to keep momen- tum going in favor of freer trade. Ongoing negotiations also make it easier to keep protectionist legislators in check by invoking the possibility of upsetting delicately poised talks, while in the United States the delegation of authority to the executive to negotiate tariff reductions depends on rounds of trade negotia- tions (Snape 1987). In short, the smart way to keep freer trade going is to keep negotiating it: as with a bicycle, if you stop moving, you fall off. Second, it is evident that the ability of the United States to bring the MTN to a start in Uruguay was a result not of consensus on the issues but almost entirely of the fear on the part of the trading partners of the United States that there was a real likelihood that the protectionist forces would triumph in the U.S. Con- gress if the efforts to launch the MTN failed. This fear was not sufficiently strong when the earlier U.S. efforts to start trade talks failed at the General Agreement on Tariffs and Trade (GATT) ministerial meeting in November 1982. Third and equally important, the MTN also represent part of the protection- containment strategy of the United States in a different sense. In a pluralistic framework, lobbying pressures for protection are usually more potent than pressures for freer trade, since typically the benefits of protection accrue to a few (producers) while the costs are spread over many (consumers). Given the strength of the protectionist sentiment in recent years, the Reagan administra- tion has been looking for "countervailing" lobbies that favor trade rather than protection, thus assisting in the growth and success of the export lobbies for sectors such as agriculture and services. Aside from bilateral moves in these areas, the United States has seen the MTN as the principal arena where these protrade interests can be given play. III. UNIQUENESS OF THE URUGUAY ROUND FOR DEVELOPING COUNTRIES The Uruguay Round, however, is unique from the viewpoint of the developing countries. It marks a sufficiently radical departure from the earlier GATT rounds in that, more than ever, it calls for the developing countries to engage actively in the negotiations. Historically, the developing countries were not particularly active participants in the tariff-cutting exercises of the postwar period. With the GATT'S well-known Bbagwati, Krueger, and Snape 541 focus on first-difference reciprocity in trade negotiations,4 and the exemption traditionally accorded to developing countries from reciprocity, it was inevitable that the incentive afforded to developing countries to engage actively in the GATT rounds was small. This incentive was reduced further by the effective exclusion from the GATT negotiations of major agricultural products, important exports for mnany developing countries. The possibility of concessions by developed countries on agriculture could have induced greater interest in the GATT and possibly reciprocal bargaining by some developing countries; the waivers from the provisions of the GATT given to the United States and European countries in the late 1950s for their agricultural production effectively closed this door. Developing countries therefore did not participate actively in the earlier GATT rounds.5 The benefits of generalized tariff cuts came to them simply by way of their most favored nation (mfn) rights as GATT members. The failure of developing countries to offer reciprocal tariff cuts or similar measures to increase access to their markets may have contributed to the bias of the tariff cuts away from the labor-intensive manufactured exports of the devel- oping countries.6 Finger (1979) has produced data on how U.S. imports were affected by Kennedy Round concessions7 which show that the lower the partici- pation by a group of countries in the Kennedy Round negotiations, the lower the share of U.S. imports from that group that benefited from U.S. concessions: Affected 1 964 U.S. imports as a Country group percentage of imports from each country group Major participants 70 Other developed-country participants 49 Active developing-country participants 33 Other developing countries 5 Moreover, aside from nonreciprocity in tariff bargaining, the developing countries generally enjoyed Special and Differential (S&D) treatment in two ways: (i) they were granted preferential entry under the Generalized System of Preferences (GSP) on their exports to developed countries, their market access being thus exempt from the MFN obligation of GATT membership; (ii) they were 4. By first-difference reciprocity, we mean reciprocity or balancing of trade concessions at the margin to be contrasted with full reciprocity or the "level playing fields" approach (see World Bank 1987, box 8-6). 5. Additional factors, such as the asymmetry of market-access obligations, are discussed below. 6. Cheh (1974) argued that exemptions from tariff cuts in the Kennedy Round largely centered on labor-related variables and therefore that the government could be minimizing "short-run adjustment costs,' reflecting a "noneconomic" objective. (The absence of developing countries from the reciprocal negotiating table made it easier for developed countries to grant this assistance to labor.) Such a nonutili- tarian objective was introduced into an augmented social utility function by Bhagwati and Srinivasan (1969) (following on the earlier work of Corden 1957 and Johnson 1965) in which consumption and sectoral allocation of labor were also considered as such noneconomic objectives. An alternative, but less plausible, interpretation seems to be that labor was an effective lobbyist perhaps because a "benign" government does have adjustment-cost-minimization for displaced labor in its objective function. 7. Tariff reductions and bindings. 542 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 permitted extensive exemptions from obligations to provide access to their do- mestic markets (termed reverse market access) on either infant-industry or bal- ance of payments grounds, implicitly or explicitly under GATT Article XVIII. An important consequence was an asymmetry of obligations between the developing and the developed countries under the GATT. This reinforced the lack of incentive to bargain in the GATT rounds. Equally, a chief consequence was the ease with which the developing countries could sustain unity in their trade positions. For as long as S&D was the ruling principle, the governments of the developing countries had a commonality of interests and a fairly unified agenda. They could have preferential access to external markets while choosing their own preferred level of openness or closedness without needing to offer reciprocal access. This convenient deal and the resulting unity of the developing countries had origins and rationales that have now seriously weakened, resulting in both a differentiation of their trade interests and the need to participate actively in the Uruguay Round. The rationale for preferential market access for the exports of developing countries was varied. It was sometimes justified as an extension of the infant- industry argument: preferential access implied that the infant was protected in foreign, not just domestic, markets. The primary rationale, however, was that it would be a form of "disguised" aid. The GSP schemes were part of a wider agenda of mechanisms designed to transfer more resources to the developing countries as budget-financed foreign aid programs faltered. The special drawing rights-link proposal was one such device. The GSP schemes were another. The implicit argument went something like this. Assume that the proposed benefi- ciary of a GSP program would like to export a certain amount. Given a preferen- tial duty, this beneficiary then enjoys a terms of trade gain in not having to cut its export price by the amount of the preference, and the importing country offering the preference correspondingly loses tariff revenue.8 But if GSP schemes were intended to be such a mechanism for transferring aid, it equally meant that the GSP in practice would be characterized by the politics of aid. And that is precisely what happened. These preferential-entry schemes were tightly controlled by beneficiary and product. Equality of treatment on prefer- ences was not available for all developing countries nor for all among them who were GATT members. Nor would there be bindings. The GSP would be subject to periodic renewal: beneficiaries, product coverage, and degree of preference would all be at the discretion of the importing country exactly as aid flows are. Economists have debated whether the GSP produced significant welfare gains. The consensus seems to be that, granted as an "unrequited transfer" (that is, without direct reciprocal benefits), these schemes have generally produced few benefits in total-though there are significant gains for some countries. In gen- 8. The actual economics is of course more complicated. What is stated in the text is the implied argument underlying the case for GSP as an aid mechanism. Bhagwati, Krueger, and Snape 543 eral, preferences are low and the coverage limited, often excluding items in which the developing countries have significant export advantage while includ- ing items such as jet aircraft. Of the United States's $120.3 billion in imports from developing countries in 1981, the total from GSP beneficiaries was $68.5 billion, and of this only $8.4 billion entered duty-free.9 Again, analyses of the European Economic Community's GSP show that imports from nonbeneficiaries grew faster than those from beneficiaries (World Bank 1987, p. 167). Several economists have wondered thus whether it was sensible for the devel- oping countries to have pushed for the GSP benefits rather than for nonpreferen- tial, generalized barrier reductions on products of interest to them. To suggest that the latter would have produced greater results in practice (without reciproc- ity) seems debatable, to say the least, and the evidence on the evolution of the Multifibre Arrangement would seem to militate against such a notion. Also unpersuasive is the criticism sometimes advanced that the GSP may have encour- aged protectionism by prompting the protectionist lobbies to cite it as an added and obnoxious source of market disruption. While U.S. congressional testimo- nies to this effect are not lacking, can one seriously maintain that this does little more than lend some color to the protectionist outcry? We now turn to the rationale for the lack of reverse market access to the developing countries themselves. Here, the economic philosophy of infant indus- try protection and the theory that balance of payments difficulties justified ongo- ing exchange and trade restrictions helped to legitimize the general exemptions from conventional GATT rules that the developing countries were granted, under Article XVIII (B) and Part IV in particular. Given these economic theories, which were widely shared at the time, it is arguable that the developed countries saw the accommodation of asymmetric obligations as desirable and legitimate, even though the GATT is a contract, the essence of which is the notion of symmetric rights and obligations (Bhagwati and Irwin 1987). But aside from these economic perspectives, it may be argued that the willing- ness of the developed countries, and particularly of the United States, to permit these asymmetries reflected two other political-economy-theoretic elements. First, the developing countries' share of world trade was sufficiently small for the question of reverse market access to be considered relatively unimportant. The transaction cost of insisting on reciprocal obligations was not commensu- rate with potential benefits. Second, the United States can plausibly be regarded as the guarantor of the liberal international economic order, embodied in the Bretton Woods institutions and the GATT, and therefore to have been willing (albeit reluctantly) to accom- modate the developing countries in their desire to maintain asymmetric rights under the GATT by appropriate modifications such as Part IV. Such an accommo- dation would permit the GATT'S continued comprehensive membership, while not seriously compromising its integrity, since the developing countries were not 9. Billion is 1,000 million. 544 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 yet important actors on the trade scene.10 The acceptance of developing coun- tries as "free riders" in the GATT system would therefore have been logical. By now, however, the rationales for asymmetry of obligations, which are aid transfer and balance of payments considerations, have weakened. The effective rationale for GSP was aid transfer. But this implies that beneficiary developing economies that have ceased to require aid are likely to be forced out of the schemes. This is the "graduation" issue now on the negotiating agenda. The candidates for such (involuntary) graduation are those which are successful exporters, are advanced industrially, or have reached high standards of living. Brazil, the Republic of Korea, and Taiwan are among the candidates. For the heavily indebted among them, with necessity to find markets abroad to earn foreign exchange for their debt service, this raises problems since loss of GSP benefits is tantamount to having tariffs raised against them. The developing countries' rationale for denial of access to their own markets has weakened even further. The economic theory of balance of payments man- agement has changed with the advent of flexible exchange rates. It is not possible to argue convincingly that a developing country needs to be wedded to long-term controls to ration the use of foreign exchange when exchange rates can be changed. Nor can many adherents be found for the view that such controls offer a feasible way of permanently curing payments imbalances that reflect sustained macroeconomic imbalances. The economic theory underlying GATT Article XVIII (B) is outmoded. Equally, the major developed countries are not willing any longer to see the more successful developing countries get by without reciprocal opening of their markets. The growth of reciprocitarian ideas in the United States has been particularly dramatic, with politicians and industry demanding "level playing fields" from the East Asian and many other larger developing economies. These markets are now substantial; the rise of successful new competitors crowds the producers of the "diminished giant" that the United States is now in the world economy; and the long period of U.S. trade deficits has fueled the exaggerated sense that Japan, Korea, Taiwan, and other larger developing economies have prospered by closing their markets and enjoying access to the open markets of the United States. Reverse market access is therefore a major new issue that the newly industrialized economies face today. The issue is likely to become impor- tant for more developing economies in the longer run. IV. THE ERA OF NEGOTIATIONS The Uruguay Round opens with an altogether new scenario for the developing countries. Issues which were taken for granted over the last two decades are now on the agenda. The agenda includes new sectors, such as services, for which the 10. This view fits in with the Gramscian thesis that a hegemonic power seeks such marginal accommo- dations with the nonhegemonic nations during its ideological primacy. Bhagwati, Krueger, and Snape 545 developed countries mainly seek well-defined GATT-style access to foreign mar- kets as part of the overall conception of level playing fields or reciprocity. It also includes a serious attempt to address agricultural protection, an issue of export interest to many developing as well as developed countries. Agricultural ex- porters have been particularly frustrated in the GATT in the past, frustration that has discouraged them from offering concessions in GATT negotiations. In addi- tion, questions about asymmetry of obligations, and about graduation on both market access and reverse market access, have become key. The effect of these critical developments is to drag the developing countries, however unenthusiastically, into the MTN, where they will have to give in order to receive. But this also means that their interests will diverge depending on the issues and sectors in question, whereas simply receiving (S&D) united them with considerable ease in the past. These divergent interests will surely criss-cross both developing and developed countries, making likely the formation of negoti- ating groups that embrace coalitions of like-interested countries, whether devel- oping or developed. Thus the Group of Ten (led by Brazil and India) opposed the inclusion of services on the agenda; many of the smaller developing countries deserted this position and went instead with major developed countries in the Group of Forty-eight coalition. And in the Cairns group of "fair-trading" agri- cultural exporters, Australia and Canada are joined by Argentina, Thailand, and Uruguay (among others) as forceful members. V. THE BANGKOK CONFERENCE AND THIS SYMPOSIUM ISSUE The Bankgok conference was designed to identify these interests within the framework of the sectoral and systematic issues likely to emerge at the MTN. The focus was on analyzing these interests rather than on suggesting specific pre- scriptions on issues, coalitions, and negotiating strategies. The conference papers selected for this symposium issue of the Review fall into two broad groups. The first set deals with major new sectors, services (Bhagwati) and agriculture (Valdes, Sathirathai and Siamwalla) and with one perennial sector, textiles (Cable), in which the Multifibre Arrangement has just been renegotiated, which is of importance in the overall framework of the world trading system.1" The second set deals with systemic issues. In view of the critical relevance of the S&D issues, Wolf addresses the underpinnings of this asymmetry of obliga- tions, whereas Anjaria considers in depth Article XVIII (B) and the attendant balance of payments justification for denying reverse market access. These sys- 11. The Sathirathai and Siamwalla case study of cassava exports from Thailand shows how critical is legal and negotiating expertise in handling trade issues (and/or improved dispute settlement procedures in the GATT), and how the GATT may need to be strengthened to assist the smaller developing countries. As developing countries enter the era of trade negotiations, this question will surely assume more importance. 546 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 temic questions impair the ability of the developing countries to participate symmetrically in the GATT as originally conceived. They are directly at issue in the intellectual debate on the world trading system and are explicitly or implic- itly on the MTN agenda. The critical systemic issues of the developed countries are altogether different. The developed countries do not resort to trade restrictions to deal with payments difficulties; nor do they comprehensively invoke infant-industry and related ar- guments for systematically rejecting foreign access to their markets. But their failure to adhere to the original GATT rule of law comes rather from their inability to deal with the sectoral politics of market disruption, reflected in part by their tendency to bypass the safeguards clause (Article XIX) in the GATT by resorting instead to bilateral arrangements such as voluntary export restraints (VERS). Hindley's paper addresses in a novel way the question of reforming Article XIX while keeping developing-country interests explicitly in the fore- front. Yet another worrisome development has been the tendency in developed coun- tries to use the countervailing duty and antidumping mechanisms, designed to offset so-called unfair competition under free trade, as protectionist instruments instead. By harassing foreign suppliers with frivolous complaints, for instance by invoking these processes for minuscule subsidies, these instruments can be cap- tured by protectionist interests and utilized to intimidate foreign competitors into accepting VERS and similar trade restrictions. Finger and Nogues, as also Nam, explore the facts on these mechanisms in depth; the former paper also offers an assessment of the harassment issue. The issue does not offer a comprehensive guide to all the topics on the MTN agenda. It omits, for instance, protection of intellectual property rights and trade-related investment rules. But it should provide a handy and illuminating guide to the principal questions that developing economies must ask and answer as they negotiate at the MTN. REFERENCES Balassa, Bela. 1986. "The Importance of Trade for Developing Countries." Paper pre- sented at the Bangkok conference. Bhagwati, Jagdish. Forthcoming. "Export-Promoting Trade Strategy: Issues and Evi- dence." World Bank Research Observer. Bhagwati, Jagdish, and Douglas Irwin. 1987. "The Return of the Reciprocitarians: U.S. Trade Policy Today." World Economy 10, no. 2 (June): 109-30. Bhagwati, Jagdish, and T. N. Srinivasan. 1969. "Optimal Intervention to Achieve Non- Economic Objectives." Review of Economic Studies, January: 27-38. Cheh, John. 1974. "United States Concessions in the Kennedy Round and Short-Run Adjustment Costs." Journal of International Economics, November: 323-40. Corden, W. Max. 1957. "Tariffs, Subsidies and the Terms of Trade." Economica, August. Finger, Michael. 1979. "Trade Liberalization: A Public Choice Perspective." In Ryan Bhagwati, Krueger, and Snape 547 Amacher, Gottfried Haberler, and Thomas Willett, eds., Challenges to a Liberal Inter- national Economic Order. Washington, D.C.: American Enterprise Institute. Johnson, Harry G. 1965. "Optimal Trade Intervention in the Presence of Domestic Distortions'" In R. Baldwin and others, Trade, Growth and Balance of Payments. Amsterdam: North-Holland. Krueger, Anne 0. 1983. Trade and Employment in Developing Countries. Vol. 3, Syn- thesis and Conclusions. Chicago: University of Chicago Press. Nurkse, Ragnar. 1959. Patterns of Trade and Development. Wicksell Lectures. Stock- holm: Almqvist and Wiksell. Snape, Richard H. 1987. "Bilateral-Multilateral Tension in Trade Policy." Paper pre- sented to the meetings of the Western Economic Association, Vancouver, July 7-10. World Bank. 1987. World Development Report 1987. New York: Oxford University Press. THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4: 549-569 Trade in Services and the Multilateral Trade Negotiations Jagdish N. Bhagwati The dispute between developed and developing countries over the inclusion of services in the Uruguay Round of trade negotiations reflects critical differences in perspective on substantive issues. In particular, these substantive divisions arise from the differences between services and goods in matters such as regulation and the requirement in many instances offreedom to move productive factors across national boundaries-for exam- ple, the "right to establish" that would permit the provider of services to get to the user. In addition, developing countries see the developed countries as seeking concessions on service trade in exchange for removal of the latter's existing and potential barriers on trade in goods, rather than establishing quid pro quos within the service compact itself. Developing countries have possible export advantages in the service sector and have much to gain by joining actively in negotiating a services compact that permits them to exploit these advantages. The question of inclusion of services in the Uruguay Round was a principal source of discord between the Group of Ten (G10), led by Brazil and India, and the developed countries,-led by the United States in the negotiations prior to the Punta del Este meeting.' In between these two "hard-line" groups2 were doubt- less other developing countries who shared G10-type concerns. Nonetheless, they felt sufficiently pressured by the ballooning protectionist threat in the 1. The G10 was the group of developing countries consisting of Argentina, Brazil, Cuba, the Arabic Republic of Egypt, India, Nigeria, Peru, Tanzania, Vietnam and Yugoslavia. 2. It has become customary in some sections of the press to describe the G10 as "hard-line" developing countries and the G48 as being led by "moderate" developing countries and "medium-sized" developed countries, when equally accurately the latter could be described as the "medium-sized" developing coun- tries and the "moderate" developed countries. The dialogue between the two sides with opposed view- points is hard enough to manage without the addition of such pejorative characterizations of the principals. The author is a professor of economics and politics at Columbia University and currently consultant to the World Bank. Thanks are due to Brian Hindley, Ambassadors Shukla and Batista at the GATT, and Ambassador Muchkund Dubey for helpful conversations. The comments of Harvey Bale, Narongchai Akrasanee, David Lee, Tony Lane, Paul Leuten, Martin Wolf, and others at the Bangkok conference where the paper was presented, have also led to necessary revisions. Copyright © 1987 by the International Bank for Reconstruction and Development / THE WORLD BANK. 549 550 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 United States and the energetic and relentless diplomacy of its negotiators, to become with the European Community (EEC) the "moderate" brokers of a com- promise solution at Punta del Este. But the compromise merely clears the way for the trade talks to be launched despite the discordant views on services. The compromise relates to procedures on which the contending parties fought because, as I shall explain below, they symbolized substantive differences. These differences are serious and they raise both broad conceptual questions and narrow negotiating issues. This article seeks to address these issues and to define the possible agenda that the develop- ing countries may seek in the service negotiations that are now to begin. I. THE QUESTION OF TRACKS: FORM AND SUBSTANCE The procedural issues that divided the United States from the G10, if we may confine ourselves to the principals, related to two questions: 1. Would the General Agreement on Tariffs and Trade (GATT) be augmented to handle a service compact, or would there be a separate institution or agreement to oversee and regulate world commerce in services, 2. Would the negotiations for arriving at such a compact be conducted under GATT auspices or independently; by contracting parties or by a different group; and parallel to the next round of talks on goods or disjoint therefrom? The U.S. position at the outset was to augment the GATT to include services, leaving the form of such an augmentation to the negotiations themselves. That shape may, as a witticism went, be simply to add to the GATT Articles the two words "and services" wherever the word "goods" appeared, or alternatively, taking the cue from the conventional Oxford English Dictionary where "man" embraces "woman," to declare that "goods" imply "services" in the Agreement. But, as often, good wit is bad economics; and services raise issues that go well beyond the scope of the GATT as it currently stands. It followed equally that the United States wanted the new round of trade talks to include the negotiation of the services compact. The so-called single track was therefore the preferred U.S. option. By contrast, Brazil and India, and the G10 as a whole, wished to delink the GATT from a potential services agreement and derive comfort rather than suffer- ing embarrassment from the fact that the acronym for the General Agreement on Services would be GAS. In turn, therefore it was reasonable for them to seek a neat separation in the negotiating procedures for goods and for services: this was the dual-track procedure proposed by Brazil in June 1985. The negotiations, according to this formula, would be distinct for services, would be undertaken by governments rather than GATT contracting parties, need not be parallel to those in goods, would not be under GATT auspices, and would lead to a services compact outside the GATT. What transpired at Punta del Este was a compromise between these two Bhagwati 551 opposed procedural designs. The dual track was preserved in that the "contract- ing parties" would negotiate on goods but would change their hats to "govern- ments" when they negotiated on services. But the G10 yielded to the extent that both groups would operate under the aegis of the Trade Negotiating Committee, to which they would take their recommendations; and the question of whether the GATT would be augmented or bypassed via a separate services compact was deliberately avoided. Why all this fuss? Was it really a "farce," as U.S. ambassador Yeutter is reported to have remarked? As it happens, it was not. Underlying these proce- dural issues is a key, substantive source of discord. The United States, and lately the EC, have given the impression that they would trade concessions on their imports of goods, for concessions on their exports of services. Recent U.S. Section 301 actions (which are trade actions directed at what are deemed unfair practices affecting U.S. exports) have even explicitly followed this type of link- age with a degree of energy that leaves little doubt of U.S. earnestness in the matter. The linkage has been formulated not merely in terms of "rollbacks" of barriers against developing countries' exports of goods in exchange for access to developing-country markets in services. More seriously, the linkage has been made in terms of denying "standstills" and hence added protection being threat- ened, for developing-country exports of goods if they did not offer "reverse" market access on services. Opposed to this approach has been the position of the G10 that most "roll- backs" and "standstills" on goods merely call for the contracting parties to conform to explicit GATT rules. The demands of the developed countries such as the United States that goods and services should be linked are seen as offering conformity to GATT rules on goods as an exchange for developing countries opening up new areas such as services to market access. This is considered to be unfair and wrong. In short, the U.S. position is construed as a demand for an unrequited concession by developing countries on services masquerading as a quid pro quo trade of concessions by developing and developed countries. The single-track and dual-track modalities are therefore not superficial phe- nomena but reflect the desires of their respective proponents to choose bargain- ing procedures that reflect and hence enhance these substantive positions. Single-track negotiations do underline linkage; dual-track negotiations do not. II. SERVICES VERSUS GOODS: CONCEPTS AND CONSEQUENCES It is important, at the outset, to recall the conceptual advances that interna- tional economists, following as usual in the footsteps of activist policymakers, have now made in the matter of defining services. This conceptualization should provide the underpinnings for the positions that governments must consider in formulating the general principles of a services compact, just as the theory of trade and welfare provides the underpinnings for the general principles that 552 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 underlie GATT and for the impulse to trade liberalization that informs current World Bank conditionality. How, then, are services to be defined? Or how are they different from goods? Adam Smith, John Stuart Mill, and many others raised these questions, but perhaps the earliest answer to them was attempted by T. P. Hill (1977) only recently. Hill focused on the fact that producers cannot accumulate a stock or inventory of services, stressing that services must be consumed as they are pro- duced. This key element will not characterize all items which we customarily define as services: for example, "answering services" do store messages. But such exceptions do not detract from the usefulness of a definition of services that characterizes them as nonstorable because they require the simultaneity of provi- sion and use.3 Services Requiring Physical Proximity versus "Long-Distance" Services If services must be used as they are produced, then there must of necessity be interaction between the user and the provider of the service. A producer of goods, by contrast, can produce but store, and generally transact with users at any subsequent time. But this interaction, in turn, implies that we can contem- plate two essential categories of services: first, those that necessarily require the physical proximity of the user and the provider; and second, those that do not, though such physical proximity may be useful. I noted this important distinction sharply in a recent article (1984): Basically one has to draw a distinction between services as embodied in the supplier of the services and requiring their physical presence where the user happens to be and services which can be disembodied from the supplier and provided without a physical presence being necessary."4 Physical proximity essential. The class of services where physical proximity is essential is usefully thought of as consisting of three categories based on the mobility of the provider and user of the services. The first category is mobile provider, immobile user. This class of services requires that the provider go to the user, while the reverse mobility is physically 3. Another characteristic of services which is necessary is that services occur between different eco- nomic agents or otherwise all activities and value added would collapse into the service sector. An implication of this characteristic is that the definition of services reflects economic organization or "market structure." If Mr. Smith paints a car on the assembly line inside your auto plant as your worker, then his wages are part of goods production and value added. But if he does the same job from his own establishment, his wages or income are part of service production and value added. For a detailed discussion of this question, see Bhagwati (1987a, 1984). 4. In my 1984 paper, I focus on the latter class of services (which I now call "long-distance" services) discussing how the "disembodiment" effect can frustrate the intention of immigration restrictions on skilled labor. Conversely, Gary Sampson and Richard Snape (1985) have drawn on this twofold distinc- tion in my 1984 article to explore further the former dass of services, where physical proximity of the provider and the user is involved, with a valuable classification of such services which I use, with some simplification, below. Bhagwati 553 infeasible. If an Indian or Korean firm had won the bid for construction of the Connecticut Turnpike, unskilled Indian or Korean labor services could have been provided only by moving them to Connecticut. Supplies of brute, Ricar- dian-style labor services must be relocated to the user's locale, as we have seen in the Middle East since the 1970s. The second category is mobile user, immobile provider. This is another impor- tant class of services in which the user must move to the provider. Open-heart surgery cannot currently be done in Zaire because, even though Dr. Cooley can go from Houston to Kinshasa, there is no way the necessary support services and hospital care can be duplicated there. In this class of services, some key elements are simply not transferable geographically to the user's location. The third category is mobile user, mobile provider. For this range of services, mobility is symmetrically possible. Haircuts, tailored suits, and lectures are the type of services which are in principle transmittable between user and provider in either's location, the only difference being the cost of so doing. The generic class of services, where the provider must move to the user, as a sheer physical necessity (as in the first category above) or because of overwhelm- ing economic advantage in so doing relative to alternative means of effecting the service transaction at long distance (as discussed immediately below), I call 'temporary-factor-relocation-requiring" services. Physical proximity inessential: the "long-distance" services. In the second broad class, which I call "long-distance" services, physical proximity between providers and users may be useful, but it is not necessary. Live music concerts and data transmission "over the wire" are obvious examples. Traditional bank- ing and insurance services fall into this category, in principle, since loans could be secured by mail or phone, and insurance policies are often so purchased. The scope for long-distance service transactions will increase with the advance of technology (see Bhagwati 1984). This has important implications for broader issues such as the trend effect of immigration restrictions on the relative wages of skilled and unskilled labor since skilled services may increasingly be transacted "long-distance" whereas the latter cannot. Physical proximity between provider and user in many services (especially in banking) does involve substantially greater efficiency, however, and at times may allow a wider range of possible transactions even when long-distance or arm's length transaction is feasible. Technical change which has opened up product diversification in banking, for instance, has reinforced this aspect. In legal serv- ices, continuous interaction between local client and overseas lawyers is deemed essential for efficient service and has fueled lobbying efforts by multinational legal firms to secure ways of establishing physical proximity to their clients abroad. The vast majority of service providers are likely to require and therefore press for physical proximity. The question of devising a service compact, whether as part of an augmented GATT or outside the GATT, is thus inextricably bound up with the question of provider-mobility across national borders. 554 TIAEWORLD BANKECONOMIC REVIEW, VOL.1, NO.4 The negotiations on services must therefore come to terms somehow with the implications of this essential connection, in many services, between international factor mobility and international trade. While we have accepted the distinction between these two phenomena since the founding of both economics and the GATT, it vanishes for these, indeed the preponderant class of, services. Factor mobility and trade are simply two integral aspects of the service transaction. For this reason, I prefer to talk of service transactions rather than service trade, so that we do not lose sight of this dual nature of the services that do not fall into the "long distance" mold. This essential connection of services with international factor mobility has critical implications for government restrictions on service transactions. If serv- ices require factor mobility, then the ability of governments to exclude or impede service transactions does not depend altogether on restrictive border measures on trade. Restrictions on factor inflow can suffice for this purpose. Hence arises the immediate and compelling need to go beyond the conven- tional focus on border trade measures such as tariffs or nontariff barriers (NTBS) for services. This fuels the demands for the "right to establish" domestic outlets.5 But the phrase "right to establish" conceals a continuum of factor-mobility phenomena which embrace both capital and labor mobility. It can cover the right of an American bank to establish a branch in Bombay, implying foreign investment, and the right to employ foreign personnel locally, implying skilled and semiskilled importation of labor. It can extend to a Korean firm's right to construct a road or a harbor by importing skilled and unskilled labor, both constituting (according to sound economic theory as spelled out above) an inte- gral component of the service transaction in that sector. Equally, it can extend to an English multinational legal firm setting up an office in Tokyo, with local personnel but with English barristers or American lawyers who fly in and out to work with multinational Tokyo-based and other local clients. It could include hospital management contracts with short-term inflows of managerial personnel. In short, factor mobility can be complex, not fitting into any particular mold. What is certain however is that the concept of the "right to establish" cannot meaningfully or justifiably be circumscribed to exclude the inward mobility of foreign labor and its services. And the problem that this raises cannot be dis- missed simply by saying, "Oh, we cannot dismantle immigration restrictions and have free mobility of labor across national borders." For, as I have argued earlier (Bhagwati 1987a), the concept that we can work with is that of "temporary- factor-relocation-requiring" services and hence of temporary residence by for- eign labor to execute service transactions. For example, Korean construction 5. At the time of U.S. treasury secretary Connolly's efforts to "open up Japan," the "right to establish" question applied to goods trade. It was then believed that, unless Japan permitted U.S. goods exporters to set up their own retail outlets, market access to Japan could not be effective. The economic implications of this issue have been discussed and modeled in Bhagwati (1982a). Bhagwati 55S firms would bring in workers to build a turnpike; when the task was completed, the workers would return to Seoul. Or an Indian legal firm would have lawyers come from New Delhi for specific assignments or predetermined periods, the firm then rotating the personnel as necessary to avoid permanent residence (for example, immigration) of specific individuals. Conceptual clarification of the nature of service transactions therefore has led to a keen awareness that freeing trade in services and the associated "right to establish" question, will raise serious questions relating to labor relocation as well. As long as the "right to establish" was regarded as simply a question of U.S. banks, insurance companies, and multinational professional firms setting up branches in Bangkok, Dar-es-Salaam, and Tokyo, there was at times a certain sense of patronizing disdain for the hesitations of the countries that found the factor-mobility aspects worrisome. As Hindley (1987) shrewdly remarks, however, a certain ambivalence has apparently crept into the U.S. negotiating attitudes, now that the labor-mobility issue suggests that openness to foreign services may create immigration problems for the United States. On the one hand, the impression has been given by U.S. officials at times that the overall services compact should confine itself to long- distance and arm's-length transactions, ruling out "right-to-establish" questions and hence the corresponding enormous range of services that require such estab- lishment. On the other hand, since the powerful U.S. lobbies from the service sector continue to clamor for the "right to establish," some official spokesmen have instead tended to opt in favor of an emasculated (and conveniently self-serving) notion of the "right of presence" or "right of market access," euphemisms which are designed to ensure artfully that the labor-mobility aspects of the "right to establish" questions will be soft-pedaled.6 Exclusion of services that require significant temporary relocation of labor, however, would rule out of the compact a range of services in which some of the principal developing countries that have been skeptical about or opposed to negotiating services happen to have sufficient skills and endowments to consider developing exports of such services.7 Except for a handful of developing coun- tries such as Singapore and Hong Kong, which entertain offshore banking and insurance establishments without hesitation, "such a definition of 'services' . . . excludes any substantial export interest on the part of developing countries" (Hindley 1987, p. 4). It would also necessarily reinforce the position of those developed countries which seek concessions on services from developing coun- tries in exchange for their (real or apparent) concessions on goods to the devel- oping countries. The question therefore is pertinent to the issues that divide the G10 and the United States. 6. Compare with Hindley's (1986a) penetrating discussion of this issue. 7. This implication was noted earlier in Bhagwati (1987a, 1985a, 1985b) and has been further discussed by Hindley (1987), among others. S56 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 Regulation Another aspect of the difference between services and goods is the much more pervasive application of regulation to services than to goods, and the rare har- monization of regulatory provisions across national boundaries. The critical difference, however, arises from the fact that these regulations often apply to the provider of the services while their intent is to protect the user of the services, whereas with goods the regulations apply to the product itself. Thus, with trade in goods, it is possible for foreign suppliers to meet national regulations by manufacturing to necessary standards. While it is not uncommon to hear complaints about how different health, safety, and human rights tradi- tions and standards result in "unfair" competition, it is conventional with goods not to be bothered by the behind-the-trade-scene regulations as they differen- tially affect rival producers in competing countries. With services, this detach- ment is often impossible. The regulations imposed on the provider can critically affect the service transaction, as for instance with reserve requirements that an insurance company has to meet before it is allowed to even begin to attract any customers. This regulatory difference between services and goods implies that, while local establishment by a foreign provider to supply a service will permit the fulfillment of local regulatory criteria, sale of such services from a base abroad where the regulatory criteria are less strict, will not. This difficulty with regulation arises with "long-distance" or arm's-length transactions, whereas the difficulty (identi- fied earlier) with service transactions requiring physical proximity between pro- vider and user arose where such long-distance transactions were infeasible or significantly inefficient. The nonharmonization of regulatory systems has led to major difficulties with service trade liberalization in the EEC (Hindley 1986a). The EEC does not lack for the "right to establish ." But the incapacity to sell services from a base abroad, where the regulatory regimes are dissimilar, has been a major obstacle to liberali- zation and accounts for the miniscule progress observed to date. A gung-ho reaction to this issue would be to permit regulatory systems to "compete through their outputs." A less demanding or restrictive system would then prosper at the expense of ones that restrict or regulate more. In their present deregulatory mood, U.S. officials may then see a triumph of the more efficient resulting over the less efficient. It is unlikely that others will see the matter this way, however, any more than the U.S. would if the shoe were on the other foot. Within the EEC freer service trade has not been permitted to transpire; and successful efforts have not been made to harmonize the service trade regimes. It is unlikely that the developing countries, where regulation sometimes tends to be stiffer, will be enthusiastic about these matters either. Between the hesitations over the "right to establish" and the desire to emascu- late it to developed-country advantage, on the one hand, and the hesitations over the indirect competition between unharmonized regulatory regimes that the Bhagwati 557 developing countries with greater attention to the role of the State must fear when services are transacted without the benefit of local establishment, on the other hand, it seems as if progress toward the general principles underlying a service compact is likely to be slow. Infrastructure, National Security and Other Constraints on Liberalization Overlaying these difficulties is the fact that some of the service sectors (for example, banking) are regarded by the hesitant developing countries to be part of their infrastructure which they feel they must control for political reasons much as, say, the U.S. restricts ownership by foreign nationals in its media (services) sector. Again, transborder data flows and the information sector are regarded as sensitive areas that raise issues bordering closely on "national secu- rity" for the "middle powers" such as Argentina, Brazil, and India. In these areas it is therefore difficult to urge the developing countries to discard such notions altogether, especially when these types of asymmetrical views about some services and many goods are held by many influential citizens within the developed countries themselves. (As I argue later, however, their fears and concerns are greatly exaggerated and need to be carefully evaluated in their own interest.) Consider the following impassioned pronouncement: We ought to be exporting computers, not shares of IBM. We should seek to sell more, not sell out. To accept the de-industrialization of [our nation] while exulting in the growth of foreign ownership and influence in our domestic institutions could be an unwitting prescription for slowly becoming an economic col- ony again. It came, not from Prime Minister Rajiv Gandhi or from President Alfonsin. The author was U.S. Representative Jim Wright, majority whip in the U.S. Congress, writing in the Wall Street Journal (October 3, 1985). III. COMPARATIVE ADVANTAGE IN SERVICES, COST OF PROTECTION, AND POLICY-MIX SOLUTIONS The foregoing analysis highlights the difficulties that emerge for the impend- ing service negotiations, especially as they reflect the special characteristics that serve to set services apart from goods. But before I turn to the prospects for different solutions to these difficulties, it is necessary to speculate on where the comparative advantage in service transactions may lie, especially between the developing and the developed countries. Several observations on that issue are in order. First, while the trade data for services are extremely unreliable, Sapir's (1985) careful analysis of what is available underlines strongly what common sense would suggest: many traded services tend to be intensive in the use of technology 558 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 Table 1. Trade between the Industrialized and Developing Countries, 1980 (billions of dollarsa) Category of Industrialized-country exports Developing-country exports merchandise and services to developing countries to industrialized countries Merchandise trade, of which: 277 385 Fuels 6 258 Other primary products 44 67 Manufactures 227 60 Service trade, of which: 72 30 Transport 35 10 Travel 14 12 Other private services 23 8 a. Billion is 1,000 million. Source: Sapir (1985, table 2); data for merchandise trade are based on GATT (1983); and for service trade, on own estimates. and of capital, whether human or physical. This should give the developed countries the competitive edge since they are abundant in the endowment of human and physical capital. It is suggestive that, when Sapir (p. 37) looks at the balance of trade in services, it is the advanced newly industrializing economies such as the Republic of Korea, Singapore, and Taiwan that come out with small positive or negative balances rather than the large deficit of many developing countries. 8 However, it would be totally wrong to infer that developing countries simply cannot find traded services that they can export successfully. Table 1, compiled by Sapir, gives an aggregated and very rough picture of service trade among the two groups of industrialized and developing countries for 1980. The data can be read two ways. On the one hand, they show that the service exports of develop- ing countries are a substantially smaller fraction of their total exports than is the case with industrialized countries' share of service to total exports. On the other hand, the developing countries' service exports are by no means negligible, as recorded, and seem to reflect earnings not only from tourism and transport but also from "other private services" (which include professional, design, construc- tion, and related services). Second, detailed studies further underline the export possibilities that the energetic, outward-oriented newly industrializing economies have in services. Thus, for example, table 2 suggests that the earlier U.S. domination of the world market for international construction may have diminished with the medium- level developed countries and a newly industrializing economy such as Korea taking significant shares in the 1980s. A non-negligible share of the developing countries is evident from table 3. In the more complex field of international design contracts, again the data show a sizable share of contracts being awarded to firms from Brazil, India, Korea, Lebanon, and Taiwan (Sapir 1986, table 3). 8. In itself, however, the trade balance would be, of course, an inconclusive piece of evidence on the issue. Bhagwati 559 Table 2. Market Share of International Construction Measured by New Contracts Awarded to the Top Two-hundred Fifty International Contractors, 1980-84 (billions of dollars) Country 1980 1981 1982 1983 1984 United States 48.3 48.8 44.9 29.4 30.1 (45%) (36%) (36%) (31%) (38%) France 8.1 12.1 11.4 10.0 5.4 (7%) (9%) (9%) (11%) (7%) Germany, Fed. Rep. 8.6 9.9 9.5 5.4 4.8 (8%) (7%) (8%) (6%) (6%) Italy 6.2 9.3 7.8 7.2 7.8 (6%) (7%) (6%) (8%) (8%) United Kingdom 4.9 8.7 7.5 6.4 5.7 (5%) (6%) (6%) (7%) (7%) Other European 9.2 12.6 10.3 9.1 7.2 (8%) (9%) (8%) (10%) (9%) Japan 4.1 8.6 9.3 8.7 7.3 (4%) (6%) (8%) (9%) (9%) Korea, Rep. 9.5 13.9 13.8 10.4 6.8 (9%) (10%) (11%) (11%) (8%) All other 9.4 10.5 8.6 7.0 5.9 (9%) (8%) (7%) (7%) (7%) Total 108.3 134.4 123.1 93.6 80.5 Source: Various issues of Engineering News Record; from ongoing studies by U.S. Office of Technology Assessment. Third, there is little doubt that the broader group of newly industrializing economies, not just the super exporting economies like Korea but also the tradi- tionally inward-looking ones like India, have the skills to develop export advan- tages, not merely in computer software (a good, not a service) and in an increasing range of "on-the-wire" services that new technologies make possible, but also in the services that imply temporary relocation of skilled labor. I would expect that legal and professional services, with right of establishment, exhibit a mutual rather than one-sided export advantage for developing and developed countries. The developing countries must not be misled into thinking otherwise simply because the initiative to include such trade in a services compact comes almost wholly from multinational firms in the developed countries. Why? The reason is that such services are not homogeneous. It is necessary to think of "dualistic" structures here (Bhagwati 1986d). The advantage in tendering services at the multinational level is likely to inhere in developed countries: in fact, these multinationals are piggybacking on their multinational clients in other sectors that have operations abroad. Only as the developing countries expand their own multinationals in nonservice sectors, as is beginning to hap- pen, will they begin to develop some "linked" advantage in professional services. At the other end of the spectrum, however, the advantage must belong to lawyers, doctors, and accountants in the developing countries, because, while they are equally competent, they can work more cheaply and offer a range of 560 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 Table 3. Cumulative Foreign Awards of Top International Contractors by Economy, 1978-83 (billions of dollars) Economy Awards All countries, 566.6 of which: All developing countries, 99.2 of which: Korea, Rep. 56.2 Turkey 10.0 Yugoslavia 7.4 Brazil 5.8 India 3.9 Taiwan 3.4 Philippines 3.2 Argentina 3.0 Lebanon 1.4 Pakistan 1.3 Kuwait 0.7 Singapore 0.6 Malaysia 0.5 Panama 0.4 Mexico 0.2 Thailand 0.2 United Arab Emirates 0.2 Colombia 0.1 Indonesia 0.1 Note: Countries are ranked according to the foreign contract values of their top firms. Until 1980, the top 200 firms were surveyed; since then, this number was raised to 250. Source: Compiled by Sapir (1986, table 2) from Engineering News, various issues. services where price competition is decisive.9 If they are allowed to enter under "4temporary-factor-relocation" visas to make service transactions possible, I see no reason why they cannot increasingly take a sizable fraction of the market at that level.10 Such a "dualistic" view is quite consonant with mutual trade in "similar prod- ucts.' A product is a vector of characteristics, and different countries can have advantage in some and not in others. In service transactions, physical proximity accentuates such differential elements and can lead to mutual comparative ad- vantages within a sector for suppliers from different countries. Fourth, the export possibilities become even more compelling for developing countries if the issue of unskilled labor mobility in the execution of specific short-term contracts (as in the Middle East), is resolved in favor of its inclusion in the "right to establish." It is already within the realm of probability, thanks to 9. The question of whether they would be allowed to indulge in price competition is critical. Attempts by professional associations to regulate minimum prices would then be in restraint of trade. 10. It is important not to confuse the "brain drain" question with the issue of temporary relocation of labor. I have discussed the contrasts in Bhagwati (1985c). Bhagwati 561 the widespread use of such unskilled labor, including that by U.S. international contracting firms, during the 1970s and 1980s. It also has legitimacy in the practice of Western Europe in its postwar "guestworker" systems and in the latest U.S. legislation which permits over 300,000 workers to be imported for specific types of short-term work (that is, in U.S. agriculture)."1 Fifth, it is important for developing countries to recognize that a great number of traded services are intermediates. Protecting banking and insurance sectors for example, increases domestic prices of these services. As they are inputs into other goods, this can raise prices of export goods and undermine export pros- pects. The effects of protecting intermediate services are similar to those that result from increasing the cost of intermediate goods such as steel.12 But the adverse effects on exports of goods are more serious in the present instance because, in denying the domestic exporters of goods access to efficient banking services, the protective policies succeed in denying access to more than cheaper credit. More important, exporters are denied access to the entire vector of services that mod- ern international banks can provide to facilitate international commerce. The prot-ection of intermediate services, in the interest of goals such as political control, therefore has costs that are not negligible and have presumably not been properly assessed by the developing countries. Finally, it is important to recognize that policies such as the protection of locally produced computer hardware in the telematics and information sectors may represent an unnecessarily expensive way of securing one's objectives. If the objective is to build up national technological know-how through "learning by doing" (rather than to develop the industry itself), then the cost of such a policy is to spread computer illiteracy in the population and high costs to producers who must make do without lower-cost access to modern information technology in the production process.13 A country such as India (and possibly Argentina and Brazil as well) has the possibility of using an alternative policy instrument to achieve the desired mas- tery of know-how without these costs. Remember that the know-how is em- bodied in one's citizens. If one then looks at the national-origin composition of scientists in only the artificial intelligence, robotics, and computer science labs 11. In the spirit of the U.S. service sectors' demands, the developing countries may well ask for equal access to these jobs instead of having them de facto assigned to applicants south of the Rio Grande. 12. The successful outward-oriented regimes have managed to ensure that internationally traded intermediate goods are available to domestic producers at world prices. Similar logic should apply to internationally traded intermediate services as well. 13. In India, import-substitution in computer hardware has led to higher costs, unavailability, and enormous lags in the use of computers in tourism, the judiciary and the schools. The import-substitution policy manages to distance greatly even a highly educated population and skills-endowed economy such as India from the modern world outside. It also inhibits the rapid adoption of modern information- technology-based processes that are essential to absorbing high-productivity economically-efficient ad- vances in the manufacturing sector. 562 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 and institutes in the United States, it is possible to find numerous Indian mathe- maticians and scientists, even in leadership positions. These Indians embody know-how in these fields at the very cutting edge of technology. Since the sociology of international migration of professional classes increas- ingly permits immigrants to retain ethnic ties to their countries of origin, and therefore the "diaspora" model has increasingly come into its own, the Indian government has the option of utilizing this U.S.-based resource any time it wishes to do so.14 Going the protectionist route will yield a lower level of embodied technology in resident nationals (who may also leave anyway) and will sacrifice computer literacy and efficiency in production. Permitting cheap imports at world prices avoids these costs, and utilizing the superior know-how embodied in one's nationals abroad also secures know-how at its best and cheap- est.'s To put it differently, the two objectives of (i) spreading computer literacy and encouraging adoption of efficient production processes, and (ii) building up technical know-how among one's nationals are impossible to achieve with one policy instrument, namely, protection. They are achievable, and are in effect Pareto-dominated in outcome, by the use of two policy instruments: (i) world- price imports of computers and related technology; and (ii) an open-door policy on emigration combined with a policy to utilize the know-how embodied in one's nationals abroad. Such a policy mix breaks from the protectionist mold and requires an imagina- tive and simultaneous use of policies from what are generally considered to be unrelated areas of governmental intervention. But they do offer the prospect of a superior approach for those countries such as India which have the talents and the skills in the field of information to make such an approach feasible. IV. DEVELOPING COUNTRIES: BARGAINING OPTIONS AND STRATEGIES What positive approaches emerge that the developing countries may take in the forthcoming negotiations on services? Developing countries cannot be expected to opt en bloc for any one approach on services any more than we can expect them to have identical positions on agricultural liberalization or the developed countries to agree on the optimal redesign of safeguard procedures. Thus, Hong Kong and Singapore can be expected to be agreeable to the more "hard-line" developed-country positions on 14. When a people are geographically dispersed but ethnically linked, this new reality has several important implications for a variety of other policy issues such as the appropriate exercise of income tax jurisdiction on one's nationals when they are internationally mobile. See Bhagwati (1982a) and Bhagwati and Wilson (1987). 15. Although I talk of "nationals,' there is little doubt that even those who change nationalities today often have attachment to their countries of origin. Nonetheless, this distinction in turn raises the issue whether developing countries ought not to consider permitting their nationals to hold dual nationality as part of the policy option that I advocate above. Bhagwati 563 services. Brazil and India can be expected to oppose them. They, and also the developing countries that initiated the Uruguay Round under the G48 umbrella, will have to decide what kind of game they want to play now that the players are coming onto the field. The options that they must consider are best determined by the demands that the developed countries, especially the United States, have been making.i6 These options will have to be defined in terms of the responses that the developing countries make to these demands or negotiating positions, as they have been indicated so far. Let me begin with what are generally understood to be the broad outlines of the current U.S. positions, however negotiable they may turn out to be in the course of the Uruguay Round itself. Generally Perceived U.S. Positions Inclusion of services in the Uruguay Round, and indeed of other "new" sectors and areas such as intellectual property and trade-related investment rules, is considered to be part of a "grand tradeoff" where these new areas benefit the United States. In return, the United States is willing to consider rollbacks and standstills (consistent with the exercise of trade-affecting, GATT-compatible actions such as countervailing duties, antidumping, and Section 301 actions) on goods (see also, however, the discussion below). The "grand tradeoff" is seen in terms of both cosmopolitan interest (that is, what international economists describe as "world welfare") as well as U.S. inter- est (that is, what international economists describe as "national welfare"). The former position emphasizes that an efficient world allocation of resources re- quires that "everything be put on the table": the outmoded GATT must be rede- signed, augmented in scope, and brought up to date to embrace new realities. The latter position is developed in terms of U.S. comparative advantage having shifted to the new areas so that, if United States is to yield on goods, it is fair for it to ask others to yield on services and new issues. A brief tabular arrangement of United States-perceived losses and gains in relation to those of the developing countries in terms of the mercantilist logic of trade-barriers-bargaining is pre- sented in table 4. A third argument in the United States in favor of this grand tradeoff is that the current presidential administration is too beleaguered to hold protectionists at bay in Congress unless the advanced developing countries (and, of course, Japan and the EEC) open up their markets to U.S. exports of services as a quid pro quo.17 These countries are faced with what could be construed as a rather 16. While the discussion below focuses on the United States, it is clear that, unlike in the 1982 GATT Ministerial meeting, the EEC also perceives export competitiveness for itself in services and hence is closer to the U.S. positions on it than before. See, for example, the recent statements of Willy de Clercq (1986) to this effect. Table 4 on the United States (below) therefore could be readily modified to one for the EEC, with agriculture being considered as a "loss" instead of a "gain." Japan, with its enormous surplus, also sees clear comparative advantage in the financial services area. 17. The U.S. government has encouraged such export-seeking lobbies as a political countervailing force to the trade-threatening protectionist lobbies. 564 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 Table 4. Perceived U.S. Benefits and Losses in Relation to those of Developing Countries from the Prospective Liberalization of Trade in the Uruguay Round U.S. "Benefits" U.S. "Losses" 1. Services 1. Rollback of the Multifibre 2. Intellectual property Arrangement (MFA) and other 3. Trade-related investments voluntary export restraints (VER) 4. Reverse market access to and Organized Marketing developing countries Arrangements (OMA) on goods 5. Agriculture, 2. Standstill on VERS, OMAS on goods 3. More stringent use of safeguards and tighter rules to prevent abuse of countervailing-duty and antidumping actions 4. Improved structural adjustment a. Agriculture is included as a benefit because agricultural liberalization in cereals is expected to favor U.S. exports, mostly at the expense of the EEC and Japan but, depending on the final package, even at the expense of some developing countries. difficult situation: trade concessions appear to be demanded of them as a way of ensuring that market access for their exports is continued. But this, in turn, reflects a substantial shift in U.S. positions in trade negotia- tions from what I have called GATT-style "first-difference" reciprocity to "full" reciprocity."8 The United States has increasingly looked at, not the balance of advantages from cbanges in trade barriers, but the balance of advantages from the trading system in toto. Doubtless this attitude stems from the macroeconomic difficulties that the overvalued U.S. dollar entails and the resulting substantial adjustments forced on the traded sector. It also stems from the "diminished giant syndrome" that has affected the United States as its effortless postwar hegemony in the world econ- omy has been threatened by the relentless advent of the Pacific Century."9 But overlaying these two factors has been the fundamental fact that the GATT's basic conception, and indeed that of the United States as its leading founder, was always based on contractual and (fully) reciprocal rights, with member states enjoying symmetrical rights and obligations. The United States, which emerged as the force majeure in the 1940s, permitted Western Europe effectively to get away with nonreciprocity while it worked through the 1950s to achieve current account convertibility, and agreed to special and differential treatment for devel- oping countries until now. The current U.S. insistence on full reciprocity can then be seen as an inevitable return to the original symmetrical conception of the world trading order. Hence, 18. I note with pleasure that Brian Hindley, in his article in this volume, has embraced this terminol- ogy. The use of the phrase "aggressive reciprocity" to denote full reciprocity is inappropriate: full reci- procity could equally be pursued in a tranquil way. 19. On this, see Bhagwati and Irwin (1986) on the parallels between late nineteenth-century Britain and the present-day United States in the rise of "fair trade" movements. Bhagwati 565 it is not a position that the developing countries (or Japan, which is alleged, rightly or wrongly, to offer less-than-symmetrical access to its own markets) are likely to be able to challenge with success, much as they consider it to be unfair from the perspective of first-difference reciprocity. My judgment therefore is that the developing countries must proceed from the unhappy premise that the United States, especially its Congress, cannot be expected to trade access to its markets any longer without significant elements of reciprocity from the develop- ing countries, even if the balance-of-trade deficits are somehow eliminated. The Developing-Country Options From the viewpoint of the hesitant developing countries, the U.S. position presents one major difficulty even if they are prepared to accept the reality of full reciprocity and yield on their sense hitherto that the so-called bargain being offered to them simply is not one. It is unclear what the United States and the EEC, can offer by way of standstills and rollbacks on goods if these developing countries offer concessions on services. I quote one influential commentator, who was a member of the U.S. administration: The issue for the United States is whether a meaningful standstill and rollback commitment would apply to existing U.S. restrictions in sugar, meat imports, textiles, steel, automobiles, etc. as well as the use of future 301 actions in both goods and services trade. Ideally, the United States would like this commitment to apply only to new measures, not existing restrictions or extensions of existing programs (such as another VRA [vol- untary restraint agreement] in steel or tightening sugar quotas under the existing programs). According to U.S. interests, it would not apply at all to trade legislation consistent with GATT (201, CVD [countervailing duty], and AD [anti-dumping] provisions and national security), and it would mean submitting 301 cases to GATT but only in goods. In new areas- services, etc.-the United States would remain free to retaliate under 301, including retaliation in goods areas, without submitting to GATT rules. [Nau 1986, pp. 22-23] This has been the sense of the remarks reported in the U.S. press by Ambassador Yeutter on his return from Punta del Este. This is also consonant with the substances of Martin WolPs "Europessimistic" argumentation on special and differential treatment presented in this volume: few meaningful concessions on rollbacks and standstills on goods can really be expected and the developing countries ought to yield on reverse market access largely because it is good for them to liberalize as suggested by the export-promoting strategy.20 Doubtless, as I have already emphasized in section III, even unilateral trade liberalization in intermediate services should have big payoffs for the developing countries. But if only we could persuade trading nations to accept such compelling arguments, 20. On the merits of this strategy, see the extended review in Bhagwati (forthcoming). 566 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 we would not have to worry about rollbacks and standstills either. The develop- ing countries cannot realistically be expected to be less mercantilist than those who preach free trade but practice mercantilism themselves. This is a pity, but a reality too. This reinforces, in my judgment, a suggestion I had made earlier (Bhagwati 1985c), that the developing countries ought to participate actively in the service negotiations instead of rejecting them on grounds of first-difference-reciprocity unfairness. They should then seek quid pro quo (in terms of export possibilities) within the service sector itself. Not merely is it risky to establish linkage between goods and services when the goods "benefit" is less likely than the services "loss." It is also silly to let the developed countries define the service compact all by themselves in a way that can then be fully expected to serve their own narrower, export interests rather than reflecting more fairly and adequately the general principles as set out in recent analyses and recapitulated in section II of this article.21 The latter would also serve the interests of developing-country exporters. As I argued in section III, quid pro quo within the service sector certainly exists for the skill-abundant, newly industrializing countries, and especially so if the temporary-factor-relocation-requiring labor- and skilled-labor-intensive services are not omitted in the formulation of a services agreement.22 The difficulties that I detailed in section II that plague rapid progress in service liberalization also imply that the Uruguay round is unlikely to yield anything more concrete than a code or an agreement of principles. It is improbable that actual service liberalization under the code will emerge during the course of the Round itself. This prospect also underlines the wisdom of a strategy in which the develop- ing countries offer to discuss services, thus assuaging the desire to begin bringing them under trade discipline and hence helping to head off protectionist pressures on goods trade. At the same time, they can use the opportunity to ensure that their export prospects are adequately reflected in the service code. My suspicion is that, while this multilateral "constitution making" on the 21. It seems probable that the failure of the developing countries to be actively involved in the Tokyo Round negotiations on the subsidies code, for example, until fairly late may have caused the code to be written against their interests (for example, in the blanket restrictions on export subsidies) and hence have led to widespread refusal by the developing countries to sign it. 22. Feketekuty's (1986) extended analysis and documentation of U.S. visa practices in regard to domestic entry for temporary business purposes needs to be read by the skeptical among the developing countries. Evidently, there is far more room for active diplomacy and negotiations here than is commonly believed. Also, the reader should consult the entire issue of the Chicago Legal Forum (1986, vol. 1, no. 1), which deals with the question of trade in legal services across nation states and which contains the Feketekury (1986) and Bhagwati (1986c) papers, especially the papers by Barton, Cone, Noyelle, and Rossi. Needless to say, the negotiators will have to address complex issues which get even worse in dealing with unskilled labor mobility. For example, while firms can bring in professionals of different nationalities under the temporary visas, could a U.S. firm bring in Bangladeshi construction labor to Dusseldorf? Bhagwati 567 principles of a services Agreement will go on, the United States will continue to use bilateral approaches with the aid of Section 301 to pry open selected service sectors in selected countries. This is probably unavoidable, given the immense Congressional and lobbying pressures to produce quick results. To some extent, the U.S. Trade Representative (USTR) can be expected to ensure that these bilateral approaches are adopted to "set useful precedents" for the wider, multilateral code. At the same time, there is some cause for apprehen- sion that the sectoral lobbying pressures to produce results may lead to "quan- tity" rather than "rule" outcomes (as strongly suggested in an insightful study by Cho [1986] of the United States-Korea 301 episode on the opening of the Korean insurance market).23 This tendency to substitute "quantity" outcomes in favor of U.S. export sectors rather than to secure rule-oriented liberalization abroad is a peril that has not been easy to avoid. This was evident in the case of beef quotas in Japan in which the United States reportedly wanted a larger quota rather than genuine Japanese liberalization which would have allowed Austra- lian beef to triumph over that both from the United States and Japan. Similarly, in negotiations on trade in semiconductor chips, an assured market share in Japan for U.S. firms was actively urged and, while not finally in the pact, is still the basis by which Japanese "performance" on the pact has been judged to be inadequate and hence to require the retaliatory tariffs imposed by President Reagan in April 1987. This is such an interesting innovation in trade policy that I have recently (Bhagwati 1987b) christened it as VIES (voluntary import expan- sions).24 There is little that the developing countries targeted for bilateral negotiations will be able to do since it is evidently a case where the strong prevail over the weak. This is the oldest argument in the book for resort to multilateralism, which has been regarded as the only shield of the weak. Evidently, as such bilateral targeting and quantity targets multiply the wisdom of the developing countries joining in devising a multilateral compact will become increasingly evident.25 Yet another compelling reason for the developing countries to join in writing the multilateral rules is that rules written between "equals" will tend to under- play the problems that "unequals" face in service liberalization. One would have 23. Thus, one of Cho's central conclusions is that "both governments approached the case with the perception that the main issue of negotiation is the sharing of profit [rents] in Korea's insurance market. In the process of negotiation, both governments (especially U.S.) basically represented the interests of their insurance industries. The effect of the results of the negotiation on other sectors and efficiency of the economy as a whole has not been an important consideration." (1986, p. 17). 24. See also the discussion of this issue in Bhagwati and Irwin (1987) in the context of recent U.S. trade policy. 25. For those in the United States who believe therefore that such bilateralism, or "plurilateralism"- what a lovely euphemism for "regionalism" and other nonmultilateral arrangements-, is only a tactical device to get rapidly towards multilateralism, it would be wise to remember that such arrangements create vested interests against new entry, no matter what you write into the rules! S68 THE WORLD BANKECONOMIC REVIEW, VOL.1, NO.4 to be deranged to imagine the largest American banks taking over wholly from the largest five British banks in the United Kingdom if banking were fully liberal- ized. Yet such fears are routine in New Delhi and Dar-es-Salaam. The "political control" issues take an added significance in the context of such fears. I spend a fair amount of time arguing with the risk-averse that such scenarios do not make much sense for New Delhi either, that it would be an act of insanity for a large American bank to open branches in India's vast hinterland, that its clientele and operations would most likely be in international transactions. But it is evident that the unequals have fear; and, as the Russian proverb goes, fear has big eyes. So, we will need to incorporate, at least for developing coun- tries, some quantity-safeguards, just as we have GATT Article XIX as a safeguard on goods. These would have to be more generous for the developing countries, subject to eventual and negotiated erosion with "graduation," perhaps even slower-paced than as now discussed for goods. In essence, therefore, we should contemplate freer, not free, trade in services, and, contrary to the conventional rules of the strange English language where "freer" should mean more free than "free,' the developing countries should remember that it is just the other way around. But these explicit safeguards to assuage their fears will not arrive like manna from heaven. The developing countries will have to argue for this; they cannot do it if they do not actively participate in the rulemaking. Everything therefore points to one simple piece of advice for the hesitant developing countries: get into the negotiations on the code and exercise a voice; to exit is certainly the inferior alternative. REFERENCES Bhagwati, Jagdish. 1982a. "Structural Adjustment and International Factor Mobility: Some Issues." In Karl Jungenfelt, ed., Structural Adjustment in the World Economy. London: Macmillan. -. 1982b. "Taxation and International Migration." In Barry Chiswick, ed., The Gateway: U.S. Immigration Issues and Policies. Washington, D.C.; American Enter- prise Institute. . 1984. "Splintering and Disembodiment of Services and Developing Nations." The World Economy, 7, no. 2, (June): 133-143. . 1985a. "GATT and Trade in Services: How We Can Resolve the North-South Debate." Financial Times, November 27. -. 198Sb. Opening Up Trade in Services: U.S. Should Heed Third World De- mands." New York Times, November 10. . 1985c. "Trade in Services: How to Change Indian Strategy." Economic Times (India), December 2. . 1986. "Economic Perspectives on Trade in Professional Services." Chicago Legal Forum 1, no. 1 (March): 45-56. . 1987a. "International Trade in Services and Its Relevance for Economic Devel- opment." In Orio Giarini, ed., The Emerging Service Economy. Services World Econ- omy Series 1. New York:Pergamon Press. Bhagwati S69 . 1987b. "VERs, Quid Pro Quo DFI and VIEs: Political-Economy-Theoretic Analyses." International Economic Journal (Seoul) 1(1) Spring, pp. 1-15. . Forthcoming. "Export Promoting Trade Strategy: Issues and Evidence." World Bank Research Observer. Bhagwati, Jagdish, and Douglas Irwin. 1986. "Lessons from Nineteenth Century Britain: Fair Trade Could Trap the Democrats." New York Times, June 22. . 1987. "The Return of the Reciprocitarians: U.S. Trade Policy Today." The World Economy 10, no. 2 (June): 109-30. Bhagwati, Jagdish, and John Wilson, Eds. 1987. Income Taxation and International Personal Mobility, Amsterdam: North-Holland. Cho, Yoon Je. 1986. "U.S.-Korea Disputes on the Opening of Korean Insurance Market: Some Implications." Washington, D.C.: World Bank: Development Research Depart- ment. Processed. de Clercq, Willy. 1986. "The European Community and GATT Negotiations on Trade in Services." Speech at Lugano Economic and Financial Symposium, May 27. Feketekuty, Geza. 1986. "Trade in Professional Services: An Overview?" Chicago Legal Forum 1, no. 1 (March). GATT. 1983. International Trade 1982-83. Geneva: GATT Secretariat. Hill, T. P. 1977. "On Goods and Services." Review of Income and Wealth 23, no. 4 (December). Hindley, Brian. 1986a. "Introducing Services into GATT." Paper prepared for the Euro- pean meeting on the position of the European Community in the New GATT Round convened by the Spanish Ministry of Finance and the Economy and the Trade Policy Research Centre, Residencia Fuente Pizarro, Collado-Villalba, Spain, October 2-4. . 1986b. "Liberalization of Service Transactions." Washington D.C.: World Bank Development Research Department. Processed. . 1987. "A Comment on Jagdish Bhagwati's Geneva Association Lecture." In The Emerging Service Economy, Services World Economy Series 1. Orio Giarini, ed., New York: Pergamon Press. Nau, Henry. 1986. "Bargaining in the New Round: The NICs and the United States." Georgetown University Paper for the Quadrangular Forum Meeting, Vermont, July 18-20. Sampson, Gary, and Richard Snape. 1985. "Identifying the Issues in Trade in Services." The World Economy 8, no. 2 (June): 171-82. Sapir, Andre. 1985. "North-South Issues in Trade in Services." The World Economy 8, no. 1 (March): 27-41. . 1986. "Trade in Investment-Related Technological Services." World Develop- ment 14, no.5: 605-22. THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4: 571-593 Agriculture in the Uruguay Round: Interests of Developing Countries Alberto Valdes Agriculture has been favored and protected in developed countries while trade policies in developing countries frequently support industry at the expense of exportables and unprotected importables in agriculture. This protection constrains expansion of tem- perate and subtropical agricultural exports from developing countries. Several studies have estimated the effects of liberalization of trade restrictions on world prices, export earnings, and import costs. While developing countries generally would benefit from having the agricultural policies of the most powerful countries bound by international rules on trade, there are differences among the developing countries as to which prod- ucts should be liberalized. Even if such conflicts did not exist, politically feasible means to obtain such compliance are elusive. The potentially most feasible approaches for developing countries to obtain some measure of liberalization in the Uruguay Round of trade negotiations are discussed. This article first highlights some of the traditional and emerging issues of interest to developing countries related to trade policies and negotiations on agricultural products in the framework of the General Agreement on Tariffs and Trade (GATT). This is followed by a presentation of the findings of several studies on trade liberalization in agriculture. The complex pattern of agricultural protec- tion in the countries that belong to the Organisation for Economic Co-operation and Development (OECD) and its implications for the diverse exports of develop- ing countries are central themes in the second section of this article. The last section presents some thoughts on what developing countries should ask for and offer in the Uruguay Round of the Multilateral Trade Negotiations (MTN). 1. BACKGROUND: AGRICULTURE IN THE GATT SYSTEM It is now generally recognized that developed countries give significantly greater protection to agriculture than to manufacturing. While the degree of protection tends to rise and fall with world market prices, domestic food prices The author is at the International Food Policy Research Institute. He wishes to thank Jagdish Bhagwati, Suzanne Gnaegy, D. Gale Johnson, Panos Konandreas, Edward Rossmiller, and Maurice Schiff for helpful comments on an earlier version, and Jim Gilmartin for data processing. Copyright © 1987 by the International Bank for Reconstruction and Development / THE WORLD BANK. 571 572 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 in Western Europe and Japan are often twice as high as international prices (Anderson and Hayami 1986). In most industrial countries, agricultural trade policy has become an instrument to validate price-support policies aimed at redistributing income to agriculture. 1 In many developing countries, however, agriculture is taxed and manufactur- ing is usually protected from import competition. But there is a further contrast. In developed countries agriculture is explicitly protected, and it was excluded from the discipline of the GATT from the beginning. Conversely, many develop- ing countries have followed an import substitution strategy, with agriculture often neglected because industrialization was explicitly or implicitly favored. This intraindustry contrast will be important later when we discuss "reciprocity" in trade negotiations. Trade in temperate agricultural products has been a problem since the begin- ning of multilateral trade negotiations. As raw materials, many tropical prod- ucts face relatively low levels of protection in OECD countries, and there are many opportunities in these commodities to expand exports. As such, a large volume of agricultural trade in tropical products operates under GATT rules. Exporters of temperate and subtropical products, however, face several restric- tions on market access to OECD countries. This comes from protection to reduce substitutes in consumption; from tariff escalation as the level of processing increases; and from the fact that several products, such as sugar, livestock and rice, are produced in both temperate and tropical regions. Much of the trade in the main temperate and subtropical agricultural products is beyond GATT rule. The United States in the past and the European Economic Community (EEC) and Japan today have insisted that domestic farm policy measures should not be subject to international limitations and scrutiny. Exemp- tions have frequently been sought from the GATT disciplines, such as the waiver granted to the United States in 1954, and more recently the tacit acceptance of the Common Agricultural Policy (CAP) of the EEC. During the last few decades, agricultural trade has been drifting toward bilat- eral agreements and market-sharing arrangements, with the tacit acceptance of generous "safeguards" against so-called unfair practices. There has been a dra- matic increase in subsidies to agriculture in industrial countries compared with the levels prevailing ten years ago. Farming support in the United States has increased from $2.7 billion (billion is 1,000 million; dollars are U.S. dollars) in 1980 to a record $25.8 billion in 1986 and EEC taxpayers spent around $21.5 billion on farm support in 1986, up from $6.2 billion ten years ago (Economist, November 15, 1986). Levels of protection of world agriculture are higher than at the beginning of the Tokyo Round in 1973, and the major actors in trade have not shown a commitment to freer multilateral trade, nor to a stronger GATT with the power 1. There are good reasons for doubting the effectiveness of current policies in raising small farmers' incomes. Empirical work supporting this conclusion is found in GATT (1986, p. 29). Valdes 573 to establish rules and enforce them. The relative ineffectiveness of the GATT in dealing with temperate and subtropical agriculture products in the past has been recognized by most analysts. For example, the 1985 "Leutwiler Group" Report, written by an independent study group appointed by the Director-General of the GATT, contains fifteen principal recommendations on trade reforms that govern- ments need to address in the coming years (GATT 1985). The recommendations include "clearer and fairer rules for agricultural trade, with no special treatment for particular countries or commodities,' and "greater integration of developing countries into the trading system, with all the accompanying rights and responsi- bilities." While there is hardly any disagreement on the need for reform of agricultural policy, the definition of and approach to agriculture in trade negoti- ations remains a major challenge. Cereal imports by developing countries are the principal growth area in world agricultural trade while more than 65 percent of developing countries' agricul- tural export revenues came from exports to the wealthier OECD countries (Vald6s and Gnaegy 1984). More generally, the rate at which many developing countries grow is a function of their export earnings, most of which are from agriculture. A harsh reality for developing countries in agricultural trade is that adherence to multilateral trade rules by major economies (the EEC, United States, and Japan) has far greater ramifications than adherence by smaller economies (most developing countries), because the consequences of their actions on agricultural trade flows and world prices are so much more widespread. In no other major sector is such a high proportion of production sold on world markets at less than domestic prices. This is, to a great extent, a direct effect of domestic farm policies of developed countries. What can developing countries do? Because developing countries individually have little market power, and there is no unified developing country representa- tion in trade negotiations, it is in the interest of the weaker countries that the binding of the strong by international rules be accepted. In this sense, the GATT could serve as a safeguard for the weaker powers. It would appear that develop- ing countries have no real alternative to the GATT system. Looking beyond Measures Applied at the Border The analysis of nonborder policies affecting agricultural trade has two related dimensions. One is defining what constitutes a "trade" measure, considering the various forms of domestic assistance to agriculture. The other relates to the increasing influence on agriculture of macroeconomic policies affecting the ex- change rate and thus the competitiveness of agriculture. In the past, much of the international effort to improve the trading system has focused on measures applied at the border, such as tariffs, quantitative import restrictions, and export subsidies. The distinction between border and nonbor- der domestic policies breaks down, however, since domestic subsidies and taxes affect trade flows (Blackhurst 1981). This is particularly relevant for agriculture, 574 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 in which farm products are influenced by domestic price policy including finan- cial assistance, input subsidies, and tax policies. Trade policies support domestic prices; without frontier barriers on the same products, nonfrontier policies such as production subsidies would tend to have rather limited effects on trade flows in the long run.2 Up to now, domestic policies have not been effectively subject to the GATT'S trade rules, although the trade effects of subsidies received much attention in the Tokyo Round. Thus, even if the negotiations were to focus on the border measures required to operate domestic policies, eliminating border measures is a threat to domestic programs; the negotiators therefore inevitably face the question of whether domestic policy can and should be the subject of international negotiations, and what mechanisms can be devised to offset the impact of national policies on world markets where such effects are seen to be negative. In recent years, most new instruments to influence trade flows have been nontariff barriers (NTBS), which often create a problem of negotiability since their effects on trade flows are hard to assess. Several studies have documented the widespread use of NTBS in agricultural trade, although NTBs are in violation of GATT principles when used as protection devices (see, for example, Australian Bureau of Agricultural Economics 1985; Nogues, Olechowski, and Winters 1986; and World Bank 1986). There has been an upsurge in the proportion of agricultural trade subject to such NTBS as quantitative restrictions, government procurement policies, technical barriers to trade (including health and sanitary, packaging, and labeling regulations), customs valuations and nomenclature, and more recently, voluntary export restraints. Many of these NTBS, including state trading, could render most of the current rules of the GATT virtually ineffective. Given the form that protection is taking, predominantly as NTBS, and the lack of consistency of many of them with GATT rules, the GATT system is not only having increasing difficulty in furthering trade liberalization but also in safeguarding previously negotiated levels of market access (GATT 1986). NTBS are more diverse and less visible, add a considerable uncertainty to exports, and are more selective than tariffs. Trade negotiations conducted on the basis of trade flows-as implied by reciprocity-are unlikely to be productive when attempting to reduce NTBs. Reciprocity requires some kind of valuation of the mutual concessions being offered, but with so many NTBS in effect, it would be impossible to quantify concessions on either side (Dell 1986). A second major nonborder influence on agricultural trade is macroeconomic policy, which influences exchange rates and thus agricultural markets and trade. As an illustration from the experience of industrial countries, the rise in the value of the dollar in 1983-85 brought international wheat prices close to the artifi- 2. Snape (1987) makes the case for concentrating international negotiations on frontier barriers, which for agriculture means nontariff barriers-rather than on domestic subsidies, (a) because trade barriers are more harmful policy instruments for other countries than nonfrontier barriers partly because they but- tress internal price-support measures; and (b) because of difficulties of agreement on the definition and measurement of the trade effects of nonfrontier policies. Valdes 575 cially higher domestic prices of the EEC (in EEC currencies), thereby reducing their rates of protection and their export subsidies, and relieving the financial pressure on the CAP. In this case, macroeconomic policies pursued in the United States had a significant impact on the EEC'S grain policy. This raises the question of whether to use an index of protection in the negotiations. Farmers are likely to resist adoption of measures which may be favored by economists that make the extent of protection transparent, however, such as nominal rates of protection or producer subsidy equivalents (PSES). They would argue that they buy and sell in their local currencies and hence the prices they face should not be based on measures which depend so much on the exchange rates at the time of measurement. This objection is not, of course, specific to agriculture. More generally, I would anticipate that the link between protection and exchange rate misalignment will become a major issue particu- larly for developing countries as they become more integrated into the GATT. II. THE INCIDENCE OF AGRICULTURAL PROTECTION IN OECD COUNTRIES: FINDINGS OF DIFFERENT STUDIES AND IMPLICATIONS FOR DEVELOPING COUNTRIES The direct effect that industrial countries' farm policies have on other coun- tries has three dimensions: they depress world prices and thus developing coun- try export revenues; they result in savings in developing countries' import costs; and they induce greater instability in world prices. Most economists agree that to achieve more rapid economic growth, developing countries' trade regime must be made more neutral among industrial and agricultural tradables. In the medium term, however, the current slow growth of the international economy, combined with the high rates and the unpredictable nature of protection in the principal agricultural markets, makes it difficult for developing country policy- makers to hold the line against domestic pressure groups who demand a more inward-looking trade policy. There are no real prospects for developing coun- tries to reduce their foreign debt substantially without a rapid expansion of foreign exchange receipts, which come from agriculture in many countries. Most heavily indebted developing countries cannot further restrict their imports to the amount necessary to generate a larger export surplus to pay for the foreign debt. The trade restrictions imposed by developed countries include tariffs and NTBS, and they vary considerably in severity among countries and products. They all tend to lower world prices by artificialy reducing domestic consumption and raising domestic production. As a consequence, the volume of exports from nonsubsidizing countries is reduced. Price and volume effects together translate into a loss to developing country exporters of foreign exchange and welfare. On the other hand, many developing countries have benefited from trade restric- tions on cereals in developed countries as protection has led to lower world prices of their cereal imports. A few studies available now (table 1) have assessed the effects of agricultural 576 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 Table 1. Agricultural Trade Liberalization Literature in the 1 980s Trade barrier Authors Year Liberalizing Crops reduction Years of study published area liberalized (percent) covered, Valdes and 1980 OECD 99 commodities 50 1975-78 Zietz Tangermann 1982 Global Beef 25, 50, 100 1977-79 and Krostitz Koester 1982 EEC Cereals 100 1975-77 Koester and 1982 EEC Sugar 100 1978-79 Schmitz Roberts 1982 EEC Sugar 100 1968-81 Kirmani, 1984 Canada, EEC, Meat, sugar, and Molajoni, Japan, United cereals 100 1977-81 and Mayer States Matthews 1985 EEC Cereals, sugar, oilseeds, dairy products, and meat 100 1981 Schiff 1985 Argentina, Wheat __b _b Australia, Canada, EEC, Japan, U.S.S.R. Zietz and 1986 OECD Wheat, maize, Valdes beef, and sugar 100 1979-81c Tyers and 1986 (1) Global Grains, sugar, Anderson (2) OECD and livestock (3) Developing products 100 1980, 1982 countries Parikh and 1986 Global Cereals, beef, Tims and dairy products 100 1980d Note: All studies liberalize tariffs and nontariff barriers. a. Years covered in trade flow matrix and serving as base for protection levels. b. Estimates free trade, market-clearing world price based on econometrically estimated trade equations for 1960-80. 6. c. Effects based on 1983 protection levels were also c6mputed. d. Based on data for 1961-76, updated to 1980 where necessary. protection on world market prices, export earnings, and import costs, and on the resultant welfare gains and losses. Some have also analyzed the benefits and costs of trade liberalization and have identified the products with the greatest potential for export growth. Several problems arise when comparing such studies. Commodity definitions often differ. Kirmani, Molajoni, and Mayer (1984) used broad aggregates such as meat or cereals rather than specific commodities such as beef or wheat, and Valdes 577 Tyers and Anderson (1986) study coarse grains rather than maize. Valdes and Zietz (1980) distinguish between raw and refined sugar, although their 1986 study uses only total raw sugar equivalents. Also, few studies use the same base period, and prices for different years can vary widely. In 1980, for example, the world and New York spot prices of Caribbean sugar were approximately 29 and 30 cents per pound, respectively. In 1985 the world price was 12 cents and the New York spot price was 27 cents. The work of Valdes and Zietz (1980) is based on 1975-77 averages, as is that of Koester (1982); Matthews (1985) takes 1981 as his reference year; and the Zietz and Valdes (1986) study uses an average of the years 1979-81, as do Kirmani, Molajoni, and Mayer (1984). Another problem is that large differences in the level of protection can show up, as Kirmani, Molajoni, and Mayer (1984) reported for a number of surveyed studies. Finally, use of different methodologies between studies can make com- parison of results meaningless. In the light of the above problems, most of the comparisons that follow have to be viewed with caution. This author and J. Zietz analyzed a hypothetical 50 percent reduction in trade barriers for 99 commodities in 17 developed countries (table 2). The effects on the export earnings and import costs of 56 developing countries were quantified, and the most promising products for these countries were identified. Given this 50 percent tariff cut, developing countries would have increased their export revenue by 11 percent or nearly $6 billion in 1985 prices, and the Table 2. Effects of a 50 Percent Decrease in OECD Tariff Rates on Export Revenues and Import Costs for Selected Commodities of Developing Countries, 1975-77 All Middle- and developing Low-income high-income Commodity countries countries countries Change in export revenue (millions of 1985 dollars) Sugar 2,108 394 1,714 Beverages and tobacco 686 191 495 Meats 655 33 620 Coffee 540 123 417 Vegetable oils 400 60 339 Cocoa 287 21 265 Temperate-zone fruits and vegetables 197 60 137 Oilseeds and oil nuts 109 19 90 Other 883 96 788 Total increase of all exports 5,866 998 4,867 Change in import costs (millions of 1985 dollars) Cereals -876 -530 -345 Other -497 - 152 -345 Total increase of all imports -1,373 -683 -690 Note: "Developing countries" include those with populations of more than 4 million in mid-1975; country classifications as defined in World Bank (1986). 578 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 export revenues of low-income countries separately would have increased by 8.5 percent. Trade flows and OECD protection have increased since 1977-79, (the base for these calculations) so that the benefits of liberalization would be sub- stantially greater in 1985. For most commodities the price change ranged between 2 and 10 percent. As expected, the world price change was nil for the few commodities that faced little or no protection in the OECD markets, including cotton lint, jute, natural rubber, sisal, and hemp tow. At the other extreme, the calculated price change of wine, roasted coffee, malt, and cocoa paste cake fluctuated between 10 and 15 percent. Apart from the change in world price, which reflected the degree of protection, changes in export revenues were determined by the initial market share of developing countries and their relative export-supply elasticities. The study identified many other commodities with significant export poten- tial, including green coffee, wine, tobacco, and maize. Two critical commodity groups, capturing approximately 47 percent of the potential increase in export revenues for developing countries, were sugar and its derivatives (36 percent) and meats (11 percent). Latin America would have captured 63 percent of the total benefit from sugar and its derivatives, followed by Asia with 34 percent. Smaller countries not included in the sample, such as Cuba, Jamaica, and Mau- ritius, would also benefit substantially from liberalization in these areas. It should be noted that the developing countries studied capture 50 and often as much as 70-80 percent of the additional overall trade resulting from trade liberalization in agriculture. The developed countries exports that expand are commodities such as wheat, pork, and mutton and lamb. In a more recent study reported in table 3, protection and trade values were updated for wheat, maize, beef, and sugar; adjustments were introduced to allow for countries which change from being net exporters to net importers when protection is reduced; and a hypothetical removal of trade barriers (in- stead of a 50 percent reduction) was examined (Zietz and Valdes 1986). Sensitiv- ity of the results to different elasticities was also tested. The price increases predicted by the model for sugar vary between 13 and 30 percent, depending on the supply elasticity used. The price changes predicted using 1983 protection levels are more than twice as large as those estimated using the 1979-81 protection levels. Even more dramatically, world beef ex- ports would more than double following this liberalization. From 1979-81 protection levels, trade liberalization in all four commodities would be predicted to create an increase of approximately $10 billion per year in foreign exchange earnings (in 1980 dollars). The considerable increase predicted in the export earnings of developing countries reported in tables 2 and 3 does not imply that all developing countries share equally in the gains from trade liberalization in either absolute or relative terms. One way to break down the distributive effects is presented in table 2, which analyzes the potential effects of liberalization on a group of low-income countries separately (defined as those with a 1981 per capita gross national product of $400 or less). Table 3. Changes in Prices, Export Revenues, and Welfare Due to Trade Liberalization in Sugar, Beef, Wheat, and Maize: Varying Domestic Supply Elasticities Absolute change (billions of 1980 dollars) Percentage Developing Developing Developing change Developing country country country country Commodity and country World World foreign exchange exporters import net supply elasticity price exportsa earnings welfare bill welfare Sugar 0.6 for all countries 16.7 12.4 2.75 0.60 -0.33 0.08 0.06 for EEC members 13.6 10.4 2.19 0.46 -0.31 0.03 6.0forall EEc and 4.0 forallotherdeveloped 29.4 31.3 S.11 1.25 -0.42 0.39 1.2 for all developing 12.9 16.8 3.04 0.49 -0.48 0.09 Beef Bycountry(0.38-1.02) 18.5 167.7 5.10 0.54 -0.33 0.32 0.4forEEcandJapan 16.2 143.2 4.38 0.43 -0.28 0.22 Wheat By country (0.4-0.9) 12.7 10.2 1.17 0.13 -0.35 -0.66 0.8 for all developing 11.5 10.5 1.37 0.13 -0.58 -0.58 Maize By country (0.19-0.91) 11.7 35.6 0.61 0.14 -0.57 -0.07 0.08 for all developing 10.8 35.3 0.84 0.14 -0.74 -0.04 a. The sum of net exports of all net exporting countries Source: Zietz and Valdes (1986). 580 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 As suggested in table 2, low-income countries as a group would be hurt by trade liberalization in cereals. Table 2 also suggests that a small percentage of the foreign exchange gains from meats accrues to low-income countries. Liberaliza- tion for beef exports is mainly in the interest of middle-income developing countries, particularly those in Latin America, although the increase in export earnings is split about equally between developed and developing countries. The gains to both low and middle income developing countries from trade liberaliza- tion in sugar are large, and only a small fraction of the total increase in export earnings attributable to sugar is captured by developed country exporters. Most of the gains from liberalization of trade in wheat and maize would accrue to Australia, Canada, and the United States at the expense of the EEC, and to Argentina, Thailand, and a few developing country cereal exporters which are the exceptions. Table 3 shows the net welfare changes resulting from liberalization of four commodities. The welfare benefits from trade liberalization calculated here rep- resent a transfer of real income to developing country exporters equal to the increase in export earnings less the resource costs of increased exports. That is, the welfare gains to developing countries come from an increase in the value of the original volume of exports and the increase of producer surplus in the additional production for export. Similarly, the welfare loss results from the increased value of the lower volume of imports plus the loss in consumer surplus from the reduction in imports. The results shown suggest how interests among developing countries diverge according to the commodity being considered, as they do on other fronts in the MTN. In their study of the beef market for the Food and Agriculture Organization (FAO), Tangermann and Krostitz (1982) find that a complete removal of trade barriers would result in a world price increase of 47 percent, a value consider- ably greater than that reported in table 2 (a comparison of several studies and some of their calculated world price increases are presented in table 4). In addition, they estimate that world exports would increase by 300 percent, which is about double the values found in the Zietz and Valdes (1986) study. One reason for the larger percentage changes predicted by Tangermann (1980) is his assumption that trade barriers are removed in OECD and developing countries. Koester (1982) uses the same methodology and data base as Valdes and Zietz (1980) but concentrates his analysis on cereals. Assuming a complete removal of trade barriers in the EEC, Koester predicts increases in the world prices of wheat and maize of 9.6 and 2.2 percent, both less than those shown in table 3. They are based on the assumption that trade barriers are removed only in the EEC however, rather than in all OECD countries. Koester also finds that trade liberali- zation in cereals as a whole results in a net welfare loss to developing countries. Koester and Schmitz (1982) examine the effect on all beneficiary African, Carribbean, and Pacific (ACP) countries of the protection given to sugar produc- tion in the EEC (the Valdes and Zietz studies are limited to those countries with 5 million or more inhabitants). Given the gains from higher export prices under Valdes 581 Table 4. Estimated Effect on World Prices of Trade Liberalization in Selected Commodities Estimated increase in world price (percent) Source Cereals Meat and dairy Sugar Others Valdes and Wheat (5) Beef and veal (7) Raw (8) Roast coffee (11) Zietz Wheat flour (7) Beef preparations (4) Refined (6) Cocoa powder (1980) Maize (2) Pork (9) Confectionary (9) (14) Rice (0.4) Mutton and lamb (4) Wine (16) Barley (8) Chicken (3) Soybeans (1) Sorghum (0.6) Tea (3) Palm oil (3) Tangermann Beef (7-47) and Krostitz (1982) Koester Wheat(10) (1982) Barley (14) Maize (2) Rye(9) Oats (20) Sorghum (0.6) Koester and Sugar (12) Schmidt (1982) Roberts Sugar (7.2-11.3) (1982) Matthews Wheat (0.7) Beef (4) Sugar (11) (1985) Barley (3) Pork(4) Maize (0.5) Mutton (5) Rice (0.1) Poultry (3) Butter (11) Milk powder (8) Schiff Wheat (15) (1985) Zietz and Wheat (12) Beef (16-18) Sugar (13-29) Valdes Maize (11) (1986) Tyers and Wheat (2) Beef and lamb (16) Sugar (5)b Anderson Coarse grains (1) Pork and poultry (2) Sugar (3.2)c (1986)a Rice (5) Dairy products (27) Parikh and Wheat (18) Bovine and ovine Tims Coarse grains products (17) (1986) (11) Rice (21) a. In a more recent study, these authors have significantly revised their estimates of the price effects of protection. The same authors provide estimates of price effects in their 1983 study. b. Liberalization in industrial market economies only. c. Liberalization in all developing economies only. 582 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 the EEC Sugar Protocol, Koester and Schmitz conclude that a removal of EEC trade barriers would result in a net loss to the ACP beneficiaries except for India and Kenya. A study by Roberts (1982) of EEC sugar trade calculates that devel- oping countries as a whole could expect an increase in welfare of between $370 million and $570 million, compared with a loss to the ACP countries under the Sugar Protocol of about $170 million. Roberts's results cannot be compared directly with the figures in Zietz and Valdes (1986) because of different assump- tions regarding the scope of trade liberalization. More recently, the World Devel- opment Report 1986 (World Bank 1986) presents estimates of the effects of the Lome Convention on the income transfer derived from EEC sugar quotas to ACP countries. The studies reported above isolate the effects of liberalization by developed market economies. Tyers and Anderson (1986), and Parikh and Tims (1986) try to quantify the likely effects of global trade liberalization including liberalization of developing countries' agricultural trade. The results are generally similar to those of Valdes and Zietz (1980) and Zietz and Valdes (1986) for OECD trade liberalization. Tyers and Anderson and Parikh and Tims, however, find no loss of welfare to developing countries from liberalization of cereals, and they predict a substantial loss in foreign exchange earnings rather than an increase. This is a result of the positive nominal protection rates for cereals found in several devel- oping countries. Positive nominal protection in many developing countries, however, can be viewed as partial compensation for significant discrimination against agricultural import-substitutes resulting from an overvalued exchange rate, high levels of industrial protection, and macroeconomic policies that favor industry over agriculture (Valdes 1986). While the selection of which countries liberalize their agricultural trade makes a major difference in the potential trade effects, all studies that analyze OECD trade liberalization in agriculture predict an increase in the world price if part or all of the barriers to trade are removed (table 4). The extent of this increase is highest for the most protected commodities, such as sugar, beef, and dairy products. Sugar prices are predicted to increase between a low of 5 percent (Tyers and Anderson 1986) and a high of 13 to 29 percent (Zietz and Valdes 1986), and both studies suggest a 16 to 20 percent increase in the world price for beef. For dairy products, Tyers and Anderson (1986) put the increase at 27 percent. These world price increases accord well with the results reached by the linked system of national agricultural policy models developed at the International Institute for Applied Systems Analysis (IIASA) (Parikh and Tims 1986). Using 10 commodity classifications, the IIASA model predicts a long-run increase in world prices of about 9 percent. This figure translates into a foreign exchange gain for developing countries of about $7 billion for all commodities taken together. Compared with Zietz and Valdes (1986) and Tyers and Anderson (1986), this overall foreign exchange gain seems low. Both these studies arrive at a figure of around $8 billion, for Valdes 583 sugar and beef taken alone, evaluated at 1980 prices. According to Tyers and Anderson (1986), trade liberalization in dairy products would generate an addi- tional $7.8 billion of foreign exchange for developing countries. Another reason why Parikh and Tims may underestimate the effect of liberalization is that neither they (nor Tyers and Anderson), analyze the potential benefit to develop- ing countries in the major nonstaple export crops, as opposed to the subset of foods and feeds from temperate zones. As Valdes and Zietz (1980) have shown, developing countries are likely to realize substantial gains in foreign exchange if trade barriers were lowered or removed for such products as tobacco, roasted coffee, coffee extracts, cocoa derivatives, or oils and seeds. The authors have shown that the likely gains on these products (for developing countries as a whole) would also more than compensate for the losses developing countries could expect from the price increases in cereals, which many of them currently import. Because of the limited commodity coverage of the IIASA and Tyers/ Anderson models one cannot conclude from them that developing countries would suffer a net welfare loss from OECD agricultural trade liberalization. A second benefit likely to be obtained through liberalization is a reduction of world price instability. This effect results from the removal of some NTBS, specif- ically variable levies, which insulate domestic markets from border prices. Both Schiff's recent study of wheat in the EEC (1985), and the Tyers and Anderson study (1986) arrive at the conclusion that changes in the system of protection would contribute significantly to world price stability. For example, Schiff (1985, table 25) estimates that the variability of world prices of wheat would fall from a coefficient of variation of 0.46 to 0.32 if the EEC alone were to remove its trade barriers on wheat. If less price instability is valued highly, as the discussion of commodity price stabilization and buffer stocks seems to indicate, then the increased price stability resulting from liberalized trade should be included as a benefit in an overall evaluation of trade liberalization. There are several limitations to these trade liberalization models. While none of them includes adjustments for the exchange rate, some policy simulations may affect the exchange rate, which could in turn affect the domestic responses to the initial change in world prices. Supply elasticities tend to be much larger in the long than short run, and permanently reducing trade barriers in developed countries would lead develop- ing countries to develop new export products and to expand their processing operations. In addition, it would probably encourage developing countries to direct more resources toward increasing agricultural production, helping to break the current climate in developing countries of "export pessimism" that inhibits the adoption of export-oriented policies in agriculture, as well as in other sectors. The increased uncertainty of world prices and of market access due to current OECD policies may also affect the trade response of developing countries. If OECD countries were to liberalize their agricultural markets, there would be lower and perhaps less stable food prices in the highly protected industrial countries but S84 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 higher and more stable food prices in the rest of the world. This change may stimulate developing country governments both to reduce their explicit or im- plicit taxation of agricultural prices and to expand public investments in agricul- ture in order to take advantage of the more profitable and less risky opportunities to earn export income in international agricultural markets. III. THE RECENT GATT COMMITTEE'S PROPOSAL ON MINiMuM-AccEss IMPORT REQUIREMENT AND PRODUCER-FINANCED EXPORT SUBSIDIES There are two potentially important proposals by the GATT'S Committee on Agricultural Trade that deserve special attention because of the way they relate market share agreements to domestic prices and the financing of subsidies. These two proposals, submitted to the GATT Council in 1984 and 1985, can be understood as attempts to end the developed countries' increasing reliance on tighter import restrictions or subsidized exports to manage their internal supply problems. These proposals were the subject of a quantitative evaluation by Valdes and Zietz (forthcoming). Export subsidies, according to the Committee's report, should only be main- tained if they are financed by the producers, thereby diminishing the net price they actually receive. Exactly how this might work depends on the particular scheme that would be adopted to collect the funds to finance the export subsidy. The Valdes and Zietz (forthcoming) study assumed that an ad valorem tax would be placed on total production, which would reduce total production. The second proposal would require that countries maintaining import restric- tions allow imports to reach some minimum percentage of domestic production, which was taken to be 10 percent for illustrative purposes. All import-restricting countries that do not already import at least 10 percent of domestic production in the initial situation would be subject to this "minimum-access" requirement. This rule would apply equally to countries that are net importers initially as well as to countries that are net exporters (because of import restrictions rather than because of comparative cost advantages). Four alternative scenarios and their respective quantitative implications were investigated in the Valdes and Zietz study. The effects shown by the model simulations of these proposals on the world price and the foreign exchange earnings of 58 developing countries are summa- rized in table 5. As one might expect, removal of trade barriers in the OECD countries results in an increase in world prices in each scenario, as the world import demand curve shifts to the right. The size of the predicted increase depends not only on the type of measure proposed, that is producer-financed export subsidies or minimum access, but also on the type of domestic adjustment one hypothesizes for those countries directly affected by the changes in the trade regime. The assumed effects of the GATT proposals were compared with results in the Zietz and Valdes (1986) study, which assumed a reduction of tariffs and NTBS in Valdes 585 Table 5. Effect of Proposed Minimum-Access Import Requirement and Producer-Financed Export Subsidies on World Price and Developing-Country Foreign Exchange Earnings of Sugar Change in developing country foreign exchange earnings EEC Change in Constant nominal Proposed world price millions protection measure (percent) of dollars Percent coefficient Producer-financed export subsidies 0.73 105.6 4.0 0.57 Minimum access Constant protection level 1.91 276.8 10.4 0.57 Constant domestic price, 2.30 334.2 12.6 0.54 Constant absolute surplus' 2.10 305.9 11.5 0.39 Domestic market clearingc 6.74 1,018.7 38.4 0.007 a. Countries reduce their level of protection in the face of a rising world price so as to keep the domestic price level constant in absolute terms. b. The level of protection is reduced so that the domestic surplus in the final equilibrium just equals the initial surplus before minimum access. c. The protection level is reduced to a level that just equilibrates domestic demand and supply. Source: Valdes and Zietz (forthcoming). a "most-favored-nation" approach. There is a significant discrepancy between the effects of these two approaches because both producer-financed export sub- sidies and the minimum-access rule are tied to conditions that a priori relieved many highly protected OECD countries from the disciplines of the proposed GATT schemes. Only two developed countries on the export scheme and three countries on the minimum-access rule (including the EEC) are directly affected. Countries with highly protected trade regimes in sugar such as Canada, Japan, Switzerland, or the United States would not be forced to reduce their protection under the GATT proposals.3 This is because they either do not export sugar and hence are not subject to the export subsidy regulations, or they import an amount equal to or in excess of domestic production and hence evade any discipline under the minimum-access rule. None of the other developed countries either exported sugar or imported less than 10 percent of domestic production during the years 1979 to 1981. The results hardly change if, say, the 10 percent rule is simply replaced by a 20 or 30 percent rule. Preliminary work on beef seems to lead to similar conclusions. The results may be more significant for other commodities (dairy products, for example). The Valdes and Zietz evaluation of the GATT proposal concluded that, al- though developing countries are divided as to which of the two schemes is preferable (depending on whether they are importers or exporters of the prod- ucts in question), for developing countries and consumers in the affected OECD 3. By 1990 however, the United States may import less than 10 percent of its sugar consumption and thus the proposal on imports could have a marginal effect. 586 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 countries alike, minimum access with "constant absolute surplus" or "domestic market clearing" seems to be preferable. This is based on the assumption that countries refrain from increasing their exports by an amount equal to or close to their minimum import requirement. Should this assumption not hold, then one could not expect an increase in world price, or any resultant foreign exchange gains for developing countries. However, a problem with this conclusion is that while trade regimes may be influenced by the GATT, the domestic reaction of governments to the ensuing changes may not be. So unless the proposal is complemented with restrictions on protection, even if "minimum access" should be adopted as part of a new revised GATT code on trade in agriculture, there is still a wide variety of ways in which individual countries can adjust, only a few of which would be clearly in the interest of consumers and developing countries. One may conclude that changes in the GATT code with regard to export subsidies and a minimum-access requirement are a step in the right direction but are not a substitute for a general reduction in trade barriers. The analysis also suggests that if any such scheme were implemented in practice, disagreement among the affected countries on the interpretation of the accompanying rules is very likely. It reinforces the view that there is an urgent need to strengthen the present GATT rules and disciplines on subsidies and quantitative restrictions, and to improve dispute settlement procedures. IV. APPROACHES TO NEGOTIATIONS ON AGRICULTURAL TRADE IN THE URUGUAY ROUND: WHAT DEVELOPING COUNTRIES SHOULD ASK FOR AND OFFER In light of the discussion up to this point, one might then ask, what developing countries should be seeking from the MTN, what they could offer, and what they should avoid. However, a reference to "developing country views" should be interpreted with caution-like industrial countries, developing countries have conflicting interests with regards to particular commodity markets. In this sec- tion, three sets of issues are discussed: (1) direct actions to increase market access to developing countries, (2) strengthening GATT rules and disciplines, and (3) reciprocity and what developing countries could offer. Direct Action to Increase Market Access Several implications emerge from the quantitative evidence presented in sec- tion II of this article of relevance for negotiating strategies for developing coun- tries. Broadly, the survey presented in section II underscores the need for developing countries to ensure that agriculture is high on the agenda in the Uruguay Round. Another feature worth highlighting is that the literature on the trade liberalization models suggests that the adjustment costs to trade liberaliz- ing countries would be lower under multilateral liberalization, as compared with unilateral reduction in trade barriers. The loss in producer welfare in the liberal- Valdes 587 izing countries is smaller if other countries liberalize at the same time, a fact which is politically important. The larger the number of protected markets liberalizing trade in a given product, the larger the increase in the world price and so the smaller the reduction in the producer price required in the protected markets. On the questions of who liberalizes (OECD, EEC, Japan, developing countries), what trade barrier is lifted (quantitative restrictions and/or tariffs), and which crops are affected, two major implications from the previous sections arise. One is that what really matters for developing economies, and for world trade in agriculture more generally, is trade liberalization in industrial countries. Trade policies of two other actors are also influential in the final outcome. The U.S.S.R., not a member of the GATT, has become a major importer in the world markets for grains, butter, sugar, and to a lesser extent poultry, beef, and pork. Developing economies are now the fastest growing market for grains, at what I would argue have been low levels of protection, with exceptions like the Repub- lic of Korea and Taiwan. But barriers to imports in developing economies are influential in several other products (jute, fruits, and others), and their trade policies should be addressed in the Uruguay Round. Still, as far as we can tell from the available research, the industrial countries are the dominant actors in most agricultural trade and are the principal contributors to the "disarray" of world agricultural markets, to use Johnson's (1973) all too appropriate term. High domestic prices are the basic problem behind the current overcommitment of resources to agriculture in OECD countries. A second principal implication of the studies examined is that the importance of NTBs and nonborder types of intervention strengthens the argument for con- centrating less on trade flows and more on domestic price levels. Ideally, empha- sis at the MTN should be on NTBS and on all domestic policies that may have trade effects. Export subsidies, variable levies, and other interventions are an integral part of domestic policies aimed at farm income support. To effect any substantive change, negotiations must not be restricted to tariffs. The "tariff escalation" phenomenon is important for developing countries, however, and thus tariff reduction in semiprocessed foods, for example, is an important ele- ment on the agenda. What is needed is an overall measure of protection and, for the next three to four years, a binding of domestic supports based on a direct comparison of domestic and international reference prices,4 to be followed by a gradual annual percentage reduction of protection. Measures of nominal protection, or of pro- ducer and consumer subsidy equivalents, are possible indexes of protection, capturing the effects on domestic prices of import quotas, variable levies, tariffs, 4. Such a suggestion was made by the EEC at the Kennedy Round in the "moutant de soutien" or margin-of-support scheme. It was rejected by the United States because it did not contain minimum- access guarantess to the EEC market. 588 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 acreage diversion payments, and export subsidies. While input subsidies cannot be captured by measures of nominal protection, direct measures of the levels of nominal protection on tradable inputs could be estimated. This would be desir- able if measures of effective rates of protection (ERPS) are unsatisfactory or agreement on an appropriate system is difficult to reach. Some direct income transfers to farmers (such as social security but not deficiency payments as currently used in the United States) could be left out of the calculations, to the extent that they do not influence the level of protection to producers or prices paid by consumers. T. Josling's (1982) work on consumer- and producer-subsidy equivalents (CSES; PSES) is relevant as a possible measure of overall intervention. More re- cently, the OECD'S Trade Mandate Study on agricultural protection among OECD member countries derived measures of CSES and PSES. The United States and EEC are currently in "hot debate" internally about how to use these measures in trade negotiations. The question of commodity coverage in the negotiations raises a thorny issue for developing countries. The results in section II show how interests among developing countries diverge with respect to which commodities should be given priority. Most developing countries would benefit from priority being given to products other than cereals, such as sugar and meat. The most controversial commodity group among developing countries is cereals, because of the depen- dency of so many of them on cereal imports. The developing countries' negotiat- ing positions could be weakened if they each press individually for their "preferred" products in the negotiations, however, and liberalization of cereals trade appears to be high on the priority list of developed countries in the negotia- tions (and is a high priority for a few developing countries such as Argentina, Thailand, and Uruguay). The conflict regarding cereals raises the question of whether the negotiations should proceed commodity by commodity, which is likely to be the preference of some developed countries. Sugar, cereals, and dairy products are often singled out as obvious candidates. Such an approach is attractive as a practical way of getting the process started, when engaging in a battle against protectionism across many sectors simultaneously appears politically very difficult. The results for developing countries, however, are likely to be poor. Individual commodity conventions have been attempted in the past (such as in wheat, dairy, and meats), and they have been unsuccessful. They do not allow for negotiations across sectors (a point examined below) and therefore are unlikely to be given a high level of political commitment. As a result, commodity conventions are negotiated by sectoral representatives, whose political support comes from the farm lobbies in their own countries rather than from a broader range of pro- ducer and consumer interests. Furthermore, such an approach increases the likelihood of deals among the major industrial countries, with little attention being paid to the needs of developing countries. Valdes 589 Strengthening GATT Rules and Disciplines Rulemaking for agricultural trade in the 1990s is expected to become a critical issue at the Uruguay Round, requiring agreement on new issues and new rules to be followed. This includes deciding on abolition of permanent agricultural ex- ceptions, and whether to ban or limit certain policy instruments, such as subsi- dies. The GATT negotiations have been successful in reducing industrial tariffs and enforcing these reductions. Unfortunately, this has not been the case with non- tariff barriers, which are the dominant form of protection in agriculture. Critical problems for agricultural exporters have been quantitative restrictions (QRS), and domestic subsidies, selective safeguards targeting specific exporting coun- tries (which are historically damaging to developing countries), and the weak- ness of GATT mechanisms for surveillance and for dispute settlement. The disputes panel, which is charged with handling enforcement of decisions, has no real power to put the recommendations into effect. Strengthening the GATT'S capacity and authority to evaluate, expose, and sanction violaters of agreed- upon accords would become a very important means to increase market access. Developing countries should be particularly interested in becoming active participants in an international system which provides a framework of norms, rules, and procedures. The alternative could easily be discriminatory protection and bilateralism, which is much less promising for small economic actors. A strong monitoring mechanism within the GATT could implement surveillance of national trade policies, increasing each government's accountability for its own trade-related practices. This would be particularly useful for small countries and for most developing countries. Reciprocity and What Developing Countries Could Offer The major trading powers-the EEC, Japan, and the United States-have to reach agreement among themselves if anything significant is to happen during the MTN. The risk for developing countries is that the three big powers could strike a deal among themselves before starting negotiations with the developing countries, which could be highly deleterious to the long-term interests of devel- oping countries and to the GATT system as a whole. Developing and developed countries (other than the EEC, Japan, and the United States), must offer some incentive to the trading powers in order to become influential. I believe this incentive could come from two fronts. First, developing countries must be pre- pared to reciprocate in trade concessions, on the premise that developed coun- tries would then be more willing to make additional concessions. Second, they must be willing to compromise on preferential treatment and graduation, issues to be examined briefly below. The application of GATT rules is more lenient for developing countries than for industrial countries, allowing developing countries to enjoy the benefits of re- 590 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 duced protection in developed country markets while not making concessions themselves. This has been supported by a tacit acceptance of import substitution as a development strategy. The economic importance of developing countries in trade negotiations has been diminished as a result. Industrial protection, commonly adopted in developing countries, helps in- dustry at the expense of agriculture. A policy that protects industry raises the cost of tradable inputs, such as fertilizers, machinery, and other inputs, and the resulting changes in the exchange rate penalize producers in other import-com- peting sectors, as well as those producing exportables. As a result, resources move from the traded agriculture sector to the protected and nontraded sectors (Valdes 1986b). The losers from industrial protection are consumers of indus- trial goods and the nonprotected activities. This depresses the potential of agri- culture, which is a highly tradable sector compared with the rest of the economy in most developing countries. Liberalization of industry could be used to provide additional leverage for developing countries in agricultural negotiations during the MTN, and in the process, help to promote the growth of their agriculture. Greater freedom in agricultural trade could be the price demanded by developing countries in ex- change for the prospect of freer trade in their industry (and services?) and better terms for foreign direct investment in their economies. A direct implication of the above is that agricultural negotiations should not be self-contained. In order to gain in agriculture, developing countries would have to provide bargaining chips from the nonagricultural sector, which is not compatible with single-com- modity negotiations. The second incentive developing countries could offer in the MTN relates to preferential treatment. A critical issue is whether the growth and adjustment of developing economies call for more or less preferential treatment in trade policy. Some developing countries still believe that their leading objective in any GATT negotiation should be to preserve their preferential access to developed country markets. Most trade economists argue, however, that greater benefits to devel- oping countries have come from the reductions in tariff barriers under the GATT, rather than from preferences under the Generalized System of Preferences (GSP). The GSP is restrictive at best, and many agricultural products are excluded in the U.S. scheme, while most are excluded by both the EEC and Japan. The effect of the Lome Convention on sugar exports is significant for only the smallest econo- mies among the eighteen ACP countries that have quotas to export sugar to the EEC. The Leutwiler Report (GATT 1985), like others, concluded that special and preferential treatment has been of limited value for most developing countries. And there is a risk that insistence on preferential treatment could have a high opportunity cost in terms of concessions made and that it would only detract attention from other more important issues. Preferential treatment of the small- est and poorest, however, seems appropriate. Perhaps rather than emphasizing a political alignment on North-South lines, a fresh approach would be for devel- Valdes 591 oping countries to focus negotiations on issue alignment, with the decision on where to align based on the selection of those specific measures and rules of high priority for developing countries' own self-interest, such as export subsidies and quantitative restrictions. In this regard, the initiative by the Cairns Group (which includes some developed countries) is a very positive development for agricultural-export-oriented developing countries, both as a mechanism to help keep the negotiations within a multilateral framework and to focus the negotia- tions on an issue alignment. This approach implies differentiating the rights and obligations of smaller, low-income developing countries from those of large, middle-income developing countries, including newly industrialized countries. The argument is based on the fact that, first, the rent these small, low-income developing countries derive from trade preferences represents a significant amount of their foreign exchange supply. Second, these economies are highly vulnerable to foreign trade fluctua- tions in one or two commodities. Third, the cost to preference-giving countries is relatively small. Fourth, the opening of the trade regime in the graduating developing countries would, it is presumed, represent an additional force to help these poorest developing countries. The graduation from trade preferences to reciprocity for middle-income devel- oping countries would probably influence coalition building and may create divisions between them and low-income developing countries. Is this inevitable? This differentiation between low and middle-income countries does not imply that low-income ones have no incentives to support middle-income ones in their new (hypothetical) stance in the negotiation. If the same level of implicit income transfer from nonreciprocity and preferential treatment is now reserved exclu- sively for low-income countries, this would presumably be considered a benefit for them. Furthermore, the product mix of their exports is changing and the pressure of middle-income countries for more access to OECD markets and to other imiddle-income countries, for products such as sugar and derivatives, vege- table oils, beverages, tobacco, and fruits is also of great potential interest to low- income countries as exporters. Agricultural trade is only one area of concern for developing countries, however, and the formation of coalitions involves many. Ultimately it can be expected that developing countries will find greater benefit from negotiating in concert than from taking on their largest trading partners alone. REFERENCES Anderson, K., and Y. Hayami. 1986. The Political Economy of Agricultural Protection. Boston: Allen and Unwin. Anderson, K., and R. Tyers. 1983. "European Community's Grain and Meat Policies and U.S. Retaliation: Effects on International Prices, Trade and Welfare?" Center for Economic Policy Research, Discussion Paper 83. Canberra: Australian National Uni- versity. October. 592 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 Australian Bureau of Agricultural Economics. 1985. Agricultural Policies in the Euro- pean Community: Their Origins, Nature, and Effects on Production and Trade. Policy Monograph 2, Brisbane: Watson Ferguson. Blackhurst, Richard. 1981. "The Twilight of Domestic Economic Policies." World Econ- omy 4, no. 4 (December): 3 57-74. Dell, Edmund. 1986. "Of Free Trade and Reciprocity." World Economy 9, no. 2 (June): 125-40. GATT. 1985. "Trade Policies for a Better Future: Proposals for Action." Leutwiler Group Report. Geneva. - .1986. "Prospects for International Trade,' GATT/ 1392, Geneva, September. Johnson, D. G. 1973. World Agriculture in Disarray. London: Fontana. Josling, Timothy. 1982. "Developed-Countries Agricultural Policies and Developing Country Supplies: The Case of Wheat." Research Report 14, Washington, D.C.: Inter- national Food Policy Research Inititute Kirmani, N., P. Molajoni, T. Mayer. 1984. "Effects of Increased Market Access on Exports of Developing Countries.' IMF Staff Papers (December): 661-84. Koester, Ulrich. 1982. Policy Options for the Grain Economy of the European Commu- nity: Implications for Developing Countries. Research Report 35, Washington, D.C.: International Food Policy Research Institute. Koester, U., and P. Schmitz. 1982. "The EC Sugar Market Policy and Developing Coun- tries." European Review of Agricultural Economics 9: 183-204. Matthews, Alan. 1985. The Common Agricultural Policy of the Less Developed Coun- tries. Dublin: Gill and Macmillian. Nogues, Julio J., Andrzej Olechowski, and L. Alan Winters. 1986. "The Extent on Nontariff Barriers to Industrial Countries' Imports:' World Bank Economic Review 1, no. 1:181-99. Parikh, K. S., and W. Tims. 1986. "From Hunger amidst Abundance to Abundance without Hunger." Summary Findings by the International Institute for Applied Systems Analysis. Laxenburg, Austria, February. Roberts, I. M. 1982. "E.E.C. Sugar Support Policies and World Market Prices: A Com- parative Static Analysis." Australian Bureau of Agricultural Economics, Working Paper 8213. Canberra. Schiff, M. 1985. An Econometric Analysis of the World Wheat Market and Simulation of Alternative Policies, 1960-80. Washington, D.C.: U.S. Department of Agricultural Economic Research Service, October. Snape, Richard. 1987. "The Importance of Frontier Barriers." In Henryk Kierzkowski, ed., Protection and Competition in International Trade: Essays in Honor of W Max Corden. Oxford: Blackwell. Tangermann, Stefan. 1980. "Protectionism in the Livestock Sector." Rome: FAO CPP MEAT/4. October. Tangermann, Stefan, and W. Krostitz. 1982. "Protectionism in the Livestock Sector with Particular Reference to the International Beef Trade." Paper prepared for Ninth Session of U.N. Food and Agriculture Organization's Intergovernmental Group on Meat, Rome, December. Processed. Tyers, R., and K. Anderson. 1986. "Distortions in World Food Markets: A Quantitative Assessment." Paper prepared for World Bank, World Development Report 1986. Processed. Valdes 593 Valdes, Alberto. 1986a. "Exchange Rates and Trade Policy: Help or Hindrance to Agri- cultural Growth,' In A. Maunder and U. Renborg, eds., Agriculture in a Turbulent World Economy, London: Gower. - 1986b. "Impact of Trade and Macroeconomic Policies on Agricultural Growth: The South American Experience,' In Economic and Social Progress in Latin America. Washington, D.C.: Inter-American Development Bank. Valdes, Alberto, and Suzanne Gnaegy. 1984. Trends and Structure of Agricultural Trade among Developing Countries, 1962-1979. Washington, D.C.: International Food Policy Research Institute. Valdes, Alberto, and Joachim Zietz. 1980. Agricultural Protection in OECD Countries: Its Cost to Less-Developed Countries. Research Report 21. Washington, D.C.: Inter- national Food Policy Research Institute. Forthcoming. "Export Subsidies and Minimum Access Guarantees in Agricul- tural Trade Negotiations: A Developing Country Perspective.' World Development. World Bank. 1986. World Development Report 1986. New York: Oxford University Press. Zietz, Joachim, and Alberto Valdes. 1986. The Costs of Protectionism to Developing Countries: An Analysis for Selected Agricultural Products. World Bank Staff Working Paper 769. Washington, D.C. THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4: 595-618 GATT Law, Agricultural Trade, and Developing Countries: Lessons from Two Case Studies Surakiart Sathirathai and Ammar Siamwalla This article examines the relevance and applicability of the law of the General Agree- ment on Tariffs and Trade (GATT) to two specific problems faced by Thailand: the negotiations of a voluntary export restraint agreement on cassava with the European Economic Community and the increased subsidies on rice given by the United States under the Food Security Act of 1985. In the case of cassava, Thailand appears to have had parts of GATT law on its side, but the government was very reluctant to use the law to its own advantage. This reluctance was due to unclear procedures under GATTas to how to make effective use of these legal advantages. On the rice subsidy issue, conversely, the substantive law is unclear and provides limited protection for competing exporters. The wider lesson drawn from the two cases is that GATT's law should be modified and its role reevaluated so that both developed and developing countries can participate more fully in the GATT system. This will be necessary if the GATT's laws are to become useful instruments in the hands of developing countries in their bilateral negotiations with contracting parties which are the more powerful economically. Agricultural trade has the pride of place in the new round of multilateral trade negotiations (MTN), but in the brave new world that is to come after this latest round, the landscape of international agricultural trade will remain substantially unchanged. No doubt, concessions will be made during the round and some order put into the present chaos of subsidies, but there will still be voluntary export restraints (VERS), and the subsidy rules will still be full of exceptions, or worse, ambiguities. After and even during the upheaval of the MTN, normal life will continue, bilateral deals will be cut and adjustments will be sought by some countries against actions that seriously affect their economic interests. The GATT is seen in this article as the backdrop against which a number of Surakiart Sathirathai is lecturer, Faculty of Law, Chulalongkorn University, Bangkok. Ammar Siamwalla is program director, agricultural and rural development, Thailand Development Research Institute, Bangkok. The section in this article on cassava trade was based on extensive research carried out by the Thailand Development Research Institute and financed by the Cassava Trust Fund of the Cassava Trade Association, Bangkok. The authors wish to express their gratitude to Robert Hudec for his constructive comments. Copyright © 1987 by the International Bank for Reconstruction and Development / THE WORLD BANK. 595 596 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 bilateral trade negotiations take place. The importance of GATT law, inasmuch as it is implicitly accepted as a constraint on the behavior of sovereign nations, lies in the ever-so-slight shift in bargaining power that occurs as a result of its existence.' Consequently, the next round of MTN should examine the workings of GATT law in this (usually neglected) context. The task of this article is to persuade readers to pay attention to the "normal" type of international agricultural negotiations, which are usually bilateral and commodity-specific. From our point of view, a successful MTN round is one in which the conditions of "normality" improve the negotiating or trade position of the weaker developing countries. The mode of analysis we shall use is the case study method. We examine in depth (in sections I and 11) two problems of agricultural trade faced by Thailand in the last few years, one involving trade with the European Economic Community (EEC), proponent of the conciliatory school of GATT law, and the other involving the United States, a proponent of the adversarial school. The first case concerns a VER while the second involves an export promoting subsidy, the legal status of which is still untested. Both issues are of concern in the (MTN). The two cases studied allow us to examine problems arising out of the application of GATT law in the final section of the article. 1. THAI-EEC CASSAVA NEGOTIATIONS Thai-EEC Cassava Trade until 1982 Since the inception of its Common Agricultural Policy (CAP), the EEC has allowed its cereal prices to stay well above world price levels. The mechanism used to maintain these prices is a variable import levy or, if the commodity is exported, a variable export restitution (that is, a subsidy for exports). Between 1977-78 and 1979-80, the nominal rate of protection for cereals averaged about 90 percent (Australia 1985, table 4). These high prices have discouraged consumption of cereals in the EEC, particularly as animal feed. It was very quickly found that noncereal substitutes could be used as animal feed just as well. Thus, 0.77 kilogram of cassava, which is almost entirely starch, can be mixed with 0.20 kilogram of soybean cakes, a protein-rich sub- stance, to replace 1.00 kilogram of barley in pork rations (Koester 1982, p. 35). Similar proportions apply for other cereals in other rations. Furthermore, both cassava and soybeans are subject to the relatively low import levies of 6 and 0 percent respectively. The EEC bound these tariffs in the Dillon and Kennedy MTN rounds. The cassava trade with the EEC was initiated by Brazil and Indonesia, both of which have large cassava production, but they were quickly overtaken by Thai- 1. The view that international law can serve as a constraint on powerful states and thus provide bargaining power for developing countries is in line with the approach taken by various legal scholars. For representative examples see Schachter (1977) and Friedmann, Henkin, and Lissitzyn (1972). Sathirathai and Siamwalla 597 land, which until then had been a relatively small producer. Between 1967 and 1980, production of cassava roots in Thailand doubled every three to four years. By 1982 the EEC market for cassava was worth 770 million U.S. dollars to Thailand, making cassava its second largest commodity export after rice. Within the EEC, most Thai cassava ends up in the Federal Republic of Ger- many and the Netherlands. It is used entirely for animal feed, principally for pigs although dairy cattle take up about 30 percent of the total. These two countries are the principal users of Thai cassava because they are Europe's major pork producers, they (particularly the Netherlands) have good inland waterway sys- tems that connect the ports to the feed compounding mills, and they peg their green exchange rates at a more depreciated level than the official exchange rate and thus shifted relative prices in favor of cassava usage as against cereals.2 The surge of cassava imports reduced the demand for cereals, imposing addi- tional costs on the EEC budget through lower import levy revenue or larger export restitution payments on cereals. The cheap feed benefits the livestock producers (particularly pork producers) in Germany and the Netherlands at the expense of France and Italy, which resulted in French pressure on the EEC to restrain trade. Once the pressure began, the Thais tried to obtain the support of the beneficiaries in their negotiations with the EEC, with meager results. The emergence of Thailand as a dominant supplier of cassava to the EEC can be explained by its comparative advantage, not so much in production as in transport. Cassava is a plant that can grow on relatively poor lands, which many tropical countries have in abundance. As a relatively low-valued commodity, however, efficient transport is essential for the cassava trade to grow. With an excellent road network, Thailand was in a good position to keep transport costs low. In 1982, farmers in the main producing area about 240 kilometers (150 miles) from the port were able to get 76 percent of the port price, this for a commodity that starts out its journey as fresh roots fetching only 3 cents per kilogram. As the scale of the trade grew, Thailand was able to reinforce its comparative advantage. Large investments were made in loading facilities as well as in pro- cessing. The cassava exported is actually a crude product. After harvest, the cassava is broken into smaller pieces, called chips, and then sun dried. With the increasing scale of trade, the chips were "pelletized,' a process in which chips are squeezed through holes so that they emerge roughly the shape and size of ciga- rettes. This additional transformation of the product is solely to economize on transport costs: more can be packed into ship holds, and less dust is generated at the time of loading and unloading. 2. The green exchange rate is the currency used to translate the administered prices for agricultural goods set by the EEC as a whole (expressed in European Currency Units), into national prices (expressed in domestic currencies). Administered prices include threshold, intervention, and target prices, which together set the structure of prices in the EEC. When a particular green exchange rate is set differently from the official exchange rate, it allows the levels of agricultural prices in that country to differ from the European norm (see Ritson and Tangermann 1979). 598 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 The Thai share of total European cassava imports in 1980-82 amounted to 87 percent. But even then, new exporters were entering the scene. In 1981, China, for example, exported 607,000 tons or 9 percent of the total, reaching that level after only entering the trade in 1979. In the final analysis, the cassava trade is an artificial one, made possible only by the EEC policy of keeping cereal prices high. Thailand could find very few other outlets for its cassava pellets. Indeed, Thai animal feed compounders found that, at the prices the pellets were fetching in the EEC, it was better to allow the cassava to be shipped to Rotterdam than to use it as feed themselves. For them, using cereals is more economical. The cassava trade to the EEC has always posed a dilemma for the Thai govern- ment. It is clearly a very profitable trade, but also a perilous one, given its dependence on policy decisions made in a single market. Many voiced their concern and suggested some method to prevent too heavy a reliance on this one commodity. For although at their height cassava exports were about 13 percent of total exports, they formed a very important part of the cash incomes of farmers in the Northeast, Thailand's poorest region. Against this, it may be noted that cassava is a relatively easy crop to grow, requiring little cash outlay in its cultivation. Consequently there is considerable flexibility on the part of the farmers, and estimates of supply elasticity usually yield figures of around 1. The investments in transport and processing men- tioned earlier are not specific to cassava, except for the pelletizing plants which contribute less than 10 percent of the final value of the exported cassava. Despite the EEC'S budgetary problems, it has always seemed unlikely that there would be a rapid dismantling of the protection policy for cereals in the EEC. If a decline in protection were to be gradual, Thailand would be able to cope with the adjust- ment costs fairly easily. The Thai government opted to do nothing and let the trade grow until it was forced to restrain exports by the EEC, for the European response to the cassava trade was not to dismantle the protection on cereals, but to erect barriers against cassava. The 1982 Cassava Trade Agreement In 1982, the EEC entered into major agreements with each of the three cassava- exporting countries, Brazil, Indonesia, and Thailand. Because the EEC is bound by its tariff concession commitment under the Kennedy Round in 1967, it had to follow procedures set forth in Article XXVIII ("Modification of Schedules") of the GATT before it could modify such concessions.3 Article XXVIII requires the concession granter who wishes to modify the concession to negotiate and seek agreement with the principal supplier, the initial supplier with which the granter negotiated the concession, and any suppliers who may have a substantial inter- 3. These tariff concessions generally called "bindings;' are contained in a list termed "tariff schedules;' which under Article 11 of the GATT are legally binding. Sathirathai and Siamwalla 599 est. The negotiation process may also include compensatory adjustments made by the party requesting the modifications.4 In 1982 Brazil and Indonesia, both GATT contracting parties, were treated as the initial supplier and principal sup- plier, respectively, under the GATT. Because Thailand was not at that time a contracting party to GATT, the EEC was not obliged to follow Article XXVIII procedures with respect to Thailand. The essence of the agreements with these two countries is that both temporar- ily permit the EEC, within the life of the agreements, to unbind its tariff conces- sion obligation under the GATT in order to set up a tariff quota system (EEC 1982a, 1982b). The agreement allows a prohibitively high tariff to be levied on the amount in excess of the quota. The agreement allocated generous and rising quotas to GATT contracting parties between the years 1982-86 (588,235 metric tons in 1982 rising to 970,590 tons in 1986) and assured Indonesia of 85 percent of all quotas allotted to GATT contracting parties. The increased quota allowed seemed to be a part of the compensatory adjustment under Article XXVIII of the GATT. Both countries also agreed that the EEC would be the import licensor. The two agreements were due to expire on December 31, 1986, but could be ex- tended for another three years unless denounced by either party to each agree- ment one year before the expiration date. In 1982 the parties to both agreements notified the GATT that both Brazil and Indonesia had agreed to the modification of the EEC'S tariff commitment (unbinding) under the Kennedy Round of 1967. The EEC convinced Thailand, at that time not yet a GATT contracting party, to enter into a voluntary export restraint agreement with it. The agreement is euphemistically entitled the "Cooperation Agreement between the Kingdom of Thailand and the European Economic Community on Manioc Production, Marketing, and Trade" (hereinafter referred to as the 1982 Agreement; (EEC 1982c). The major features of the 1982 Agreement are described briefly as follows. As the EEC entered into the agreement with Thailand as a non-GATT contracting party, there were no Article XXVIII negotiations involved. The quota agreed upon was a declining one: 5 million tons in 1983 and in 1984 (with an extra 500,000 tons allowed for either of the two years) and 4.5 million tons for 1985 and 1986 (also with an extra 450,000 tons allowed for either of the two years). The Agreement allowed Thailand to issue the export licenses, upon presentation of which, the EEC issues import licenses for the Thai products. The EEC agreed to provide assistance for rural development and crop diversification projects in Thailand of 75 million ECU'S (European Currency Units) for the period 1982-86. The EEC also loosely incurs an obligation to secure Thailand's position on the EEC cassava market against increases of cassava imports from other countries, and to "bear in mind" the import of other carbohydrate prod- 4. In the so-called Chicken War between the United States and the EEC, when Germany raised its tariff against import of poultry from the United States upon joining the EEC, the United States was allowed to withdraw concessions on a volume of trade equal to the volume that would have occurred under the poultry concession in a representative period (see NValker 1964; Hartwig and Tangermann 1987). 600 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 ucts which could compete directly with cassava. While, the 1982 Agreement was scheduled to expire in December 1986, it could be extended for a further three- year period if neither party denounced it one year before the expiration date. The Thai negotiating power prior to the signing of the 1982 Agreement, and Thai officials' perception of the trade benefits and security provided under the agreement, played an important role in their decision not to accept simple exten- sion of this 1982 Agreement for three years after December 1986. First, as Thailand once exported as much as 7.3 million tons of cassava to the EEC in the absence of the quantitative restriction, the agreed quota was relatively small. The officials appear to have taken the view that the restricted and declining quota was balanced out by the arrangement which made Thailand the export licensor, allowing the quota rent accrue to it. Further, since the EEC financial assistance to Thailand is given through the Ministry of Agriculture and Cooper- atives, there was strong support for the 1982 Agreement from that ministry. Finally, there was a widespread perception that, as Thailand was not a GATT member, it had little bargaining power against the EEC. It was felt that the terms of the agreement, although not as good as a free trade regime, were probably the best Thailand could obtain, for a probable alternative was a unilateral move by the EEC. There was some opposition to this perspective within Thailand. Many felt that the government should delay the agreement as long as possible, at least until Thailand became a GATT member. However, the EEC strongly hinted that they would apply Article XXXV (which provides for nonapplicability of specific GATT oDligations between particular contracting parties) if Thailand did become a GATT member without signing the VER. Some Thais felt that the declining quota was not fair, in the sense that a most-favored-nation (MFN) treatment (to which Thailand is entitled from an earlier agreement between the EEC and the Association of South East Asian Nations) of a tariff quota can be construed to require that Thailand at least should have maintained its share of EEC imports vis-a-vis Indonesia and Brazil.s Finally, the obligation on the EEC to "bear in mind" the impact of cassava and other carbohydrate products imported from other countries on Thailand was thought to be too vague. This concern is supported by the fact that while the EEC has been restricting the cassava exported from Thailand, it has done very little to restrict the U.S. exports of citrus pulp and corn gluten, two other major cereal substitutes. These perceptions and arguments relating to the 1982 agreement were to play an important role in 1985 when the decision to extend or denounce the Agree- ment again became a public issue in Thailand. A point of note here is that during the 1982-86 period, Brazil and Indonesia did not fill their annual quota alloca- tions while the quota for Thailand was binding. Although the EEC used the GATT 5. To apply the MFN concept to a tariff quota, the principle implicit in the GATT'S Article XIII (2) (nondiscriminatory administration of quantitative restrictions) is applied. The agreements between the EEC and the three countries started at the same time in 1982, and the previous representative period of three years prior to 1982 would have supported the belief that this was an unfair quota allocation. Sathirathai and Siamwalla 601 to its own benefit in 1982, in 1985 its representatives told the Thai negotiators frequently that appealing to the GATT is "legalistic" and therefore inadvisable. The Negotiations of 1985-86 The issue in 1985 was whether Thailand should renounce the 1982 Agree- ment, due to expire in December 1986. The key element for the decision lay in Thailand's legal position consequent upon its entry into the GATT. The issue became a public debate when some government officials and Cabinet members who were to negotiate the VER said that Thailand should extend the 1982 Agreement for three more years. To them, the Agreement provided certainty, for it guaranteed the amount Thailand could export to the EEC. The GATT was seen as a paper tiger not to be relied on; a bilateral pact was viewed as more depend- able than the EEC's and Thailand's commitments to a multilateral GATT. Others took the view that, since Thailand had become a contracting party to the GATT in 1982, it should be able to benefit from GATT principles, in particular the bound concession given in the Kennedy Round to allow cassava to enter the EEC subject to 6 percent tariff without any quantitative restriction.6 They argued that the 1982 unbinding, negotiated with Brazil and Indonesia, was temporary and was due to end in 1986. Under Article XXVIII of the GATT, further continu- ation of the unbinding could only be reached after negotiations with all contract- ing parties having rights under the bound tariff. Those of this view argued that since Thailand was now a GATT contracting party, it had rights to the binding, replacing Indonesia as the principal supplier of cassava to the EEC.7 Thailand's rights under the binding would then have been its main negotiating weapon. This view of Thailand's rights under the GATT was advanced during the prelimi- nary negotiations conducted at the technical (nonministerial) level in Bangkok in October 1985. As for the EEC'S attitude before and after this meeting, in both formal and informal sessions, a somewhat legalistic position appears to have been taken by the EEC, despite its often reiterated suggestion to the Thai government not to be too legalistic. For example, the EEC argued that the permission granted to the EEC by Indonesia to unbind its obligation under the Kennedy Round was a permanent one, and there was thus no need for it to renegotiate with Thailand or any GATT contracting parties. To us, this argument is untenable. The EEC- Indonesia Agreement of 1982 was clearly temporary in nature, its continuation being contingent upon the absence of denunciation by either the EEC or Indone- 6. Thailand is entitled to this binding through the Most-Favored-Nation Principle of Article I of the GATT. Hudec suggested that there could have been a reservation or side agreement concerning cassava at Thailand's accession to the GATT. While the EEc might have made a reservation on its cassava concession, the schedule indicated no such reservation. Our inquiries with Thai negotiators involved in the accession to the GATT and the 1982 Agreement indicated also that no oral or written side agreement was arranged. 7. Thailand's position as the principal supplier of cassava was confirmed by Ake Linden, legal advisor to the director general of the GATT in his letter to the Thai ambassador to the United Nations at Geneva, dated November 22, 1985. 602 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 sia one year before the expiration date in December 1986. In the letter to the GATT, the EEC used the phrase "negotiations were conducted in accordance with Article XXVIII of GATT with an aim of suspending temporarily the Community's tariff concession . . ." (emphasis added), (EEC 1982a, 1982b). In the annex to a letter to the Director-General of the GATT signed jointly by the EEC and Indone- sian Delegates, there is a statement "Concessions to be suspended until 31st December 19862' No other statement in the letter or the annex indicates that the suspension is permanent or indefinite (GATT 1982). And Article 2 of both Agree- ments with Indonesia and Brazil stipulates that the arrangements will remain in force until December 31, 1986 (EEC 1982a, 1982b). EEC officials also dropped hints that the binding in the Kennedy round applied only to cassava chips and not to pellets. This possible interpretation is countered by the Custom Cooperation Council Nomenclature (CCCN) and the EEC'S own customs law. While the 1967 binding under Subheading 07.06 of the CCCN did not explicitly include cassava pellets, the Explanatory Note of the CCCN issued in 1968 states that Subheading 07.06 includes cassava pellets. This Explanatory Note is supported by EEC practice and the 1982 inclusion of cassava pellets into its own Common Custom Tariff (CCT) Subheading 07.06 A (II). Thailand did not strongly dispute the EEC'S arguments, nor did it assert its principal supplier's rights under Article XXVIII of the GATT. It did request that its quota be enlarged to include the potential imports of Spain and Portugal, who were then becoming members of the EEC and therefore subject to the cassava tariff quota restriction if an agreement regulating cassava trade were to con- tinue. The EEC'S (correct) response was that under GATT Article XXIV (5) and (6) it was under no legal obligation to increase the quota for Thailand because neither country had any tariff binding on cassava imports before joining the Community.8 In contrast to the EEC'S legalistic approach to the negotiation (when such an approach suited its position), Thailand was reluctant to apply GATT law to strengthen its position in the trade negotiations or to achieve its objective. The question, often raised in domestic discussions, as to whether one could be confi- dent that Thailand would "win the case" in the GATT'S formal panel proceedings shows a misperception of the use of GATT mechanisms in trade negotiations. There is a difference between the use of the GATT dispute settlement mechanism to reach a satisfactory agreement and the winning or losing of a domestic court decision. By its reluctance to employ GATT law, the Thai government relin- quished its position as a negotiating partner and ineluctably sank into the role of supplicant. Later in 1985, the Thai minister of commerce went to Brussels to negotiate a 8. Article XXVIII provides that parties that withdraw negotiated concessions may be required to compensate other contracting parties. The establishment of a customs union affecting concessions adopted previously by new members must also conform to the compensatory procedures of Article XXVIII. Since Spain and Portugal had not negotiated concessions on cassava, they were not required by Article XXIV (6) to follow Article XXVIII procedure. Sathirathai and Siamwalla 603 new cassava agreement. A Draft Protocol was initialed extending the 1982 Agreement from 1987 to 1990, with the trade quota increased from 4.975 to 5.25 million tons per year from 1987 to 1990. The EEC acceded in writing that the 6 percent tariff under the CCCN subheading 07.06 (or 07.06A of the Com- mon Custom Tariff) cover both Thai cassava chips and pellets. The EEC noted, however, that such acceptance of the product coverage under these tariff head- ings holds only during the life of the Protocol. More important, Article 3 of the Draft Protocol was phrased so that it appears that Thailand accepted the ECC'S claim that its cassava pellets are made from flour or meal, which would subject Thai cassava exports to a very high tariff level.9 The EEC could then claim that Thailand had accepted in writing that the Draft Protocol includes chips and pellets in 07.06 only as an exception. Finally, the Draft Protocol took into consideration that since entry into force of the 1982 Agreement, Thailand has become a contracting party to the GATT. That was the last time the GATT was mentioned in the draft: there was no recognition of Thailand's claimed position as a principal supplier. The outcome of the negotiation once again demonstrates Thai reluctance to insist on what many felt were clear and favorable legal principles under the GATT. The government immediately accepted the increased quota allowance, despite the fact that Thailand had another year (1986) to negotiate the quota and other issues. Given the estimated 8 million tons per year of possible exports in the absence of quantitative restrictions, it is arguable that a quota of only 5.25 million tons warranted further assertion of Thailand's rights under the GATT, or compensation for the lost quantities due to the restriction. That Thailand did not press the EEC to recognize Thailand under the GATT as a principal supplier shows again the Thai government's perception of the insignificance of GATT provisions in strengthening its negotiating position. Conversely, the Thai position on the tariff classification dispute shows that, after a series of domestic discussions, the government was more aware of the implications of the nomenclature issue. The EEC finally accepted the view that the 6 percent tariff should be applied to both cassava chips and pellets, but it did not extend the acceptance beyond the life of the Protocol. In response to concerns regarding these issues, the Thai Cabinet instructed the commerce minister in May 1986 to seek agreement from the EEC to amend the Draft Protocol, though not the quota, as a precondition to the signing of the new cooperation agreement. The EEC Commission refused to modify the Draft Proto- col, arguing that the Council of the EEC did not mandate any change. The parties decided on "Agreed Minutes,' signed May 23, 1986, the legal status of which is less binding than an agreement. 9. Article 3(i) of the Draft Protocol states that Thai hard pellets are "those hard pellets resulting from a process in which manioc roots are first reduced to flour/meal before pelletizing." While the EEC draft claims that "pellets of flour" appear under subheading 11.04 of the CCCN, 11.04 deals with flour/meal but not pellets. Moreover, it is technically impossible to use flour to make cassava pellets, which are merely chips broken in pieces before being pelletized, in one continuous process. 604 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 In the Agreed Minutes, the "Commission took note" of Thailand's declaration that Thai cassava pellets are not made from flour, contrary to Article 3 of the Draft Protocol. The words "took note" do not seem to place any legal obligation on the EEC, although they do permit Thailand to raise the issue in other forums without an EEC claim of estoppel. The Agreed Minutes do not resolve to Thai- land's benefit, however, the issue of Thailand's status as the principal supplier. While the EEC recognizes Thailand on a bilateral level as the principal supplier of cassava under the Agreed Minutes, that is not a recognition under GATT Article XXVIII. Thailand agrees until 1990, however, not to object to the EEc-Indonesia ex- tended tariff unbinding agreement which recognizes Indonesia as a principal supplier, which will bar Thailand from claiming its principal supplier's status in 1989 when the EEc-Indonesia agreement expires. More important, since the EEc-Indonesia three-year extension can be renewed for another three years, it can be expected that the EEC will argue that such terms indicate a permanent unbinding permission, and that the EEC does not need to negotiate with Thai- land to continue its modification of the tariff concession. An EEC letter informing the GATT that the tariff unbinding agreement with Indonesia will be extended sine die was a clear manifestation of the EEC'S legal position (GATT 1986a, 1986b). The letter was circulated by the GATT; the Thai government sent a protest note, which was also circulated. It can be argued that Thailand's hesitation to apply GATT substantive provi- sions stems from its lack of confidence in their effectiveness, from a misconcep- tion of GATT dispute settlement as a court of law, and from an insufficient understanding of how GATT procedure works, both as it appears in written documents and in customary practices. Those espousing this view also suggest that without more expertise in international trade law, Thailand will too readily accept a bilateral settlement of trade relations, rather than use a legal framework to confront the economic, diplomatic, and legal pressures from developed coun- tries. Part of the problem faced by Thailand is the inadequate training of govern- ment staff for this work. An ability to effectively negotiate within the GATT framework, however, also requires an understanding of the potential benefits and costs of GATT law by those to whom the negotiators are held accountable- ministers, parliament, and the public. Part of the problem must also be the GATT'S abstruseness, which demands a level of expertise that developing coun- tries, including Thailand, can ill afford. II. THE RICE TITLE OF THE U.S. FOOD SECURITY ACT OF I985 World Rice Trade Prior to 1985 Two exporters, the United States and Thailand, have dominated world rice trade since the late 1970s. Table 1 sets out the volume and relative shares of the Sathirathai and Siamwalla 605 Table 1. Relative Shares of the United States and Thailand in World Rice Exports Year United States Thailand Total U.S. and (three-year Percentage Percentage Thai share moving Thousands of world Thousands of world of world average) of tons exports of tons exports exports 1976 2,118 24.13 1,906 21.72 45.85 1977 2,187 22.94 2,119 22.23 45.17 1978 2,265 21.12 2,395 22.33 43.46 1979 2,503 21.89 2,323 20.32 42.20 1980 2,751 21.83 2,815 22.34 44.18 1981 2,824 22.50 3,123 24.88 47.38 1982 2,608 21.24 3,456 28.15 49.39 1983 2,315 18.57 3,949 31.67 50.23 1984 2,153 17.38 4,159 33.58 50.97 Source: U.S. Department of Agriculture. two in total world rice trade. Rice prices peaked in 1980 and 1981, (see figure 1), ending the period since 1973 in which nominal rice prices rose sharply. In retrospect, the miniboom of 1980-81 can be seen as a fluctuation caused by a shortfall in production in the Republic of Korea. More important, Indonesia, the largest rice importer throughout the 1970s, was so successful in its produc- tion drive after 1977 that its imports dropped to zero by 1984. The peak prices of 1980 and 1981 hid these weaknesses in export markets, particularly for the U.S. rice industry, but they became apparent in the 1980s. The rising nominal prices fueled a boom in land prices in the United States, which could not be sustained when rice prices decreased. Meanwhile in Asia, where 90 percent of rice is produced and consumed, the increased output caused by the Green Revo- lution continued. In 1981, the U.S. Congress did not recognize these fundamental changes in international rice markets, and in passing the quinquennial farm legislation it set the rice support price (the loan rate discussed below) at a level which was not unreasonable given past market performance. By mid-1982, however, U.S. prices fell to the level of the effective support price (see figure 1). Unlike in the EEC, the United States had no export subsidy scheme, so that U.S. export quota- tions remained at the support price level. At the same time, Thai export prices fell continually, the only action taken by the Thai government being a steady reduction in its export tax. The differential between Thai rice (100 percent 2nd quality) and U.S. rice (No. 2) widened from about 15 percent to 80 percent. As a consequence the share of U.S. exports gradually declined while the Thai share increased. The Rice Title of the U.S. Food Security Act of 1985 The framers of the rice title of the U.S. Food Security Act of 1985 (United States 1985) appear to have had one central aim: to prevent the lessons of the Figure 1. U.S. Paddy and Thai Rice Prices and the U.S. Rice Loan Rates 17 (A 1 6- U.S. farmgate price (left scale) - 800 is 5 14 - 700 K~~~~~~~~~~~~~~~~~~~~~~ 13 3 12 6 600 -O~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~r o5 ~ ~ ~ ~~~z . 1 17 17 1 0 -z z~ ~~~~~ 10 500 _~~ r_j 5o(s FoegUArclur.evieS. . Deparmnrte (lft Agrcaltue) rightraeofTaiad 4 200~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 0 0~~~~~~~~~~~~~~~~~~~~~~~I- Note Hunredeigh 10 pouds.The .S.pric is he armgte riceforlonggran pady (oug ric), hilThaiThaexport isf.pricengk(rightiledscale) o Sourcs: Foeign gricuture ervic, UUS. lopantm rate (lft sgrcaltue) orfTaeo hiad Sathirathai and Siamwalla 607 1981-85 market which would have implied an exodus from rice production, from reaching American rice farmers. The language of the Act (particularly Section 1165) advocated that the United States should reclaim specifically from Thailand, its share of the world rice market, it being implied that Thailand was unfairly subsidizing its rice exports. As Congress was deliberating on the farm bill, the U.S. Rice Millers' Association brought a countervailing duty action against Thai rice. It was established that the Thai government subsidized rice exports by 0.75 percent.10 The cost structure of the U.S. rice production is relatively rigid and allows very little downward flexibility, and U.S. rice exporters can compete in world markets only by increasing government subsidies. This increase is what the rice title of the 1985 farm act set out to achieve. Three administered prices are central to the U.S. subsidy system. 1. Target price. This is the price set to assure farmers that they will receive an adequate level of income. In exchange for a uniform percentage reduction (not to exceed 35 percent) in the rice area base for each farm, rice farmers belonging to the government program are assured that the rice that they produce from the permitted growing area will yield the target price. For the 1986 crop year the target price for rice was set at $11.90 per hundredweight of rough rice, with a gradual 2-3 percent per year decline thereafter. If a farmer produces the rice and submits it as collateral for a loan (see 2 below) or sells it in the market and receives less than the target price, he is paid the difference in the form of a deficiency payment from the U.S. Treasury (not to exceed $50,000). 2. Loan rate. Farmers can pledge their rice as collateral for a loan from the government; part of the loan is advanced at the time of planting. Farmers can avoid repayment by letting the government keep the rice. The collateral value of the rice paid to farmers is called the loan rate, being $7.20 per hundredweight of rough rice for the 1986 crop. In the previous legislation the loan rate then became, in effect, the support price in the sense that the U.S. market price for rice cannot fall below it. But that role is now taken by the loan repayment rate. 3. Loan repayment rate. The farmer is not required to pay the full amount that he had borrowed from the government ($7.20 per hundredweight in 1986), but only the portion determined by the loan repayment rate. This loan repay- ment rate is the greater of (i) 50 percent of the loan rate ($3.60 per hundred- weight in 1986, to increase by 70 percent by the 1989-90 crop years) and (ii) the "world price of rice,' a notional price announced weekly by the U.S. Department of Agriculture (USDA). Since the implementation of the 1985 Act on April 15, 1986, the loan repayment rate has fluctuated between $3.50 and $4.12 per hundredweight of rough rice (long-grained). Under the prior system, a farmer would only sell his rice in the market if he were able to get a price greater than the collateral value ($7.20 per hundredweight in 1986) which would then be 10. The CVD procedure does not allow the export tax, then amounting to about 5 percent of the value, to be offset against the subsidy. 608 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 due. Under the new "marketing loan program" if a farmer is able to obtain a price greater than the repayment rate, he will profit by selling the rice and repaying at the lower rate. This then allows the U.S. floor price to decline to the level of the loan repayment rate. The counterpart of this is that the new level of the rice subsidy (the target less the repayment rate) totaled $11.90 less (around) $4, which is $7.90 per hundredweight or about 198 percent of the market price. While the deficiency payment (equaling the difference between target price and loan rate) is subject to a $50,000 limitation, the difference between the loan and loan repayment rates was not subject to any limitation. In 1986, Congress did pass an amendment which imposed a $250,000 limit (per farm) on all government payments other than the deficiency payment. To implement the marketing loan program, the USDA is required to make weekly announcements of "world market prices" for long-, medium-, and short- grained rice. Given the thinness of the world rice market, and the very broad range of qualities of rice that are actually traded in that arena, to summarize the various actual transacted prices into three statistics implies a generous amount of discretionary power for the USDA. Despite the claims made by the U.S. Agricul- tural Attache in Bangkok that it has not and will not use that power in a "predatory manner," it seems clear (and was admitted by the same official) that the USDA wishes the American rice exporters to recapture the "lost" market share. The weekly announcements of the world market price are one of the means to that end (Slayton 1986). Two additional aspects of the rice title have a bearing on the analysis below. Because the loan repayment rate cannot decline below 50 percent of the loan rate, this would have imposed a new floor to the U.S. price ($3.60 per hundred- weight in 1986) and could possibly have made its rice uncompetitive again. Bent on avoiding the mistake of the 1981 Act, Congress introduced the "export marketing certificate." The administration is somewhat complicated, but this is a clear export subsidy given to exporters in an amount equal to the difference between the loan repayment rate and the world market price. This program is similar to the variable export restitution program of the EEC. The Food Security Act of 1985 (sections 1165 [a], 1127 [a][1] and [a][3][A][i]) also contains an "export enhancement program" which allows the secretary of agriculture to embark on a subsidy war in specific markets against any country deemed to be practicing unfair trade. So far this program has not yet been applied to rice, but there is no assurance that it will not be, particularly as the U.S. Rice Millers Association has convinced the U.S. Congress that Thailand has already been proven to be subsidizing its rice exports-albeit by three- quarters of one percent. Is the U.S. Rice Program Subject to International Economic Regulations? Remedies for injury caused by agricultural subsidies to countries exporting similar products can be based on paragraphs (1) and (3) of Article XVI of the Sathirathai and Siamwalla 609 GATT, as well as on the Subsidies Code which was negotiated during the Tokyo Round of multilateral trade negotiations and which elaborates on these provi- sions. As Thailand has not signed the Subsidies Code (although the United States has) and as the benefits under the Code are being granted by the United States only to other signatories, Thailand cannot expect to use the provisions of this Code. (It may be noted, however, that this selective application of the Subsidies Code is not accepted by all countries; some argue that the Code should be applied to all members of the GATT.) The United States has argued that "grandfather rights" exceptions to the GATT allow their agricultural export subsidies. "Grandfather rights" here mean that all trade laws existing before the GATT came into force in 1947 remain effective even though they may be otherwise inconsistent with the GATT. It is argued that the 1985 Farm Act, seen as another five-year extension of the original Agricul- tural Adjustment Act of 1933, has validity despite any conflict with the GATT. This claim has little substance because, in addition to the domestic agricul- tural subsidies that existed under the old Act, the 1985 Farm Act has added significant new mechanisms for subsidies covered by Article XVI (1) which cannot really be regarded as part of the old law covered by the grandfather rights exception. Furthermore, as Article XVI (3) on export subsidies was added in the late 1950s, it can be argued that it cannot be used as a base for grandfather rights exceptions. Some parts of tlhe price support under the Farm Act may be regarded as "domestic" rather than export subsidies. However Article XVI (1) of the GATT covers all forms of subsidies that adversely affect other countries. The Subsidies Code also covers all forms of export-promoting subsidies although it distin- guishes between export subsidies and other subsidies (Hudec 1984). U.S. law has several provisions for remedies against the export-promoting subsidies of other countries, whether or not they are "domestic" or "export" subsidies. (Jameson 1985; Section 771 (5) (B) of Tariff Act 1930 [as amended]; United States, 1973). It is quite clear that under U.S. law, the GATT, and the Subsidies Code, the forms of price support given to U.S. rice farmers could legitimately be subject to countervailing import duties by an importing country whose domestic rice industry had been injured by imports of U.S. rice. But Thailand is a rice exporter: the key issue is whether agricultural exporters such as Thailand can rely on international trade rules to help negotiations with the United States. Using GATT Law to Tackle Export Subsidies Since Thailand is not a party to the Subsidies Code it cannot expect to draw on its provisions. Reliance must then be placed on Article XVI of the GATT. Article XVI (2) of the GATT seems to suggest that export subsidies should be avoided because they may have "harmful effects" for other contracting parties and may "cause undue disturbance to their normal commercial interests." Article 610 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 XVI (1) requires subsidizing countries to notify the GATT contracting parties"1 in writing of the extent, nature, and estimated effect of such a subsidy, and upon request, to meet with other contracting parties concerned in cases where such subsidization causes or threatens serious prejudice to their interests. With regard to the export subsidy of a primary product having an impact on normal competi- tive conditions in the world market, Article XVI (3) prohibits a contracting party from granting an export subsidy which operates directly or indirectly to increase the exports of a primary product resulting in that contracting party "having more than an equitable share of world export trade in that product." In determining the "equitable share," it states that account must be taken of "the shares of the contracting parties in such trade in the product during a previous representative period, and any special factors which may have affected or may be affecting such trade in the product." It is very hard to pin down the scope of "more than an equitable share of world export trade." The United States has argued that its mechanisms under the Farm Act and the reduction of its rice export price to the calculated "world price" are intended merely to recapture the market it has lost to Thailand since 1982. Although the United States did lose its market share, the gain in the Thai mnarket share was not due to a Thai export subsidy. Rather, it was due to the U.S. price support policy that lifted the prevailing U.S. price well above the Thai export price. Articles 10 (2) and (3) of the Subsidies code attempt to shed some light on the language of Article XVI of the GATT. They read: 2. For purposes of Articles XVI: 3 of the General Agreement ... (a) more than an equitable share of world export trade shall include any case in which the effect of an export subsidy granted by a signatory is to displace the exports of another signatory bearing in mind the develop- ments on world markets; (b) with regard to new markets traditional patterns of supply of the prod- uct concerned to the world market, region or country, in which the new market is situated shall be taken into account in determining equitable share of the world export trade; (c) a previous representative period shall normally be the three most re- cent calendar years in which normal market conditions existed. 3. Signatories further agree not to grant export subsidies on exports of cer- tain primary products to a particular market in a manner which results in prices materially below those of other suppliers to the same market. [Em- phasis added.] The terms "to displace the exports of another signatory" seem to be the key to understanding the "equitable share" concept. Still, a brief survey of GATT prac- 11. The form "contracting parties" refers to GATT members acting individually. "Contracting Parties" is used in this article in place of the official GATT usage, "CONTRACTING PARTIES," to refer to actions by signatory countries as a group. Sathirathai and Siamwalla 611 tices demonstrates confusion and uncertainty in the application of the equitable share concept. There have been four GATT panel decisions on the issue of "equitable share" (Coccia 1986 pp. 16-21). The first case involved an Australian complaint in the late 1950s against the export of French wheat flour in which the panel decided that a system of refunds employed by the French acted as export subsidies (GATT 1959, p. 46; Coccia 1986, p. 16). The panel opined that "[though] there is no statistical definition of an 'equitable' share in world exports, subsidy arrange- ments have contributed to a large extent to the increase in France's exports of wheat and wheat flour." The panel concludes that "the present French share of world export trade, particularly in wheat flour, is more than equitable" (GATT 1959; cited in Coccia 1986). This case seems to give us some hope that "equita- ble s!hare" is something one can rely on, yet three subsequent cases discourage one's expectation. Parallel disputes arose between the EEC and Australia and Brazil due to EEC refunds on sugar exports in the late 1970s. The GATT panels refused, however, to reach a definite conclusion because it was difficult to rule on a causal link between the increase in the EEC'S market share and the decrease in the complain- ants' shares (GATT 1978-79; p. 290; 1979-80, p. 69; Coccia 1986, p. 16). The panel on the 1983 U.S.-EEC disputes on wheat flour also followed these prece- dents (Coccia 1986; pp. 17-22). It determined that the EEC share was larger because of its export subsidies payment, yet given the difficulties in application of the concept of "more than equitable share,' it concluded that "despite conside- rable increases in EEC exports, market displacement in the sense of Article 10(2)(a) [of the Subsidies Code] was not evident in the seventeen markets pre- sented by the United States and examined by the panel" (cited in Coccia 1986; p. 19). In the case of the U.S. Rice Title, the issue of share in specific markets, as distinct from global shares, may also warrant some discussion, although GATT practice does not clarify this issue (Coccia 1986, p. 19). Although the panel in the French wheat flour case ruled that the share refers to world export trade and not individual markets, (GATT 1959; p. 52) the panels on the sugar cases, while reaffirming this rule, seriously considered the effect on individual markets in determining the displacement of exports. (Coccia 1986, p. 19). In the case of Thai rice exports, it is not at all clear which market or markets the GATT would examine in determining whether or not the United States has acquired more than an equitable share. The foregoing overview of GATT practices raises many other uncertainties. Does "displace" mean displace partially or absolutely? Does "displace" refer only to the loss in quantity or does it encompass a situation in which Thai exporters had to sell a larger volume of rice but with a much lower price than it could have otherwise obtained, in order to survive in the export business? Furthermore, Article X (3) prohibits subsidies on certain primary products (including rice) which results in prices materially below those of other suppliers in the same 612 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 market. It does not deal with the situation in which, although the U.S. price is not below Thai price, the difference between the two prices has been sharply reduced. It is evident that even if Thailand were a signatory to the Subsidies Code it would not receive clear and firm legal support from that Code. The situation Thailand is facing at present is one in which subsidies from a developed country are having a clear negative effect, both in the short and long run, on Thai rice exports. The problem posed in rice is different from that in cassava. For cassava, the substantive laws are relatively clear (at least to us) that Thailand has certain specific rights, but the problems lie in the lack of confi- dence on how such rights are to be enforced, both at the negotiating table and in a GATT panel. The reluctance to take action over rice derives mainly from the ambiguity and uncertainty of the substantive rights provided by the law. The ambiguity of the law, the lack of understanding of the GATT system, inexperience in resorting to the GATT, and the lack of expertise among govern- ment staff in this area, all discourage Thailand from raising the rice issue at the GATT or from relying on international trade regulations when it negotiates this case with the United States. Although the Thai government decided several months ago to bring the rice issue to the GATT, up until the time of this writing (April 1987) Thailand has taken no action in the GATT. In the absence of confidence in the GATT system, the traditional approach of diplomatic talks with the U.S. Administration and the State Department seems to override alternative forms available by which to resolve disputes in international trade. III. WHAT Is To BE DONE? The case studies discussed above argue for an adjustment in the GATT legal system. We suggest below changes in GATT law, practice, and attitudes necessary to support equitable agricultural trade. A Change in GATT Laws: Increasing Clarity Thailand's experience with cassava and rice demonstrate that the ambiguity in GATT substantive laws on agriculture discourages agricultural exporters and particularly developing countries from resorting to GATT laws when negotiating with other countries. A clear example is provided by the provisions concerning subsidization for which the Subsidies Code is of little help, even for signatories. On the procedural side of the law, there should be clarification of steps to be taken to obtain enforcement of substantive rights stipulated by GATT laws. For example, a GATT Code could spell out steps negotiating parties must take to modify their commitments in accordance with Article XXVIII. When and how should parties inform the GATT of their agreement under Article XXVIII? What if the party wishing to modify concessions seeks agreement with a country which is not a principal supplier? Will the GATT reject the notice because that party is clearly not the actual principal supplier? Does the principal supplier have to take the initiative to protest to the GATT? If the GATT circulates a letter of protest and Sathirathaiand Siamwalla 613 the party modifying the concession does not respond, what is the status of each party in that situation? These questions reflect points of concern of the Thai government in the cassava case. Another example often cited in current GATT literature is the unclear dispute settlement process. The Thai and other governments have serious doubts about how to implement the principles and procedures embodied in Articles XXII (consultation) and XXIII (nullification and impairment) of the GATT. Greater clarity and discipline are necessary to increase confidence in GATT rules and procedures. Moreover, there should be a time limit stipulated for dispute settle- ment by conciliation or panel decision. The delay in the judicial procedure, for example, is extremely costly for traders and itself becomes a barrier to trade. In the cassava case, fears that delays in the GATT legal process would damage Thai interests discouraged Thailand from even advocating the Thai position in the GATT. The reporting and recording of the results of consultations and of the rationale of panel decisions could be greatly improved. This does not imply that the doctrine of precedent is introduced because trade issues are generally ac- cepted as having a very ad hoc nature. Rather, such rationales should serve as authoritative interpretation or guidelines for the application of GATT laws and thus as a basis for trade policy planning. International customary law-that which requires a proof of the state's prac- tices accepted as law-could play a more active role in the GATT legal frame- work. Article 38(1)(b) of the Statute of the International Court of Justice states that customary law is one of the sources of international law. International customary law has served as the basis for normative rules, especially in those areas where a written law (for example a treaty) does not exist or exists but is regarded as a "soft" law. The law of the merchant (Lex Mercatoria), which derives from customary practices among traders, is an extremely important source of law governing commercial transactions (Berman and Kaufman 1978, pp. 224-29). Customary law would not change, but merely supplement the law of the GATT. Related to this, the GATT Contracting Parties, with help from experts or panels, could play an active role in interpreting ambiguous provisions or rules. Such interpretations could be binding upon the contracting parties. The Con- tracting Parties adopted a procedure to allow the director-general of the GATT to use his good offices to facilitate a solution to a dispute between a developed and a developing country (GATT 1966, p. 18). The GATT Secretariat also could be formally encouraged to assist in the dispute settlement process. A Change in GATT Practice: Dealing with Voluntary Export Restraints As developed countries increasingly demand that they be allowed to withdraw tariff concessions on agricultural (and other) products, an effective process is needed for dealing with the demands for VER which frequently arise with these modifications. While the GATT has ignored VERS in the past, new provisions are 614 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 needed to strengthen the bargaining power of developing countries facing pres- sures to adopt VERS. Some authors have proposed the outlawing of VERS(GATT 1985) by the GATT. Others have proposed the legalization or regulation of VER within the GATT framework. Both proposals could be counterproductive to the interests of devel- oping countries. The elimination of VERS is an unrealistic aim and would take away some economic benefits that can go to developing countries in the absence of a free trade regime. (Hindley 1980; see also his article in this volume). Con- versely, the legalization of VERS under the GATT could eviscerate the legal frame- work of the GATT, or "gut the GATT," in Wijkman's words (Wijkman 1986). It would undermine the nondiscrimination principle which has been a GATT cor- nerstone and superimpose another trade regime on the GATT legal framework (as the Multifibre Arrangement has done). The Thai-Eec cassava case suggests that a link between VERS and the GATT could benefit developing countries. For example, greater clarity regarding the steps to be taken under Article XXVIII, to determine the principle supplier and to reject an incorrect procedure would have assisted Thailand in its negotiations with the EEC. Thailand could then have sought better terms for its VER if the EEC had still sought to modify the concession. Good faith negotiations would have been facilitated had the GATT decided officially that Thailand was the principal supplier of cassava, and that the EEC unbinding of 1982 was temporary. A Change in GATT Attitudes Our final proposal concerns the need for a change in the GATT's approach to exceptions to GATT laws and to the style of negotiation and dispute settlement it encourages. Both attitudinal changes would benefit developing country parties, the first allowing them to attain compensation for exceptions, and the second increasing the discipline of and the number and type of forums for negotiation. The role of exceptions. Many have complained that exceptions have smoth- ered principles in the GATT. Teese observed that the GATT "[is] no longer a legal framework for international trade but rather a set of principles more honored in the breach than in the observance:" (Teese 1982). Exceptions are important in international system. They constitute a means to build up more rules (Wijkman 1986, p. 39). A country is willing to undertake and honor principles if it knows that it can waive its commitments when circum- stances change.'2 Article XXVIII is a good example of this. In this sense, excep- tions help increase the individual states' willingness to accept rules and thus increase their participation in the GATT system. At present, exceptions to the GAIT are tolerated and obligations under the exceptions are generally expected 12. The principle of change of circumstance derives from the universal rebus sic stantibus principle, which allows a state to renegotiate or to break a treaty if there is a situation which, had it existed at the time of the signing, would have precluded the treaty (or contract) (see also Vienna Convention on the Law of Treaties 1969, in Brownlie 1978). Sathirathaiand Siamwalla 615 to be upheld. This tolerance of exceptions reflects the norms of the Contracting Parties. For developing (and other) countries to be successful in agricultural trade negotiations, they have to learn to utilize exceptions to the GATT to their benefit. The U.S. grandfather rights exception to the GATT, the VER arrange- ments by the EEC on agriculture with developing countries or the blind eye turned toward outright violations of GATT principles cannot be terminated over- night. The GATT should increase the costs of invoking an exception, providing bar- gaining power for those adversely affected by these exceptions. For example, the compensatory adjustment under Article XXVIII should be obligatory, and the obligation to seek agreements with the three groups of suppliers before any unbinding can be made should be supervised and monitored. Amendments should probably also be introduced to allow discriminatory retaliation against unilateral withdrawal of concessions limiting the retaliation to the withdrawing country. Generally, mechanisms outside the defined scope of exceptions should be compensated. Conciliatory versus adversarial styles for the GATT. Perceptions of the GATT vary a great deal, ranging from the view of the GATT as a court in international trade, to GATT as a "General Agreement to Talk and Talk." The first views GATT dispute settlement procedure as adversarial and somewhat courtlike. This is the position usually advocated by the United States. The other views it as concilia- tory and accommodating. This latter view is supported by the EEC. (Ehrenhaft 1986; Phan 1986). Despite those differences, both views do suggest that the GATT appears to give something to countries that they would not otherwise obtain. As for the Thai government, although it is inclined toward the EEC view, it appears to dismiss the GATT as merely "a paper tiger,' as was expressed during debates over the cassava issue. The GATT is extremely weak in some areas, but it is also true that there are many aspects of the GATT that contribute significantly to international trade. There have been an increasing number of GATT panels over the last decade, and countries including the United States, the EEC, and Japan, are willing to abide by at least some panel decisions (Ehrenhaft 1986, p. 149).13 Despite vast differences, the central goal for the GATT, in most eyes, remains to set guidelines for international trade and resolve problems arising therefrom. We do not wish to emphasize the enforcement aspect of international law be- cause its "horizontal enforcement" nature (enforcement among equals) is neces- sarily more complex than and different from the "vertical enforcement" nature of the domestic legal system (which invokes the higher power of the state). Rather we would like to see a more disciplined yet not always adversarial dispute settlement procedure. It should be sufficiently disciplined to ensure that an 13. For example, the EEC adopted the panel findings that it had violated the Government Procurement Code. The United States accepted that the Manufacturing Clause of 17 USC Section 601 on the printing of copyrighted books violated the National Treatment principle set forth in provisions of GATT Article 111. 616 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 economically weaker country can clearly foresee possible benefits from settling disputes in the international forum, especially where that country has a strong legal position. For the GATT to serve its purpose and achieve that goal, both GATT and participants in the GATT system need to adjust some of their attitudes. First, just because the GATT is more than a messenger between parties in a trade dispute does not make the GATT a world trade court. It is one thing to act as a judge in a case which has winning and losing parties. It is quite another thing to give a ruling to facilitate good faith negotiations. The ruling may cover such things as who is the principal supplier or whether a particular unbinding is temporary or permanent. Such facilitation should become a part of the GATT framework for trade negotiations. Second, the GATT does not have to define its role precisely. Robert Hudec once observed: [T]he force of GATT legal rules has always been particularly dependent on the force of the normative consensus they represent . .. Having been cast in a less rigorous form, GATT rules must necessarily draw the major part of their authority from their consistency with the norms and values current in the GATTcommunity. [Emphasis added; cited in Phan 1986.] Accommodating many views does not necessarily imply intellectual flabbi- ness. The GATT could command the faith and confidence of its members by providing them with precisely outlined choices of dispute settlement processes. The fully flexible consultation process in Article XXII could still exist, but with more specific guidelines regarding procedure. The role of the director-general in helping to solve disputes, similar to the U.N. charter provisions for its Secretary- general, could be spelled out. Choices of mediation, conciliation, and arbitra- tion which require third party involvement could be given, and the procedures for each clearly laid down. Finally, a rigid procedure involving an adversarial forum or a courtlike dispute settlement can also be contemplated in the frame- work of GATT panels. Such changes will never be accomplished if both the GATT and its members do not adjust their attitudes toward the GATT system. The GATT does not have to be one thing or the other, and cannot be all, yet the GATT can provide choices encompassing most views that are evidently the "values" in the GATT community at the moment. A country having faith or experience in a particular approach can follow that route. The key is that the system must provide a certain level of certainty as to where people go in each choice, and how the GATT can assist in such processes. Here again, the GATT'S willingness and ability to state what its law means or says substantively and procedurally could be of great help. In summary, we wish to emphasize the necessary changes which will make specific, bilateral agricultural trade negotiation under the GATT (not in the MTN), and probably other negotiations as well, more beneficial to all parties. There is a need to clarify some unnecessarily vague substantive provisions, to widen choices of dispute settlement, to be willing to give judgment on some technical Sathirathai and Siamwalla 617 legal or factual issues, and to regard GATT provisions as negotiating tools inter alia in the context of VERS. These are issues that the new MTN round could accornmodate its agenda. The suggested changes not only require amendments of GATT provisions, they also require more complicated and sensitive changes in practice and attitudes. These cannot change overnight by issuing a new law, but require the participation of developing countries whose view of law may be different from that of the developed countries. Finally they require that the GATT mobilize opinions internally and among its members before changes can be adopted that will reflect internationally shared values and norms. REFERENCES Australia, Commonwealth of. 1985. Agricultural Policies in the European Community: Their Origins, Nature and Effects on Production and Trade. Bureau of Agricultural Economics Policy Monograph 2. Canberra: Australian Government Publishing Ser- vice. Berman, Harold J. and Colin Kaufman. 1978. "The Law of International Commercial Transactions (Lex Mercatoria)." Harvard International Law Journal 19, no.1. Brownlie, Ian. Ed. 1978. Basic Documents in International Law. 2d ed. London: Oxford University Press. Coccia, Massimo. 1986. "Settlement of Disputes in GATT under the Subsidies Code: Two Panel Reports on EEC Export Subsidies;' Georgia Journal of International and Com- parative Law 16: 16-22. Ehrenhaft, Peter D. 1986. "A U.S. View of the GATT." International Business Lawyer 14, no. 5 (May): 147-49. European Economic Community. 1982a. "Agreement in the form of an exchange of letters between the European Economic Community and the Federal Republic of Brazil." OfficialJournal of the European Communities L 219 (July, 28 1982): 58J-59. - 1982b. "Agreement in the form of an exchange of letters between the European Economic Community and the Republic of Indonesia." Official Journal of the Euro- pean Communities L 219 (July 28): 56-7. - 1982c. "Cooperation Agreement between the European Community and the King- dom of Thailand on Manioc Production, Marketing and Trade." Official Journal of the European Communities L 219 (July 28): 53-4. Friedman, Wolfgang, Louis Henkin, and Oliver Lissitzyn. 1972. Transitional Law in a Changing Society. New York: Columbia University Press. GATT (General Agreement on Tariffs and Trade). 1959. "French Assistance to Exports of Wheat and Wheat Flour." Basic Instruments and Selected Documents (7th supple- ment). .1966. Special Dispute Settlement Procedure. Basic Instruments and Selected Docu- ments (14th supplement). - 1978-79. EEc-Refunds on Exports of Sugar-Complaint by Australia. Basic In- struments and Selected Documents (26th supplement). - 1979-80. EEc-Refunds on Exports of Sugar-Complaints by Brazil. Basic Instru- ments and Selected Documents (27th supplement). 618 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 1982. Letter from the Delegates of the European Economic Community and Indo- nesia to the Director-General of the GATT Secretariat, dated July 20, 1982. .1985. Trade Policies for a Better Future: Proposals for Action, Geneva. - 1986a. Circulation Documents No. 121 SECRET/259/ADD.4. Dated March 6, 1986. - 1986b. Circulation Documents No. 121/SECRET/259/ADD.4/CORR 1. Dated March 27, 1986. Hartwig, Bettina, and Stefan Tangermann. 1987. Legal Aspects of Restricting Manioc Trade between Thailand and the EEC. Kiel: Wissenschaftsverlag Vauk. Hindley, Brian. 1980. "Voluntary Export Restraints and the GATT'S Main Escape Clause." World Economy 3, no.3 (November): 313-41. Hudec, Robert E. 1984. "Regulation of Domestic Subsidies Under the MTN Subsidies Code." In D. Wallace, F. Loftus, and V. Krikorian, eds., Interface Three: Legal Treat- ment of Domestic Subsidies. Washington, D.C.: International Law Institute. Jameson, Paul W. 1985. "The Administration of the U.S. Countervailing Duty Law with respect to Domestic Subsidies: Where It's Been, Where It Is, Where It May Go." Syracuse Journal of International Law and Commerce 12, no. 1 (Fall): 59-124. Koester, Ulrich. 1982. Policy Options for the Grain Economy of the European Commu- nity: Implications for Developing Countries. Research Report 35. Washington, D.C.: International Food Policy Research Institute. Phan van Phi, R. 1986. "A European View of the GATT." International Business Lawyer 14, no.5 (May): 150-51. Ritson, Christopher, and Stefan Tangermann. 1979. "The Economics and Politics of Monetary Compensatory Amounts." European Review of Agricultural Economics 6, no.2: 119-64. Schachter, Oscar. 1977. Sharing the World's Resources, New York: Columbia University Press. Slayton, Thomas. 1986. "Effects of Farm Act on Thai Rice." The Nation, July 3, 1986. Teese, Colin F. 1982. "A View from the Dress Circle in the Theater of Trade Disputes." World Economy 5, no.1 (March): 43-60. United States. 1973. "X-Radial Steel Belted Tires from Canada." Federal Register 38, p. 1018. -. 1985. "Proceedings and Debates of the 99th Congress, First Session." Congres- sional Record 131, no. 175, part II, section 17. Walker, Herman. 1964. "Dispute Settlement: The Chicken War." American Journal of International Law 58, no.3: 671-85. Wijkman, Per Magnus. 1986. "Informal Systemic Change in the GATT." World Economy 9, no.1 (March): 37. THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4: 619-646 Textiles and Clothing in a New Round of Trade Negotiations Vincent Cable Because developing countries generally have a comparative advantage in the production of labor-intensive textiles and clothing, the liberalization of trade in these products is critical to their prospects for increasing foreign exchange earnings. The new round of trade negotiations is likely to be less important for trade in textiles and clothing, however, than were the recent renegotiations of the fourth Multifibre Arrangement (MFA). As the MFA remains in effect until July 1 991, this limits the range of measures which can be altered in the interim. Nonetheless, textile trade will influence the process and outcome of the current negotiations because of the overlap of textile trade disputes with other broader trade issues. These include tariffs, the rollback of tariffs and quotas, voluntary export restraints, and other nontariff barriers, all of which have been applied to textiles trade at various times. Each of these measures has an influence on trade flows and a cost to both the importing and exporting countries. Any proponents of liberaliza- tion of trade in textiles and clothing must not only be familiar with these costs but also must be knowledgeable about the economic and political forces which have initiated and sustained the protective measures. The following discussion suggests that there are groups and transitional approaches which may favor some progress in textile trade liberalization in the current round of negotiations. The Multifibre Arrangement (MFA), the international rules which govern much of the international trade in textiles and clothing, imposes particular costs on the developing countries for it not only restrains trade but is discriminatory in its application. It is designed to restrict low-cost exports of clothing and textiles to developed-country markets. Of the $104 billion (billion is 1,000 million) in 1985 world trade in textiles and clothing, almost all of the $29.5 billion of exports from developing and Eastern Trading Area countries to developed coun- tries is subject to negotiated quota controls, either under the MFA or outside it, or to the uncertainty posed by the imminent imposition of such restraints in almost any area of textile or clothing trade (see table 1). In contrast, most of the $44 billion trade among developed countries is traded freely, albeit subject to tariffs. Only a small amount is quota-controlled-mainly Japanese exports to the United States. The negotiations which will take place on textiles in the current multilateral Vincent Cable is special adviser to the Commonwealth Secretariat. Copyright ( 1987 by the International Bank for Reconstruction and Development / THE WORLD BANK. 619 620 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 Table 1. An Overview of World Textile Trade To developing countries and To developed the Eastern countries trading area From developed countries 44.5 12.2 From developing countries and the Eastern Trading Area 29.5 18.2 Note: 1985 trade flows in billions of dollars; textiles and clothing combined. The estimated 1985 world textile and clothing trade of $104 billion to $105 billion had broadly three main components: (a) Of the $44.5 billion trade between developed countries, the majority-$21.6 billion of intra-EEC and $8.7 billion of EEC-EFTA (European Free Trade Association) trade-is almost entirely free of quotas and tariffs bar some transitional quotas on exports from Portugal or Spain. A small part is quota-controlled: the $1.25 billion of Japanese exports to the United States. The rest is restricted only by tariffs. (b) $29.5 billion is the value of exports from developing countries and the Eastern Trading Area-including China-to developed countries. Almost all is potentially covered by the MFA or other quota systems (for example, non-MFA quotas in Taiwanese exports; industry to industry voluntary export restraints [vERS] rather than MFA quotas on imports by Japan). Excepted, until MFA4, were $3 billion of goods made of non-MFA fibers (such as jute, silk, and linen). Not all potentially controlled trade is actually regulated and not all quotas are fully used; but mechanisms exist to stop rapid import growth. (c) The residual $30.4 billion is imports into and trade among developing and Eastern Trading Area countries. With the exception of imports by Hong Kong ($4 billion in 1985), very little is freely traded. Source: GATT data. trade negotiations are overshadowed in many respects by the multilateral and bilateral negotiations conducted under the fourth phase of the MFA. While this may suggest that little can be done in the Uruguay Round to improve the position of the developing countries in textiles trade, any protracted trade nego- tiations will eventually overtake textiles arrangements and will have important implications for the MFA as well as being influenced by it. This article examines the subject of textiles and clothing trade in broader terms than is encompassed by the new round alone, including the background to the current MFA, a sum- mary of its economic effects, and how the new round might be the basis for some progress in liberalization of textiles and clothing trade. I. ORIGINS OF THE CURRENT ARRANGEMENTS Discriminatory protectionism in textiles has been around a long time, which perhaps accounts for some of the fatalism which surrounds discussion of it. In modern times the problem can be traced back fifty years to the proliferation of quotas on textile exports from Japan and from other (then) developing coun- tries. The institutionalization of restrictions on a multilateral basis dates from 1961 and 1962, when the Short Term and then the Long Term Arrangement regarding International Trade in Cotton Textiles (STA and LTA) were negotiated (see Kees- ing and Wolf 1980; and Aggarwal 1985). The LTA allowed developed countries Cable 621 to impose restrictions (either unilaterally or through a negotiated voluntary restraint agreement) on imports from developing countries which were consid- ered to be a source of actual or potential market disruption; in the case of potential market disruption, restrictions could not be imposed unilaterally. De- veloped countries preferred this to taking safeguard action under the GATT, which would have required them to restrict textile trade among themselves, and allowed retaliation. GATT Article XIX also differed from the LTA in that it required proof of "serious injury" rather than "market disruption." The Arrange- ment thus represented a fundamental breach with the nondiscriminatory princi- ple of the GATT. For developing countries, the LTA was meant to offer a transparent set of rules concerning market access, including a guaranteed in- crease in quotas (of 5 percent per year in most cases), which was considered advantageous to them relative to having their exports subjected to a series of ad hoc, restrictive measures. In addition, the LTA required importing countries to undertake adjustment measures with the objective of restructuring their indus- tries and returning international trade in textiles and clothing to GATT rules. The LTA was extended in 1967 and again in 1970. In 1974 it was replaced by the arrangement governing International Trade in Textiles, known as the Multi- fibre Arrangement, which was extended in 1977, 1981, and 1986. The MFA has broadly similar objectives to the LTA and represents an extension of restrictions to non-cotton textiles due to cotton textile exporters' diversification into syn- thetic textiles (see Pelzman 1984). At the same time, the condition of market disruption was more precisely defined, growth in market access was set at 6 percent per year, and various elements were introduced to make the quotas more flexible. These included a "swing" provision which allows an exporting country to shift trade into another unfilled category, and allowance for countries to "carry forward" unfilled quota balances from one period to the next. Few would now seriously contend that the MFA'S proclaimed objectives-of temporary regulation, adjustment, and overall liberalization-have been real- ized (this is confirmed by exhaustive studies in GATT 1984 and OECD 1983). The bilateral agreements reached under the successive MFAS, instead of liberalizing access, have grown progressively more restrictive. Annual growth rates permit- ted within quotas have generally been below 6 percent; the number of product categories subject to restrictions has been increased; quota fragmentation cou- plecl with revision of the original "swing" or "carry forward" provisions has reduced quota utilization; requirements of proof of injury to domestic producers have become minimal; and many very small suppliers have been controlled. Despite almost twenty-five years of protection, there is little evidence of devel- oped country textile and clothing industries being prepared to recognize the essentially transitional character of the Arrangements and compete with unre- stricted imports from developing countries. In the Tokyo Round (1973-79) the subject of textiles and clothing was raised but only given peripheral treatment, primarily in regard to tariff liberalization and "safeguards." Textiles were systematically marginalized in the 1960s in the 622 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 Kennedy Round and in the Tokyo Round with the "snap back" clause, which made acceptance of the MFA a prior condition for tariff negotiations on textiles. After this assurance cuts in tariffs on textiles and clothing were made, with annual reductions to start in 1982 instead of 1980 when cuts on most other products began, and with an average total reduction of 17.5 percent instead of the 34 percent average for all products. Tariffs remain two to three times higher than the average for manufacturing, and there is tariff escalation on finished items. The discussion on safeguards in the Tokyo Round broke down on the issue of selectivity, which contributed to the continuation of separate arrange- ments for textile trade. At the 1982 GATT Ministerial meeting, Contracting Parties called for a study on the textiles and clothing trade and the impact of the MFA.I They agreed on the need to examine ways of liberalizing textiles and clothing trade in order to bring it under GATT rules, and subsequently set up a working party for this purpose. Despite several meetings that went beyond its original November 1984 deadline, the working party was unable to fulfill its mandate. Various options for trade liberalization were discussed, but no specific recommendations made, and the 1986 renegotiations did not become a major break with the past. Background to the 1986 Negotiations The MFA is nominally concerned with regulating the pace of long-term struc- tural adjustment in textile trade. In practice the terms of each successive MFA and of the bilaterals agreements have been governed more by current circumstance than by any historic progression. Much of the impetus behind tightening the MFA'S controls in the late 1970s came from the EEC, as a result of political pressures originating in a fall in textiles and clothing output after 1973 and from the crisis of overcapacity in synthetic fibers. But prior to the negotiation of MFA4, there had been some recovery from recession in the EEC economies. Textiles production at the end of 1985 was 10 percent higher than two years earlier; clothing output had increased by 13 percent. Profits were up, investment increased roughly 25 percent between 1983 and 1985, and business forecasts in 1986 were generally favorable. The Com- munity then signaled its willingness to consider some degree of relaxation in the MFA. By contrast, a boom in U.S. production was over by the time of the negotia- tions. Between the last quarter of 1982 and the first quarter of 1984, output of textiles and clothing rose over 20 percent, followed by a 24 percent increase in investment in 1984. The boom was fed by rapid growth in domestic consump- tion. But under the combined influence of sharply rising home demand and an appreciating dollar, exports declined steadily after 1981 and import volume 1. The form "contracting parties" refers to GATT members acting individually. "Contracting Parties" is used in this article in place of the official GATT usage, CONTRACTING PARTIES," to refer to actions by signatory countries as a group. Cable 623 almost doubled between the beginning of 1982 and mid-1984. Even though external influences on the U.S. textile industry are not as large as it claims, the deteriorating trade balance triggered a fall in textiles output of 10 percent in 1984 and a much smaller decline in clothing. Both largely recovered in 1985. Although developed countries were a major source of U.S. import growth in 1984-85, MFA suppliers bore the brunt of increased U.S. protectionist actions. Additional criteria were introduced for the "presumption of market disruption,' so that byJuly 1984, after being in operation for only six months, over 100 calls had been made for restraint against more than 20 developing countries. In December 1984 and January 1985, ten and out of thirteen countries were found to have used subsidies on their exports to the United States and were thus subject to countervailing duties. In 1985 monthly limits were added to the prior annual limits on imports which were considered capable of causing "market disruption if permitted uncontrolled entry." More restrictive "rules of origin" were adopted which counted an import against the quota of the country where the item last underwent a substantial transformation. Finally, more stringent bilateral agree- ments with Hong Kong, the Republic of Korea, and Taiwan were negotiated, even though existing agreements were not due to expire for another two years. By superimposing tight group ceilings on collections of specific product catego- ries, the agreements effectively stifle the flexibility which previously existed. Perhaps the most restrictive trade legislation posed in the U.S. Congress was the so-called Jenkins Bill, which sought to roll back imports of textiles from developing countries to their 1980 level. According to the United States Depart- ment of Commerce, it would have led to a 27 percent reduction in total U.S. textile imports, with larger cuts for countries which recorded above average growth in their exports to the United States in recent years. For example, India's exports would have been cut by two-thirds and Hong Kong's by 14 percent (its MFA exports by 12 percent and its non-MFA exports by 70 percent). In the aftermath of the MFA renegotiation, the bill just failed to achieve the two-thirds majority needed to override a presidential veto. The U.S. administration had been able to deliver an MFA which satisfied at least some of the industry's de- mands. The Terms of the 1986 Renegotiations The MFA renegotiations under GATT auspices produced a framework within which bilateral agreements are determined. It is in the bilaterals that the real substance is negotiated which determines the ease of market access. The GATT Conference. Compared with the first three MFAS, the new MFA maintains one important line of continuity: the basic stated objectives are re- tained and reproduced, word for word. The more restrictive changes expanded the type of textiles covered, added rules against false declarations, and strength- ened the hand of importers using nonbilateral quotas. At the insistence of the United States, fiber coverage was extended from cotton, wool, and synthetic fibers to all vegetable fibers and silk blends (not pure 624 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 silk) except for some goods "traded in commercially significant quantities before 1982 such as bags, sacks, carpet backing, cordage, luggage mats, mattings, and carpets typically made from fibers such as jute, coir, sisal, abaca, maguey, and henequen.' The target of this extension is primarily the ramie-based goods (sweaters especially) exported from China. But the extension is a flat contradic- tion of the preamble and Article I of the original MFA, and creates legal and practical ambiguity which threatens to envelop a range of product areas hitherto immune from MFA controls. What is a "blend"? India is developing a jute-based garment fabric: will jute used in new forms be restricted? These changes further diminish the limited degree of flexibility allowed to reflect demand and techno- logical changes. Only some hair fibers (such as cashmere and angora) and min- eral fibers (such as glass and asbestos) are now outside MFA coverage. MFA4 also tightens restrictions against "false declaration" of country of origin. False declaration commonly involves a country's attempt to avoid its quota limit by having some superficial processing or finishing done by a country which is not near its limit. Under the restrictions, exporting countries may collaborate, but importing countries have the discretion to take "appropriate action" in response. Article 3 (which governs quotas other than those forming part of bilateral agreements) has been tightened. This strengthens the capacity for holding down imports and for fixing a low base for subsequent bilateral agreements with small and new suppliers. This was demonstrated in the recent use of Article 3 powers by the United States against Nepal. There are some features which have possibilities for increasing both liberaliza- tion and restrictions. The "reasonable departures" clause which legitimized the strengthening of restrictive provisions under MFA2 is resurrected but in a much diluted form. In particular, it proscribes "negative growth" of imports and any tightening of flexibility provisions which had been applied to major suppliers. The United States insisted on maintaining the "antisurge" formula introduced by the EEC into the MFA3 (although it had not been used). This formula was justified by the claim that "real difficulties may be caused in importing countries by sharp and substantial increases in imports as a result of sufficient differences between larger restraint levels and actual imports." The formula now provides for consul- tation with exporters, however, and appears to strengthen the basis of "equitable and quantifiable compensation." And in other respects there are modest improvements from an exporter's standpoint. A commitment has been made to exclude "least developed coun- tries" from control, subject to the proviso that if controls are found necessary, treatment "should be significantly more favourable." (Bangladesh should be the main potential beneficiary, but it is far from clear how it would be treated if its exports were to rise rapidly beyond the levels at which it attracted quota action in the EEC [1984] and the United States [1986].) Consistently underutilized quotas "will be removed on request " The EEC has agreed to scrap 25 percent (about 600) of its quotas, but there has been concern by exporters as to the level Cable 625 at which scrapped quotas might be reintroduced later. There is an explicit com- mitment, albeit heavily qualified, to the "final objective" of an "application of GATT rules to trade in textiles," but no timetable or end-date is admitted. And the next MFA was scheduled five years hence, not four, as were its predecessors. Assessment. Industrial country lobbies have expressed outrage at the failure of their negotiators to secure a tightening of restrictions. Developing-country reac- tions have emphasized that it could have been worse and that the damage was limited. There is a broad concensus that the results of the renegotiation were roughly neutral in the sense that concessions between developed and developing countries balanced out. While developing countries have won some improved wording, however, the industrial countries gained more concretely, with wider coverage of restrictions on textile imports. Developing-country negotiators were pushed onto the defensive, trying to argue against the introduction of a "social clause" to protect wages and social conditions in exporting countries, the "protection of intellectual property rights," more bilateral reciprocity and a new U.S. invention: a wider ranging clause permitting import curbs where trade "destabilises domestic industry." Moreover, the view of the "neutrality" of the outcome is very static; the mere preservation of the MFA in its preservation of the MFA in its present form is a major setback for developing countries in terms of their original demands and the powerful intellectual case mounted for liberalization, not least by the Orga- nization for Economic Co-operation and Development (OECD) and the GATT. The fact that the developing countries (and the cause of freer trade) suffered a setback can be attributed to two principal factors. The first was the highly protectionist stance of the U.S. administration, and congressional insistence on tighter textiles restriction. The EEC found itself in the unaccustomed position of being relatively (but not very) liberal on textiles, though it was guided through- out by a determination to ensure that the United States was able to achieve its main negotiating objectives. The common developing-country position broke along familiar lines. Many of the minor exporters, especially the less competitive Latin American and East European exporters, lack militancy since they see the MFA as providing a guaran- teed market share in a field they would otherwise find difficult to enter. Some major producers, notably Hong Kong, are able to take maximum advantage of current quota flexibility, and are sufficiently well-organized to maximize unit values, including quota rent, from existing opportunities. Hong Kong had al- ready reached a key bilateral agreement with the United States before the MFA was signed, in order to protect its position in the U.S. market. This left at greatest disadvantage the populous low-income Asian countries, notably India and China, and even they signed at the end. Bilaterals. The Geneva MFA essentially provides a legal basis for bilateral agreements the terms of which are rather more important than the MFA itself. The United States had already negotiated a major (six-year) bilateral agreement with two of its three leading suppliers-Hong Kong and Taiwan-a month 626 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO.4 before the Geneva negotiations were completed incorporating the main features of the subsequent MFA, notably fiber extension. Agreement with the Republic of Korea was reached later. Of the three, Hong Kong got better terms-higher growth rates and the security of a longer, six-year, agreement-reflecting, it is said, a wish to reward Hong Kong for its absence of domestic protection. But the U.S. agreements incorporate less flexibility and much lower growth rates than were achieved under MFA3-around 1 percent a year on average. A similarly restrictive approach has been taken by Canada. The European Community had already renegotiated the basic terms of half its bilateral agreements before MFA4 was finalized and was primarily concerned to ensure that the MFA provided a legal basis for validating what had already been decided in principle. Most of the EEC'S bilateral agreements fall into three broad categories: major suppliers (Hong Kong, Korea, Macao, and Taiwan) which are to be allowed only 1 percent growth on the eight "sensitive" product categories and existing growth rates on others; most other countries which are allowed annual growth in the 4 to 6 percent range depending on product or country; and "specially favoured" countries, allowed up to 7 percent growth. The Community has tried to observe a generally more liberal approach by scrapping a quarter of its quotas (those which were not in use) and dispensing with quotas for minor and "least developed" suppliers. II. THE ECONOMIC CONSEQUENCES OF CURRENT TEXTILE ARRANGEMENTS There is a substantial measure of agreement, at least among economists, about the damaging economic effects of the quota regimes which operate under the MFA. Evidence has been produced on the price-raising effect of quotas and their costs to importing countries as a whole, to their consumers, and to devel- oping country exporters. The arguments are summarized here but the emphasis is less on historical analysis of the economic effects, than on those economic factors that have an impact on current policy and that are likely to have rele- vance over the period of the new round. What is the Extent of Protection? There are two basic approaches to refining the answer to the above question. One is to look at quota coverage. About a quarter of world trade is directly controlled by quotas, though another quarter is potentially subject to MFA re- straints. The amount and proportion of trade from controlled sources would be much higher, however, if restrictions had not diverted trade into other channels. Not all commodities potentially restricted are subject to binding quota limita- tions since quotas may be underused. A study by Koekkoek and Mennes (1986) focuses on groups I and II, which are the most sensitive items and account for 50 percent of the EEC'S MFA textile imports and over 90 percent of MFA clothing imports. Of the Group I and II quota categories, imports reached more than 70 percent of the quota limit in only 42 percent of the categories in 1978, and that Cable 627 share had dropped to 38 percent in 1983. But this relatively small share of the number of categories accounted for 80 percent of the actual import value regis- tered under Groups I and II during this period. Underutilization does not mean that quotas do not constrain imports; it may occur, for example, because the system does not allow for the necessary degree of flexibility in meeting changing fashion needs. The impact of quotas is not one of blanket protectionism but a subtler deterrent to trade which is very different in its incidence among suppliers and products. Another approach is to try to calculate the tariff equivalent of quotas. Data on quota premiums have been used for this purpose, but these are only available for a small number of major suppliers and fluctuate enormously over short periods of time between products and between importing and exporting countries. Crude approximations based on trends suggest that the protective effects of MFA quotas are at least as important on average as tariff protection-and, of course are superimposed upon it (see Hamilton 1984; Jenkins 1980; and Cable 1983b). The Effect on Trade Flows The declared objective of the MFA is to secure "a substantial increase in the export earnings of developing countries from textiles and a greater share in world trade." This provides two general criteria for evaluating its effects, though neither of these would reflect the major process of trade diversion which has occurred among developing countries, usually from more to less efficient sources. The MFA originally specified 6 percent annual import growth as representing a reasonable minimum expansion (for those items subject to quota control). While it is difficult to come up with overall figures in real terms for textile and clothing import growth, it would appear that at least in the main markets, the EEC and the United States, import growth has been kept to within 6 percent for most of the MFA period for those suppliers covered by controls. In Martin Wolf's (1986) analysis of some important changes in real growth trends, he notes that percep- tions about the MFA have been colored by two separate sequences of events. The first was the very sharp cutback in MFA imports, especially in the EEC, in the late 1970s, (see table 2) with the implementation of the first set of bilateral agree- ments under MFA1 and the negotiation of increased protectionism under MFA2. It Table 2. Real Growth of Textile and Clothing Imports in Developed Countries from Developing-Country Exporters (percent per year) Category 1963-76 1976-78 1978-84 Textiles 7.2 4.6 3.7 Clothing 20.9 4.8 10.9 Total 14.1 4.8 9.0 Source: Wolf (1986). Table 3. Import Penetration Ratios in Selected Developed Countries (percent) Group, year, and United Belgium and Fed. Rep. United isic code Canada States Japan Australia Luxembourg Finland France of Germany Italy Netherlands Norway Sweden Kingdom World 1975 3 29.75 7.01 4.94 22.78 64.56 29.40 17.91 24.25 21.92 55.37 43.84 35.12 21.95 321 27.72 3.67 6.10 34.02 78.23 39.05 19.04 28.06 17.85 100.20 54.21 54.98 22.45 322 14.83 9.80 8.32 21.67 80.44 31.96 16.57 44.91 17.61 86.82 66.24 62.00 27.48 1981 3 32.02 9.51 5.35 27.69 87.07 28.90 24.29 32.75 30.61 65.51 42.53 40.92 19.12 321 25.20 5.54 7.54 43.30 104.60 45.68 93.04 39.43 22.97 107.70 54.18 67.38 37.03 > 322 21.08 18.41 15.27 27.44 91.47 67.68 30.67 67.93 47.83 89.11 81.83 85.96 44.93 0 1983 3 28.17 10.28 5.26 23.45 100.3 30.12 26.21 35.11 31.19 67.10 44.23 44.92 29.32 321 24.74 5.30 6.92 39.35 102.90 48.52 98.51 41.52 25.05 111.70 58.17 75.67 38.77 322 19.92 20.27 13.04 23.66 90.16 55.25 33.10 73.17 67.58 90.86 86.49 90.93 39.97 Non-OECD total 1975 3 1.78 2.16 2.02 3.55 5.00 4.93 2.04 3.64 3.22 5.71 3.40 4.21 3.54 321 3.85 1.62 3.74 10.76 6.61 3.86 2.20 5.07 4.14 8.81 4.35 8.58 4.67 322 8.84 7.89 6.25 15.52 8.08 7.70 4.64 18.63 5.61 22.52 11.57 18.65 15.53 1981 3 2.35 3.21 2.26 5.67 7.51 5.5 3.46 5.38 5.51 9.98 3.48 5.05 2.69 321 5.06 2.90 4.56 17.40 10.76 5.93 13.42 8.96 5.50 12.35 5.96 11.75 6.18 322 15.89 16.47 11.61 22.27 12.75 24.87 10.65 32.83 20.92 30.03 12.83 28.11 25.86 1983 3 2.25 3.63 2.11 4.68 9.72 5.73 3.66 5.74 6.25 10.84 4.34 6.04 3.99 321 5.67 2.71 4.32 16.62 11.43 7.19 13.76 9.37 6.29 13.15 5.81 12.49 6.35 322 15.56 18.05 10.10 18.92 10.74 17.93 10.88 34.83 26.49 27.41 12.73 29.33 20.73 Developing countries (African and American, and Asian newly industrializing) 1975 3 1.32 2.01 1.68 3.21 3.29 1.33 1.34 2.42 2.09 4.04 2.19 1.97 2.57 321 3.16 1.56 3.65 9.90 5.44 1.84 1.71 4.28 3.40 7.13 2.37 6.68 3.98 322 8.23 7.73 6.20 15.19 6.24 6.60 3.33 14.68 3.20 16.66 9.37 16.20 13.70 1981 3 1.97 2.98 2.03 5.24 4.90 1.46 2.47 3.69 4.7 6.44 2.46 3.08 2.13 321 4.41 2.81 4.43 16.31 9.92 2.83 11.20 7.58 4.61 10.58 4.43 9.50 5.17 322 15.23 17.82 11.57 21.92 10.11 18.08 9.13 26.95 15.26 24.39 10.83 25.05 22.40 1983 3 2.08 3.44 1.89 4.33 6.01 1.71 2.74 4.02 4.73 6.47 2.77 3.22 3.12 321 5.10 2.63 4.14 15.80 10.13 4.66 11.43 7.97 5.20 11.19 4.38 10.23 5.25 cr, 322 15.06 17.82 10.08 18.65 8.69 14.03 9.30 28.70 20.24 22.34 11.37 26.81 17.41 to Newly industrializing countries 1975 3 0.79 0.88 0.68 1.27 0.79 0.39 0.30 1.16 0.66 1.47 0.87 0.84 0.84 321 1.63 0.72 1.92 4.62 2.74 1.01 0.43 1.37 1.56 3.59 0.87 2.63 1.89 322 6.94 6.32 5.13 9.23 4.19 5.21 0.78 13.06 1.36 13.87 8.13 13.23 2.67 1981 3 1.33 1.63 0.91 2.97 1.43 0.64 0.83 1.70 1.27 2.43 1.09 1.55 0.99 321 2.30 1.24 2.38 9.24 2.33 1.61 3.04 1.90 1.22 3.90 1.78 3.54 2.25 322 12.22 12.10 9.12 13.87 4.99 10.29 2.79 18.56 4.48 16.54 7.03 18.27 18.04 1983 3 1.48 2.06 0.86 2.55 1.50 0.74 0.87 1.87 1.44 2.61 1.60 1.78 1.44 321 3.03 1.36 1.64 9.81 2.49 2.80 2.84 2.21 1.45 4.15 1.64 3.97 2.13 322 11.94 12.79 7.46 12.38 4.07 7.31 2.41 20.00 4.60 15.28 8.25 19.10 13.60 Note: isic categories: 3 = manufacturing; 321 = textiles; 322 = clothing. Source: OECD (I1986). 630 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 was this which gave rise to some of the most severe criticism of the protectionist character of the MFA, especially by developing countries. The second critical trend was the period of more rapid import growth in the 1980s, especially in the United States, despite the existence of quotas. It is this which has impinged rather more in the current policy context. The real growth rate of clothing and textile imports into industrial countries had declined sharply by the early phase of MFA2, 1976-78. Under the Long Term Arrangement (1963-73), clothing and textiles imports had grown at 21 and 7 percent per year, respectively. Under MFA1 these rates fell to 14 and -0.4 percent per year for clothing and textiles, and by 1976-78 the rates were 4.6 and 2.5 percent. This change was most striking in the EEC, and for developing country exporters to it. There was a less abrupt change in the United States, where imports from MFA suppliers grew less than 6 percent in the late 1970s but took a rapidly expanding share of imports. In the MFA3 period (the early 1980s), however, imports from the developing countries once again grew rapidly. This was most striking in the case of the United States, where imports from MFA suppliers in 1983 and 1984 grew by more than 20 percent per year (though imports from the rest of the world grew even more rapidly). Much of this growth could be attributed to sufficient flexi- bility within the United States system of bilaterals to permit growth of imports, mainly in the form of products outside the MFA such as ramie and silk blends. Exporters were able to take advantage of the overvalued dollar and high growth in consumer spending in the United States. But there was also a sharp rise in import growth in the EEC, especially from developing countries. This suggests that even such a complex and formidable system of controls as the MFA still retained enough flexibility to permit some response to market forces. Information on the aggregate growth of imports does not allow differentiation between the element of import growth due to an increase in domestic demand as compared with the extent of market penetration. By analyzing the level and growth of imports in relation to domestic absorption, it appears that rather rapid growth in market penetration of industrial countries by developing coun- try textile and clothing exporters occurred over the 1975-83 period, particularly in relation to developing country exports of manufactures as a whole (table 3). The relatively higher growth of textiles imports was true for nine out of thirteen importing countries, and for nine out of fourteen countries in clothing imports. The rate of growth of market penetration in clothing by developing countries is one of the largest of all manufacturing industries, despite the MFA. While striking, these conclusions need to be treated with caution. The fact that market penetration is estimated on the basis of values rather than quantities means that it incorporates changes in relative unit values as well as volumes. In the case of the MFA, increased "market penetration" is partly accounted for by the influence of quota premia and higher unit values within quotas. There is a general conclusion which can be drawn, however: the MFA has not so far been successful in preventing significant growth in market penetration by developing- Cable 631 country textile and clothing exporters, although the growth would doubtless have been higher without these controls. As for the MFA commitment to allow developing countries an increasing share of world trade, the evidence is generally positive, albeit with qualifications. Except in the case of the United Kingdom, Belgium, and Norway (for garments) all developed countries experienced more rapid market penetration by develop- ing countries than by imports in general (table 3). In the EEC, developing-country MFA suppliers slightly increased their share of textiles and clothing imports dur- ing MFA3, (1982-85) reversing a downward trend in MFA2 (see table 4). In the United States over the same period, however, the most rapidly growing source among generally rapidly growing imports was Western Europe, and MFA suppli- ers lost some of the import market share they gained in the 1 970s. Another measure of the success of the MFA in meeting its declared objectives is the overall sector trade balance (table 5). This is a crude and possibly misleading indicator, however, which has to be used with some care as the sector boundaries are arbitrarily drawn, and the overemphasis on sector balances can lead to unjustified normative implications. While this measure indicates that developing countries are still able to enlarge their trade surplus in textile and clothing products, it also brings out the fact that while developed-country imports may be growing quite rapidly, exports are also growing. It is also apparent that the large and growing surplus of developing-country trade in clothing is mainly due to U.S. imports; nominal balances have fallen in the EEC and Japan. Developing countries run a deficit with the developed in textiles principally because of Italian and Japanese export surpluses, and the decline in the deficit in the 1980s was also due to U.S. imports. If the sectoral balance is more broadly defined to include textile machinery, synthetic fibers, and dyes, the surplus enjoyed by developing countries is re- duced. At a time when Western Europe and the United States are cutting syn- thetic fiber capacity and Japan is not expanding its capacity, however, there has been a major expansion of capacity in China, Taiwan, and Korea. This con- tinues a shift over the last decade of production of synthetic fibers away from the developed world to developing Asia. An additional consequence of the MFA is trade diversion from more to less restricted sources among developing-country exporters. The OECD notes that "trade diversion effects have been widespread in the past" (OECD 1985, p. 110). In the EEC, substantial trade diversion occurred to some low cost non-MFA suppliers in the Mediterranean basin whose exports to the EEC have consistently grown faster than those from countries with MFA agreements. And within the MFA agreements themselves, there has been considerable differentiation in treat- ment with trade diverting effects. There has been a good deal of trade diversion in the form of "quota hopping." For example, the large-scale overseas investment by the Hong Kong clothing industry has been partly motivated by the pursuit of lower costs but also by a wish to evade quotas (for example, in Macao in the mid-1960s, Mauritius in the Table 4. Textile and Clothing Imports of the Developed Countries: Share by Exporters (percent) Clothing exporters Textile exporters Centrally Centrally Developing Southern planned Developed Developing Southern planned Developed Period countries Europe economies countries countries Europe economies countries 1963 19 3 1 77 15 3 2 80 1973 31 8 5 56 13 5 3 79 1976 40 8 5 47 15 5 4 76 1978 38 8 5 49 14 5 3 78 1980 39 5 6 50 15 4 4 77 1981 44 5 6 45 15 4 5 76 1982 45 6 7 43 15 5 4 76 1983 46 6 7 42 16 5 4 75 1984 47 7 7 38 17 6 5 73 Note: Based on imports of the developed countries of Western Europe and North America. Centrally planned economies include China. Source: OECD (1986), based on GATT data. Cable 633 Table 5. Net Trade Balances with Developing Countries (billions of dollars) Economies and sector 1975 1977 1979 1981 1983 1984 All developed (total) -2.14 -5.00 -9.82 -9.17 -11.44 -16.38 Textiles 1.14 2.05 1.23 3.42 2.22 1.27 Clothing -3.28 -7.05 -11.05 -12.55 -13.66 -17.65 EEC (total) -1.06 -3.05 -5.39 -3.92 -3.74 -3.97 Textiles 0.23 0.02 -0.72 0.35 0.01 -0.12 Clothing -1.29 -3.07 -4.67 -4.27 -3.75 -4.09 United States (total) 1.31 -2.69 -3.68 -5.95 -8.30 -11.78 Textile -0.05 - 0.23 0.06 -0.47 -0.90 Clothing -1.26 -2.69 -3.91 -6.01 -7.83 -10.88 Japan (total) 1.14 2.69 1.17 2.03 2.10 1.47 Textiles 1.34 3.02 2.38 3.00 2.81 2.55 Clothing -0.20 -0.33 -1.21 -0.97 -0.71 -1.08 All developed in constant 1985 prices (total)' -2.4 -4.38 -6.88 -6.00 -8.00 -11.85 Textiles 1.14 1.92 0.85 2.45 1.69 1.01 Clothing -3.28 -6.30 -7.73 - 8.45 -9.69 -12.84 Note: Developing countries do not include those in Eastern Trading Area. a. Price deflation using index of developed-country manufactured exports. Source: Compiled from GATT International Trade Yearbooks. early 1970s, Sri Lanka and Indonesia in the late 1970s, and more recently in the Maldives and, on a much larger scale, China; Young and Hood 1985). But the increasingly all-embracing character of MFA controls (at least among developing countries) has now so narrowed the scope for further trade diversion among developing countries that its restrictive effects are more likely to be reflected in future in the growth of total developing country exports. The Effect on Developed Country Economies Basic principles and a good deal of empirical work both show that sectoral protection creates higher prices for the protected good and an increase in output and employment in the protected industry in the short term. There is little consensus on the magnitudes of price increases. The price effect depends in part on the tariff equivalent of quota protection (which is subject to serious estimat- ing problems), on the extent to which a "law of one price" operates that trans- mits higher border prices to protected domestic producers' prices, and on the extent to which retail and distribution margins reflect movements in relative producer costs. Silbertson (1984) suggested that average U.K. retail prices of imported and domestically produced textiles and clothing would be 5 to 10 percent lower relative to prices of other goods, were the MFA abolished. And the price effects are larger for lower quality items. In evaluating the effects of sectoral protection on output and employment, problems of methodology and interpretation center on the extent to which trends in output-labor ratios can be expected to remain stable with a higher level 634 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 of protection (or whether further capital deepening will occur); and the costs of adjustment avoided by protection which are a function of the character of the labor force and the state of national or regional labor markets affected. The Silbertson study (1984) estimated that, depending on the assumptions used, over the period of the forthcoming MFA4, complete liberalization of quotas would displace 10,000 to 50,000 jobs altogether in the U.K. textile and clothing indus- tries (from levels of around 500,000) but that 20,000 a year would be lost anyway through productivity losses, even if the MFA was preserved. Moreover, the cost to the U.K. economy of saving a job is considerably more than the average wage. The same conclusion is reached by other studies, particularly in the United States, which suggest that the annual cost per job protected ranges from two to eight times the annual wage in the industry (see Hufbauer, Berliner, and Elliot 1986; and Tarr and Morkre 1984). These sector-specific considerations have to be considered in an economywide framework. Restrictions on imports raise the equilibrium exchange rate, offset- ting the inflationary impact of protection but reducing prices and thus output and employment in other traded good sectors. Conversely, quota premiums accruing to exporters raise the cost of imports in the import-restricting country with the opposite effect on the exchange rate. If wage bargaining and levels of employment are linked, price increases will either reduce real wages or, if wages are raised, reduce employment. Clearly, the magnitude and direction of the secondary effects depend heavily on the macroeconomic assumptions used in any particular model, and no claims to precison can be made in this area (Cable and Weale 1985). Some of the most serious consequences of MFA protection relate to the dy- namic effects, and the impact on the process of adjustment. "Adjustment" to changes in trade patterns and increases in imports in an MFA context was clearly intended to mean exiting from production, diversification or closure by firms, and switching to other activities by workers. To a significant extent, adjustment of this kind has been taking place in industrial countries. The number of textile and clothing producing firms has dropped substantially; employment in the sector has fallen (from 4.5 million in the EEC when the MFA was introduced to under 3 million a decade later); and output has declined even where there has been some demand growth (end-1985 textiles output in the EEC was over 10 percent down on 1980 levels, and clothing output was 18 percent down). At the level of individual firms, it is possible to see evidence of major efforts to restruc- ture from textiles and fiber production (as Courtaulds has done in the United Kingdom-moving toward other chemicals and chemical end uses); the develop- ment of new fibers such as Tyvek and Kevlar with a whole range of new non- woven applications (Dupont, Enka, and the U.K. fiber groups); of successes in those segments of the industry where low wage competition is less plausible (for example, Vantona and household fabrics); of successful specialization in high- quality fabrics and garments less vulnerable to price competition (common among German, Italian, and Swiss firms); and of a switch from production to Cable 635 distribution or overseas offshore production (as Japanese firms have done on a large scale) (Cable 1983b, chap. 6; Shepherd 1981; de la Torre 1984). There are, however, forms of adjustment occurring which do not entail exiting from lines of production in competition with MFA exporters; indeed, they are designed to overcome it. Protection, by raising profits for producers, creates an incentive to keep resources in and to make new investment in the industry which may not be competitive at world prices. One form of investment being encour- aged is the use of faster and more flexible machines to cut labor costs. This has been occurring for decades in the textiles industry where such innovations as the introduction of shuttleless looms (from the 1960s) and open-ended spinning (from the early 1970s) have radically transformed the cost structure of produc- tion. The textile industry of industrial countries appears now to be more capital- intensive than manufacturing as a whole in terms of capital per worker. Heavy investment of a capital-deepening form has taken place, especially in the textile industries of France, the United Kingdom, and the United States, and costs have been further reduced by relocation to low-wage areas (in the United States) and vertical integration (in the United Kingdom and, to a lesser extent, in France). The rapid advance of microelectronic innovations has given textile producers a new self-confidence; the first wholly integrated spinning and weaving mill since World War II has just been installed in the depressed textile town of Rochdale, England. Skill and capital intensive techniques have become common for finish- ing both synthetic and blended fiber fabrics. In practice, and as the theory of the product cycle might have suggested, however, automated mass production of standardized goods using widely available technology is not a secure form of adjustment for developed country producers. The dependence of these firms on high break-even points in highly price-competitive markets has left them ex- posed to severe loss of market share when relative price factors-notably ex- change rates-have been adverse. The latest stage of technical innovation in labor-saving equipment has impor- tant implications for clothing as well as textiles. Clothing has so far remained a relatively highly labor-intensive industry (except in one or two segments such as hosiery) though important changes are now taking place. Lasers have been introduced into cutting and, together with computer aided design and auto- mated handling of materials, many operations are now being automated, at least for long production line items and in larger firms where heavy capital investment is economic (Disher 1986). The main operation in garment making is sewing and despite a steady increase in machine speeds this has remained a labor- intensive operation except for a few standardized garments such as jeans. That may, however, be changing. Efforts are now being made in the United States, the EEC, and Japan to automate the sewing operation, with substantial government backing in the EEC and Japan. Although there are many technical problems in such automation, mainly in terms of material handling to ensure a steady flow of faultfree cloth, these are being overcome. Sewing is also being replaced by 636 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 machine welding of seams. There is evidence, too, that other labor-intensive operations, such as embroidery, are now being brought within the range of mechanization, thanks to computer-controlled devices. An electronic sock ma- chine has been introduced. Widespread diffusion of commercially viable applica- tions is still some way off, and there are many barriers to diffusion, however, including the financial weakness of many small firms, the lack of trained person- nel, and the remaining technical problems. The early phase of textile automation represented a cost cutting approach for standardized output which was relatively insensitive to demand considerations, and in particular to the opportunities for new products and high quality lines. But the new technologies permit these concerns to be incorporated. Many com- panies in industrial countries are now developing what is called a "quick re- sponse" approach to take advantage of geographical proximity to retailers and to keep ahead of more far-flung producers in meeting fashion demand. Machine flexibility is an important element in this approach. For example, the use of advanced computer technology, allied to low-cost outworking, has been the hallmark of the most successful of the Italian knitwear goods producers, such as Benetton. They make extensive use of computer aided design, robotic cutting, and computer systems to monitor stock and sales, and they can use their com- puter system to produce individual items for order. Were the new technologies to be widely diffused, there is the potential of a quite fundamental shift in compara- tive advantage. Hoffman and Rush (1983) suggest that even if protection were given to the industry, however, it could be twenty years or more before the new technologies were widely used in the clothing industry. And it is far from clear that the heavy costs to the community of protection would ever be repaid. Effects on Developing Countries We have already noted that in aggregate terms, MFA controls have not pre- vented developing countries from increasing their market penetration of indus- trial countries at a rate comparable with manufactures generally. And there is a fair degree of underutilization of quotas, other than among the dominant suppli- ers. Nonetheless the magnitude of quota premiums clearly indicates that exports are considerably less than their potential. Various attempts have been made to estimate the magnitudes involved by simulating the effects of complete tariff and nontariff liberalization on developing country exports. In one estimate by Inter- national Monetary Fund (IMF) researchers, imports into the main OECD markets would rise by 82 percent for textiles and 93 percent for clothing given the assumption of infinitely elastic supply (a reasonably plausible assumption in all except the very short term) (Kirmani, Molajoni, and Mayer 1984). UNCTAD (1986) estimated that complete nondiscriminatory liberalization could raise de- veloping country exports of clothing by 135 percent and of textiles by 78 per- cent. It should be stressed that these are rather crude estimates which ignore the fact of a significant degree of quota underutilization by many suppliers. Ex- porters may not be willing to invest to produce up to the limit given uncertainty Cable 637 about future quota levels, and because small incremental investment may not be possible, making it unprofitable to invest to fill the 10 or 20 percent of a quota level remaining. There could be significant time lags, therefore, before exports could respond to liberalization. The growth of quota premiums have also had unintended consequences. It has led to political "rent-seeking" behavior, which can have major distributional effects where quota rents account for a significant share of an exporting coun- try's gross domestic product (GDP) (as in Hong Kong, where they are estimated by WVolf [1986] at about 5 percent of GDP). Premiums also have also led to distortions in resource use by providing artificially high profits, and to market rigidities, as historic market shares determine which firms will obtain the premi- ums. A more positive indirect result of the premiums has been financing for diversification. This diversification, however, can be in the direction of future comparative advantage, or it can be used in inappropriate, highly capital-inten- sive ventures. There is the unquantifiable but undoubtedly important effect of premiums on expectations, influencing the willingness of decisionmakers in de- veloping countries to face the adjustment costs of adapting to a more export- oriented trade policy. The impact of MFA restrictions on the orientation of currently small but potentially large suppliers such as Bangladesh, India, Paki- stan, and Sri Lanka is of particular importance. If textile and garment quotas were to be removed, there would be major distributive implications as between developing-country exporters. Such is the degree of regulation at present that it is difficult to say which countries would benefit most from the increased opportunities, and which would experience a loss of market share. The main implications are for garment trade, which is larger than that for textiles and more tightly restrained. It would seem that the main casualties would be the middle-income countries which do not have well- developed textile export industries, have relatively high wages, and have ex- panded or maintained exports mainly through taking advantage of gaps in the quota regime such as many of the Latin American exporters, Malaysia, and Singapore. What is less clear is whether, under the influence of market forces alone, the center of garment production would shift to the countries with low labor cost production (especially South Asia); to countries with proximity to EEC and U.S. markets; or to those countries where high technology can be best integrated with low labor cost outworking and fashion sophistication, as cur- rently occurs in Hong Kong. On the few occasions when it has been possible to simulate market conditions (for example, the adoption of global quotas by Norway), there was a strong shift, in the short run, to Hong Kong. III. TOWARD LIBERALIZATION The Political Economy of Protectionism In making an assessment of the likelihood or form of any liberalization, it is necessary to take into account the forces which have contributed to the current 638 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 arrangements. A great deal of analytical work has been carried out over the last five years which has contributed to an understanding of the politics as well as the economics of protectionism. (A good summary in relation to the U.S. literature is found in Magee 1986.) While much of the policy-oriented literature and the rhetoric of policymakers centers on problems of "adjustment"-how to manage redundancies and factory closures-this is probably less important than the political pressures created by intramarginal producers in protected industries that enjoy higher profits and wages, together with exporters that gain quota rents under the voluntary export restraints (VERs). These protectionist interests will not be weakened by adjustment policies designed, for example, to facilitate absorption of displaced workers into the labor force. The most intense pressure for protectionism in the United States has come at a time when overall economic activity has been rising rapidly and unemployment falling, a context in which adjustment should be relatively easy. It is often the interests of capital and management, rather than labor, which are predominant. Experience suggests that pressures for protection sharply de- cline where manufacturers-especially in the more capital-intensive, upstream operations-can adjust through diversification, or by switching into profitable importing and retailing activities, or by offshore processing of domestic fabrics and fibers. By contrast, reinvestment in lines which directly compete with im- ports creates an enduring interest in protection even if the labor force is consider- ably reduced (see, for example, Cable 1983a). The current move toward investment in automated garment manufacture is ominous in that respect. Experience of other industries (shoes, consumer electronics, cutlery, toys, leather goods) suggests that there is nothing inevitable about protectionism. In most industrial countries, pressures for MFA-type quota systems have been re- sisted and firms and workers have adjusted to import competition. The textiles and clothing industry is significantly larger, and effective coalitions have been formed between fiber, textile, and clothing interests. Moreover, current arrange- ments could not have been sustainable without active support or passive acquies- cence of many of the developing-country participants in the MFA. To these interests and the wider question of MFA renewal and renegotiation we now turn. The Prospects The powerful forces maintaining the status quo should not be underestimated. Producers in developed countries fear uncertainty and competition, and pro- ducers in both developed and developing countries derive economic rent from protection. Many have entered production and expanded capacity in response to profitable opportunities provided by regional protection, especially Southern European exporters within the EEC. Some in developing countries value the predictability of quotas above the opportunities of open markets. A substantial number of bureaucrats are employed to operate the systems. And there are negotiators who see no better way of reconciling diverging interests. Substantial numbers of developing country governments have never pushed their criticism of Cable 639 the MFA to the extent of mounting a serious assault on it, which suggests some degree of implicit endorsement. Nonetheless, there are some long-term forces making for liberalization. There is recognition in industrial countries that protectionism does impose consider- able costs on importing economies by worsening the inflation-employment tradeoff. Developing countries will need to be able to expand export volumes more rapidly to maintain interest payments if the current approach to debt is to be sustained. And pressures for liberalization are mounting as a result of IMF and World Bank adjustment programs in developing countries which have led to the adoption of more export-oriented policies. The interest of a growing number of developing country exporters in enlarging their export volume growth, espe- cially among low-income developing countries, is one factor making for a much more assertive approach by developing countries to the MFA., Some developed countries, notably those that belong to the EEC, seemed in MFA4 negotiations to have come to terms with the need for liberalization of the MFA and the eventual application of GATT rules to textiles and clothing. Other developed countries, such as Australia, have been pressing for trade in textiles and clothing to be returned to GATT rules and liberalized forthwith. In the short run, the main scope for liberalization lies in measures which create greater flexibility within the framework of bilateral agreements, or at least those still open to continuing negotiation. These measures may include: * Improvement and gradual widening of carryover and swing provisions to allow exporters to carry forward unused quota allowances and move be- tween quota categories * Gradual elimination of controls on very small exporters * Widening quota categories ("broad banding") to eliminate unnecessary pro- liferation (for example, women's/girls', men's/boys', knitted/woven dis- tinctions could be discarded) • Elimination of underused quotas and those in industries where import penetration is close to 100 percent * Removal of quotas on textiles which have been made largely redundant by technology change and represent an unnecessary cost to the garment indus- try. Outward processing quotas allow sewing or other labor-intensive as- pects of the production process of a domestic country firm to be undertaken abroad and permit reimport of the product under a specific or expanded quota limit. This is one way of restoring the balance of interests between clothing and textiles producers. Textile Issues in the New Round If more far-reaching changes are to be accomplished in the new round, a protocol amending the MFA will be needed which clearly signals the end of the current arrangements. The Punta del Este Declaration refers to textiles in the following terms: "Ne- 640 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 gotiations in the area of textile and clothing shall aim to formulate modalities that would permit the eventual integration of this sector into GATT on the basis of strengthened GATT rules and disciplines, thereby contributing to the objective of further liberalization of trade." But textiles and clothing are relevant to or are subsumed by other issues, such as tariffs, safeguards, differential treatment and reciprocity, "standstill," "rollback" of trade barriers, and nontariff measures. Tariffs and nontariff barriers. On tariffs, there are two points to consider: first, whether tariff reductions would help to promote trade in textiles and clothing, and second, whether quotas under the MFA could be converted to tariffs which are then gradually reduced, as part of a return to GATT rules. Weighted average "most favored nation" (mfn) tariff levels for textiles and clothing remain high in the developed countries, even after the Tokyo Round cuts, at 11.5 percent in Japan and the EEC, 19.0 percent in the United States, and 21.5 per cent in Canada. In contrast, weighted average mfn tariffs for manufac- tures (excluding petroleum) in these markets are now 5.5, 6. 5, and 8.5 percent, respectively. Most favored nation tariffs on textiles and clothing generally are now between two and three times higher than tariffs on manufactures as a whole. Levels of effective protection are even higher, as a result of tariff escala- tion. As such, tariffs remain an important barrier to trade between most devel- oped countries. The major exceptions are intra-EEC and EEC-EFTA (European Free Trade Association) trade, which are duty-free. Tariffs are of relatively little consequence for developing countries, however, as voluntary export restraints under the MFA are the main constraint on their textile and clothing exports. There is already some tariff liberalization under regional preferential arrangements, such as the EEC-ACP (African, Carribbean, and Pacific) Lome Convention, the EEC's agreement with Mediterranean coun- tries, and some schemes under the Generalized System of Preferences (GSP), although the GSP provides the smallest proportion of coverage to textiles and clothing of any industrial sector. Only 71 percent of tariff lines on textiles and clothing are covered by the GSP; the average is 90 percent. Some regional prefer- ential agreements also have limited provisions for textiles and clothing. The EEC and Japan have introduced unlimited duty-free treatment, however, for the least developed countries. Nontariff barriers (NTBs) are the major obstacle to developing country textile and clothing trade. UNCTAD (1986) has estimated that if the EEC, Japan, and the United States removed all such NTBS, this would generate additional developing countries' exports of $11.8 billion, or 75 percent of their 1983 levels. If quotas were lifted, the subsequent removal of all tariffs on a preferential basis would generate an additional $5.8 billion in export earnings, while on a mfn basis, the increase would be $5.3 billion. Earlier, Craig McPhee estimated a $10.7 billion gain to developing countries in these three markets from tariff elimination on a preferential basis (based on 1980 trade flows, and assuming removal of the MFA), while on a mfn basis, the gains were estimated at $6.8 billion (UNCTAD 1985, p. 78). For developing countries in general, quotas, not tariffs, are the key issue. Cable 641 An exception to the rule that tariffs are of secondary importance is Australia which, following its departure from the MFA in 1977, replaced its VERS with high tariffs. In addition, Australian imports of most clothing and a third of textiles items above a certain volume (that is, tariff quota) are subject to a duty ("pen- alty"). This, in ad valorem terms, effectively discriminates against lower unit value imports from developing countries. The high tariffs involved appear to have had the same protective effect as quantity restrictions in other countries. But developing countries have increased their share of Australia's clothing im- ports because of the annual increase in the tariff quota-about 2 percent plus the rate of market growth, which gave a weighted average increase of 15 percent in 1985 over 1984. New Zealand maintains similar high tariffs (96 percent on clothing, 15.5 percent on fabrics) though it also operates quantitative restric- tions (QRS), with global quotas on a third of textile imports and 90 percent of clothing. But its use of QRS iS falling as the government has sought to implement its commitment to replace import licensing with tariffs. One way of phasing out the MFA could be for other countries to follow the Australian example, to convert VERS or QRS to equivalent tariffs which could be gradually reduced. This would also help to make the costs of protection more transparent and their reduction, therefore, more politically feasible. It could also serve as a model for the liberalization of trade by developing countries them- selves, since there are similar problems of dismantling complex quota regimes in the face of political resistance. The technical difficulties of agreeing on a set of tariff equivalents to existing quotas and then binding them in a GATT context should not, however, be underestimated. Safeguards. The GATT'S Article XIX allows countries to adopt tariffs as safe- guards against injury to domestic producers, under conditions which developed countries consider to be overly restrictive (see Hindley, this issue). Developed countries see a reformed safeguard clause as a prerequisite to the removal of the MFA. Essentially the issue is whether quota action in the case of demonstrated market disruption should be allowed on a selective basis or whether (as develop- ing countries insist) the nondiscriminatory character of Article XIX should be preserved. This issue is discussed more fully below. Differential treatment. The issue of differential treatment is central to return- ing textiles and clothing trade to GATT rules. The MFA has led to differentiation at three levels-between developed and developing countries; between develop- ing countries who are not MFA signatories and those which are; and between different groups of MFA signatories. Under this system, nOn-MFA member coun- tries frequently receive preferential treatment. For example, the EEC confers liberal terms on imports from Cyprus, Malta, and other Mediterranean basin countries. As part of its Caribbean Basin Initiative, the United States has intro- duced a new regime for imports made from cloth woven and cut in the United States, imported from sixteen Caribbean countries. This will still limit the amount any single beneficiary can export to the United States. Differential treat- ment among MFA members has involved giving the most established suppliers the lowest quota growth rates, with the stated objective of redistributing some of 642 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 their market share to other countries. Under MFA4 the EEC has proposed annual growth rates of 0-1 percent for the four "dominant" suppliers, 6-7 percent for the least developed, and 4-6 percent for others. While the developing countries have called for an end to any discrimination in favor of developed countries, they have endorsed the need for special treatment of the least developed countries. They have also called for differential treatment of other groups of MFA developing countries in the phase-out period in recogni- tion of their differing interests. Thus, the future of the textiles regime is inex- tricably bound up with the related issues of graduation, differentiation, and reciprocal obligations. Reciprocity. Following the trend in discussion on trade liberalization in other sectors, a number of developed countries have raised in an MFA context the issue of reciprocity from developing countries in return for more favorable treatment for their textiles and clothing exports. The United States has stated that it would seek reciprocal commitments from exporting countries on measures to open their markets for both textiles and clothing, as well as commitments to reduce subsidies or other trade-distorting measures on their exports. The EEC has adopted a similar position, restricting its demand for reciprocity to the more advanced exporting countries for whom annual quota growth rates will be determined according to the "openness" of their markets. Hong Kong (in the United States) and Singapore (in the EEC) have been accorded special status on these grounds in recent bilaterals. Such demands are likely to be pursued in any further discussions on the liberalization of textiles and clothing in the context of a new round. In addition, reciprocity may be sought in the sense of a tradeoff with other areas. EEC sources have suggested that developed countries might try to secure advance in the treatment of services in return for textile liberalization. The developing coun- tries' formal position has been to reject this approach outright. Not only are they opposed to the general principle of reciprocity, but in particular they have re- jected the concept of any concessions in return for the liberalization of restric- tions which they see as a fundamental departure from GATT principles. In practice, however, some form of linkage, even if only implicit, may very well be sought as a way of liberalizing textiles. Integrating Textiles into the New Round On both a formal and a practical level, the textiles issue is in abeyance for the next few years. It will not figure prominently, if at all, in the first stages of multilateral negotiations. Nonetheless, it is of considerable importance politi- cally and in decisions about negotiating modalities. Politically, a commitment to further liberalization, leading to a phasing out of textiles restrictions would have a large weight in making the round ultimately acceptable to most developing countries. It is the only product area outside primary commodities where developing countries' revealed comparative advan- tage has manifested itself in a trade surplus, and it accounts for no less than 25 Cable 643 percent of developing countries' (nonoil) manufactured exports, and for over 20 percent of total exports for some (Bangladesh, Cyprus, Hong Kong, India, Korea, Malta, Mauritius, Pakistan, Sri Lanka, and Turkey). As the OECD (1985) acknowledges, "The expansion of textiles and clothing exports had become for the developing countries an increasingly important determinant of their eco- nomic development." If the new round is to satisfy the large majority of its participants, some forward momentum on textiles would seem to be essential. In negotiations, the textiles issue will influence discussions because of the implications for a return of textiles trade to normal GATT rules in the 1990s. There is also considerable overlap between the textiles question and the discus- sion on safeguards and tariffs, and on more general themes such as reciprocity and differentiation. Developed countries will not be able so easily to make further extension of the MFA a prior condition for agreement in these areas as they were able to do with tariffs in earlier rounds. In the light of these factors, a few general suggestions are offered as to how the round might deal with textiles. The underlying assumptions are, first, that devel- oping country preoccupations have to be given considerably more attention than in earlier rounds reflecting their greater weight in the trading system and in the GATT. The second is that the format of negotiations will have to change signifi- cantly from earlier rounds. Experience has shown the difficulty of attempting to achieve a single package from one major set of negotiations. This ungainly and unbalanced approach was barely workable in the Tokyo Round and is even less likely to produce results in the current, more difficult, environment and with a larger number of countries seeking effective participation. A round cannot make progress if each issue is to be indefinitely deferred until agreement on some final single package in which all the issues are resolved. But despite its extreme diffi- culties, the latter approach is not without its supporters, particularly in the EEC, and there will need to be specific proposals to counter it if the negotiations are not to be effectively stalled for a considerable time. One has only to consider, for example, how long it might take to achieve some directly linked progress on textiles, services, and agriculture, to foresee the scope of this problem. A new approach to negotiations which seems to offer a path forward is phased discussions. This would envisage a series of negotiations which would be drawn up based on the prospects for achieving useful results. It implies that eventually all issues should be incorporated into rolling negotiations but that this could be done in such a way as to provide for different approaches and timing. Implementing a Phased Approach One design for a phased approach could be the following, presented in a much simplified form and highlighting those elements which relate to textiles. The process must begin with a "stdndstill" agreement, with substantive lan- guage to halt further imposition of protective measures (to the extent that they are not covered in the Punta del Este communique). This would reinforce pres- sures on MFA signatories not to increase the restrictiveness of quotas under MFA4. 644 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 Phase I would then be designed to restore confidence in the GATT system, through quick resolution (say within twelve months) of issues on which the debate is relatively advanced, which are not technically difficult, or where the outline of common positions is already discernable. It is possible to envisage such agreements in the following areas: disputes settlement, surveillance and transparency of all barriers and procedures, tropical products, treatment of least developed countries, and natural resource products. Negotiations would also start on two other issues-tariffs and rollback of trade barriers not consistent with the GATT-but these could continue through subsequent phases. Prepara- tory work could also start on such issues as agriculture. But the key to this first phase would be safeguards, agreement on which could unlock other problems, including textiles. There has been, over the last few years, a convergence of views on several aspects of safeguards including the desirability of adequate transparency of measures, consultation between trading partners before adoption of these measures, and multilateral surveillance of agreements; compensation for and retaliation against new protection; limiting the duration of safeguard action; and the progressive liberalization of applica- tion. The sticking point is the insistence of the EEC that the importing country be allowed to apply Article XIX actions selectively against specific country ex- porters with other parties-developing and developed-favoring nondiscrimina- tory mfn treatment in applications of the Article. The contracting parties have been through the arguments on all the issues so often that progress is now a matter of political will. In the main, progress will require movement by the EEC away from selectivity; this would provide the impetus to resolving most of the other issues relatively quickly and certainly within the period suggested for phase 1. As the great majority of GATT members now recognize the need for urgent action on safeguards, both as a crucial issue in its own right and as a major element in restoring the GATT'S credibility, its resolution should not prove impos- sibly difficult. In phase II, negotiation would involve some of the more difficult issues, focusing on reintegration of textiles trade into the GATT, but including liberaliza- tion of the highest tariff items; infringement of intellectual property rights; barriers to agricultural trade (rather than the more difficult problem of domestic distortions); and antidumping, countervailing duties, and government procure- ment. A principal objective over the approximately two-year period, would be to develop a regime to allow the expiration of MFA4 without its extension. In each case, however, developing-country negotiators would have to decide how far to press for further liberalization if the price to be paid was a greater reluc- tance by industrial countries to negotiate a timetable for phaseout of the MFA. It may be worthwhile, therefore, to create some linkage between the textile trade regime and the "new" issues in which developed countries have initiated de- mands, such as the liberalization of services trade. The main purpose of phase III would be to finalize (or at least make substan- tial progress in) treatment of the most difficult issues, such as trade-related Cable 645 aspects of investment, intellectual property rights and especially services, and to complete negotiations on those items which have been programmed to cover more than one phase-notably agricultural trade and rollback of barriers. Agreement would also be sought on the treatment of countertrade in the GATT, the process of integrating developing countries into the multilateral trading sys- tem would be continued, and any issues unresolved from earlier phases would be concluded. A three-year process might be envisaged, ending in December 1992. Within this phase the new textiles regime would come into effect and the spirit and content of that regime could have an important influence on perceptions, especially those of developing countries, in this concluding phase. At this stage it is extremely difficult to anticipate the shape of any pOSt-MFA arrangement. The most plausible mechanism for liberalization would be one which incorporated rising quota increases and the sweeping away of controls on the smaller and low-income suppliers. The difficulty facing those wanting an end to the MFA will be to find a transitional arrangement which does not merely become another MFA; that is, a slightly more liberal version of present arrange- ments. Past experience might suggest to the cynical that little real progress will be made on textiles in the foreseeable future given the accretion of vested interests in both developed and developing countries. More positively, a slow process of intellectual osmosis concerning the merits of economic liberalization in general and trade liberalization in particular seems to have had some impact in stem- ming the tide of protectionist thinking. There is no better place to exploit any changes for the better than in relation to the MFA. REFERENCES Aggarwal, V. 1985. Liberal Protectionism: The International Politics of Organized Tex- tile Trade. Berkeley: University of California Press. Cable, V. 1983a. Economics and the Politics of Protection. World Bank Staff Working Paper 569. Washington, D.C. 1983b. Protectionism and Industrial Decline. London: Hodder and Stoughton. Cable, V., and M. Weale. 1985. "The Economic Costs of Sectoral Protection in Britain.' World Economy 6, no. 4 (December): 421-38. De la Torre, J. 1984. Clothing Industry Adjustment in Developed Countries. Thames Essay 38. London: Trade Policy Research Centre. Disher, M. 1986. "High Technology in the Clothing Industry." Textile Outlook Interna- tional. London: Economist publication. September. GATT. 1984. Textiles and Clothing in the World Economy. Geneva. .1985-1986. International Trade Yearbook. Geneva. Hamilton, C. 1984. "Voluntary Export Restraints on Asia: Tariff Rents and Trade Bar- rier Formation." Stockholm: Institute for International Economics Studies Paper 276. Hoffman, K., and M. Rush. 1983. "Microelectronics and the Transformation of the Clothing Industry." International Labour Organisation Working Paper. WEP 2-22/ WP 138. 646 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 Hufbauer, G. C., D. T. Berliner, and K. Elliott. 1986. Trade Protection in The United States: 31 Case Studies. Washington, D.C.: Institute for International Economics. Jenkins, G. P. 1980. Costs and Consequences of the New Protectionism: The Case of Canada's Clothing Sector. Ottawa, Canada: North-South Institute. Keesing, D., and M. Wolf. 1980. Textile Quotas against Developing Countries. London: Trade Policy Research Centre. Kirmani, N., P. Molajoni, and T. Mayer. 1984. "Effects of Increased Market Access on Exports of Developing Countries." IMF Staff Papers (December): 661-84. Koekkoek, K. and L. Mennes. 1986. "Liberalising the Multi-Fibre Arrangement." Jour- nal of World Trade Law (March/April). Magee, S. 1986. "The Political Economy of Protection." Kiel Conference on Free Trade in the World Economy. June. OECD. 1983. OECD Textiles and Clothing Industries. Paris. 1985. The Costs and Benefits of Protection. Paris. 1986. "Compatible Trade and Production Data Base, 1970-80." Department of Economics and Statistics Working Paper 31. Paris. Pelzman, J. 1984. "The MFA and Its Effect on the Profit Performance of the U.S. Textile Industry." In R. Baldwin and A. Krueger, eds., The Structure and Evolution of Recent U.S. Trade Policy Chicago: Univ. of Chicago Press, 111-41. Shepherd, G. 1981. Textile Industry Adjustment in Developed Countries. London: Trade Policy Research Centre. Silbertson, A. 1984. The MFA and the U.K. Economy. London: Her Majesty's Station- ery Office. Tarr, D. G., and M. Morkre. 1984. Aggregate Costs to the United States of Tariffs and Quotas on Imports: General Tariff Cuts and Removal of Quotas on Automobiles, Steel, Sugar and Textiles. Washington, D.C.: U.S. Federal Trade Commission. UNCTAD. 1975. Operation and Effects of the Generalized System of Preferences. TD/B/ C.5/42. New York: United Nations. 1986. "Protectionism and Structural Adjustment.' TD/B/1081. New York: United Nations. Wolf, M. 1986. "The MFA-An Obstacle to Development." Kiel Conference on Free Trade in the World Economy. June. Young, N., and S. Hood. 1985. "Foreign Direct Investment and the Hong Kong Textile and Clothing Industry." Textile Outlook International. London: Economist publica- tion. September. THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4: 647-668 Differential and More Favorable Treatment of Developing Countries and the International Trading System Martin Wolf This article argues that "differential and more favourable treatment" of developing countries in the General Agreement on Tariffs and Trade (GATT) has been a logical consequence of their own inward-looking policies and the GATT'S implicit mercantilism, the latter implying that liberalization, being costly, should not be demanded of rela- tively poor countries. Time has, however, reduced both the appeal of the protectionist model of development and the willingness of developed countries to accord differential treatment. The upshot has been pressure on more advanced developing countries to 'graduate" and a growing literature recommending fuller and more equal participation of developing countries in the GATT. The case for fuller and more equal participation is not self-evident. It needs to be assessed on its merits in terms of the prospects for improved market access abroad and more efficient policy at home. The analysis indi- cates that the potential benefits should not be oversold. On balance, however, the most advanced developing countries would probablygainfrom active and more equal partic- ipation in both GATT and the multilateral trade negotiations while the remaining devel- oping countries would benefit from graduation by the more advanced. (iv) Contracting parties agree that the principle of differential and more favourable treatment embodied in Part IV and other relevant provisions of the General Agreement and in the decision of contracting parties of Novem- ber 28, 1979 on differential and more favourable treatment, reciprocity and fuller participation of developing countries applies to the negotiations. In the implementation of standstill and rollback, particular care should be given to avoiding disruptive effects to the trade of less-developed contract- ing parties. (v) The developed countries do not expect reciprocity for commitments made by them in trade negotiations to reduce or remove tariffs and other Martin Wolf is director of studies at the Trade Policy Research Centre, London. The views expressed do not represent those of the staff and council of the Trade Policy Research Centre, and result partially from a research program on the participation of the developing countries in the international trading system. Particularly important are two papers prepared under the program by Hudec (forthcoming) and Riedel (forthcoming). The issues addressed in this article have also been considered in a previous paper by the author (1984, pp. 201-29). The author is grateful to participants at the Bangkok conference for their comments and especially to Jagdish N. Bhagwati, Muchkund Dubey, Anthony Lane, and Vincent Cable. Copyright ( 1987 by the International Bank for Reconstruction and Development / THE WORLD BANK. 647 648 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 barriers to the trade of developing countries. Developed contracting parties shall therefore not seek, neither shall less-developed contracting parties be required to make, concessions inconsistent with the latters' development, financial and trade needs. (vi) Less-developed contracting parties expect that their capacity to make contributions or negotiated concessions or take other mutually agreed action under the provisions and procedures of the General Agreement would improve with the progressive development of their economies and improvement in their trade situation and they would accordingly expect to participate more fully in the framework of rights and obligations under the General Agreement. (vii) Special attention shall be given to the particular situation and prob- lems of the least developed countries and to the need to encourage positive measures to facilitate expansion of their trading opportunities. -From the Ministerial Declaration at Punta del Este, Uruguay, announcing agreement to start the "Uruguay Round" of multilateral trade negotiations within the framework of the General Agreement on Tariffs and Trade, September 1986. The theme of this article is the relation between two sets of ideas. The first of these is the approach to trade policy that derives from the view of economic development adopted by the governments of most developing countries since the end of World War II. The underlying vision is one of market failure and of the need for active government intervention in the market, particularly with regard to transactions with other countries. The second set of ideas is that of the General Agreement on Tariffs and Trade (GATT) itself. Those ideas may be described as "disciplined mercantilism" or, better perhaps, the "mutual disarma- ment approach to trade liberalization." Protection is appropriate, according to this vision of the world, in the absence of an acceptable reciprocal bargain. Such a vision makes it easier to justify giving in to protectionist pressure, while still retaining some curb upon it, than would a consistent attachment to the principle of free trade, an attachment which existed in the United Kingdom, for example, during much of the nineteenth century. The issue to which much of the present article relates these ideas is the "differ- ential and more favorable treatment" of developing countries within the GATT. The concept of more favorable treatment derives from the attempt to accommo- date the interventionist view of the relation between trade and development within the GATT system. The article argues that this process of accommodation has created problems because the economic ideas that underlie both the interven- tionist approach to trade policy and the GATT are unsound. The focus of the present discussion is this broad theme, not precise negotiating tactics and options.1 The argument is structured as follows. First, the meaning of "differential and more favourable treatment" will be explored, along with the 1. Those negotiating issues and options are carefully considered in Balassa and Michalopoulos 1985; see also Scott and others (1984). Wolf 649 way in which the place of developing countries in the international trading system has become a vexed issue. Second, the origin of the calls for special treatment of developing countries will be traced, focusing on the economic ideas and experience that persuaded the representatives of developing countries to advance these arguments. Third, the discussion will turn to how those claims have interacted with the principles of the international trading system itself. The way in which economic developments have undermined the postwar view of the link between trade and development and the implications for the trade policies of developing countries is the next subject. The principal conclusion is that the developing countries have won a series of largely pyrrhic victories. The discussion concludes with a consideration of the issue of graduation and exam- ines what the developing countries should do now and what benefits they might derive from alternative approaches to the issue of "differential and more favor- able treatment" in the Uruguay Round. I. ROLE OF DEVELOPING COUNTRIES IN THE TRADING SYSTEM "Differential and More Favourable" Treatment The essence of the exceptional status desired by developing countries is quite simple: the developed countries should grant to the developing countries more favorable market access than they do to one another, while conceding to the developing countries very considerable freedom in their own trade policies. To clarify the dimensions of that special position, it is useful to start with the Framework Agreements of the Tokyo Round. In a decision of the GATT's con- tracting parties of November 28, 1979, entitled "Differential and More Fa- vouirable Treatment, Reciprocity and Fuller Participation of developing Countries ' the first article in the general statement states: Notwithstanding the provisions of article I of the General Agreement, contracting parties may accord differential and more favourable treatment to developing countries, without according such treatment to other con- tracting parties. [GATT 1979, pp. 5-7] This article is then explained as having four aspects: (i) the Generalized Sys- tem of Preferences (GSP); (ii) "differential and more favourable treatment" of developing countries in the context of codes on nontariff barriers negotiated within the GATT; (iii) permission for developing countries to create regional and global trading arrangements that do not conform to Article XXIV of the GATT; and (iv) particularly special treatment for least developed countries. The nar- rowest definition of "differential and more favorable treatment" would look only at item (ii) above, namely, treatment of developing countries within the codes on nontariff barriers. A somewhat broader definition would include all the four enumerated exceptions. There is, however, more to the special position of developing countries. As is 650 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 made clear in Article 5 of the decision of November 28, 1979, reciprocity is not expected of developing countries, an exception that had already been admitted in the GATT'S Part IV (Article XXXVI [8]). More precisely, "The developed countries do not expect the developing countries . . . to make contributions which are inconsistent with their individual development, financial and trade needs." The waiving of reciprocity, though not strictly part of "differential and more favourable treatment," cannot be ignored, since the GATT is a framework for mutual concessions. Mertion should also be made of Article XVIII of the GATT, which covers infant industry protection and restrictions for balance of payments purposes, the latter being much the more important. In this case, however, it is not the position of developing countries de jure that is "differential and more favourable,' as the developed countries have the largely equivalent Article XII. What is special is how the developing countries have been treated. In practice, Article XVIII, Section B, on balance of payments restrictions has provided developing countries with what amounts to a carte blanche for quantitative restrictions on imports that in some cases have lasted for years and in a few for decades.2 Accordingly, in the present discussion "differential and more favourable treat- ment" is taken as referring to everything mentioned in the Decision of November 28, 1979, plus the freedom of developing countries to introduce quantitative restrictions for balance of payments purposes. While this is not, perhaps, the legal definition of the term, it is the one that makes practical sense. The position of developing countries in the GATT system and the problems associated with it are the product of all aspects of their special treatment. So far as their own policies are concerned, developing countries have not reciprocated in tariff bargaining, their tariffs are virtually unbound, they can readily impose quantitative restrictions for balance of payments purposes and they have been relatively unconstrained-at least until recently-in the use of domestic subsidies and other nontariff measures. Starting in the second half of the 1970s and continuing in the 1980s, the enthusiasm with which the United States of America has pursued so-called unfair trade practices has grown sub- stantially, one result being a great increase in countervailing duty cases against exports from developing countries. These cases have affected not merely the use of export subsidies by developing countries, but almost equally that of domestic subsidies (though countervails of such subsidies are, at least arguably, contrary to the Code on Subsidies and Countervailing Duties agreed in the course of the Tokyo Round). (On this development and its implications for the developing countries, see the papers by Nam and by Finger and Nogues in this volume.) This new limitation on the freedom of action of developing countries has oc- 2. According to a paper by Richard Eglin of the GATT Secretariat, seventeen countries were invoking Article XVIII (B) in 1985, many of these countries having consulted in the GATT'S Balance of Payments Committee repeatedly in the previous ten years. In fact, fourteen countries consulted six or more times between 1974 and 1985. (See Eglin 1987, table 1). On the general issue of quantitative restrictions for balance of payments purposes, see also Anjaria in this volume. Wolf 651 curred not so much because of changes in their formal obligations under the GATT but rather because of changes in the domestic policies of the GATT'S most important single contracting party. In other words, in the area of subsidies the de facto curbs on the developing countries are now tighter than those in effect de jure. So far as the policies of developed countries are concerned, however, the principle of preferential access to the markets of developed countries was for- mally accepted as a permanent part of the GATT after a long rearguard action by the United States. Difficulties with the Position of Developing Countries It had been thought by developing countries that favorable treatment would be granted because of a consensus that they have special problems. Despite apparent agreement on the principle, the developing countries have found them- selves exposed to discriminatory export restraint arrangements and other trade distorting measures in many industries in which they have shown a comparative advantage, agriculture, textiles and clothing, and steel being the most important examples. Moreover, the incidence of these restrictions has not been declining as the developing countries have secured concessions-quite the contrary (see World Bank 1986, pp. 23-24, and Nogues and others 1986). At the same time, the preferences granted to the developing countries have been so hedged about by safeguards that their value has been widely questioned (for an assessment of the GSP, see Langhammer and Sapir, forthcoming). From the point of view of the developed countries, the "concessions" they have granted to the developing countries have carried only modest domestic political costs, largely because they have had little disruptive economic effect at home. The developed countries might feel thankful that the existence of highly protected markets in developing countries provides a convenient excuse for their protectionism against developing countries, even if it is not the reason for it. Nevertheless, not all in developed countries are protectionist. Some incline to liberalism. Consequently, to the evident dissatisfaction of the governments of developing countries must be added the dissatisfaction of those in the developed countries who would like to see a more liberal international trading order. The question naturally arises whether it is not possible to do better. II. TRADE POLICY AND THEORIES OF DEVELOPMENT The cornerstones of the Havana Charter of 1948 were the removal of quanti- tative restrictions, nondiscrimination, and reciprocal bargaining over tariffs. Representatives of the developing countries rejected these ideas, tabling some eight hundred amendments to the draft Charter, demanding, in effect, that they should be free of obligations which would limit their autonomy in trade policy. They wished to protect infant industries with measures not otherwise per- mitted; they wished to be permitted to receive new tariff preferences from 652 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 other developed or developing countries; they wanted the right to benefit from developed country tariff concessions without having to offer equiva- lent tariff concessions of their own . . . About the only element of current developing country policy . . . that did not appear at Havana was the call for systematic tariff preferences by the developed countries ... There were demands that the ITO rules should permit the granting of such preferences, but there was no demand for a legal obligation requiring developed coun- tries to do so. [Emphasis in original; Hudec, forthcoming] As Hudec has shown, it took a long time for the developed countries to grant freedom of action formally, but de facto even those relatively few developing countries that joined the GATT in its early days had that freedom from the beginning (part I). Most of the already independent developing countries either continued with or reinforced the protectionist policies of the 1930s and 1940s. The principal colonial governments had also adopted protectionist policies on behalf of their dependencies during the 1930s and World War II. Following independence these countries generally went still further along the same path. India, in particular, which was probably the most influential to the newly inde- pendent countries, showed a strong and growing attachment to an inward- oriented development strategy. Self-fulfilling Prophecy of Foreign Exchange Scarcity The thinking underlying this inward-oriented development strategy was that import substituting industrialization would be required both to accelerate eco- nomic development and to save scarce foreign exchange. There were, in other words, both microeconomic and macroeconomic arguments for protection. On the microeconomic level, rapid industrial growth was assumed to be the defining characteristic and sine qua non of economic development. There was a hope that protection would accelerate the development of infant industries by internalizing the externalities presumed to be involved in industrialization. On the macroeconomic level, it was widely believed, both by the governments of the developing countries and by many distinguished academic commentators, not only that commodity exports faced stagnant demand and falling relative prices, but also that developing countries had little chance of competing successfully in world markets for exports of manufactures.3 Given such export pessimism, foreign exchange was presumed to be in permanently scarce supply. Import substitution plus aid were then the only feasible ways of avoiding the foreign exchange famine with which the developing countries were believed to be threat- ened. In short, whether the argument began with "linkages," "externalities," or "self-reliance," the desirability of protection was the conclusion. 3. These ideas are discussed in Little (1982, chap. 4) and Riedel (forthcoming). Some of the classic references include Nurkse (1953), Prebisch (1964), and Myrdal (1968). Prebisch's book was the intellec- tual starting point for the first United Nations Conference on Trade and Development, but it built on already influential ideas that Prebisch had promoted in previous years. Similarly, Myrdal's work synthe- sized previous analyses of the obstacles to economic development. Wolf 653 Protectionist pressure in developing countries has been further increased by the tendency to expand budget deficits in the name of development finance. Aid donors can be criticized for encouraging development strategies that included the deliberate planning of resource gaps to be filled by aid.4 The outcome has been pressure on the balance of payments, and inflation. Currencies have tended to become overvalued and import controls have had to be introduced, increasing protection in a haphazard, even chaotic, manner. A tragic element in this all-too-common history is that the idea of a foreign exchange shortage is generally self-fulfilling. Import restrictions usually start with supposedly nonessential consumer goods, but as they are tightened, the supply of imported basic intermediates and capital goods is also curtailed. As the level of protection increases so does the implicit taxation of exports, and export performance tends to deteriorate (see Clements and Sjaastad 1985). In most developing countries the result has been slower growth of output and, in some cases, a decline. Since the current account reflects the difference between na- tional output and expenditures, declining output (absolute or in relation to trend) often offsets the effects of any policies to constrain expenditures and the balance of payments position worsens. It can become impossible to combine the desired rate of growth with a manageable balance of payments, however high the level of protection. Few countries have managed successfully to reverse their direction once far down this protectionist road due mainly to the role of interest groups. On the whole, the protectionist trade policies of developing countries were not created by industrial interests, for those interests, like the industries themselves, did not exist. On the contrary, it was the protection that created the interests, but they have come to represent a powerful obstacle to liberalization. In the absence of international commitments (other than intermittent pressure from international lending agencies like the International Monetary Fund and the World Bank), those government elements wanting to liberalize have usually found themselves opposed both by other parts of government and by affected producer groups. Preferences as Panacea By the mid-1 950s it was becoming obvious that trade liberalization among the developed countries was associated with an explosive growth of trade. At this stage, there were no miraculous examples of export growth of manufactures from developing countries, even Hong Kong's performance being barely notice- able, while the developing countries were losing market share in the exports of many primary commodities. The international order did not appear to be satisfying the needs of developing countries. In 1957 in response to these concerns the GATT commissioned a report 4. Little (1982 p. 53), remarks that "it was often the need to justify aid that produced plans"; see also Johnson (1967, p. 4). These plans were often, perhaps usually, overambitious, so that the one part of the economy over which government has some control, public expenditure, was the most likely to expand in line with these overambitious expectations, creating serious public finance problems. 654 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 entitled Trends in International Trade, which was drafted by a distinguished group of economists chaired by Gottfried Haberler.s The report revealed the failure of the exports of developing countries to grow as rapidly as those of the developed countries and argued that an important reason was the obstacles in developed countries to imports of products in which the developing countries had a comparative advantage. General trade liberalization by the developed countries on products of interest to the developing countries was recommended. The report led to the Program of Action advanced by developing countries in 1963, hut no general liberalization of obstacles to exports from developing countries was achieved, the principal (and disappointing) outcome being the GATT'S Part IV on "Trade and Development."6 The first United Nations Conference on Trade and Development (UNCTAD) took place in 1964. Its intellectual father was Raul Prebisch (see Hudec, forth- coming, part I). Prebisch's analysis revealed that import substitution had proved to be insufficient on its own; there had to be more exports as well. But, as Harry Johnson noted: The most important points of [Prebisch's] indictment [of import substitu- tion industrialization] relevant to development through exports are the effects of import substitution in raising the cost of production above world market costs and the effects on competition and efficiency of protected production for a small market. Prebisch blames this situation on the un- willingness of the developed countries to open their markets to industrial producers in less developed countries, instead of on the maintenance of overvalued exchange rates by protectionist import-substitution policies. The solution he recommends is preferential entry, whose high costs would be borne by the consumers of the developed countries, instead of a combi- nation of exchange rate adjustment and import liberalisation by the less developed countries. [1967, pp. 73-74] Johnson also described Prebisch's recommendation as "the internationalisation of protectionism, or the inversion of protection in developed countries in favour of the less developed countries" (1967, p. 28). Why did preferences become the principal positive demand of developing countries at UNCTAD (apart, that is, from commodity agreements)? The answer is not self-evident, for, as Johnson noted, the logic underlying Prebisch's recom- mendation is unclear. Perhaps the failure of the Action Program convinced devel- oping countries that nondiscriminatory liberalization by the developed countries was politically infeasible. What is evident, however, is that the central case for preferences actually presented at UNCTAD derived from the infant-industry argu- 5. Other members of the Panel of Experts were Roberto de Oliveira Campos, director of the Brazilian National Bank; James Meade, professor at Cambridge University; and Jan Tinbergen, professor at the Netherlands Institute for Advanced Economic Studies, Rotterdam. 6. On the attempt to deal with the problems of developing countries, see Dam (1970, chap. 14) and Johnson (1967, especially chap. 1). Wolf 655 ment for protection (Johnson 1967, p. 181). Use of the infant industry argument to justify preferences implies that exports of manufactures from developing countries were deemed to be uncompetitive without a degree of special assis- tance. Nondiscriminatory liberalization by the developed countries could not provide infant industry assistance to the exports of manufactures from develop- ing countries since those exports would then be subject to equal competition from all other suppliers. Much of the pessimism about the capacity of the new industries of developing countries to compete on equal terms derived naturally from one of the most important consequences of inward-oriented trade regimes, namely, that highly protected import-competing industry looks highly uncompetitive, because the domestic prices of protected products are inevitably above (sometimes well above) those of imports. It must have appeared self-evident that such products could be exported only if they enjoyed a preferential advantage over products coming from competitors. Equal treatment of unequals then appeared to be not merely unfair in the abstract, but unworkable. Under that principle, the perfor- mance of developing countries, which virtually everybody agreed to be unsatis- factory, was unlikely to improve. In consequence, it was argued that developing countries needed positive discrimination in their favor if their infant exports of manufactures were to be encouraged. By the early 1960s there was in place an apparatus of mutually reinforcing self-justification. Many developing countries pursued inward-oriented policies, the consequences of which were poor export performance and increasing scar- city of foreign exchange. The poor export performance could be blamed, with considerable plausibility, on the protectionism of developed countries, the reac- tion being insistence on improved market access. Furthermore, it was argued, infant industry assistance required preferential access, and so long as preferences were granted only grudgingly and protection of the most vulnerable industries of developed countries continued in force, it was always possible to blame external circumstances rather than internal policy for the continuing failure. III. GATT MERCANTILISM AND DEVELOPING COUNTRIES There was a very good reason for the developing countries to think in the way they did: the developed countries appeared to as well. The GATT is based on the same protectionist ideas. The French nineteenth-century pamphleteer, Frederic Bastiat, has put the point as follows: But, they say, free trade must be reciprocal. If we lowered the barriers we have erected against the admission of Spanish goods, and if the Spaniards do not lower the barriers they have erected against the admission of ours, we should be victimised. Let us therefore make commercial treaties on the basis of exact reciprocity, let us make concessions in return for concessions, let us make the sacrifice of buying in order to obtain the advantage of selling. 656 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 People who reason in this way, I regret to say, are, whether they realise it or not, protectionists in principle; they are merely a little more inconsistent than the pure protectionists, just as the latter are more inconsistent than the advocates of total and absolute exclusion of all foreign products. [Bastiat 1964, pp. 67-68] The principal idea in GATT negotiations is that the liberalization of imports is the price of improved access for exports. While some negotiators might say, when outside the negotiation, that liberalization is in the interests of the liberal- izing country, such ideas are quickly dropped when the practical reality of trade negotiations arrives. Then the fear of going naked into the conference chamber ensures that there will be a plentiful supply of protection at home with which to bargain.7 All too often one hears the arguments that unilateral liberalization is like unilateral disarmament and that protection must be kept as a bargaining chip for the negotiations. The fact that the GATT is presented as a mutual disarmament treaty for mer- cantilists has great importance for relations between the developed and the developing countries. It has made it effectively impossible for developed coun- tries to insist on equal participation by the developing countries. What is more, it has made such unequal participation attractive to many in the developed countries. The unequal role of developing countries in the GATT, indeed the very insis- tence that equal treatment of unequals is unfair, was a natural corollary of the ideas of the GATT itself. It took time and much rearguard action by the United States, in particular, before the inherent logic of the situation became reality, but by the end of the Kennedy Round the outcome was already quite predictable. These ideas are still very much alive, as can be seen from the ministerial declaration at Punta del Este. The declaration insists that reciprocity shall not be required of developing countries and, in general, that they should not make "concessions inconsistent with [their] development, financial and trade needs." The declaration goes on to remark that "the capacity to make contributions or negotiated concessions . .. would improve with the progressive development of their economies and improvement in their trade situation." Liberalization is presented as a policy that only the relatively rich and successful can adopt. What alternative to this way of incorporating the developing countries into the GATT could there have been? The alternative consistent with the nondiscrimi- nation and liberalization principles of the GATT would have involved an absolute 7. It can be argued that the mercantilism of the GATT is the outcome of a domestic political process in which all parties see their interests "rationally" so that "incorrect" ideas then necessarily imply lying or "false consciousness." This type of reductionism is not very helpful, at least in the absence of a specified theory of bow conflicts of domestic interests generate mercantilist ideologies. In the absence of such a theory, it seems easier to believe that people are influenced by what they say they believe. Mercantilism is still the best description of the behavior of the states involved, even if there are no mercantilists. If it looks like a duck and quacks like a duck, it is a duck (though it may, in fact, be a goose with false conscious- ness). Wolf 657 commitment by the developed countries to nondiscriminatory treatment of de- veloping countries, that is, to the most-favored-nation (MFN) rule, and a willing- ness on their part to negotiate a reduction in protection of particularly sensitive industries of importance to developing countries, such as agriculture, textiles, or clothing. These were the most important immediate export interests of the devel- oping countries during the late 1950s and 1960s, when most of the exceptions regarding the place of developing countries in the trading system were agreed. Strict adherence by the developed countries to MFN combined with liberaliza- tion of the most vulnerable industries was not a plausible alternative. The talk of "market disruption" by low-wage producers that led to the Short Term Arrange- ment Regarding International Trade in Cotton Textiles in 1961 (followed by the Long Term Arrangement in 1962), made it clear that the MFN rule would not be applied to such "sensitive" products (see Dam 1970, chap. 17). Similarly, the birth of the Common Agricultural Policy in the early 1960s, itself following the agricultural waiver granted to the United States in 1955, made it evident that liberalization of agriculture was not to be expected. The importance of these developments in agriculture can be judged from the fact that both the terms of reference of the report of the panel of experts, Trends in International Trade, and the report itself focused almost entirely on primary commodities. The generally negative reaction of the developed countries to the Action Program confirmed their unwillingness to offer much liberalization to developing countries. In addition, the developing countries had, in any case, very little with which to bargain in the standard reciprocal framework. Accord- ingly, the alternative route of full and equal membership, to the extent that it was considered at all (which is unclear), would have been rightly judged to be quite unpromising by the developing countries. For the developed countries, acquiescence in the demands for special treat- ment was less politically costly than the alternative of significant nondiscrimina- tory liberalization on products of interest to developing countries. Those advantages bear elucidation. In the GATT framework the only effective way to bring liberalization is through reciprocal bargaining. In the nonreciprocal relationship between the developed and developing countries, therefore, there was little likelihood of any politically effective demand that the developed countries liberalize imports of the more sensitive products. By conceding the right to protect to the developing countries, the developed countries assured themselves of what turned out to be a persuasive excuse for not granting liberal access to the exports of the developing countries. They could, indeed, complain about the injustice of the protectionism of more ad- vanced developing countries. By concentrating positive assistance to developing countries in preferences, developed countries were able to minimize the potential inconvenience to them- selves, for preferences are not bound under GATT. Not being the result of recip- rocal bargaining, they are regarded as ex gratia favors. Accordingly, developed 658 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 countries have constructed a protectionist system of preferences, removing them whenever the result might be uncomfortable for competing domestic interests.' The special position of developing countries has followed naturally from the economic ideas of both developed and developing countries, at least as they appear in the GATT. The combination has been both a compellingly logical and politically cheap way for the developed countries to accommodate the develop- ing countries within the GATT system. IV. ECONoMIc DEVELOPMENTS It is particularly ironic that a system of ideas was to be enshrined in the international trading order-in Part IV of the GATT, in the waiver for the GSP, and in the general grant of differential and more favorable treatment-just when economic development itself was to demonstrate the fallacies underlying it. Unlike the developed countries, in whose trade policies the reciprocal bargaining of the GATT has tended to bring about conformity, the developing countries have adopted a variety of trade policies and that very variety has given rise to one of the rare, relatively controlled, experiments in economic life. Hong Kong was the first free trader, largely by necessity, followed by Singa- pore, Taiwan, and the Republic of Korea.9 These economies experienced what appeared to be almost miraculously rapid growth, both of exports of manufac- tures and of their economies. The supposedly crippling obstacles to export- oriented manufacturing turned out to be illusory. None of these successes had much, if anything, to do with the special position of developing countries in the GATT. They had much more to do with the GATT'S central achievement, the liberalization of the trade policies of the major devel- oped countries. Preferences did not exist at first, and when they did they were of marginal significance, and either the governments of the more successful devel- oping country exporters did not use the freedom to protect domestic markets or offset the adverse effects by means of export promotion. The result was that unilateral liberalism worked, at least in an environment in which the major developed countries continued, on the whole, to adhere to the generally liberal policies agreed in the framework of the GATT. 8. The idea of preferences was much more acceptable to the European Economic Community than to the United States, for few member countries have been concerned about nondiscrimination, and a system of preferential trade with former colonies and Mediterranean countries was already being developed. There may also have been recognition of the value of preferences as an easy alternative to more liberal market access. The "Brasseur plan" for selective preferences, proposed by the Belgian representative to the GATT in 1963, seems to have been an important step on the way to generalized preferences. (See, on this, Dam 1970, p. 248 and, on the general attitude of the Europeans, Johnson 1967, p. 180.) In practice, the European Economic Community seems to have been concerned to avoid excessive erosion of the value of preferences accorded to associated states. 9. Southern Europe, Greece, Portugal, Spain, and Israel all liberalized to a considerable extent in the 1950s or 1960s, and all enijoyed relatively favorable performance. These are important cases, but their example was probably somewhat less influential than that of the four East Asian economies. Wolf 659 The significance of these developments began to become evident to econo- mists in the late 1960s. By the 1970s a number of major research projects had recorded and explained what was happening."0 While a relatively small number of developing countries took the greatest advantage of the opportunity, many began to put more emphasis on exports and some were influenced to try some degree of unilateral liberalization. As a result of these developments and basic changes in the pattern of comparative advantage, developing countries experi- enced major changes in export structure during the 1960s and 1970s, the most important aspect being the rapid growth and increasing importance of exports of manufactures. l These developments have implications for the appropriateness of the institu- tional and conceptual legacy of the 1950s and 1960s. (1) The implicit assump- tion that developing countries would, in general, have little to gain from reciprocal bargaining has become less compelling, given the height of their own trade barriers and the diversity of their exports of manufactures. (2) The belief, expressed again in the ministerial declaration at Punta del Este, that liberaliza- tion can be afforded only by the more advanced is not merely wrong in theory but has been shown to be wrong in practice. (3) Relatedly, the view that develop- ing countries can compete with developed countries only if they enjoy special favors has been demonstrated to be mistaken. (4) The major issue in the external environment for a large proportion of developing countries has clearly become security of access for exports that have the potential to displace production in developed countries and are therefore likely to create what the governments of developed countries would perceive as injury. In short, there is no obvious reason why the problems of developing countries should be regarded any differ- ently from those of smaller developed countries. V. DEMAND OF DEVELOPED COUNTRIES FOR GRADUATION The alteration in the prevailing view of the link between trade and economic development has led a number of scholars to question the basis on which the developing countries had been incorporated into the GATT. At the same time, with the success of a number of developing countries, especially their success in world markets, attitudes among the governments of developed countries appear to have changed. These new ideas and attitudes provide the framework 10. It is impossible to provide a complete bibliography. The more important publications were Little, Scitovsky, and Scott (1970); Balassa, (1971); Donges (1976); Bhagwati (1978); and Krueger (1978). There is a particularly lucid recent discussion of the evidence on the issue of export promoting trade strategy in Bhagwati (forthcoming). 11. The implications of this development for economic relations between developed and developing countries is the principal theme of Riedel (forthcoming). In the author's judgment, the "engine of growth" vision which lay behind much of the theorizing on trade and development in the 1950s has been entirely overtaken by events. 660 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO.4 within which developing countries will have to take decisions, whether individu- ally or collectively, about their role in the Uruguay Round. Within the GATT'S intellectual framework the demand of developed countries for graduation by the more successful of the developing countries is inevitable. As countries show themselves to be competitive and economically successful, developed countries conclude that they do not warrant preferential access and are no longer entitled to exemption from the obligations of membership. Such countries are seen as free riders on the willingness of others to bear the onerous burden of liberal trade policies (and these are politically, if not economically, burdensome). It is this notion that underlies clause (vi) of the Uruguay declara- tion, itself not a new one, which states that "less-developed contracting parties expect that their capacity to make contributions or negotiated concessions or take other mutually agreed actions . . . would improve with the progressive development of their economies and improvement in their trade situation." The demand for graduation is not radical, since it retains the basic dividing line between developed and developing countries, but it indicates that a number of countries should move from below to above the line. Emphasis on graduation reinforces the belief that the right to protect is a benefit to be given up only by the more successful, and then only in return for a suitable consideration (see Hindley 1986). Four questions arise regarding the demand for graduation. (1) Would any changes in the GATT and its codes be required for there to be graduation? (2) To what extent is agreement actually required to make the demand effective or can graduation be imposed unilaterally? (3) Do developed countries want the devel- oping countries to graduate or is the demand rather an excuse for protection against them? (4) Can the developing countries use the demand for graduation for their own benefit? These questions are considered in turn. Prima facie graduation can take place without any change in GATT rules. There can be a purely ad hoc accommodation. Graduation would require con- troversial changes only if it were to be made automatic, for this would require a definition of the term "developing country." There has never been agreement on such a definition, and it is not easy to believe that there could be agreement now. Part of the demand for graduation can be (and is now being) imposed by developed countries, since they can make unilateral changes in eligibility for GsP treatment, if they wish. Developed countries can also try to be stricter in review- ing the developing country import restrictions that are justified by the need to manage the balance of payments. Nevertheless, within the framework of the GATT, which stipulates nondiscrimination, they can do little about the trade regimes of developing countries without their agreement. A substantial part of what developed countries appear to want is in the hands of developing countries so there is apparently room for negotiation. Developed countries state firmly that graduation is necessary if developed countries are to maintain low barriers to exports from advanced developing countries, let alone to liberalize further. It is asserted that it is difficult to justify Wolf 661 low barriers to imports from successful developing countries that not only re- ceive GSP treatment but also have both high tariffs and highly restrictive non- tariff barriers against imports. It is unclear, however, whether the demand for graduation is made by those inclined to keep trade barriers low or by those who hope that graduation will be refused, so justifying further protection. Even completely liberal economies like Hong Kong suffer from discriminatory protec- tion against their exports. Accordingly, it is a matter of judgment whether the developed countries are offering anything in return for graduation or are merely proposing to graduate countries into a limbo of adverse discrimination. So far as the last question is concerned-whether there is an opportunity here for developing countries-the only way of finding out is to bargain. There is no overwhelming prima facie case for taking the protestations of developed coun- tries seriously. At the same time, the developed countries may be prepared to improve their policies or at least make a serious effort to hold the line against increased protection in return for fuller and more equal participation by the advanced developing countries. With that thought in mind, the discussion turns to the options and opportunities now facing developing countries. VI. DIFFERENTIAL AND MORE FAVORABLE TREATMENT IN A NEW ROUND Some observers conclude from experience with the demand for differential and more favorable treatment, that a far better alternative would be full and equal participation, especially for the newly industrializing countries. The right of developing countries to protect themselves freely is, they would argue, no more than the right to shoot themselves in the foot. The preferences that they have obtained are so surrounded by safeguards as to be of no more than modest value, and they are very unlikely to improve. Would it not be worthwhile to try the alternative tack of equal participation, now that experience has shown both that the economic problems of developing countries are really not so very differ- ent from those of the smaller developed countries and that the attempt to codify a special position has yielded but modest fruit? The argument is popular among scholars and has been made by, among many others, the present author (1984). But, while from this perspective the ideas that underlie the demand for differential and more favorable treatment are wrong, so also are those underlying the GATT. It is not self-evident that the developing countries would benefit from changing (what these writers perceive to be) falla- cies in midstream. I consider below the two benefits one might obtain out of fuller and more equal participation: improved market access abroad and more liberal and efficient policies at home. But first it is necessary to consider what fuller and more equal participation might mean and for whom. Defining Fuller and More Equal Participation Three elements can be envisaged in fuller and more equal participation. First, developing countries might participate, either individually or in groups, both in 662 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 tariff bargaining on a reciprocal basis and in bargaining on nontariff barriers (if there were to be direct reciprocal bargaining of these barriers). In the case of tariffs, any bargaining is likely to start with a generalized tariff-cutting formula, over which the developing countries are likely to have little influence. The participation of developing countries would, therefore, have to be mainly in the negotiation of exceptions. Developing countries would presumably attempt to negotiate away the peaks of their own tariffs for the relatively high protection in developed countries of industries of interest to developing countries, like textiles and clothing, that have been made exceptions to past tariff-cutting formulae. In the case of nontariff barriers, however, there is not yet any plan for a direct negotiation. In the past the GATT process has dealt with nontariff barriers either through bans or codes. Second, developing countries might agree to apply GATT articles and codes more stringently on, for example, quantitative restrictions for balance of pay- ments purposes, subsidies, and emergency protection. They would try, in other words, to abide by such discipline as is left in these various GATT articles and codes and would negotiate new codes, for example, on services and safeguards, with the expectation that they would apply equally to themselves. Third, some developing countries might accept the withdrawal of GSP privi- leges in return for nondiscriminatory and liberal protection in areas of particular interest to themselves. Changes in all these aspects could be made de jure, to apply to all developing countries, but in practice this is unlikely to happen. Many developing country contracting parties will certainly resist so radical a formal change in their place in the GATT. It is therefore best to envisage the options discussed below as applying to those developing countries that are particularly interested in improving their position through bargaining, many of whom are under strong pressure to grad- uate. Improving Market Access The first and most obvious question to be asked by developing countries considering fuller participation is what it would bring from their trading part- ners. Even if the main objective were improved policy at home, it would be difficult to sell these changes in policy if they had little impact abroad. Unfortunately, while the understanding of the economic position of develop- ing countries and their competitiveness in exports of manufactures, in particular, has greatly improved in the past quarter century or so, the capacity and willing- ness of developed countries to liberalize in relation to these exports has not. The chances that the developed countries would liberalize significantly in the inter- ests of developing countries without reciprocity are negligible. But the chances that they would do so with reciprocity are probably little better. Consider the situation of the principal developed countries. The United States is afflicted by strong protectionist pressures associated with a very large current account imbalance. Given the fact that the presidential veto of the highly protec- Wolf 663 tionist Jenkins bill on textiles was nearly overridden, the main objective for the Uruguay Round, so far as American policy is concerned, is probably to avoid increases in protection. The European Economic Community has high unem- ployment and needs to absorb countries of southern Europe that are direct competitors of many developing countries. The Community is also preoccupied with the liberalization of the internal market. Furthermore, the larger the num- ber of cooks-now twelve-the more difficult it will be to prepare a meal con- taining anything other than a minimalist position on trade liberalization. In addition, the EEC may be adversely affected by the adjustment of the American balance of payments. Whether covertly protectionist or not, Japan's willingness to import, regardless of formal liberalization, is widely doubted. It may be noted that Japan has started to negotiate voluntary export restraint arrangements with some of its suppliers, particularly Korea and China in textiles. It has to be doubted whether there is anything that developing countries can offer in standard bargaining that would induce these countries to liberalize significantly, even though taken together the developing countries accounted for 20 percent of world imports in 1985, more than the United States (GATT 1986, table A12). In general their markets are too small or uncertain, individually, while coordination of the offers and demands of many developing countries together would be difficult, because their interests are so diverse. Willingness to enter into reciprocal bargaining with the developed countries may therefore have not much more than an atmospheric effect, though it might produce the important benefit of slowing the deterioration in the policies of developed countries. There may also be benefits for one another from liberaliza- tion of the markets of developing countries. There could well be some important indirect effects of liberalization by devel- oping countries on market access in developed countries, especially if that liber- alization were combined with adherence to codes on subsidies and more limited use of quantitative restrictions for balance of payments purposes. High protec- tion against imports is often offset by generous incentives for exports and, even where there are no such incentives, goods protected by quantitative restrictions on imports are often exported at prices below those in domestic markets. For both reasons exports from protected economies are extremely vulnerable to countervailing or antidumping duties imposed by developed country importers, especially the United States. It is noteworthy that exports from Hong Kong are so far unaffected, while harassment by fair trade law has become a major con- cern of exporters from Korea. For this reason there may be a valuable indirect effect on market access from reductions in import barriers at home (see Finger and Nogues, this volume). What return might be derived from a willingness of newly industrializing countries to acquiesce in graduation from preferences? The answer, unfortu- nately, is "probably none." Preferences are, and are likely to remain, a useful diversionary tactic for the developed countries, but it is unlikely that abandoning this claim would encourage much liberalization by developed countries in sensi- 664 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 tive areas. At the same time, developing countries should recognize that if aid is to be a part of trade, the obvious countries to get it are those that appear to need assistance to compete in world markets, among whom Brazil and Korea cannot be numbered. What would be desirable would be for developing countries to exchange a precise and relatively narrow definition of the appropriate beneficia- ries (along with rules for graduation) for a commitment by developed countries to make preferences available to beneficiaries without restriction. Such changes might bring no benefit to the more advanced newly industrializing countries, but would presumably ensure that the one positive benefit the developing countries have obtained would in fact be of use to those most in need of assistance. Improving Domestic Policy While not often made explicit, a principal reason for participation in GATT negotiations is not just to obtain improved access abroad but to make it politi- cally easier to liberalize at home. Large countries cannot only secure greater market access abroad but can also use it effectively to secure liberalization at home. For them, reciprocity is twice blessed. Small countries, however, can hardly argue convincingly that liberalization at home is the price of improved market access abroad because their influence on policies abroad is transparently negligible. Rightly, Hong Kong and Singapore did not retain barriers to imports in the hope that they could be bargained away for more favorable access. The GATT does not care for the very small. One recalls the warning example of Australia, whose tariffs on manufactures remain high and largely unbound in part because it has waited for reciprocal liberalization of barriers to its agricultural exports. There is the very real danger for small countries that the emphasis on reciprocity will not only misinform their public on the true gains from liberalization but also will fail to bring the ex- pected benefits. For them, in other words, reciprocity may be twice cursed. Reciprocity works best from the point of view of domestic liberalization when the other side is prepared to play the game. But for reasons already discussed, it is questionable whether the developed countries are so prepared. Nevertheless, perhaps a few large developing countries could make something of reciprocal bargaining with developed countries, individually, or, alternatively, through co- ordination of offers. It would be misleading, however, to oversell these possibili- ties. In some large countries where strong nationalistic sentiment exists, the fact that liberalization is a result of a bargain with developed countries may actually make it less politically acceptable. In these circumstances, it is likely that liberali- zation will be attacked as having been conceded under the duress of inimical foreign powers. In such circumstances, reciprocal bargaining would appear to have little, if anything, in its favor.'2 At the same time, international agreements, especially commitments to other 12. The point is owed to Cable. Wolf 665 governments, have sometimes helped small countries to liberalize.13 Experience suggests that governments help one another by making their commitments to liberalize more credible. The pressure of international organizations is often too intermittent, while purely internal commitments to liberalize, on the whole, lack credibility. The cases of the agreements of Greece, Israel, Portugal, and Spain with the EEC all suggest the importance of international commitments in shield- ing the process of trade liberalization from day-to-day political pressures. With a credible international agreement, trade policy becomes a fixed point, to be changed only at great cost, not a weather vane to be moved with the slightest protectionist wind. An open question is whether an enfeebled GATT can serve the purpose of providing such a credible commitment. At least the larger and more significant trading countries, if convinced of the value of liberalization for themselves, might be able to construct an international agreement in the Uruguay Round that supports their liberalization efforts at home. It would be difficult to judge what might be obtained abroad against what would be politically persuasive at home. Countries must, however, avoid the danger of convincing interests at home that liberalization is only acceptable if reciprocated and then finding that nobody wants to reciprocate. It would be worse than futile to postpone liberalization in the hope of a liberalization abroad that then never comes. Large and Small, Richer and Poorer Developing Countries Such limited possibilities as exist of fruitful and active engagement in the GATT process appear to be restricted to the larger and more advanced developing countries, perhaps fifteen or so countries in all. This list would include the newly industrializing countries and such giants as China and India. What of the rest? There is little they can achieve directly, though they could achieve something indirectly. Graduation of others is in the interests of those not being asked to graduate, since it would both exclude the most competitive present beneficiaries from the GSP and open up some of the world's most dynamic markets to imports. It is puzzling, therefore, that most developing countries have not welcomed the idea with enthusiasm. Beyond this, the smaller and poorer countries are likely to be affected most significantly by whatever is agreed in the area of agricultural trade, some benefiting from liberalization and others presumably being harmed. Their capacity to influence the outcome in that crucial area, however, must be very small. VII. EQUAL OR DIFFERENTIAL TREATMENT? At least some developing countries have a choice, and it is not an easy one to make. Differential and more favorable treatment has little economic logic under- 13. This is one of the conclusions the author has drawn from participation on the external panel of the World Bank comparative studies research project on timing and sequencing of a trade liberalization policy (see Wolf 1986). 666 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 lying it and is also unlikely in practice to bring the developing countries major improvements in market access abroad. Unfortunately, the same may well prove true of fuller participation in the GATT itself. Even a willingness to abandon preferences is unlikely to gain much return from the developed countries in their present mood. The most important reason for participating in international negotiations on trade is that in certain circumstances they can help to discipline domestic politi- cal processes. Developing countries have found themselves at least as vulnerable to protectionist lobbying as the developed, so that the relative freedom from external constraints on their policies has borne bitter fruit in the chaotic and costly trade regimes of so many countries. With imagination and cunning a government that wants to liberalize and recognizes the significance of the domestic pressures against it may be able to use the forthcoming GATT negotiations for its own ends. In so doing it must aim for improvements in access to markets abroad that are feasible objectives for the negotiation and are also judged to be worthwhile gains by their citizens at home. The search for such bargains is the most obvious task of the governments of at least the more advanced developing countries. At the same time, a situation worse than failure to obtain improved market access is threatened. There is a chance that, succumbing to the GATT'S own mercantilism, governments of major developed countries will react to a refusal of the more successful developing countries to participate more fully and equally in the GATT process by increasing their level of protection. There is certainly considerable resentment expressed in many developed countries against "free riders"-not all of which are developing countries. The developing countries that are most dependent on the open international trading system may well consider their own interest in securing the vitality of the principles that underlie it. Nevertheless, the real challenge is now to the developed countries, who have lived very comfortably with the outsider status of the developing countries. The developed countries are likely in practice to find an effort by successful develop- ing countries to achieve a truly equal status very unwelcome, preferring their present grounds for complaint to any remedy. If any developing countries decide to challenge this position, they must retain the by-now widespread understanding in developing countries that liberalization is not just a favor to others. If they can promote ordered, liberal relations among countries, without succumbing entirely to the mercantilist ideology in which those values are packaged, they may be able both to contribute to and gain from the next act in the drama of the GATT. It is, on balance, worth supping with the mercantilist devil; but one should sup with a long spoon. Wolf 667 REFERENCES Balassa, Bela. 1971. The Structure of Protection in Developing Countries. Baltimore, Md.: Johns Hopkins University Press. Balassa, Bela, and Constantine Michalopoulos. 1985. Liberalizing World Trade. Devel- opment Policy Issues Series. Washington, D.C.: World Bank. Bastiat, Frederic. 1964. Economic Sophisms. Trans. and ed. Arthur Goddard. New York: Foundation for Economic Education. Bhagwati, Jagdish N. 1978. Anatomy and Consequences of Trade Control Regimes. New York: National Bureau of Economic Research. . Forthcoming. "Export-Promoting Trade Strategy: Issues and Evidence." World Bank Research Observer. Clements, Kenneth W., and Larry A. Sjaastad. 1985. How Protection Taxes Exporters. Thames Essay 39. London: Trade Policy Research Centre. Dam, Kenneth W. 1970. The GATT Law and International Economic Organization. Chicago: University of Chicago Press. Donges, Juergen B. 1976. "A Comparative Study of Industrialization Policies in Fifteen Semi-Industrial Countries,' Weltwirtschaftliches Archiv, Band 112, Heft 4, pp. 626- 59. Eglin, Richard. 1987. "Surveillance of Balance of Payments Measures in the GATT. The World Economy 10, no. 1 (March): 1-26. GATT. 1957. Trends in International Trade. Geneva. =. 1979. "Differential and more Favourable Treatment, Reciprocity and Fuller Par- ticipation of Developing Countries," Decision of 28 November 1979. In Agreement Relating to the Frameworkfor the Conduct of International Trade. Geneva. GATT Secretariat. 1986. International Trade 85-86. Geneva. Hindley, Brian. 1986. "The Graduation Issue:' London: Trade Policy Research Centre. Processed. Hudec, Robert E. Forthcoming. Developing Countries in the GATT Legal System. Thames Essay. London: Gower, for Trade Policy Research Centre. Johnson, Harry G. 1967. Economic Policies Towards Less Developed Countries. Lon- don: Allen and Unwin. Krueger, Anne 0. 1978. Liberalization Attempts and Consequences. New York: Na- tional Bureau of Economic Research. Langhammer, Rolf, and Andre Sapir. Forthcoming. Economic Impact of Generalized Tariff Preferences. Thames Essay. London: Gower, for Trade Policy Research Centre. Little, Ian M. D. 1982. Economic Development: Theory, Policy, and International Rela- tions. New York: Basic. Little, Ian M. D., Tibor Scitovsky, and Maurice F. G. Scott. 1970. Industry and Trade in Some Developing Countries. London: Oxford University Press. Myrdal, Gunnar. 1968. Asian Drama: an Inquiry into the Poverty of Nations. London: Allen Lane. Nogues, Julio, Andrzej Olechowski, and L. Alan Winters. 1986. "The Extent of Non- tariff Barriers to Industrial Countries' Imports." World Bank Economic Review 1, no. 1 (September): 181-99. 668 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 Nurkse, Ragnar. 1953. Problems of Capital Formation in Underdeveloped Countries. Oxford: Blackwell. Prebisch, Raul. 1964. Towards a New Trade Policy for Development. New York: United Nations. Riedel, James. Forthcoming. Myth and Reality of External Constraints on Development. Thames Essay. London: Gower, for Trade Policy Research Centre. Scott, Brian, and others. 1984. Has the Cavalry Arrived? A Report on Trade Liberalisa- tion and Economic Recovery. Special Report 6. London: Trade Policy Research Cen- tre. Wolf, Martin. 1984. "Two-Edged Sword: Demands of Developing Countries and the Trading System." In Jagdish N. Bhagwati and John Gerard Ruggie, eds., Power, Pas- sions, and Purpose: Prospects for North-South Negotiations. Cambridge, Mass.: MIT Press, pp. 201-29. . 1986. "Timing and Sequencing of Trade Liberalisation." Asian Development Re- view 4, no. 2: 1-24. World Bank. 1986. World Development Report 1986. New York: Oxford University Press. THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4: 669-688 Balance of Payments and Related Issues in the Uruguay Round of Trade Negotiations Shailendra J. Anj aria Little attention has been paid to the balance of payments provisions of the General Agreement on Tariffs and Trade (GATT), despite the fact that they directly influence the trade policies of the developing countries. This article suggests that there is a need to reconsider these provisions in the context of the ongoing Uruguay Round of multilat- eral trade negotiations. The article traces the historical evolution of GATT practices on trade restrictions for balance of payments purposes. With the general introduction of more flexible exchange rate arrangements, the original rationale for temporary barriers to safeguard a country's externalfinancial position appears to have lost its force. Recent theoretical and empirical work has demonstrated that neutral or export promoting trade strategies are more effective for development than the import substitution fre- quently advocated by economists in the 1 950s and 1 960s. The current debt problems of developing countries strengthen the argument for a relatively open trade and payments regime to attain balance of payments viability. The article suggests that stronger inter- national discipline over trade restrictions for balance of payments purposes would contribute to and presuppose other necessary improvements in the multilateral trading system which are already on the agenda of the Uruguay Round. Recent years have seen an extensive, renewed discussion of the role of develop- ing countries in the multilateral trading system. New concepts such as "gradua- tion" have been advanced, while arguments for and against "special and differential treatment" of developing countries have been renewed. The subject of trade relations between developed and developing countries is one of the key elements of the Uruguay Round along with nondiscrimination, the treatment of domestic measures such as subsidies, and the future role of a strengthened GATT (see Anjaria 1986). Yet, surprisingly, the balance of payments provisions and practices incorpo- rated in the General Agreement on Tariffs and Trade (GATT), which significantly bear on the trade policies and positions of many developing countries, and hence on the role of the developing countries in the GATT system, have not been The author is in the Exchange and Trade Relations Department at the International Monetary Fund (IMF). The author would like to acknowledge helpful comments received from colleagues within and outside the IMF. The views expressed in this paper should not be construed as reflecting those of the IMF. Copyright © 1987 by the International Bank for Reconstruction and Development / THE WORLD BANK. 669 670 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 discussed in the context of the preparations for the Uruguay Round of trade negotiations. The purpose of this article is to fill this lacuna. A fundamental premise of the GATTis that domestic industries must be pro- tected only by "bound" tariffs applied on a nondiscriminatory basis. Thus tariff rates so negotiated in the GATT may not be raised without compensation to, or negotiation with, affected trading partners. Since quantitative restrictions can nullify the effects of such discipline on tariffs, the GATT includes a general pre- sumption against the use of quantitative restrictions. From the outset, however, the balance of payments situation was viewed as justifying an exception to this general presumption. The developed countries have lowered tariff and nontariff barriers since the GATT was founded and in recent years have avoided all but a strictly temporary use of balance of payments restrictions. By contrast, the developing countries have traditionally maintained high tariff and nontariff barriers, and most con- tinue to do so today. In GATT terminology, developing countries' tariff schedules are often largely "unbound"- that is, most of the tariff rates may be unilaterally raised at any time.1 Quantitative import restrictions are prevalent, although it is often difficult to determine the extent to which these are considered by the countries resorting to them as necessary for balance of payments or infant indus- try reasons. If participation by the developing countries in the GATT negotiations is to be meaningful in terms of more open markets, it will need to involve a greater degree of tariff bindings and/or a lowering of their tariff and nontariff barriers. This will require attention by trade officials to the links between trade policy, external adjustment, and development stratey. If these officials believe that de- velopment prospects or balance of payments adjustment are retarded rather than advanced by more open trade regimes, fuller participation in the world trading system by developing countries through their increased acceptance of reciprocal market-access obligations would be difficult to achieve. The developmental or infant industry provisions of the GATT allow a develop- ing country to provide governmental assistance to establish a particular industry (see Anjaria 1987). Nevertheless, the record shows that in the formal arena very few developing countries have sought to justify their infant industry protection. As of September 1986, only five countries (C6te d'Ivoire, Indonesia, Malaysia, Thailand, and Zimbabwe) had formally invoked the relevant GATT Article XVIII:C.2 By contrast, about twice as many invoked Article XVIII:B, which deals with the balance of payments provisions for developing countries. A possible explanation for this paradox is that the GATT has applied relatively 1. Typically, at least 70-85 percent of imports may be subject to an "unbound" tariff in a developing country. Two notable exceptions among the developing countries in this respect are Chile and Mexico, both of which maintain "bound" tariff schedules at maximum tariff rates of 35 percent and 50 percent, respectively. 2. In the 1950s, releases under Article XVIII:C had been obtained by Cuba, Haiti, India, and Sri Lanka, but few other countries resorted to this provision until rather recently (see Jackson 1969, p. 655). Anjaria 671 restrictive notification and compensation requirements for invoking Article XVII][:C, which may have induced some developing countries to leave certain restrictions unapproved. But a more important reason may be that some devel- oping countries find it relatively easy to obtain GATT "cover" for infant industry protection under the guise of "balance of payments" reasons which are invoked more frequently. Thus to deal effectively with infant industry protection in the new round presupposes that clearer distinctions will be drawn between balance of payments and infant industry protection. Given the prevalence of developing countries' trade restrictions for balance of payments reasons, the treatment of this issue at the Uruguay Round will be critical in shaping the future role of developing countries in the GATT system. The article analyzes this issue as follows. The historical background to and the main features of the GATT provisions and practice pertaining to balance of payments-induced trade restrictions are described in sections I and II. The main rationale for reconsidering the GATT's approach to balance of payments restric- tions is outlined in section III. Section IV suggests some issues that are likely to arise in considering a possible reform of the balance of payments provisions, and section V provides concluding observations. I. BACKGROUND TO THE BALANCE OF PAYMENTS PROVISIONS OF THE GATT Under the current provisions, subject to established consultation and review procedures by the Contracting Parties, the General Agreement permits member countries to protect their external financial situation by imposing temporary trade restrictions in the face of a balance of payments deterioration.3 To under- stand the development of this approach, it is useful to review the historical development of GATT rules and practice and the attitude of the framers of the General Agreement toward trade restrictions. From the outset it was generally accepted that quantitative trade restrictions to safeguard the international reserves and balance of payments position of a country would need to be built into the framework of the new international trade rules. A basic principle of the Havana charter of the International Trade Organization (ITO) carried over into Article XII of the GATT allows countries to use, but not abuse, import restrictions during a balance of payments crisis. Further, given the dollar shortage faced by European countries in the immediate postwar period, the GATT Articles included a provision (Article XIV) allowing discrimination in the application of trade restrictions for balance of payments purposes. In 1955, the provisions of GATT Article XII were reviewed, but this review did not result in acceptance of the proposal to introduce fixed time limits 3. The form "contracting parties" refers to GATT members acting individually. "Contracting Parties" is used in this article in place of the official GATT usage, "CONTRACTING PARTIES," to refer to actions by signatory countries as a group. The following paragraphs draw heavily upon Jackson (1969), especially pp. 673-716. 672 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 after which trade restrictions for balance of payments purposes would be ex- pected to lapse. Although several developing countries participated actively in the negotiations on the Havana charter in the late 1940s, no differentiation in the balance of payments provisions governing developing and developed countries was intro- duced until 1955. GATT Articles XII and XIV were then amended, and GATT Article XVIII dealing with governmental assistance to economic development was revised. As part of the changes, an explicit balance of payments provision relating to developing countries was introduced in Article XVIII. With the 1955 amendments to the GATT Articles, international surveillance over trade restric- tions imposed by developing countries for balance of payments reasons was loosened, and developing countries were thereafter required to consult in the GATT only once every two years rather than once a year as was expected of developed countries that invoked the provisions of Article XII. By the mid-1960s, industrial countries negotiating under the auspices of the Organisation for Economic Co-operation and Development (OECD) and the GATT had virtually eliminated reliance on trade restrictions for balance of pay- ments purposes.4 The remaining restrictions by industrial countries were no longer justified as balance of payments-induced, but either fell within one of the other GATT exceptions to the general ban on quantitative restrictions-such as for agricultural restrictions-or were submerged into a broader category of "residual restrictions,' which proved quite difficult to address despite repeated efforts. Until the mid-1970s, industrial countries introduced trade measures such as import surcharges and advance import deposit requirements for balance of payments purposes, but these were applied for temporary periods. Moreover, international surveillance over such measures was exercised relatively expedi- tiously under GATT Article XII or under other provisions and practices in the OECD designed to discourage trade and payments restrictions. In the GATT, several consultations with industrial countries applying import surcharges or advance import deposit requirements took place in special work- ing parties set up for this purpose. In part, the working parties were considered necessary since Article XII technically permitted only quantitative restrictions for balance of payments reasons, and not other measures. With the adoption of the 1979 Declaration on Trade Measures, all GATT balance of payments consul- tations were formally integrated under the responsibility of the Committee on Balance of Payments Restrictions. By contrast, the position of the developing countries evolved in quite the opposite direction. In addition to the concept of infant industry protection incor- porated in Article XVIII:C, the notion of special exemptions from rules became enshrined from the mid-1960s in Part IV of the GATT, which entered into force in June 1966. Entitled "Trade and Development," Part IV obliges developed 4. Between 1958 and 1964, the number of industrial countries invoking GATT Article XII fell from 14 to 3 (see Eglin 1987). Anjaria 673 contracting parties to extend special and differential treatment to developing countries on a "best endeavors" basis, while imposing no equivalent obligation on developing member countries. The United Nations Conference on Trade and Development (UNCTAD), with its broad mandate to discuss and advance proposals for improving the develop- ing countries' trade and development prospects, was established in March 1964, and it pressed strongly for generalized tariff preferences. In the IMF this develop- ment was reflected in the compensatory financing facility, first introduced in 1963, and then extended and liberalized in 1966. In December 1975, following extensive debate, the facility was again substantially liberalized. More broadly, the commodity price shocks of the early 1970s and the disappointing perfor- mance of aid donors led to demands by developing countries for a new interna- tional economic order which included, inter alia, suggestions for a link between allocations of Special Drawing Rights and development finance. Thus, from the mid-1960s until about the mid-1970s, trade restrictions by developing countries became part and parcel of the broader questions regarding the responsibilities of the developed countries for the trade, finance, and development requirements of developing countries as a group. In the environment then prevailing, questions about the usefulness of quanti- tative restrictions to deal with balance of payments problems were infrequently addressed at the international level. Indeed, when they were, the focus was not always on applying common understandings on which to base trade liberaliza- tion by all countries but rather on reinforcing in international rules the special treatment of developing countries.5 An important contribution of the 1973-79 Tokyo Round negotiations to the subject was the adoption by the Contracting Parties in 1979 of the Declaration on Trade Measures Taken for Balance of Payments Purposes. As described be- low, the Declaration broadened the scope of Articles XII and XVIII to cover trade measures other than quantitative restrictions, such as import surcharges and advance import deposits, and established a somewhat weak presumption- expressed only in its preamble-that "developed contracting parties should avoid the imposition of restrictive trade measures for balance of payments purposes to the maximum extent possible." In addition, the conditions for the application and surveillance of such trade measures were spelled out more clearly. However, the Declaration left unchanged the basic difference in the degree of international surveillance over such trade measures between developed and developing coun- 5. It is interesting to note, in this connection, that the Outline of Reform prepared by the IMF'S Committee of Twenty in June 1974 included explicit acknowledgment of the special position of develop- ing countries and of the need to promote their interests. In the area of trade the outline noted: "There will be a strong presumption against the use of controls on current account transactions or payments for balance of payments purposes.' It also provided: "Wherever possible developing countries will be ex- empted from controls imposed by other countries, particularly from import controls and controls over outward long-term investment. The special circumstances of developing countries will be taken into account by the Fund in assessing controls which these countries feel it necessary to apply." 674 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 tries. Indeed, the economic uncertainties of the 1970s appear to have reinforced the notion in the GATT that the balance of payments difficulties of developing countries were of a recurrent nature and that resort to trade restrictions were an acceptable means of dealing with their external payments problems. II. PRESENT BALANCE OF PAYMENTS PROVISIONS AND PROCEDURES Developed countries invoking Article XII consult in the GATT Committee on Balance of Payments Restrictions every year; developing countries invoking Arti- cles XVIII:B consult in the same committee once every two years. The first consultation on invocation is normally a "full consultation" to examine the balance of payments justification for the trade restrictions in question and the nature and modalities of the trade restrictions themselves. A statement that represents an IMF view on the balance of payments situation and prospects and on the policies being adopted by a consulting contracting party to deal with external and internal imbalances forms a main input for the full consultation, together with background documentation provided by the consulting country, the GATT secretariat, and the IMF. Subsequent consultations with developing countries are normally held under "simplified procedures." These consultations involve no discussion of the trade measures or the balance of payments justification for their maintenance and do not include an IMF statement on the situation. Rather, their objective is to pro- vide members of the Committee on Balance of Payments Restrictions with infor- mation on the balance of payments situation and measures taken by the consulting country, in order to assess whether a further full consultation is necessary. Such a decision depends on the time elapsed since the previous full consultation, the steps the consulting country has taken in response to previous conclusions by the Committee, changes in the overall level or nature of trade measures taken for balance of payments purposes, and on the balance of pay- ments situation and prospects. In most respects, consultations under the two main balance of payments provisions-Articles XII and XVIII:B-are quite similar. (For a more detailed discussion of them, see Anjaria 1987). The main differences between the two relate to, first, the frequency of the consultations and, second, the nature of the discussion and the conclusions reached by the Committee-with the Article XII consultation giving greater emphasis to early phasing out of the balance of payments restrictions. Table 1 lists the twenty-four contracting parties that have consulted with the GATT under Articles XII or XVIII:B on one or more occasions since 1974. The vast majority of consultations in the Committee on Balance of Payments Restric- tions in this period have been with developing countries under Article XVIII:B. Although many developing countries have expressed concerns about the burden imposed on them by the consultation requirements, table 1 suggests that in fact, full consultations under Article XVIII:B have been relatively infrequent for most Table 1. Consultations under Articles XII and XVIII:B of the GATT Committee on Balance of Payments Restrictions Year Country 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 Article Argentina X X Y XVIII:B Bangladesh X X X X X X X XVIII:B Brazil Y XY X Y Y X XVIII:B Chile X X X XVIII:B Colombia Y XVIII:B Egypt, A.R. X X X X X X XVIII:B Finland Y Y Y Y Xii Ghana X X X X X Y X XVIII:B Greece Y X X Y Y Y XVIII:B/XII Hungary Y Y XII India X XY X X X X XVIII:B 0>1 Indonesia X X X XVIII:B Israel YY Y Y Y Y Y Y Not stated Italy Y XII Korea, Rep. X Y X Y X X Y X XVIII:B Nigeria Y X XVIII:B Pakistan X X Y X X X XVIII:B Peru X Y X X X X X XVIII:B Philippines Y X X XVIII:B Portugal Y Y Y Y Y Y Y Y Y Y Notstated Sri Lanka X X X X X X XVIII:B Tunisia X X X Y X X X XVIII:B Turkey Y X Y X X X XVIII:B Yugoslavia Y Y X XY X X XVIII:B X = Simplified consultation. Y = Full consultation. Source: GATT secretariat. 676 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 of the countries which have claimed balance of payments justification during the whole period. For example, over the thirteen-year period covered, full consulta- tions under Article XVIII:B were held most frequently with Brazil, but only on a total of four occasions. With Korea and Yugoslavia, full consultations were held three times each, and with a number of other countries they were held only once. Colombia and the Philippines, which have acceded to GATT since the Tokyo Round, have held one and two full consultations, respectively, in the period since their accession. In the case of four countries, no full consultation has been held during the entire thirteen-year period.6 No comprehensive and definitive information is available for determining whether or not all GATT member countries that resort to trade measures for essentially balance of payments purposes invoke Article XVIII:B to justify their restrictions in the GATT. Although some informal steps taken in the Committee on Balance of Payments Restrictions are designed to encourage noninvoking countries to justify apparent balance of payments restrictions under the GATT, by and large the initiative to invoke Article XVIII:B is the responsibility of each GATT member.7 The possibility that several countries applying restrictions for balance of pay- ments reasons may remain outside GATT surveillance is suggested by table 2.8 At the end of 1985, fifteen developing countries did not invoke GATT balance of payments provisions for the import surcharges or advance import deposit re- quirements applied by them.9 Similar information is not readily available on quantitative restrictions maintained or intensified at the end of 1985 for balance of payments reasons. Even so, there is a strong case to be made for more careful scrutiny by the GATT of trade restrictions for balance of payments purposes that at present may remain outside the surveillance of the GATT Committee on Bal- ance of Payments Restrictions.10 As the maintenance of trade restrictions for balance of payments purposes may be associated with overvalued exchange rates, the IMF has a particular interest in encouraging the identification of such restrictions and their liberalization."1 Closer IMF-GATT cooperation might there- fore be especially fruitful in this area. 6. At present, full consultations are scheduled for 1987 with Egypt, India, Israel, Korea, and Peru. 7. However, the Declaration on Trade Measures Taken for Balance of Payments Purposes also provides for the reverse notifications. 8. Some of these actions may be exchange measures subject to prior approval of the IMF under its Articles. The IMF's interest in promoting an open trade and payments system, and its activities in relation to international trade, have been described most recently by Gold (1986). For an earlier account of the role of the IMF in trade policy issues, see Anjaria and others (1985), pp. 83-86. 9. The fifteen countries that are not included as consulting countries in table 1 are: Belize, Burkina Faso, Burundi, Cyprus, Dominican Republic, Gambia, Haiti, Jamaica, Kenya, Mauritius, Nicaragua, Rwanda, Sierra Leone, Suriname, and Uruguay. 10. In accordance with the 1982 Ministerial Declaration, a GATT Group on Quantitative Restrictions and Other Nontariff Measures initiated an examination of the legal basis of quantitative restrictions. With the launching of the Uruguay Round, this group's work was taken up by the negotiating commit- tees. 11. The objectives of IMF surveillance over the exchange rate policies of member countries are laid Anjaria 677 Available information regarding the number of items covered in countries that invoke Article XVIII:B is summarized in table 3. Although trade restrictions for balance of payments reasons must by definition be across the board, in most developing countries the proportion of the commodity groups covered by tariffs of which GATT is notified has been relatively small, ranging from only a handful of product categories in some cases to more than 50 percent of the number of Customs Cooperation Council Nomenclature (CCCN) categories in five cases.12 This raises the question of whether specific criteria should be developed to distinguish genuine balance of payments restrictions from other types of trade restrictions, such as those for protective or security reasons, which are more likely to be sector-specific. Although historically the GATT has been concerned about the incidental protective effects of balance of payments restrictions, no attempt has been made to establish a sharp distinction between across-the-board balance of payments restrictions and sector-specific restrictions motivated by other considerations.13 Under Article XVIII:B:10, GATT members have the ex- plicit right to vary the restrictiveness of their balance of payments measures depending upon the "essentiality" for economic development of the products in question. This provision, while not a basis for explicit sectoral differentiation in recent years, tends to cloud the distinction between infant industry and balance of payments protection in developing countries.14 At the conclusion of each full consultation, the Committee on Balance of down in Article IV of the Second Amendment of the IMF's Articles. Here it is noteworthy that one of the principles for JMF surveillance over the exchange rate policies of its members is that "the introduction, substantial intensification, or prolonged maintenance, for balance of payments purposes, of restrictions on, or incentives for, current transactions or payments . .. may give rise to the need for discussion with a member (see International Monetary Fund 1986). 12. Countertrade transactions, which are of growing importance in world trade and partly motivated by balance of payments considerations in some developing countries, are often concentrated in certain product categories. However, these have not been categorized as restrictions for balance of payments reasons in notifications to the GATT. See United States International Trade Commission (1985) for a description of countertrade practices. Huh (1983) and Gold (1986) outline IMp concerns about counter- trade, which is similar in some respects to bilateral payments agreements. 13. In 1955, the GATT approved the so-called hard-core waiver, allowing countries a transitional period of exemption from Article Xl, which establishes the basic presumption against quantitative restrictions, in "exceptional" circumstances (see Jackson 1969, p. 709). 14. In applying policies on the use of its resources, the IMF is required to make a judgment on what constitutes a trade restriction for balance of payments purposes. The IMF's approval vis-a-vis a member's representation has been described as follows: A member may declare its intention in imposing a measure is, or is not, to manage its balance of payments. The member's representation is given the benefit of any reasonable doubt, but the Fund reserves the right to pass upon the representation on the basis of the facts. These facts include the rationale offered by the member for its measures, the effect of the measures on the balance of payments and on exchange rates of the currency of the member in relation to the currencies of other members, the member's domestic and external conditions or policies that may explain its choice of the measures, expectations regarding the duration of the measures, and the prevailing practice among members in general with respect to the use of the measures for the purpose of managing the balance of payments. See Gold (1986), p. 217. Table 2. Use of Import Surcharges or Advance Import Deposits by Common GATT-IMF Members-Developing Countries Year of Use Country 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 Total Argentina X X X X X X X X X X X X 12 Belize -…… … … … ……-…-…----- ------ X X X X 4 Brazil X X X X X X 0 0 0 0 0 0 6 Burkina Faso 0 0 X X X X X X X X X X 10 Burundi X X X X X X X X X X X X 12 Chile X X X X X X X 0 0 0 0 0 7 Colombia X X X X X X X X X X X X 12 Cyprus 0 0 0 0 X X X X X X X X 8 Dominican Rep. X X X X X X X X X X X X 12 Egypt, A.R. X X X X X X X X X X X X 12 Gambia 0 0 0 X X X X X X X X X 9 Ghana 0 X X X X X X X X X X X 11 Greece X X X X X X X X X X X X 12 Haiti X X X X X X X X X X X X 12 Hungary -0-------------------------- O X X 0 2 India X X X X X X X X X X X X 12 Indonesia X X X X X X X X 0 0 0 0 8 Israel X X X X 0 0 X 0 0 X X X 8 Jamaica 0 0 0 0 X X X X X X X X 8 Kenya 0 0 0 0 0 X X X X X X X 7 Korea, Rep. X X X X X X X X X X X 0 11 Malaysia X X X X X X X X X 0 0 0 9 Malawi X X X X X X X X X X X X 12 Mauritius 0 X X X X X X X X X X X 1l Nicaragua X X X X X X X X X X X X 12 Nigeria X 0 0 0 0 X X 0 0 X X 0 5 Pakistan X X X X X X X X X X X X 12 Peru X X X X X X X X X X X X 12 Philippines X X X X X X X X X X X X 12 Portugal 0 0 X X X X X X X X X X 10 Rwanda X X 0 0 0 0 0 0 0 0 X X 4 Sierra Leone 0 0 0 X X X X X X X X X 9 South Africa 0 0 0 X X X X 0 0 X 0 0 5 Suriname - X X X X X X X 7 Thailand 0 0 0 X X X X X X X X X 9 Tunisia, 0 0 0 0 0 X X X X 0 0 0 4 Turkey X X X X X X X X X X X X 12 Uruguay X X X X X X X X X X X X 12 Yugoslavia X X X X X X X X X X X X 12 Zaire X X X X X X X 0 0 0 0 0 7 Zimbabwe -…- -- --- - X X X X X 5 Note: The common GATT-IMF members considered are the common members on June 1, 1986. The definitions of "developing countries" is in accordance with the convention used in IMF, International Financial Statistics. An "X" indicates that a country applied an import surcharge and/ or an advance import deposit requirement either through the exchange or trade system usually at the end of the calendar year for which the IMF's Annual Report on Exchange Arrangements and Exchange Restrictions (hereafter AREAER) was prepared. An "O" indicates that no such restriction was applied. A series of dashes, `-", indicates the period during which the country was not a member of the IMF. In a few cases, import surcharges are defined here to include additional import taxes not treated as "surcharges" in the AREAER. a. Tunisia acceded provisionally to GATT membership in 1985. Sources: Based on IMF, AREAER, Washington, D.C., various issues. 680 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 Table 3. Percentage of Commodity Groups Covered by Quantitative Restrictions for Balance of Payments Reasons under GATT Article XVIII:B, Selected Countries Country Year Percentage of groups affected Argentina 1986 72 Bangladesh 1986 60 Brazil 198S 32 Colombia 1986 86 Egypt, A.R. 1985 5 Ghana 1985 60 India 1985 48 Korea, Rep. 1986 20 Nigeria 1983 18 Pakistan 1985 43 Peru 1986 50 Philippines 1985 5 Sri Lanka 1986 14 Thailand 1983 0.2 Tunisia 1985 73 Turkey 1986 22 Yugoslavia 1984 19 Zimbabwe 1984 0.1 Note: Each quantitative restriction refers to an affected commodity, where a commodity is defined by one of 1,010 four-digit Customs Cooperation Council Nomenclature (CCCN) categories. The quantitative restriction may affect all or only part of the four-digit group in question. The information is assembled from notifications to the GATT by countries applying these restrictions. It excludes quantitative restrictions notified as being applied under GATT Article XVIII without further specification of whether Article XVIII:B or Article XVIII:C is invoked. It also excludes quantitative restrictions notified specifically under Article XVIII:C. Source: GATT document NTM/W/17 (9/1/86). Payments Restrictions adopts a report which is subsequently endorsed, usually without further discussion by the GATT Council. Among the points generally included in the Committee's conclusions are references to the nature of the balance of payments problem and the extent to which the external disequilib- rium is seen to be related to trade imbalances; a brief expression of the Commit- tee's general views on the other external and domestic financial policies being pursued to correct the balance of payments problem; where relevant, the appli- cation of multiple trade restrictions where a more simplified trade regime might be appropriate; and the incorporation of any announced plans by the consulting country to phase out the restrictions or, in their absence, the Committee's views on the desirability of establishing such a timetable. Following the introduction of more flexible exchange rates worldwide, the iMF finding and the GATT conclu- sions have generally sought to welcome or encourage a policy stance that is conducive to achieving medium-term viability in the external payments position. Thus in recent years unduly specific conclusions about the "need" for trade restrictions for avoiding a decline in the country's holdings of monetary reserves have been avoided. Anjaria 681 III. NEED FOR A NEW APPROACH Major changes in the world economy suggest a need to reexamine the rules governing trade restrictions for balance of payments purposes. First, there is a greater willingness among countries to use macroeconomic policy instruments to deal with external difficulties. Second, greater mobility of capital in the past decade has changed the nature of balance of payments problems facing many developing countries. Third, overlaying the discussion of the trade policies of developing countries is a considerable improvement in the understanding of the links between growth, development, and trade strategy which gives urgency to a reexamination of all arguments in favor of trade barriers. In certain important respects, the IMF'S surveillance over exchange rate policies has been adapted to reflect these changes in the 1970s, while the adaptation of GATT rules and practice has lagged behind. Modernization of the GATT balance of payments provisions would thus be a logical part of the overall strengthening of the GATT system that is now being negotiated. At the same time, it would contribute positively to the achievement of the IMF'S objectives. It is evident that the balance of payments provisions of the GATT permitting the use of quantitative import restrictions were formulated on the assumption of fixed exchange rates-or at least they presumed that exchange rate adjustments, in the absence of "fundamental disequilibrium,' were undesirable. Following the abandonment of the par value system from the early 1970s, there is a greater awareness of the role of exchange rate policy in balance of payments adjustment. Even if they do not float freely in many developing countries, exchange rates are often actively used for balance of payments adjustment. Hence, a fundamental assumption on which the GATT provisions and practice are based is no longer applicable. Moreover, at the time the General Agreement was formulated, quan- titative import restrictions were seen as an appropriate or necessary short-term device to switch the pattern of demand toward home-produced goods while policies to reduce real expenditures to bring about an improvement in the cur- rent account balance took effect. 1 Today, however, the efficacy of exchange rate adjustments to bring about the necessary reduction in real expenditures is recognized, and even among alterna- tive demand-switching policies, quantitative restrictions are regarded as an infe- rior policy instrument compared with price-related measures such as import 15. Even more than a decade ago, however, the "equivalence" of quantitative restrictions (QR) and exchange restrictions for balance of payments reasons was recognized. Thus, a rather radical view advanced prior to the Tokyo Round negotiations by Eric Wyndham-White, former Director General of the GATT, questioned the rationale for dealing with balance of payments restrictions in the GATT: The time has now come to recognize that the provisions relating to the use of QR for balance of payments reasons are inappropriate in what purports to be essentially a trade agreement. Commercial measures adopted exclusively for balance of payments reasons are essentially exchange restrictions applied by a particular technique and should be dealt with as such. See Wyndham-White (1975). 682 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 surcharges or advance import deposit requirements. There has developed a cer- tain difference between most governments and the IMF on the one hand and the GATT on the other in their respective attitudes toward the appropriate choice of policy instruments for balance of payments adjustment. Exchange rate adjust- ment supported by appropriate macroeconomic policies are generally considered as the "first-best" policy package to deal with a lasting balance of payments deterioration; yet the GATT'S mandate is limited mainly to considering the legiti- macy of quantitative import restrictions. One outcome of the disparate evolu- tion of GATT and IMF disciplines and practices-paradoxical, given the commonality of purposes of the two bodies-is that while, as IMF members, countries are called upon to explain why in particular crisis situations they prefer quantitative restrictions or import surcharges to exchange rate adjustment, as GATT members the same countries have come to expect GATT to sustain the use of such restrictions for balance of payments purposes. Perhaps the most notable change in balances of payments since the early 1970s is the profoundly greater importance of capital movements in determining the overall external sector position of any country. In the major industrial coun- tries, the rapid growth of capital markets and the responsiveness of capital flows across international borders has introduced a new element in the balance of payments adustment process. Even under a system of fixed exchange rates, a trade restriction may result in a real effective appreciation of the exchange rate through its inflationary impact. But in a world of floating exchange rates and capital mobility, attempts to improve the balance of payments position through the introduction of trade restrictions are at best unworkable, if not actually counterproductive, as they may induce a compensating appreciation of the cur- rency of the restricting country. In some cases the current account of the balance of payments may be said to be "driven" by the capital account, rather than vice versa. Restrictions imposed by a major trading nation would almost certainly prove counterproductive by provoking retaliatory restrictions by trading part- ners. The numerous governmental statements and declarations issued in the past decade testify to the awareness of policymakers in industrial countries of the dangers of resorting to trade restrictions for correcting payments imbalances. For example, in May 1974 the member countries of the OECD adopted a declara- tion expressing their determination to avoid resort to import restrictions and similar measures, a declaration that has been successively renewed since then. For developing countries, greater worldwide capital mobility has shifted the focus from balance of payments deficits per se to concerns about medium-term external viability. Many developing countries which engaged in heavy external borrowing in the 1970s found, in the worsened external economic environment they faced subsequently, that their debt burden became unsustainable. Major debt restructuring and exceptional financing were coupled with stringent adjust- ment policies to bring the external position closer to medium-term viability. But in no instance was it thought possible or desirable that the objective of medium- term viability could be achieved by increased reliance on trade and payments Anjaria 683 restrictions. Even where, in the early 1980s, adjustment packages entailed a high degree of initial import compression, reliance was placed mainly on demand management policies, rather than on intensified and prolonged trade and pay- ments restrictions. Over the medium term it was considered necessary to estab- lish the basis for permitting the growth of imports at a sustainable rate by combining exchange rate and demand management policies with appropriate, but often slower acting, supply-side policies. Thus, the "freedom" available to developing countries under the GATT to impose trade restrictions for balance of payments purposes proved to be somewhat irrelevant, as the countries con- cerned often found it feasible to go beyond reliance on trade restrictions to an active search for policy packages that reduced this reliance. Another major consideration relates to the links between trade strategy and development. A growing body of evidence demonstrates the positive experience of an outward-oriented trade strategy for developing countries' growth and development prospects. Although work on outward-oriented or export-promot- ing strategy for promoting economic development started as far back as the 1960s, it was not until the late 1970s that sufficiently extensive analysis was undertaken establishing the superiority of more open (or, strictly, more neutral) trade policy over the import-substitution policy often recommended by econo- mists in the 1950s and 1960s."6 A number of detailed studies have found that the export-promoting strategy enabled developing countries with an already estab- lished industrial base to take advantage of the postwar economic boom of the 1950s and 1960s, an opportunity that was missed by countries that pursued an import-substitution strategy. Furthermore, the former strategy was found to be superior in promoting overall economic growth as well as more rapid industrial- ization and promotion of employment.17 Given these findings, it has been in- creasingly evident that the cost of using balance of payments difficulties as a justification for maintaining trade barriers is not negligible for developing coun- tries. The view that trade barriers are an inappropriate way to address payments difficulties has thus been further reinforced. IV. SOME ISSUES FOR THE FUTURE Notwithstanding the clear theoretical as well as practical grounds for reexam- ining the balance of payments provisions of the GATT, it is premature to predict whether this reexamination will be conducted or how it will be concluded. If 16. The "export-promoting" strategy is something of a misnomer, since the literature defines such a policy as one that maintains an effective exchange rate for exports at a level not significantly different from that of imports. Thus the export-promoting strategy merely calls for eliminating the bias against exports, while the import-substitution strategy is one based on a higher effective exchange rate for imports than for exports. Of course, the maintenance of an overvalued exchange rate is not consistent with pursuit of the export-promoting strategy (see Bhagwati, forthcoming). 17. Two major studies in this area are Bhagwati (1978) and Krueger (1978). See also Bhagwati (forthcoming). 684 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 there were to develop a broad consensus for reviewing the GATT provisions and practice, the main issues that may prove difficult in the context of the Uruguay Round are the degree of discipline that should be applied on balance of pay- ments-induced restrictions and the relation of possible reforms in this area to other aspects of the GATT system and negotiations. The discussions will sooner or later involve the specific features or characteris- tics which would regulate countries' use of trade restrictions for balance of payments purposes. Among the elements that may be included are: whether temporary restrictions would be authorized for predetermined time limits; whether stricter discipline would be imposed on new restrictions than on preex- isting ones; how the generalized nature of the restrictions would be defined (for example, a balance of payments restriction must cover at least x percent of a country's imports); and whether the possible inclusion of services and trade- related investment measures in the negotiations would require a corresponding adjustment of the balance of payments provisions. How would strengthened international surveillance over trade restrictions for balance of payments purposes be exercised? From the standpoint of a country's trading partners in the GATT, it may be desirable to decide whether the country's program for phasing out trade restrictions and implementing adjustment policies is on track or off track. In practice, however, such judgments may be difficult to reach in the GATT context because they bear on macroeconomic developments and policies that are normally covered in the IMF's consultations with member countries. In the GATT context, several procedural questions would require resolution. How frequently would the GATT Committee on Balance of Payments Restric- tions make its determinations, and on the basis of what information, given that balance of payments prospects are often volatile? If a country's adjustment effort were considered "inadequate" in the GATT context, what would be the practical consequences? Could such a declaration trigger retaliatory actions by trading partners? If so, would retaliation not impede, rather than assist, in the attain- ment of the country's adjustment objectives? If the rules were too strictly formu- lated, would there be a greater risk of nonreporting of balance of payments restrictions? How would the IMF be expected to cooperate in a strengthened surveillance process, given its traditionally key role in the GATT balance of payments consultations? Two substantive issues appear critical: the timing of trade liberalization and the trade-finance link. The first issue is the optimal speed of trade liberalization. Actual experience in the developing countries suggests that no hard and fast rules can be applied a priori to the question of whether trade liberalization should be massive and quickly completed, or gradual. On the one hand, an argument can be made that liberalization of restrictions applied for balance of payments reasons requires alternative macroeconomic policy adjustment to be effective. On the other hand, there is some recent evidence that in the majority of liberalization episodes in developing countries in the postwar period, a more Anjaria 685 rapid liberalization effort than was actually implemented might have proved both feasible and more effective (Papageorgiou, Michaely, and Choksi 1986). In any event, if countries can be expected to have different timepaths for trade liberalization, how should these fit in the overall timetable of the ongoing trade negotiations? The second set of issues falls under the rubric of the "links between trade and finance,' and concerns whether some special or additional financial or trade measures are needed to encourage further trade liberalization by developing countries. The financial aspect of these links concerns the availability of external financing; the trade aspect concerns the openness of markets. Concerns have been voiced, particularly given the economic difficulties cur- rently facing developing countries, that additional external financing must be available to secure the "permanence" of trade liberalization measures by devel- oping countries. As mentioned earlier, because the "binding" of tariffs assumes a special importance in the context of GATT negotiations, it appears quite reason- able, prima facie, to argue that additional external resources should be available to finance any unexpected increase in imports resulting from a developing coun- try's binding commitments under the GATT. For example, if concessional long- term capital flows are directed toward countries which pursue appropriate macroeconomic policies combined with an outward-oriented trade policy, such additional financing may play a very helpful role. The World Bank has recently increased its emphasis on policy-based, quick-disbursing loans for structural and sectoral adjustment, in which trade liberalization often plays a very important part. Insofar as such suggestions seek to address a possible additional automatic need for balance of payments financing, however, a strong note of caution is in order. In the broadest sense, such financing by the IMF is already tailored to the objective of promoting a liberal trade and payments system. The policies govern- ing the use of IMF resources are established after a careful review of a range of factors, including the global need for balance of payments financing and the quality of programs that the IMF should seek to support. It is questionable whether it would be desirable to establish a specific balance of payments financ- ing feature linked directly to increased imports occasioned by the liberalization effort. Such an open-ended scheme would risk ignoring the essential role of macroeconomic and exchange rate policies in achieving a sustainable medium- term balance of payments outcome. Such a proposal does not appear to be desirable or feasible. Addressing the trade aspects of the trade-finance link, it has been argued that if economic growth is to be restored in the context of the balance of payments and debt problems of developing countries, specific and concrete measures must be considered by the industrial countries to improve access for developing coun- tries undertaking trade liberalization with IMF or World Bank financial support. Some policymakers, however, have questioned the justification for limiting any liberalization effort to a relatively narrowly defined group of countries such as 686 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 "indebted countries" or "countries with IMF or World Bank programs." Their argument has been that if trade liberalization is desirable and feasible, particu- larly in the so-called sensitive sectors in the industrial countries, the effort should be broad-based and not limited to a few countries. More fundamentally, ques- tions have been raised about how, under such special liberalization, the main GATT principle of nondiscriminatory treatment would be respected. The ques- tions raised by "trade-finance links" will thus not be easily resolved. Finally how would strengthened GATT provisions and practice in this area fit with the overall program of trade liberalization that is a key objective of the Uruguay Round of negotiations? While bargains will be struck with trading partners individually, developed countries' actions in sectors of special interest to developing countries will be important in influencing developing countries' will- ingness to offer liberalization of balance of payments restrictions. More broadly, since the success of balance of payments adjustment depends crucially on the assurance of open markets abroad, prospects for success in strengthening other key trading rules, such as those on safeguards, subsidies, and quantitative re- strictions, will be helpful in strengthening discipline on balance of payments restrictions. V. CONCLUDING OBSERVATIONS The issues discussed above relating to the role and interest of developing countries in the Uruguay Round form part of a complex array of questions that will be examined in the coming years. A few general ideas deserve to be high- lighted in conclusion. First, it is evident that the GATT treatment of balance of payments restrictions merits a substantive reexamination because the economic premises on which the original approach was founded have largely lost whatever merit they might originally have had. It is now well understood that recourse to trade restrictions to correct the underlying imbalance between output and expenditure may have the effect of inducing compensating exchange rate changes, distorting trade and payments positions among countries, shifting the burden of adjustment to trad- ing partners, and inviting retaliation. There is no basis for believing that trade restrictions can correct balance of payments problems. In addition, consistency of policy would argue for tighter GATT discipline against trade restrictions for balance of payments purposes, the effects of which may be quite similar to exchange restrictions which are disallowed by the IMF. Second, reexamination of the balance of payments provisions of the GATT will in practice need to proceed side by side with consideration of the related issue of infant industry protection. The rationale for long-standing restrictions in the complex and restrictive trade regimes of many developing countries is often unclear. It is difficult to see how the balance of payments provisions of the GATT could be improved and made effective without reconsidering the justification for most other trade restrictions in developing countries as well. Anjaria 687 Third, insofar as the GATT'S balance of payments provisions are concerned, the division of the world into developed and developing camps-whatever its validity in the late 1940s and the 1950s-appears to be increasingly irrelevant. Prospects for liberalization of trade restrictions in favor of developing countries in markets abroad would be enhanced if discipline on trade restrictions for balance of payments purposes were strengthened in markets of increasing impor- tance to the industrial countries. Developing countries' moves toward more flexible exchange rate arrangements and market-oriented incentives, (making possible a reduced reliance on trade restrictions and on the GATT exceptions which their trade policies receive,) would be facilitated if they expected that there would be a substantial liberalization of trade in their important foreign markets. Whether the industrial countries are prepared to concede the usefulness to them of such a quid pro quo remains to be seen, but in the author's view such a re-orientation deserves careful consideration. REFERENCES Anjaria, S. J. 1986. "A New Round of Global Trade Negotiations." Finance and Devel- opment 23, no. 2 (June). . 1987. "Balance of Payments and Related Issues in the Uruguay Round of Trade Negotiations." Washington, D.C.: International Monetary Fund. Processed. Anjaria, S. J., Naheed Kirmani, and Arne B. Petersen. 1985. Trade Policy Issues and Developments. IMF Occasional Paper 38. Washington, D.C. Balassa, Bela A. 1983. "Exports, Policy Choices, and Economic Growth in Developing Countries after the 1973 Oil Shock." World Bank Development Research Department Discussion Paper 48. Washington, D.C. Bhagwati, Jagdish. 1978. Anatomy and Consequences of Exchange Control Regimes. Cambridge, Mass.: Ballinger for National Bureau of Economic Research. . Forthcoming. "Export-Promoting Trade Strategy: Issues and Options." World Bank Research Observer. Eglin, Richard. 1987. "GATT Surveillance Over the Use of Trade Restrictions for Balance of Payments Purposes:' World Economy, March. Gold, Joseph. 1986. "Some Legal Aspects of the IMF's Activities in Relation to Interna- tional Trade." Osterreichische Zeitschrift Fiur Offentliches Recht Und V6lkerrecht 36. Huh, Kyung Mo. 1983. "Countertrade: Trade Without Cash?" Finance and Develop- ment 20, no. 4 (December). International Monetary Fund. 1974. International Monetary Reform: Documents of the Committee of Twenty. Washington, D.C. 1986. Selected Decisions. Washington, D.C. Twelfth Issue, April. Various years. Annual Report on Exchange Arrangements and Exchange Re- strictions. Washington, D.C. . Various years. International Financial Statistics. Washington, D.C. Jackson, John H. 1969. World Trade and the Law of GATT. New York: Bobbs-Merrill. Krueger, Anne. 1978. Foreign Trade Regimes and Economic Development: Liberaliza- 688 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 tion Attempts and Consequences. Cambridge, Mass.: Ballinger for National Bureau of Economic Research. 1978. Organization for Economic Cooperation and Development. 1971. Trade Measures and Adjustment of the Balance of Payments. Paris. Papageorgiou, Demetris, M. Michaely, and A. Choksi. 1986. "The Phasing of a Trade Liberalization Policy: Preliminary Evidence,' Washington, D.C.: World Bank Country Policy Department. Processed. United States International Trade Commission. 1985. Assessment of the Effects of Barter and Countertrade Transactions on U.S. Industries. Washington, D.C. Wyndham-White, Eric. 1975. "Negotiations in Prospect." In C. Fred Bergsten (ed.), Toward a New World Trade Policy: The Maidenhead Papers, pp. 321-40. Lexington, Mass.: Heath. THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4: 689-705 GATT Safeguards and Voluntary Export Restraints: What Are the Interests of Developing Countries? Brian Hindley Exports from developing countries are frequently the targets of trade protection to offset injury to domestic producers. A safeguard clause of the General Agreement on Tariffs and Trade (GATT), Article XIX, authorizes such protection, but voluntary export restraints (VERs), which are not authorized or controlled by the GATT, are often used in its place. A callfor a "comprehensive agreement on safeguards" was one outcome of the 1986 Punta del Este ministerial meeting. The spread of VERS is often taken to be a threat to the interests of developing countries. The costs of VERS to developing country exporters may have been overesti- mated, however, and as a consequence, developing countries may be at risk of conced- ing too much, perhaps in terms of a relaxation of the conditions of application of Article XIX, in an attempt to ban or directly control VERS. The central issue is the extent to which VERS are adopted to avoid invocation of Article XIX. If so, there is no valid case for developing countries to pay anything for a ban on VERS. A better course for them would be to press for more rigor in GATT articles used as threat, which would enhance their bargaining position in setting the conditions for VERS. Successful exports from developing countries are often greeted by industrial countries with an increased level of protection. Of the various means employed to achieve this result, voluntary export restraints (VERS) are probably the most important. A VER occurs when a government limits the export of some good from its territory to another country at the request of the government of that other country.1 Since VERS are bilateral agreements between states, they escape- or have so far escaped-the rules of the General Agreement on Tariffs and Trade (GATT). In the context of the GATT, VERs are typically discussed under the heading of 1. In this article the term VER will be used so as to include all related instruments and terminologies, such as orderly marketing agreements (OMAS) and voluntary restraint arrangements (VRAS). VERS some- times appear as industry-to-industry agreements, without the explicit approval of the sponsoring govern- ments. Such arrangements can give rise to acute problems of definition, but these difficulties are not central to the argument of this article. The author is at the London School of Economics and the Trade Policy Research Centre. He thanks Richard Blackhurst, Michael Finger, Sunil Gulati, Robert Hudec, Martin Wolf, and especially Jagdish Bhagwati for comments on earlier versions of this article. Copyright © 1987 by the International Bank for Reconstruction and Development / THE WORLD BANK. 689 690 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 "the safeguard issue." The GATT contains a safeguard clause-Article XIX- which permits a country to increase its level of protection against a class of imports that increases so rapidly as to cause or threaten serious injury to domes- tic producers (for a valuable discussion of the legal aspects of Article XIX, see Jackson 1986). In circumstances in which Article XIX might apply, however, the governments of developed countries have demonstrated a preference for the use of VERS. A major element of the safeguards issue is that VERS have largely replaced the use of GATT Article XIX. Wolff (1983) reports a GATT Secretariat study showing "63 cases where those 'other' safeguard measures have been used since 1978 or were still in effect during this period, compared with only 19 actions notified under Article XIX." He comments that the ratio of VERS to Article XIX actions probably is "far greater" than three to one, since, as there is "no obligation to notify the taking of the extralegal measures, presumably not all are known." This is widely taken to be a problem. The Ministerial Declaration on the Uruguay Round, issued from Punta del Este, makes clear the importance at- tached to the subject, most explicitly in the statement that "a comprehensive agreement on safeguards is of particular importance to the strengthening of the GATT system and to progress in the multilateral trade negotiations" (p. 8). Two broad approaches to a resolution of this problem have been suggested. One is reform of Article XIX and the other is direct action against VERS. The area however, is filled with pitfalls. Developing countries have a very clear interest in the form of any comprehen- sive agreement on safeguards. This article views this difficult terrain from the standpoint of developing countries. Developed and developing countries have different interests in safeguards and Article XIX. The heaviest potential users of safeguards are likely to be developed countries. The potential targets of safeguard applications are countries with a rapidly changing structure of exports, and these, in turn, are likely to be coun- tries with a rapid rate of economic growth. A number of developing countries are in this class and all developing countries presumably aspire to it. Developed countries therefore are likely to seek formulations of Article XIX that provide many rights to importers and impose few constraints on the exercise of those rights.2 Developing countries, against whom those rights would proba- bly be used, are likely to prefer an Article XIX that protects exporters and that therefore provides only minimal rights for importers and places severe restric- tions on the invocation of such rights. The presence of VERS in the world trading system, however, creates a more complex problem for developing countries. The problem emerges from the prop- osition that developed countries prefer VERS to Article XIX because the condi- 2. This is a statement about the revealed preferences of governments, not about the actual economic interests of their constituents. From the latter point of view, the case for a safeguard clause (and particularly one that authorizes tariffs and quotas) is not strong. Hindley 691 tions of application of the present Article are so restrictive (in ways that will be explored later on). If that is so, a possible means of limiting the spread of VERS- and possibly of eliminating them altogether-is to relax the conditions of appli- cation of Article XIX so as to reduce the incentive for developed countries to use VERS. This argument has been put forward by the European Economic Community (EEC), which pressed very hard for an agreement on safeguards, though without success, in the Tokyo Round (1973-79). Developing countries came very close to accepting such a major relaxation of Article XIX, which was urged upon them by observers with impeccable antiprotectionist credentials (for example, Frank 1981). In the event, however, no agreement was reached. Since that time, the fundamental importance of the problem has been restated at each GATT meeting along with a concession of a total inability to offer any solution to it. (For a summary of the substantive issues, see UNCTAD Secretariat 1984). In the absence of VERS, relaxation of Article XIX would clearly be contrary to the interests of developing countries. That they should nevertheless have con- templated acceptance of such a relaxation is testimony to the fears aroused by the spread of VERs and implies a belief that developing countries should be prepared to pay a high price to rid the trading system and themselves of VERS. That belief is not well founded, for reasons that I shall later explore. Whatever developing countries ought to be willing to pay, however, the Punta del Este meetings seemed to suggest that they would not be required to pay anything at all in terms of a weakened Article XIX, or any other concessions. The outcome of the Punta del Este meeting seemed to signal a movement away from the old battlegrounds of the Tokyo Round and to open the possibility that VERS could be negotiated out of existence. 1. THE SAFEGUARDS ISSUE AFTER PUNTA DEL ESTE The primary basis for believing that an agreement to ban VERS is now possible lies in the section of the Ministerial Declaration dealing with Standstill and Rollback (Section C, pp. 5-6). Under the subheading "Rollback," each partici- pant agrees "Commencing immediately and continuing until the formal comple- tion of the negotiations . . . to apply the following commitments": (i) that all trade restrictive or distorting measures inconsistent with the provi- sions of the General Agreement or Instruments negotiated within the framework of GATT or under its auspices, shall be phased out or brought into conformity within an agreed timeframe not later than by the date of the formal completion of the negotiations, taking into account multilat- eral agreements, . . . reached in pursuance of the objectives of the negotia- tions. (ii) there shall be progressive implementation of this commitment on an equi- table basis in consultations among participants concerned, including all affected participants. This commitment shall take account of the concerns 692 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 expressed by any participant about measures directly affecting its trade interests; (iii) there shall be no GATT concessions requested for the elimination of these measures. Section D of the Ministerial Declaration contains an explicit call for an agree- ment on safeguards. Its language is blander than that of the sections quoted above but it does appear to provide a basis for hopes of progress on safeguards. There are, however, three reasons for caution in accepting this conclusion. The first lies in the simple fact that safeguards were high on the agenda of the Tokyo Round-without any subsequent reform. The supporters of any one direction of reform in that Round appear to have preferred the present form of Article XIX to compromises that would give partial effect to the reforms sug- gested by others. Thus, like a knot between two equally matched tug-of-war teams, Article XIX remained in place. A report in the Financial Times of September 22, 1986, illustrates the second reason for caution: Mr. Clayton Yeutter, the United States Trade Representative immediately spread confusion over the scope of the standstill commitment [which was couched in exactly the same terms as the rollback agreement] by telling journalists just before he left that it did not apply to so-called "grey-areas" of protectionism. The "grey-area" refers to bilateral agreements such as voluntary export restraints or orderly marketing arrangements by which governments cir- cumvent GATT rules without actually breaching them. The problem raised by Yeutter's reported words presumably lies in the delinea- tion of the "trade restrictive or distorting measures inconsistent with the provi- sions of the General Agreement" which participants have agreed to remove. The term must cover measures which the General Agreement expressly bans (for example export subsidies). The doubt raised by the Yeutter comment, however, is whether the agreement applies to measures which the General Agreement does not authorize, but which it does not expressly ban. Yeutter's words suggest that he places VERS in this category. The actual words of Articles XI and XIII seem to leave no room for doubt that VERs are banned. According to paragraph 1 of Article XI: "No prohibitions or restrictions other than duties, taxes or other charges, whether made effective through quotas, import or export licenses or other measures, shall be instituted or maintained by any contracting party . . . on the exportation or sale for export of any product destined for the territory of any other contracting party." And paragraph 1 of Article XIII forbids any "prohi- bition or restriction . . . on the exportation of any product . . . unless the exportation of the like product to all third countries is similarly prohibited or restricted'" The problem with inferring the illegality of VERS from this is that it is ques- Hindley 693 tionable whether these Articles were framed with VERS in mind. As an historical matter, it is arguable that they were intended to protect importing countries against the disruptive actions of exporting countries. But importing countries request VERS. The issue becomes one of legal philosophy: that of the extent to which the application of GATT law should be controlled by the intentions of the framers of the GATT, rather than the words in which they expressed those intentions. Nevertheless, in the light of Yeutter's words, it would not be sensible for developing countries to assume that the VER problem is solved without cost to them and that an automatic outcome of the MTN will be a ban on VERS. Perhaps that will be the outcome. Nevertheless, the conjunction of Yeutter's statement with the Punta del Este Declaration suggests the possibility that a more complex bargaining position is being established, in which action on safeguards will be conditional on concessions elsewhere. (A similar view of the bargaining costs to developing countries is expressed in Bhagwati's article in this issue.) The third reason for caution is that there are major technical difficulties in organizing an effective ban on VERS in the GATT framework. The central prob- lem is enforcement.3 The GATT is analagous to a system of civil law. It provides legal rights that constitute a basis for one contracting party (cp) to make a claim against another. Initiation of legal action in pursuit of those claims is left to cps. As the GATT is presently constituted, there is no equivalent of a police power to make indictments on the basis of breaches of the GATT rules. That some practice of a cp fails to conform to the GATT is neither here nor there unless another cp brings a complaint about the practice in the GATT. No formal complaint against a VER has ever been lodged with the GATT. Doubt about whether VERS contravene GATT rules might explain this, and such doubt could in principle easily be resolved. A more important problem, how- ever, may be a lack of cps who are willing to argue that they have been injured by a VER. The two governments who are parties of a VER are not candidates for the role of injured plaintiff (unless the exporter complains of coercion-but then his complaint will be about coercion, not about the VER). The question therefore is whether a third country has grounds for complaint. Third countries who import the good affected by the VERS will obtain it more cheaply as an outcome of its diversion from VER-protected market. That is an economic gain, although governments of countries which contain domestic im- port-competing production will not welcome it. Whatever the assessment of their position by third party importers, however, there is no clear basis for them to suppose that they are worse-off as a result of a VER than if a GATT-approved 3. The problem of enforcement is very much complicated by difficulties in identifying VERS, which take many forms. Arrangements between firms or industries, for example, may be sponsored by govern- ments but are much less easily identified as VERS than direct agreements between governments themselves. Moreover, as Jackson (1986, p. 41) notes, the GATT does not apply to the actions of private companies. 694 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 measure (for example, an application of Article XIX) had been used to restrict imports. Accordingly, even in terms of the political costs of cheaper imports, third party grounds for complaint are limited. Third countries who export a good that is subject to a VER in a competitor exporter will lose as a result of a lower price outside of the restricted market, but will gain as a result of the higher prices in the market to which the VER applies. Overall, they might either lose or gain as compared with their position prior to the VER (see, for example, Hindley 1979). But had an Article XIX action been used inistead, they would have been even worse off. Article XIX actions limit access to the restricted market and do not produce a higher price for exporters to that market. Neither of these groups of cps, therefore, has obvious ground for complaint. A VER may be a crime without a victim who is eligible to complain in the GATT. One way to remedy this would be to introduce GATT law into the domestic law of cps, so that private citizens would have a right to bring complaints against their own government. Buyers of the restricted good in the importing country are unambiguously worse off as a consequence of a VER and to provide them with rights to remedy that situation therefore has much to be said for it. This course, however, is unlikely to be politically feasible within the time span of the Uruguay Round. A second possibility, to provide the GATT Secretariat with the power to indict cips who infringe GATT rules, also lacks feasibility. Moreover, by placing the Secretariat in a political role, and making it a potential target for political controversy, it may in the long run pose a more serious threat to the effectiveness of the GATT than VERS. To deny the possibility that satisfactory solutions can be found in the course of the Uruguay Round would be foolish. It would also be foolish, however, to rely on the appearance of such solutions-and upon the degree of consensus and the goodwill necessary to make them effective. It therefore behooves developing countries to consider the costs to them of the VER system and the value to them of the various possible reforms of Article XIX. Their hesitation in the Tokyo Round suggested that they might be prepared to make large concessions in terms of relaxing the conditions of Article XIX to halt and reverse the spread of VERS- should that be their attitude in the Uruguay Round? To approach that question, it is useful to pursue further than was done in the Tokyo Round, the idea that Article XIX and VERS are interdependent. II. THE RELATION BETWEEN ARTICLE XIX AND VERs A central issue in the political economy of VERS is the extent to which they are in fact voluntary. The possibility that they are voluntary arises from the fact that the restriction on the quantity exported is likely to raise the price of the remain- ing exports to the restricted market. The higher profit per unit sold implies the possibility that the exporting country will gain as a result of a restraint, and Hindley 695 therefore the possibility that its government will be willing to voluntarily restrain exports.4 These rents make VERs a very expensive instrument for importing countries, as many studies have demonstrated (for example, Tarr and Morke 1984; and Greenaway and Hindley 1985). From this point of view, the most obvious question about VERS is not whether or why exporting countries accept them, but why importing countries ask for them. That question will be discussed below. The more immediately relevant aspect of the issue is that one alternative to a VER as a means of blocking awkward imports is Article XIX; hence, Article XIX bears on the negotiation of a VER. Were the government of an exporting country to refuse to limit exports, an alternative course of action open to the government of the importing country is to invoke Article XIX. The government of an export- ing country will reject a VER that imposes upon it costs greater than it expects to sustain from a successful Article XIX action (unless it is subjected to some additional threat). The government of an importing country will attempt action under Article XIX unless it can achieve its objectives at a lower cost by negotiat- ing a VER. It follows from this that the VER system and reform of Article XIX cannot sensibly be discussed in isolation from one another. The existence of the VER system will affect the outcome of any reform of Article XIX and the scope of the VER system depends upon the structure of Article XIX. For example, the propo- sition made in the Tokyo Round, that VERS might be effectively eliminated by a sufficient relaxation of the conditions of application of Article XIX, probably is correct. What was less clearly stated in the safeguard debates of the Tokyo Round is that such a reform is unlikely to be in the interests of the developing countries. The reform would eliminate the rents received by exporting countries when they agree to a VER. Moreover, and possibly of more importance, the reform would reduce the costs to developed country governments of safeguard actions. It thereby plausibly would increase their number. If the alternative to VERS was more open access by developing countries to the markets of developed countries, the loss of VER rents might not weigh heavily in a decision as to which system to choose. The alternative on offer in the Tokyo Round, however, is likely to have entailed the loss of VER rents and less market access. How could such an outcome be in the interests of developing countries? One argument to this effect derived from the threat that VERS were said to pose to the "credibility" of the GATT itself, and therefore to the GATT'S potential as a de- fender of the interests of smaller countries. That two "safeguard systems" exist side-by-side, one authorized by the GATT and the other not, and that the unauthorized system is more extensively used 4. The conditions for an exporter to gain are worked out for a variety of circumstances in Hindley (1979). An alternative explanation of why an exporter might prefer a VER to an Article XIX action is presented in the "porous-protection" model of Bhagwati (1987). 696 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 apparently is sufficient to persuade some commentators that major reform of Article XIX is needed. Yet out-of-court settlements are commonplace under the civil law, and that fact is not typically taken to suggest that the civil law is in major crisis or that its immediate reform is essential.5 Nor does anyone suppose that out-of-court settlements will fail to reflect the beliefs of the parties as to what the judgment of a court would have been. Even though a case does not go through the formal court system, its outcome reflects the state of the law. Moreover, all of the disputing parties presumably feel themselves to be better off with an out-of-court settlement than if the case had come to trial. This seems to be a possible basis for thinking about the VER system. It would not be an appropriate basis, however, if VERs are not truly voluntary. A second argument for attempting to bring VERS within the GATT structure derives from a concern that small countries in particular are coerced by more powerful trading partners into yielding "voluntary" export restraints that are voluntary in appearance only and are in fact seriously adverse to the interests of the small country. This may be true. If so, however, it is this fact that makes a case against VERS, and not their mere existence or number or proportion relative to actions under Article XIX. If this was the foundation of proposals to weaken Article XIX so that it is not worthwhile for importers to seek VERS, advocates of that position supplied very little evidence on a basically factual question. It would have been useful to have instances of coercion, and information as to the coercive threats that produced them. "Coercion" and "voluntary" are words that conceal conceptual problems. In the present context, however, a clear distinction can be made between threats involving actions that would be legal under the GATT and threats of actions that are not legal under and would not be authorized by the GATT. For example, suppose that the only consequence of an exporting country refusing a VER iS that the importing country will apply Article XIX, and that the circumstances are such that it would be legally entitled to do so. In that event, the exporting country accepting a VER iS coerced into it, in the sense that it would prefer to continue to export without restrictions, rather than face either an Article XIX action or a VER. Nevertheless, it is also true that the exporting country was in effect offered a choice between a VER and an Article XIX action and accepted the former. The word "voluntary" will be used here to describe the choice of a self-administered export restraint when the alternative is an action that is legal under the GATT. Article XIX is not the only such threat available to importers, however. Dumping and contervailing actions are reportedly (and plausibly) the most com- mon basis for threats made in the course of VER negotiations. But importers might offer threats that have nothing to do with trade or any issue that falls 5. Curiously, however, the position that the mere existence of VERS threaten the viability of the GATT is sometimes taken by lawyers-for example, Wolff (1983). Hindley 697 within the competence of the GATT. Yoffie (1983) provides a valuable discussion of VER bargaining. His account strongly suggests that matters not inherently connected with trade do enter the negotiations-for example, military and eco- nomic aid (see especially pp. 145-154). Not only importers can make use of retaliatory threats, however. A recent instance is provided by the Turkish government's public linkage of expanded U.S. quotas for Turkish textiles and the renewal of U.S. bases in Turkey.6 Even accepting that extra-GATT threats enter the VER negotiating process, therefore, there is a question as to whether they lead to bargains that are more or less favorable to exporters. Yoffie notes, for example, that "in exchange for the textile agreement with some amendment of America's terms, the Koreans ex- tracted 776.3 million dollars worth of P.L. 480 [food aid] commitments" (1983 p. 151). Moreover, Hong Kong, which has accepted no VERs other than those coming under the Multifibre Agreement, seems to have been able to withstand the threats of importers without suffering major harm. Other exporters may possess fewer bargaining chips. Some "VERS" may derive from the threats by importers of actions that would not be legal within the GATT. That is very far from accepting, however, that the existence of VERS necessarily or even strongly implies that importers have exercised such coercion. There are two different theories of VERS. One emerges from the proposition that VERs result from threats of action that are not authorized by the GATT and the other from the proposition that they are voluntary (in the sense that the exporter prefers to accept a VER rather than any other action that is legally available to the importer). These theories yield different assessments of the desirability of particular solutions to the safeguard problem. If coercion is wide- spread in the negotiation of VERS, so that exporting countries are forced into arrangements that are inferior to those that would emerge from GATT-authorized measures, there is a clear basis for attempting to use the GATT as a counter- weight. This case is seriously weakened, however, if the typical alternative to a VER is an action that would be authorized by the GATT. III. THE SIMPLE ECONOMICS OF VOLUNTARY EXPORT RESTRAINTS Exposition is simplified by initially treating Article XIX as the only GATT-legal alternative that might emerge as a threat in the course of VER negotiations. Thus, if an exporting country does not grant a VER, the importing country will take action under Article XIX. Then the standard argument from voluntary exchange suggests that an exporting country government will accept a VER only if it expects to be at least as well off with the VER as with the Article XIX action.7 6. The story is headlined: "Turks Link Trade to Pact on Bases, Surprising Schultz." New York Times, March 24, 1986, p. Al. 7. There is, of course, a major problem of how costs and benefits will be defined by members of governments negotiating on behalf of private citizens. "Will reject" must therefore be treated as a state- ment about a model, not about the real world. It is the actions open to governments that are important in the present context, however. 698 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 Either a VER or an Article XIX action will reduce the output of the exporting industry below what it would otherwise have been. Either instrument will there- fore reduce the returns to factors of production employed in the industry. A VER, however, provides pecuniary compensation where an Article XIX action does not. If a country's exports are to be restricted to some level, the country cannot be worse off (and may be better off) if the restriction is achieved by a VER rather than by an Article XIX action. An Article XIX action that conforms with the GATT cannot discriminate be- tween exporters, and so will affect all other exporters. Even if no other exporter is affected by a VER, however, it is still true that the restricted exporter cannot be worse off with the VER than with an Article XIX action that cuts his exports to the same level. This is because an Article XIX action pushes the supplier back along his supply curve, while a VER achieves the same level of restriction by moving him up along his demand curve to a higher per unit price. So long as the relevant demand curve is above the supply curve, a supplier will prefer any given level of restriction of exports to be achieved by a VER. When a VER restricts a country's exports to a lower level than an Article XIX action, the circumstances in which the VER will provide an exporting country with the same level of welfare as the Article XIX are less easy to define. The VER in that event will reduce the demand for factors of production of the exporting industry to a greater extent than would the Article XIX alternative, so that they will require compensation to remain at the same level of welfare. The price of exports may rise as a result of the restriction, however, possibly providing the means for compensation. It is important, however, to be clear as to the questions at issue. If export restraints are voluntary in the sense defined here, the exporter will evaluate the effects of a VER that cuts exports by more than an Article XIX action, and based on that assessment will accept or reject the VER. If an exporting country antici- pates that a suggested VER will make it worse off than the Article XIX restric- tion, it will reject the VER and opt for Article XIX. (And, of course, the importing country may return to the suggested level of exports but with an additional side payment, or perhaps a commitment to obtain VERS from other exporters also). A good deal of discussion of VERs appears to be based on an implicit assump- tion that if VERS were not available, then importing country governments would do nothing to solve what they perceive as their problems with import-competing industries. That does not seem a plausible hypothesis. In certain cases, however, the requirement of Article XIX that the action be taken against all exporters might make it too politically embarrassing for the importing country government to apply the Article. In that event, the relevant alternative to the VER is the status quo. To avoid the awkwardness that would arise from use of Article XIX, importing country governments may be prepared to pay enough for a VER that exporters are better off absolutely rather than merely relative to the Article XIX alternative. Exporters, of course, would be Hindley 699 prepared to accept less than this if they could be persuaded that an Article XIX action was possible. Nevertheless, if the exporting country correctly assesses the situation, it might obtain a VER that improves its welfare. So far as I know, only one study of the impact of VERS on exporters is available. The results of Tarr (1987) suggest that steel exporters restricted by VERS are better off than they would have been in the absence of VERS. To the extent that VERS are voluntary, the case for GATT surveillance is weak- ened. A case for surveillance cannot then easily be based upon the effects on the parties to the agreement. If the parties to a VER prefer that arrangement to any action authorized by the GATT, what virtue would lie in an effort to force them into GATT-approved action? This is not to say, of course, that there are no losers from a VER in the countries directly involved. In particular, buyers in the importing country are worse off. They pay the rents that make the restraint acceptable to exporting country governments. A difficulty for the voluntarist theory, therefore, is to explain the revealed preference of the governments of importing countries for VERS. IV. THE DISUSE OF ARTICLE XIX: SELECTIVITY AND COMPENSATION UNCTAD (1984) provides an excellent summary of the substantive issues re- garding safeguards as they were viewed prior to the Punta del Este meetings. Its succinct conclusion is worth quotation: "The issues of discrimination and com- pensation remain central to the GATT discussions. Discussions on other aspects of the 'safeguards' system involve second-order issues, and can well have the effect of drawing negotiating energies and resources away from these two key issues" (p. 9). Some governments of developed countries claim that their neglect of Article XIX is due, first to its requirement that exporters affected by Article XIX actions be compensated and, second, to the requirement that Article XIX actions must apply to all exporters of a particular product, which means that "disruptive" exporters cannot be singled out for special treatment. These developed country governments often go on to suggest that a relaxation of these requirements would lessen their temptation to seek VERS. The compensation provisions of Article XIX are contained in Paragraph 3(a), which permits affected exporting countries to increase their protection against some export of the country applying Article XIX, or to claim from the applicant country an equivalent compensatory concession. This leads to an evident cum- bersomeness when several exporting countries are affected by an Article XIX application. In addition, the provisions are likely to increase the domestic politi- cal costs of using Article XIX-an increase in protection to one domestic indus- try must be balanced by costs imposed on other domestic industries. The second common explanation for the disuse of Article XIX, that it does not permit discrimination between different sources of imports, is sometimes 700 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 connected with the compensation requirements of the Article. It is said that discrimination between exporters provides possible means of avoiding arduous negotiations when a number of different countries are sources of the relevant import and all exercise their right to equivalent concessions. Beyond that technical problem, however, lies a more important political one. Nondiscrimination means that action taken under Article XIX must be taken against powerful trading partners as well as against weaker ones. Thus, for example, using Article XIX, the United States and the European Community (EEC) would have to act against one another's exports when an expansion of competitive imports from a newly industrializing country has "created" the problem. Tension in a central political relationship may be threatened as a result of economic events in a relationship that may be politically peripheral. Seen from another vantage point, however, the fundamental problem lies in the decision to protect the import-competing industry. This may explain why the most common defense of selectivity lies neither in the technical problem of equivalent tariff adjustment nor in the political problems of relations between trade superpowers. The most common defense lies in the ethical proposition that importing country governments have a right to prevent "market disruption" (a term used by proponents of VERs and other protective devices to describe the impact of new and dynamic exporters on established suppliers). From this dubi- ous foundation, the right of importing country governments to take action against "disruptive" imports without reference to the exporting country govern- ment is deduced. (Hindley 1979 gives a fuller discussion of this contention.) Voluntary export restraints were defined above as the exporting country's acceptance of a VER rather than an Article XIX action. This voluntarist view is used below to elucidate the issues of selectivity and compensation from what may be an unfamiliar perspective. Selectivity In the Tokyo Round, the European Community tried very hard to obtain modifications of Article XIX so that selective safeguards could be imposed at the discretion of importers (as already noted, its representatives at one point de- scribed such a reform as the "sine qua non of any overall agreement"). The rejection of this proposal was due to a failure to obtain agreement on how, or whether, the discretion of importers was to be constrained. In the Tokyo Round, it appears that agreement was almost reached on the introduction into Article XIX of consensual selectivity-that is, an amendment to permit an importing country government to discriminate against a particular exporter but only if that exporter agrees to it. Acceptance of the reform was apparently prevented by the insistence of the EEC that in specified circumstances, selectivity should be permitted without the agreement of the exporter. The quid pro quo offered by the EEC appears to have been the inclusion in the amended Article XIX of a specified limited duration for nonconsensual action; a commit- Hindley 701 ment on the quantitative impact of the application (for example, that the level of imports from the country to be discriminated against would be maintained); and the establishment of a Committee within the GATT for the settlement of disputes and the surveillance of Article XIX actions. From a voluntarist standpoint, the legal and procedural protection offered in this package has little value. It is better designed to appeal to those who believe that extra-GATT coercion of exporters by importers plays a major role in the negotiation of VERS (this belief, incidentally, does not seem to sit comfortably with the importance given to Article XIX reform by the EEC: if exporters are so easily forced into VERS, why should reform of Article XIX be a major issue for importers?) Even on the view that VERS result from extra-GATT coercion, however, very careful examination of the detail of what is offered is essential. A central element in such an examination must be the means provided to the GATT Committee to establish its authority over VERS. Without some means of achieving such control, the rest of the package favors importers. The options of importers would improve in the class of cases for which they were authorized to take selective action (which would inevitably expand without either a very careful definition or a very determined Committee, or both). For some period, importers would be able to apply Article XIX selectively without reference to the exporter (and, also, if the second major EEC suggestion were accepted, without compensation to the exporter). At the end of that period, the importer would again have the same options as at present. For exporters, Article XIX applications would become more damaging. For some class of cases and for some period, they would have no legal basis for negotiation and would, therefore, be worse off than at present. After that period exporters would face the same situation which they now face. Were the Committee unable to control VERS, the amendment would cause a deterioration in the negotiating position of exporting countries. One effect of that deterioration would be to improve the terms on which importers could obtain VERS. That, in turn, would offset the attractions of using the new Article XIX. Compensation Broadly speaking, there are two relevant alternatives to the present provisions for compensation of Article XIX. The first is to eliminate them without replace- ment-to accept the EEC idea that when a safeguard action conforms to the Article XIX rules, the applicant should not be required to pay a "penalty." The second possibility is to construct a rule for monetary compensation to exporters affected by an Article XIX action, either as a replacement for the current provisions or as an institutionalized alternative to them. The possibility will not be discussed here. Seen from this standpoint, VERS provide pecuniary 702 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 compensation to exporters. There is an obvious question of whether any simple and negotiable rule is superior to VERS for that purpose.8 The existing provisions of Article XIX provide a right to affected exporters to raise their duties on some good of interest to the Article XIX applicant. This is consistent with the GATT orthodoxy of maintaining an equality of concessions. On that view, the withdrawal of a concession under Article XIX must be matched by an equivalent alternative concession offered by the Article XIX applicant or withdrawn by an affected exporter. The retaliation provisions therefore further whatever ends are served by first-difference reciprocity.9 Introduction of the idea of penalty, however, abandons GATT orthodoxy. It fastens on the fact that the provisions make access to Article XIX more expen- sive. In doing so, however, it raises the issue of whether the compensation or retaliation provisions of the Article serve some function other than maintenance of first-difference reciprocity. In fact, it seems clear that they do. Use of Article XIX damages the interests of persons outside the applicant country-for example, those who have invested in productive equipment in affected exporting countries. That being so, there is a clear case for ensuring that the Article is not used for trivial purposes. One means of achieving that is to ensure that application of Article XIX is accompa- nied by substantial costs upon the applicant. The complaints of importing coun- try governments suggest that the existing retaliation provisions achieve that. Suppose that Article XIX were amended by eliminating any requirement to reequalize the balance of concessions. Thus, an Article XIX applicant would neither be required to offer an equivalent concession to the one withdrawn nor to accept the loss of an equivalent concession by affected exporters. Clearly, this reform would make Article XIX more attractive to importers. Hence, it might be expected to produce the same two primary effects as would introduction into the Article of selectivity: (1) it would cause substitution of Article XIX actions for VERs and hence would increase Article XIX actions relative to VERS; and (2) insofar as the costs of Article XIX actions for importers are a determinant of the bargains available to exporters who have been asked for VERS, it would lead to a shift of bargaining power from exporters to importers. Neither of these outcomes is desirable from the standpoint of developing countries. The first represents a substitution of an instrument which generates no rent for exporters for one that does. The second implies a reduction in the rents generated for exporters by VERS. A possible offset to those losses lies in the implication that a higher proportion of safeguard actions would come within the 8. Deardorff (1986) suggests global quotas with the transferrable rights to import allocated to exporters in accordance with a market share rule. Earlier, Bhadwati (1977) had also proposed an alternative reform of Article XIX. 9. Hindley (1986) discusses the function of first-difference reciprocity. Jagdish Bhagwati (Bhagwati and Irwin, 1987) appears to be the inventor of this useful term-which, when absolute reciprocity figures so largely in discussion of commercial policy, makes possible a concise discrimination between concepts often confounded. Hindley 703 GATT. The arguments made in this article suggest that that has a dubious value. Since the reform would reduce the costs to importers of safeguards, it would probably increase the total number of safeguard actions. It is therefore possible that it would increase the proportion of total safeguard actions within the GATT and increase the total number of safeguard actions falling outside of the GATT. Selectivity and Compensation from a Developing Country Perspective The different implications for policy of voluntarist and nonvoluntarist views of VERs emerge sharply from a focus on selectivity and compensation. A nonvol- untarist position suggests that developing countries should concede relaxations in the conditions of Article XIX in the hope of limiting the threat of VERS. A voluntarist position suggests the reverse. It suggests that developing countries should press for tighter and tougher conditions of application of Article XIX. A consequence of such a reform of Article XIX might be that a higher propor- tion of safeguard actions took the form of VERs. That is a cost from a nonvolun- tarist standpoint. From a voluntarist standpoint, however, if exports are to be restricted, it is better that they should be restricted by a VER than by an Article XIX action. To tighten the conditions of application of Article XIX improves the bargaining position of developing countries asked for VERS. To make Article XIX more expensive for importing countries also makes VERS more expensive. It is therefore likely to reduce the number of safeguard actions of all kinds. That a package that amounts to an overall toughening and tightening of Article XIX can be negotiated in the Uruguay Round may not be very likely. Nevertheless, this analysis suggests that before accepting any relaxation of Arti- cle XI]X, developing countries should think very carefully about what they are getting in return. Developed country protectionism is not a good thing, whether viewed from the standpoint of the developing countries or from that of the developed coun- tries themselves. Nevertheless, VERS can be viewed as an opportunity for devel- oping countries to make the most of the protectionist actions of the established industrial countries. Certainly the mere presence of VERS in the world trading system should not scare developing countries away from pursuit of their natural interest in tightening the conditions of application of Article XIX. V. OTHER THREATS OF ACTIONS LEGAL UNDER THE GATT The assumption that Article XIX is the sole source of threats of GATT-autho- rized action that might emerge in the course of VER negotiations has been useful for two reasons. First, it simplifies exposition. Second, it permits the argument to be presented against the familiar background of the safeguard issue. In fact, however, Article XIX is not the sole source of GATT-authorized alternatives to VERS. The argument based on the Article XIX alternative, however, generalizes to any other alternative that would be legal under the GATT. Insofar as the parties 704 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 to a VER regard it as superior to the GATT action, it is difficult to see that there is a case for attempting to force them to accept the GATT action. If the arrange- ments that are reached in the light of the GATT alternative are less than satisfac- tory to exporters, then the appropriate remedy is to tighten the GATT alternative. Thus, for example, if developing countries feel that they are being forced into unduly restrictive VERS by the threat of countervailing or antidumping actions from importers, their attention would be better focused on the rules for anti- dumping and countervailing actions than upon VERS. To ban VERS-even if such an act could be agreed and could be effective-would still leave developing countries facing the inappropriate rules for dumping and countervailing actions. To ban VERS would also remove the possibility of reaching an arrangement that developing countries preferred to the operation of those rules. VI. CONCLUSION: DEVELOPING COUNTRY OBJECTIVES IN THE MTN In thinking about safeguards, and the interests of developing countries in their reform, two pitfalls have not always been avoided. The first pitfall is to treat a limitation on VERs as though it is tantamount to a limitation on protectionism in developed countries. The second is to overstate the threat that VERS pose to the interests of developing countries. Regrettably, a limitation on the use of VERS may simply cause the substitution of alternative protective instruments for VERS. The effects of such a substitution on the economic welfare of developing countries obviously depends upon the alternative form of protection. In view of the call of the Punta del Este Ministe- rial Declaration for bringing "trade restrictive or distorting measures inconsis- tent with the General Agreement . . . into conformity" with the General Agreement, however, one substitution that seems especially relevant is that of Article XIX actions for existing VERS. There are good grounds for skepticism that such a substitution would be in the interests of developing countries. The point may have special relevance in the context of discussions of the safeguard issue inthe Uruguay Round. If some move toward control of VERS is made, it is not impossible that it will be accompanied by suggestions that renun- ciations of VERS by the developed countries would be facilitated by a relaxation of some of the conditions of application of Article XIX. Developing countries should assess the detail of any such suggestion very carefully indeed. A relaxa- tion of Article XIX worsens the substitution problem. The argument presented here suggests that developing country interests might be better served by a tightening of the conditions of application of Article XIX without formal control of VERS rather than by control of VERS accompanied by a relaxation of Article XIX. The second pitfall is to be overimpressed by the notion of VERS as a threat.10 10. Bhagwati's (1987) porous protection model suggests the same conclusion, though on different-but not incompatible-grounds. Hindley 705 Developed country protectionism is a threat to the interests of developing coun- tries; but no action on safeguards in the GATT will remove that threat. VERS may offer developing countries their best means of making the best of the current sad state of developed country trade policy. REFERENCES Bhagwati, Jagdish. 1977. "Market Disruption, Export Market Disruption and GATT Reform.' In Bhagwati, ed., The New International Economic Order: The North South Debate. Cambridge, Mass.: MIT Press. - 1987. "VERS, Quid Pro Quo DFI and vIEs: Political Economy-Theoretic Analyses" International Economic Journal 1, no. 1 (Spring): 1-14. Bhagwati, Jagdish, and Douglas Irwin. 1987. "The Return of the Reciprocitarians: U.S. Trade Policy Today." World Economy 10, no. 2 (June): 109-30. Deardorff, Alan V. 1986. "Safeguards Policy and the Conservative Social Welfare Func- tion." University of Michigan Research Seminar in International Economics, Discus- sion Paper 173. Ann Arbor. Frank, Isaiah. 1981. Trade Policy Issues for the Developing Countries in the 1980s. World Bank Staff Working Paper 478. Washington, D.C. GATT. 1986. Ministerial Declaration on the Uruguay Round. Geneva: GATT/1396, 25 Sept. Greenaway, David, and Brian Hindley. 1985. What Britain Pays for Voluntary Export Restraints. London: Trade Policy Research Centre. Hindley, Brian. 1979. "Voluntary Export Restraints and Article XIX of the GATT." In John Black and Brian Hindley, eds., Current Issues in Commercial Policy and Diplo- macy. London: Macmillan. Hindley, Brian. 1986. "A Comment on Jagdish Bhagwati's Geneva Association Lecture.' In Orio Gianini, ed., The Emerging Service Economy. Services World Economy Series 1. New York: Permagon Press. Jackson, John H. 1986. "The Role of the GATT in Monitoring and Promoting Adjust- ment: the Safeguards System." University of Michigan Research Seminar in Interna- tional Economics, Discussion Paper 170. Ann Arbor. Tarr, David G. 1987. "Effects of Restraining Steel Exports from the Republic of Korea and Other Countries to the United States and the European Economic Community." World Bank Economic Review 1,no. 3:397-418. Tarr, David, and Morris Morkre. 1984. Aggregate Costs to the United States of Tariffs and Quotas on Imports. Washington, D.C.: Federal Trade Commission. UNCTAD Secretariat. 1984. An Improved and More Efficient Safeguard System. Geneva. Wolff, Alan W. 1983. "Need for New GATT Rules to Govern Safeguard Actions." In William R. Cline ed., Trade Policyfor the 1980s. Cambridge, Mass.: MIT Press. Yoffie, David B. 1983. Power and Protectionism. New York: Columbia University Press. THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4: 707-725 International Control of Subsidies and Countervailing Duties J. Michael Finger and Julio Nogues The range and number of cases of administered protection in the 1 980s suggests that it has begun to play an important role in shaping international trade flows. As most of such cases are brought by the United States against developing country exporters, they are also a matter of concern for developing countries in the Uruguay Round of trade negotiations. The internationally negotiated code on subsidies and countervailing du- ties is ambiguous in its definition of "legal" subsidies, and thus in the appropriate use of countervailing duties. Because the code is applied at the national level, there is conside- rable pressure by domestic producers on administering agents to increase the use of such protective measures through the adoption of provisional measures while the inves- tigation is being conducted, and through the interpretation of criteria for coverage. Given that the most successful developing countries have been those which adopt the most neutral policies toward imported and domestic goods, and between sectors do- mestically, the very biased application of the subsidies and the trade policy distortions common in the countries examined, together with the frequency of countervailing actions against the subsidies, suggest that developing countries would benefit from abandonment of the subsidies. In September of 1986, the world's trading nations agreed to begin a new round of multilateral trade negotiations, to be called the Uruguay Round. The previous negotiations, the Tokyo Round, ended in 1979. In the six intervening years, 1980-85, seven countries and the European Community (EEC) initiated 1,155 antidumping cases. There were also 425 antisubsidy cases, half-a-hundred safe- guards cases, and several hundred other trade practice cases resulting from national attempts to interpret and enforce various international agreements (ta- ble 1). The sheer number of such cases suggests that the terms and conditions on which goods are traded across national boundaries are being affected by the outcomes of such administrative processes as much as by the gradual unrolling of the tariff reductions agreed at the Tokyo Round. The General Agreement on Tariffs and Trade (GATT) and the codes which elaborate on parts of it express the international community's agreement as to The authors are economists at the World Bank. They wish to thank Azita Amjadi for her industrious work in pulling together the information on which this article builds. Copyright ( 1987 by the International Bank for Reconstruction and Development / THE WORLD BANK. 707 708 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 Table 1. Number of Administered Protection Cases Initiated, 1980-85 Protection measure 1980 1981 1982 1983 1984 1985 1980-85 Safeguards United States, 2 6 4 2 6 3 23 Australia 1 0 1 2 0 0 4 Canada 0 1 2 0 0 1 4 EEC 3 1 1 1 1 - 7 Countervailing duties United Statesb 8 10 123 21 51 39 252 Australia 0 0 2 7 6 3 18 Canada 3 0 1 3 2 3 12 EEC 0 1 3 2 1 0 7 Antidumping actions United States' 22 14 61 47 71 65 280 Australia 62 50 78 87 56 60 393 Canada 25 19 72 36 31 36 219 EEC 25 47 55 36 49 42 254 Other unfair trading practices United Statesd 28 19 73 39 33 39 231 Total United States 60 49 261 109 161 146 786 Australia 63 50 81 96 62 63 415 Canada 28 20 75 39 33 40 235 EEC 28 49 59 39 51 42 268 -Not available. a. U.S. Trade Act, Section 201. b. U.S. Trade Act, Section 701. c. U.S. Trade Act, Section 731. d. As defined in the U.S. Trade Act, consist of unfair importing practices (Section 337), unfair trading practices (Section 301), and market disruption (Section 406). Sources: Operations of the Trade Agreements Program (OTAP), U.S. International Trade Commission (USITC), 1981-86, based on actions reported by signatories to the GATT Committee on Subsidies and Antidumping Practices; GATT, "Semi-Annual Report on Antidumping and Subsidies Measures," 1980-86; Trade Action Monitoring System, Office of the U.S. Trade Representative, April 4, 1986; U.S. Federal Register, Department of Commerce, various issues. Case numbers for U.S. countervailing duties, antidumping actions, and other unfair trading practices are based on USITC initiations rather than on the U.S. Department of Commerce listings. Australian data for 1980-81 are from "Status of Dumping Complaints,' Australian Industrial Assistance Commission Annual Report, 1983-84. GATT reports for Australia are not available for those years. Safeguards figures are from GATT, "Modalities of Application of Article XIX, which does not include the complete number of safeguard cases. EEC data are from OTAP (1985); USITC, June 1986; Official Journal of EEC, various issues; and the Annual Report of the Commission of the European Parliament, September 1983, December 1984, and June 1986. Finger and Nogues 709 when a country may take such actions. Safeguard actions may be taken when imports threaten or cause serious injury to domestic producers, and antidump- ing actions when exports are sold at less than normal value, and such goods cause injury to domestic producers when they are imported. Likewise, a country may isolate itself from the effects of another's subsidy through the use of a countervailing duty, which is a charge against the subsidized import. If country A provides a 5 percent subsidy on its exports to country B, for example, and B in turn adds a 5 percent countervailing duty to these goods as they are imported, there would be no effect on trade. The implementation of these agreements is through national, not international procedures; for example, an antidumping complaint by a U.S. industry is investigated by the U.S. government, while an import injury or safeguard complaint by an Australian industry is examined by the Australian government. Each of these processes has, as its operative instrument, the authority to restrict imports. Herein lies a problem. International agreements simply define the limits of import restrictions beyond which domestic interests do not have an internationally agreed claim. They cannot, however, constrain the extent to which import restrictions may be sought by these interests. Those who would benefit from import restrictions will pressure the agents responsible for adminis- tering them to accept their circumstances as matching the criteria of the instru- ments, or to modify the criteria to match their circumstances. Thus, overly enthusiastic application of countervailing duties (CVDS) might be as much a problem as overly enthusiastic application of subsidies. O(ur analysis begins by asking if the national use of subsidies and CVDS iS consistent with internationally agreed principles. We conclude that there are no internationally agreed principles and that this is not a multilateral, but a bilateral issue, with the United States on one side and its trading partners on the other. From there we turn to three issues related to CVD enforcement. First, we analyze concerns about harassment, concerns that CVD enforcement may go beyond the law, or that the process of CVD enforcement (apart from the final determination of a case) raises the costs of imports. Second, we discuss how countervailing duty policy, as applied by the United States, affects the national economic interest of the United States and its trading partners, on whose exports it is applied. We conclude that while U.S. counter- vailing duty policy appeals to vocal domestic constituencies, it detracts from the national economic interest of the United States. Internally, it is good politics, but bad economics. Third, we examine the effect of CVDS on trading partners. A partial review of developing countries' subsidy practices which have been countervailed by the United States suggests that these subsidies are often not consistent with an economically sound development strategy. Export subsidies which are uniform across industries and equal in magnitude to the export disincentives from import restrictions and inappropriate exchange rates can be important parts of an effec- 710 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 tive development strategy. But the actual subsidy practices which we examine cannot be defended on such grounds. I. THE INTERNATIONAL AGREEMENT ON SUBSIDIES During the Tokyo Round, the Agreement on Interpretation and Application of Articles VI, XVI, and XXIII of the General Agreement on Tariffs and Trade (the "Subsidies Code") was negotiated. The preamble to the Code reflects the desire of its signatories to bring appropriate discipline to the use of subsidies and countervailing duties. The Code recognizes that "subsidies are used by govern- ments to promote important objectives of national policy," but that "subsidies may have harmful effects on trade and production." While a stated objective of the Code is "to insure that the use of subsidies does not adversely affect or prejudice the interests of any signatory to this agreement," it also expresses a desire "that countervailing measures do not unjustifiably impede international trade." The Code does not question the propositions that one country's subsidies might cause harm to another, nor that countervailing duties impede trade. Its concern is to strike a balance on each matter. It provides certain rules for the use of subsidies and certain rules for the application of CVDS. There is, of course, an intuitive complementarity between rules for use of subsidies and rules for use of CVDS, but this complementarity is not formal. Limits that one part of the Code puts on the use of subsidies are neither the economic nor the legal inverse of provisions in another part for the use of CVDS. It is possible that a subsidy may be legal under the Code, and simultaneously a CVD against that subsidy be equally legal. Subsidy Rules Article 8 of the Code provides that, in the use of any subsidy, signatories shall seek to avoid causing either "injury to the domestic industry of another signa- tory" or "adverse effects" to another signatory. The proscribed effects could arise from (1) subsidized imports in the domestic market of the importing signatory; (2) displacement of or impediment to imports of like products into the subsidiz- ing country; or (3) displacement of exports of like products of another signatory from a third country market. Article 9 states that "signatories shall not grant export subsidies on products other than certain primary products:' "Certain primary products" comprise any product of farm, forest, or fishery in natural form or processed as is customarily required to prepare it for marketing in substantial volume in international trade. The process the Code provides for implementing its rules about the use of subsidies is not a process of "enforcement" as one would think of enforcement in the context of a penal code and a police force. There is no police force. The international track is a process intended to develop and promote a "mutually Finger and Nogues 711 satisfactory solution." Its basic parts are "consultations," "conciliation," "dispute settlement,' and "authorized countermeasures." Consultation is perhaps the major part of this track. The Code provides that a signatory may request consultations with another signatory whenever it has reason to believe that the other signatory is providing an export subsidy, or any subsidy with negative effects for the first signatory. The stated objectives of such consultations are to clarify the facts and to arrive at a "mutually acceptable solution " If such a solution is not reached in a specified number of days, either party may refer the matter to the entire membership of the Code (the "Commit- tee") for conciliation. The Committee will review the facts and use its "good offices" to encourage the signatories involved to reach "a mutually acceptable solution." If none is reached in thirty days, any involved signatory may request that a "Panel" be established. A Panel would be composed of three to five members, not nationals of the countries involved. The Panel will review the facts and present to the Committee its findings on the rights and obligations of the signatories, both under the Code and under the General Agreement. Again, the objective of the process is to generate a mutually satisfactory solution, and the details of the process are structured toward promoting such an outcome, rather than toward a formal finding of who is or is not in violation of the relevant provisions of the General Agreement and the Code. If agreement is not reached, the Panel may submit a legal finding to the Committee, which may (but is not obligated to) make recom- mendations and may, if its recommendations are not followed, authorize appro- priate countermeasures. Neither the Panel nor the Committee has any formal power to enforce coun- termeasures-to fine or otherwise punish a government for violating its Code obligations or for imposing unauthorized countermeasures. Their power is only the power of persuasion, and of appeal to an agreed international standard. Countervailing Duty Rules The code recognizes the right of a country to impose countervailing duties on subsidized imports which cause injury to a domestic industry. The relevant wording in the code is simply "subsidy,' not "illegal subsidy" or "export subsidy." The code provides considerable detail as to how a country should conduct subsidy-injury investigations. National procedures are outlined in the Code, which allows a country to determine if its imports have been subsidized and if they have brought or could bring injury to domestic interests. If the determina- tion is positive, a CVD may be imposed. The "shalls" and the "shall nots" of countervailing duties are expressed in the language of these procedures, not as definitions of "subsidy" and of "injury." From the perspective of bringing discipline to the use of CVDS, the logic of such codification is that if the circumstances under which a country may impede trade are specified, the frequency with which it does so will be less. From the perspec- 712 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 tive of promoting international agreement, formulators of the code hoped that this codification would make national procedures to impose CVDS the same in one country as in another, and standardization of CVD procedures would in turn push toward standardization of subsidy practices. This standardization would be facilitated by periodic reviews by the Code membership of subsidy and coun- tervailing duty practices which individual members had brought to the group's attention. As precedents built up, each national government would have effec- tive means for handling pressures for subsidies and pressures for protection based on the allegation that the competing imports were subsidized. A petitioner would be assured that his case was decided on the same grounds as anyone else's (as well as that these were the right grounds) and hence would be more willing to accept a possibly negative decision. Operation of the Code A generally sympathetic review, which described negotiation of the Tokyo Round codes as a "major accomplishment" (Stern, Jackson, and Hoeckman 1986, p. 1) concluded that the Subsidies Code "is generally perceived to have been working poorly" (p. 25). Rather than movement toward agreement on what constitutes an improper subsidy, the dispute settlement mechanism has served only to dramatize differences between signatories. Requests for investiga- tion of one signatory's policies often lead to stalemating counter requests. For example, a U.S. complaint against EEC subsidization of wheat flour, pasta, and poultry exports was followed by an EEC complaint of U.S. subsidization of wheat flour exports to Egypt. Panels were established, but a panel report unfa- vorable to the U.S. complaint was blocked by the United States, and the EEC blocked a report unfavorable to it. Many illustrations can be provided that the Code is not an agreement in principle on what constitutes an unacceptable subsidy or an acceptable CVD. The most obvious one is that the Code provides no definition of subsidy or even of export subsidy. The Code includes as an annex a list of practices which the signatories could only agree "is illustrative of export subsidies." There are significant differences between major signatories over what consti- tutes a subsidy and over how the magnitude of a subsidy would be determined. For example, the United States and the EEC differ on whether or not a practice must necessarily constitute a burden on public funds in order to fall under the definition of subsidy. Alternatively, any state intervention, such as price controls or regulatory standards, might be defined as a subsidy. The European Commu- nity has "consistently taken the position that a charge on the public account is an indispensible requirement of any subsidy" (Beseler and Williams 1986, p. 124) whereas "the United States authorities appear to assimilate all forms of govern- ment intervention with subsidies" (p. 123). As to an example of different inter- pretations of the relevant magnitude, under EEC practice, an export credit is a subsidy to the extent that "the rates charged are below those which the govern- ment have to pay for the funds." "The United States authorities, however, com- Finger and NoguAs 713 pute the subsidy by comparing what a company would pay a normal commercial lender with what it actually pays on the preferential loan" (Beseler and Williams 1986, p.136). The proponents of the Code realized when it was signed that it was not an agreement as to what constituted a subsidy, but they accepted it in the hope that an agreed meaning would evolve in practice. Through standardization of na- tional CVD procedures on the Code's guidelines, the international dispute settle- ment process and the various consultations that the Code provided, the international community might move toward a clearer understanding of what practices would be accepted. This has not happened. This lack of convergence is evident in the pattern of which countries initiate and which are the subject of CVD investigations. Table 2 shows that 91 percent (387 out of 425) of the CVD cases initiated in 1980-85 were initiated in two countries, Chile and the United States. Only one CVD case was initiated against the United States (in Australia) and none against Chile. Ninety-four percent (400 of 425) CVD cases were against countries which initiated no cases. Chile's cases were almost all against four countries; Argentina, Brazil, Peru, and Spain; and were nearly all initiated before March 1983, when Chile doubled its uniform tariff rate and devalued its currency.1 In brief, CVD actions are brought against a different group of countries than the countries which initiate them. Except for a brief and soon-abandoned period of activity by Chile, CVD actions are used almost exclusively by the United States. There is no balance in the international control of subsidies and of CVDs-no agreement in principle on what constitutes an improper practice, and no sem- blance of equality in the number of times the CVD instrument is applied to and by an individual country. Patrick Messerlin hardly overstates the case when he concludes, "To the United States, the Code is an instrument to control subsidies. To the rest of the world, it is an instrument to control U.S. countervailing duties" (Messerlin 1986, p. 16). II. HARASSMENT As no country wants to see its subsidies countervailed (or its exporters' dump- ing actions offset), application of such trade remedies often brings forward concerns about "harassment", applied with as much emotion to "appropriate" enforcement as to "excessive." Concerns that CVD procedures are used to "ha- rass" imports are often complaints that a CVD is applied against a subsidy which is not proscribed by the GATT or the Subsidies Code. As explained above, "legal subsidy" does not imply "illegal CVD," or vice versa. But not all such concerns can be put aside as misinterpretations of the relevant 1. Over the period 1980-85 only 1 of 135 CVD cases initiated in Chile reached an affirmative final determination. In private conversation Chilean trade officials have informed us that a major consider- ation in Chile's decision to abandon the use of CVDS was the threat of retaliation by trading partners against whom cases were brought. 714 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 Table 2. Countervailing Duty Cases: Initiating and Subject Countries, 1980-85 Cases Cases Country initiated against United States 252 1 Chile 135 0 Australia 19 2 Canada 11 9 European Community' 7 11 Japan 1 2 EEC member statesa 0 86 Brazil 0 84 Spain 0 39 Argentina 0 30 Mexico 0 28 Peru 0 21 Korea 0 18 New Zealand 0 14 Others (40) 0 80 a. EEC member states do not have CVD mechanisms; the CVD instrument exists only at the community level. Both member states and the community provide subsidies which have been the object of CVD investigations by other countries. Source: Tables compiled by the authors, available upon written request to them. legal documents. We will, in this section, develop several substantive interpreta- tions of harassment and will review relevant empirical evidence. Lower Thresholds for Provisional Restrictions When the U.S. government receives and accepts a countervailing duty peti- tion, it begins simultaneous but separate investigations to determine if the im- ported goods in question receive a subsidy and if the competing domestic industry is experiencing or threatened by material injury therefrom (table 3).2 Injury investigations in the United States are carried out in all cases concerning duty free goods and in those cases involving signatories to the Subsidies Code, Taiwan, and economies to whom the United States has extended, by treaty, unconditional most favored nation (mfn) status. In those CVD cases in the United States which do not include an injury test, the sequencing of the parts of a case are as they would be in table 3 with a positive injury determination at each stage. Such subsidy-injury investigation takes time to complete, and during this time injury to competing domestic producers might be compounded by a rush to ship in imports before the final decision is reached. To guard against this, U.S. law provides and the Subsidies Code allows for provisional measures, based on preliminary determinations of the facts of the case. The first determination that is made in a CVD case is the preliminary determination of injury. If this finding is negative, the case ends, with no resulting restriction. The next decision point is the preliminary subsidy determination. If this deter- 2. Use of "the United States" as the investigating country is for expositional convenience. The reader may substitute "EEC," "Canada;' "Chile;' or whatever country he or she wishes. Finger and Nogues 715 Table 3. Sequence of the Parts of a CVD Investigation in the United States Event Result Preliminary determinations 1. Injury Negative Case ends Positive Case continues 2. Subsidy Negative Case continues Positive Case continues and suspension of liquidation Final determinations 3. Subsidy Negative Case ends and suspension of liquidation is lifted Positive Case continues, suspension of liquidation continued or initiated 4. Injury Negative Case ends and suspension of liquidation is lifted Positive Case ends, CVD imposed if an arrangement is not agreed mination is negative, the case is not terminated. It continues to final determina- tion, but no import restrictions are imposed at the preliminary stage. However, a positive preliminary determination on subsidization brings with it "suspension of liquidation" of imports. Past the date, the importer must post a bond with the government to pay CVDS if the final determination is also positive. This is, in effect, a deposit requirement on imports, and we consider it to be a restrictive outcome (table 4). Positive determinations on both subsidy and on injury are necessary before a CVD may be imposed. Not every case ends with final or even preliminary determinations of subsidi- zation and injury. The law provides that a case may end at several points if a satisfactory "arrangement" is reached with the exporting country to end the subsidy practice in question or in some way to offset its effects. We have classi- fied such arrangements as "restrictive outcomes" in table 4. Preliminary determinations may allow administration of CVD law to restrict imports. While the rationale for a quick (though tentative) determination is to prevent an exporter from flooding the market and compounding injury, in guarding against this possibility of abuse by exporters, CVD procedures create the possibility of abuse by the administering agency. The administering agency might be subject to political pressure to restrict imports, and at the same time dependent on information supplied by the petitioner for preliminary determina- tions (because of short time limits). It is thus possible that preliminary determi- nations will be positive even though the final determination is not. We take the final determination to be the correct one. Under U.S. CVD law the accused foreign exporters (and domestic competitors) are protected from admin- istrative misapplication of the law by the right of appeal to the federal courts. (Relative familiarity with the legal system, and the relative costs to the United States and the foreign parties in a case are taken up in the second part of this section.) Table 4. Disposition of U.S. Countervailing Duty Investigations, 1980-85 Numbers of investigations Percentages of investigations Event With Without With Without injury test, injury test injury test injury test Combined Initiations 105 74 100 100 100 Preliminary disposition Restrictive, total 66 60 63 81 70 Affirmative subsidy determination (suspension of liquidation) 52 60 50 81 62 Alternative arrangements negotiated 14 0 13 0 8 Not restrictive, total 39 14 37 19 30 Negative injury determinationb 19 - 18 - 11 Negative subsidy determination, 10 8 10 11 10 Case withdrawn or terminated 10 6 10 8 9 Final disposition Restrictive, total 54 52 51 70 60 CVDS imposed 24 36 23 49 34 Affirmative subsidy determination (39) (43) Affirmative subsidy and injury determination (24) - Alternative arrangements negotiated 30 16d 28 21 26 Not restrictive, total 50 22 48 30 40 Negative subsidy determinatione 8 10 8 14 10 Negative injury determination 29 - 28 - 16 Withdrawn or terminated 13 121 12 16 14 Pending I 0 1 0 0 -Not applicable. a. Material injury investigations are carried out in all cases concerning duty free goods, and in those cases involving signatories to the code of subsidy/CVD measures, Taiwan, and the economies to whom the United States has extended by treaty unconditional mfn status. b. Investigation ends. There is no preliminary subsidy determination in such cases. c. Investigation continues without suspension of liquidation. d. Includes six agreements reached after an affirmative final decision. e. If the final subsidy determination is negative, there is no final injury determination. f. Includes one case withdrawn after an affirmative subsidy determination. Source: Tabulated from U.S. Trade Representative, "Trade Actions Monitoring System,' and U.S. Federal Register, various issues. Finger and Nogues 717 If the risk is high that a suspension of liquidation order will be in effect when goods are delivered, then buyers will begin to shift orders to domestic goods immediately upon initiation of an investigation. With a high probability that a preliminary determination will impose what is, in effect, a deposit requirement on imports, the legal cost of a CVD or antidumping (AD) petition to a U.S. industry might be less than the expected gain from sales shifted away from imports, even when calculations assume a zero probability of an affirmative final decision. Exporters, aware of this situation, would find attractive a negotiated solution (a voluntary export restraint, VER) which bought off filing of the peti- tion or included its early withdrawal. While the validity of the hypothesis that preliminary determinations are biased toward affirmative findings awaits precise evaluation, table 4 provides evidence relevant to several elements of the hypothesis. Are preliminary determinations biased toward affirmative outcomes, as com- pared with final determinations? A negative preliminary determination of injury ends a case, hence the preliminary and the final disposition of such a case is "not restrictive" (in table 4). A negative preliminary determination of subsidization is a "not restrictive" preliminary disposition, in that no "suspension of liquidation" is imposed, but the case does not end, and the final subsidy determination may be affirmative. Hence, the proportion of cases which have a restrictive final determination could be higher than the proportion with a restrictive preliminary determination. Table 4 shows, however, that preliminary determinations are more often restrictive than are, final determinations-in 70 percent versus 60 percent of investigations. It appears that U.S. domestic producers have had great success in obtaining affirmative decisions concerning the existence of subsidies on imported goods (table 5). Of 130 cases, only 18 (14 percent) received an initial negative determi- nation concerning subsidies. Despite the negative preliminary determinations, 5 of these cases (28 percent) ended with a restrictive arrangement at the next level of investigation. In most cases (87 percent), an affirmative preliminary determi- nation of the existence of a subsidy is upheld in the following determination. Table 5. Comparison of Preliminary and Final Subsidy Determinations (number of cases) Preliminary determination Final determination Affirmative Negative a. Affirmative determination 80 2 b. Negative determination 5 13 c. Restrictive arrangement 18 3 d. Not restrictive determination or withdrawal 9 0 Totals 112 18 Reversals Not restrictive outcome (b + d) 14 (13%) Restrictive outcome (a + c) S (28%) Source: As in table 4, based on authors' calculations. 718 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO.4 After final determination is reached on the existence of subsidy, the case may move to a final determination on injury. Only 20 percent of preliminary positive injury determinations are reversed. Since a negative preliminary determination of injury ends a case, those cases do not contribute to the negative findings in the final state. These reversals on positive preliminary injury determinations, com- bined with the negative final determinations on subsidies, therefore makes the proportion of restrictive outcomes lower at the final than at the preliminary stage. Data in table 4 also suggest the benefit provided to an exporting country by the injury test in limiting the number of restrictive outcomes. Thirty percent of cases without an injury test came to a nonrestrictive final determination, as compared with 48 percent of cases which included an injury determination. And in the latter cases, the negative injury determination accounted for nearly 80 percent of the nonrestrictive final outcomes, while the negative subsidy findings accounted for only 20 percent. Uncertainty and the Cost of Procedures Another perspective on harassment focuses on the uncertainty generated by the likelihood that a successful export program will be subjected to a CVD complaint, and on the relative costs of such a complaint to foreign exporters versus competing domestic firms. This sort of concern is not specific to CVD petitions but applies to all forms of administered protection, including anti- dumping and safeguards actions. From the point of view of a foreign firm interested in developing a U.S. market for its product, the threat of an administered protection (AP) petition by its U.S. competitors will increase the cost of a project to develop a U.S. market since, aside from the legal and administrative expenses to respond to the peti- tion, an additional element of risk is added to the project-the possibility of a petition and its anticipated outcome. It is not immediately apparent, however, that the cost increase will put the foreign firm at a disadvantage relative to U.S. firms. The latter will incur legal and administrative expenses in preparing and advancing their petitions. There are, however, subtle ways in which this mutual increase of costs might be to the advantage of domestic firms. First, imports account for much less than half of sales in U.S. markets, hence the foreign firm's additional expense will be distributed over fewer units than will those of the U.S. firms. Second, there may be economies of scale or economies of learning-by-doing in filing and in re- sponding to such petitions, and these economies are more likely to be captured by a domestic firm or industry group which files petitions against exports from several different countries than by firms or groups in each of the foreign coun- tries. Two factors have a bearing on whether or not the threat of an AP complaint will change the expected revenue generated by a foreign firm's project to develop a U.S. market. Whether or not the foreign firm plans to benefit from an export Finger and Nogues 719 subsidy is one factor, and the other is whether or not the foreign firm views the AP case decision processes in the United States as biased. Suppose the foreign firm views the decision processes as unbiased-it expects that a CVD will just offset an export subsidy, or a dumping duty will just offset a dumping margin. Even if the foreign firm's plan does not include dumping or the receipt of an export subsidy, however, the possibility of a CVD or AD complaint will affect the expected receipts the project will generate since this possibility will add to the riskiness of the project. In sum, it is not necessary to assume that the CVD or AD decision process is biased in order to argue that the overall investigative mechanism tends to be protectionist. Per unit costs of filing and responding to complaints are likely to be higher on imports than on domestic import replacements. Further, the mecha- nism adds to the riskiness of the expected revenue to a foreign firm from selling in the U.S. market and this would, other things being constant, tend to reduce such sales. Econometric investigations of the uncertainty generated by import relief peti- tions face the challenge of untangling the relevant cause-effect relationships. The more intense is import competition, such studies hypothesize, the more one would expect domestic industries to file petitions for protection. But if com- plaints tend to impose larger costs and uncertainties on import sales than on sales by domestic producers, then there will be a simultaneous negative relation- ship between import growth and the incidence of complaints. An econometric identification of the hypothesized negative (that is, harassment) effect of AD and CVD complaints on U.S. import growth in a study by Finger (1981) covering all AD and CVD cases filed in the United States under the 1974 Trade Act, 1974-79, found the effect to be statistically significant. In a cross-industry two-stage re- gression model, the rate of growth of imports 1974-77 was significantly corre- lated with the "complaints index," the percentage of 1974 industry imports covered by AD and CVD complaints. As to the magnitude of this effect, if AD and CVD complaints covered 1 percentage point more of one industry's imports than of another's, then, other factors being equal, the first industry's annual import growth rate would be about 0.2 of a percentage point lower than the others. Biased Economic Content of the Instruments Though the legal objective of the AD and CVD mechanisms is to police fairness of trade practices, they pursue that objective by restricting imports. While the instruments were designed to impose restrictions against unfair exporters only, they will attract not only firms and industries beset by unfair competition but also, more generally, those least favorably situated vis-a-vis their foreign compet- itors. Such agents, the logic of economic anthropology suggests, will attempt to make their needs fit the prerequisites of the instruments and will pressure to have the instrument changed to fit their needs. In corroboration, Finger, Hall, and Nelson's (1982) analysis of the factors associated with a positive dumping or subsidy determination found that among the factors examined, those which 720 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 related to comparative costs were the most powerful predictors of the outcome of antidumping and countervailing duty cases. The legal mechanics of such bias have been identified in studies of antidump- ing procedures.3 Dickey, in a study of U.S. antidumping administration (1979), explains several ways in which administrative practice is more restrictive of foreign firms' than is comparable "domestic law" of U.S. firms' sales practices. For example, sales below full cost during periods of slack demand are normally not considered an unfair trade practice under U.S, antitrust laws. But the anti- dumping law specifies that in determining the home market price of a foreign firm, observations of home market sales below cost must be disregarded (p. 245). Dumping, in concept, is selling for export at a lower price than one sells in the home market. The intent in a dumping investigation is to compare the two prices, net of selling and distribution costs, received by the seller on export and on domestic sales. Differences normally observed on sales of different amounts are usually taken into account. But prices which are normally observed include many elements of distribution and selling costs, and adjustments must be made to ensure that the comparison is made "at the same level." Christopher Norall (1987) has examined the procedures used by the Commission of the European Community to make such adjustments. He found that particularly when a for- eign firm's home market sales are made through a vertical chain of a distributor and/or retailer owned by the manufacturing firm, "various aspects of the techni- cal methodology now applied by the Commission in antidumping cases tend to make findings at significant levels automatic and inevitable and, secondly, to make it very difficult for exporters to modify or palliate the effect of antidump- ing duties by price increases after the imposition of the duties" (p. 98). Publicizing a Case for Protection: Precedent to VERS Finally, administrative mechanisms which enforce trade law may be used by a domestic industry to build a public case for protection. Filing an unfair trade practice petition is a more newsworthy event than presenting evidence of import competition, and the U.S. Congress is not likely to act to protect an industry or to apply pressure on the U.S. administration to protect it unless all routine means have been exhausted. In practice, administrative mechanisms are the "outer office" through which complaints must pass if they are to gain access to the ultimate political authority behind them. While there is always the risk that an unjustified allegation of foreign unfair trade practices will set back an indus- try's campaign to win protection, there are two reasons to discount this risk. First, at a technical level, criteria for unfair trade practices overlap considerably with the determinants of comparative disadvantage (Finger, Hall, and Nelson 1982). Second, the campaign at the level of public affairs might not be exactly parallel to the technical petition which is filed. Large cases, the facts show, almost always begin as administered protection 3. There has been no parallel research on CVD procedures that the authors could locate. Finger and Nogues 721 cases and end up as negotiated "voluntary" export restraints (VERS) (Finger, Hall, and Nelson 1982). The United States-Japan agreement on Japanese ex- ports of automobiles was preceded by a safeguards case. The network of VERS the United States has negotiated with several steel exporters had as its antecedent a series of antidumping and CVD petitions. III. EFFECTS ON THE UNITED STATES It is widely accepted in influential circles in the United States that the U.S. market is beset by foreign competitors that have an unfair advantage over U.S. firms. In 1980-81, twenty-six different pieces of legislation were proposed in the U.S. Congress to deal with such matters (Weiss 1982, appendix I), and by the end of 1985 several hundred pieces of trade legislation had been proposed. Many of these had the same intention: to widen the definition of foreign prac- tices that are "unfair" and to restrict the U.S. president's discretion to not take action when a U.S. government investigation found that an unfair practice was being employed. The number of such bills attests to the political popularity of this concern to offset unfair foreign practices-to establish a "level playing field." Foreign subsidization of goods sold in the U.S. market is one of the practices most widely accepted in the United States as being unfair, and opposition to it is good domestic politics. It is, however, bad domestic economies. Economists demonstrated more than two centuries ago that import restric- tions almost always subtract more from than they add to the national economic interest of the country which imposes them. These gains from trade stem from international prices being different from national costs. The economics of "if you can buy it for less than you can build it, you are better off to buy it" applies to a nation as well as to an individual. The exception to the generalization that import restrictions are contrary to the imposing country's national economic interest is the policing of predatory pric- ing policies by foreign firms. Predatory pricing is pricing, perhaps below cost, (supported by subsidies) which would drive domestic firms out of business and leave foreign sellers in a quasi-monopoly position. Foreign sellers would then exploit their market power, as did the domestic firms which monopolized U.S. markets in the late nineteenth and early twentieth centuries through predatory pricing. The United States would, over the long run, then end up paying more for the product than it would have if it had applied CVDS to defend domestic producers. It is not likely, however, that predatory pricing in today's international mar- kets could have this effect on the United States. The customer's best defense against predatory pricing is the availability of an alternative supplier. Likewise, a producer's best defense against being put out of business by the aggressive pric- ing of a second firm is that competition from many others will prevent the second one from seeing such below-cost pricing as a sensible business undertak- ing. 722 THE WORLD BANK ECONOMIC REVIEW, VOL. 1, NO. 4 In an open international trading system, there is little likelihood that preda- tory pricing, particularly by a small country, will eventually deprive another country of access to a particular product at a reasonable price. Even if the "predator" drives one supplier out of business, an abundance of alternative sources will remain. Yet two-thirds of U.S. CVD cases against manufactured products during 1980-85 were on carbon, steel and textiles, products available from a multitude of sources (Nam, this issue, table 4). U.S. CVD cases have been brought against forty-seven different countries, but nearly a third of them have been against Brazil, Mexico, or Spain. The large number of countries named in CVD cases, and their concentration in countries whose comparative advantage is not in industries where the number of alternative sources might be low, suggest that the trade restrictions resulting from these cases are not necessary to protect U.S. buyers from future exploitation. IV. APPLICATIONS OF U.S. LAW TO DEVELOPING COUNTRIES' EXPORTS Having examined the effects of U.S. CVDS on the interests of the United States, we look at the other side of the issue. Are the policies which the United States countervails worth fighting for? Do they augment the national economic interest of the countries which apply them? While our focus is on developing countries' policies generally, our sample is limited to the policies of Argentina, the Republic of Korea, Mexico, and Peru. A number of comparisons of developing economies have demonstrated that countries with relatively open trade policies have had better growth, better employment performance, and better income distribution performance than countries which have followed more inward trade policies (Little, Scitovsky, and Scott 1970; Bhagwati 1978; Krueger 1978, 1983; Balassa 1971). The trade regimes of the more successful countries-regimes described as Open Trade (OT) policies-have tended to be "neutral" in the sense that the amount of local currency actually received or paid per unit of foreign exchange is approximately the same for exports as for imports. Establishing such a policy regime often required the active involvement of government, to identify and scrap or neutralize disincentives to produce for export, and disincentives to use imported inputs when they were cheaper, for example. The shift away from an import substituting policy was often less a reduction of policy-based incentives for the expansion of import substituting production, than an elimination of biases in those incentives. "Policies in the successful countries have been generally supportive of industrialization and com- merce but have avoided directing that support at any particular sector or method. Decisions about what activities and what processes could be efficiently and profitably built up are left to individual firms, which succeed or fail as their decisions prove to be correct or incorrect" (World Bank 1981, p. 25). This, then, is the second dimension of neutrality in the trade regimes of successful develop- ing countries-neutrality across sectors in the availability of incentives. Korea, Finger and Nogues 723 an often cited example of success through outward orientation, ran into difficul- ties when the government attempted specifically to promote the heavy machin- ery and chemical industries. The World Development Report 1986 reviews problems a number of developing countries have also created through incentives biased against the agricultural sector. Table 6 evaluates the subsidies of Argentina, Mexico, and Perui which have been countervailed by the United States, relative to the neutrality of their trade policies. It is hard to argue that the export policies which led to these CVD actions contributed to an effective development policy. First, they violated the "neutral- ity between sectors" criterion. Traditional exports were not eligible and subsidies were provided to only a fraction of manufactured goods (Nogues 1986b). Even among the sectors judged eligible, the subsidy rates varied considerably. Second, the figures show average tariffs to be significantly higher than the CVD rates imposed. In addition to tariffs, Argentina and Mexico in 1982 and Peru in 1986 introduced across-the-board import licensing regimes. Third, the relative magni- tudes of subsidy rate and real exchange rate movements make it impossible to present these export subsidies as a compensation for changes in the real ex- change rate-in Peru, for example, an average 25 percent subsidy rate was adopted on only 3 percent of exports in the face of a 210 percent appreciation of the real exchange rate. These subsidies could not offset inappropriate exchange rates. To the extent that they were assumed to do so, or diffused political pressure to make necessary exchange rate adjustments, they did a particular disservice to the national economic interest. While Argentina's, Mexico's, and Peru's experiences indicate that the policies countervailed were mostly inconsistent with an open trade strategy, following an open trade strategy certainly does not mean that a country will escape CVD actions. Nam found, for example, that in 1980-85 the United States counter- vailed against fourteen different applications of Korean policies (this issue, table 5). But, the policies countervailed in Korea tended to provide lower and more uniform subsidy rates than those just reviewed. Against Korea, CVD rates were, with only one exception, all less than 1.0 percent. These rates document the uniformity (the intersectoral neutrality) of incentives in a successful exporting country. The low subsidy rates suggest that the administrative costs and uncer- tainty created by the threat of the CVD process may have more impact on trade flows than the final outcome of any CVD action taken. V. CONCLUSIONS AND RECOMMENDATIONS International control of subsidies and of countervailing duties is a perplexing subject area. When one looks at the details of CVD law and the Subsidies Code, things seem systematic and therefore, in the mentality of our times, sensible. But, when one takes an overview, as in the past two sections, they are not so. Export subsidies, whether or not a country's tra, partners are expected to countervail, are much more likely to compromit tional economic interest Table 6. U. S. Countervailing Duty Actions against Argentina, Mexico, and Peru Compared with Tariff and Exchange Rate Movements Percentage of exports CVD actions, 1 982-84 Import tariffs' to United States subject Real exchange Rates (percent) (percent) to CVDb rate variation, Simple Simple All 1978-85c Country Number average Range average Range products Manufactures (percent) Argentina 8 7 1-16 32 0-41 9 22 320 Mexico 20 1S 1-105 23 0-100 1 5 180 Peru 5 25 20-44 41 0-75 3 43 210 a. Figures correspond to legal tariff rates and do not include other charges paid at customs. For Argentina and Peru, the figures are representative of the 1982-84 period; for Mexico, they correspond to 1984. In this last case and in previous years, tariff rates were previously higher. b. U.S. imports for 1984 covered by 1982-84 CVD actions as a percentage of imports of the indicated base in 1984. c. The underlying real exchange rate series are measured as the nominal exchange rate times the ratio of the U.S. wholesale price index to the domestic consumer price index. The figures show the proportional variation of the maximum to the minimum real exchange rate between 1978 and 1985. Source: For Mexico and Peru, Nogues (1986a, 1986b); for Argentina, unpublished tabulations available upon written request to the authors. Finger and Nogues 725 than to augme it. ountervailing duties, whether or n_4 country's trading partners subsi re m more likely to compromis