Issues in US-EC Trade Relations 9780226036526

Citation preview

Issues in US-EC Trade Relations

A National Bureau of Economic Research Conference Report Centre for European Policy Studies

Issues in US-EC Trade Relations

Edited by

fi* 32*

Robert E. Baldwin, Carl B. Hamilton, and Andre Sapir

The University of Chicago Press Chicago and London

ROBERT E. BALDWIN is the Hilldale Professor of Economics at the University of Wisconsin, Madison. CARL B. HAMILTON is with the Institute for International Economic Studies at Stockholm University, Sweden. ANDRE SAPIR is professor of economics at the Free University of Brussels, Belgium.

The University of Chicago Press, Chicago 60637 The University of Chicago Press, Ltd., London © 1988 by The National Bureau of Economic Research All rights reserved. Published 1988 Printed in the United States of America 97 96 95 94 93 92 91 90 89 88 5 4 3 2 1

Library of Congress Cataloging-in-Publkation Data Issues in US-EC trade relations / edited by Robert E. Baldwin, Carl B. Hamilton, and Andr6 Sapir. p. cm. — (A National Bureau of Economic Research conference report) Proceedings of a conference, June 12-14, 1986, Brussels, Belgium; sponsored by Centre for European Policy Studies and National Bureau of Economic Research. Bibliography: p. Includes index. ISBN 0-226-03608-1 1. United States—Foreign economic relations—European Economic Community countries—Congresses. 2. European Economic Community countries—Foreign economic relations— United States—Congresses. I. Baldwin, Robert E. II. Hamilton, Carl. III. Sapir, Andre. IV. University of Chicago. Press. V. Centre for European Policy Studies (Louvain-la-Neuve, Belgium) VI. National Bureau of Economic Research. VII. Series: Conference report (National Bureau of Economic Research) HF1456.5.E825I87 1988 382'.0973'04—dcl9 88-1515 CIP

National Bureau of Economic Research Officers Richard N . Rosett, chairman George T. Conklin, Jr., vice-chairman Martin Feldstein, president and chief executive officer

Geoffrey Carliner, executive director Charles A. Walworth, treasurer Sam Parker, director of finance and administration

Directors at Large Moses Abramovitz John H. Biggs Andrew Brimmer Carl F. Christ George T. Conklin, Jr. Kathleen B. Cooper Jean A. Crockett George C. Eads Morton Ehrlich

Martin Feldstein Edward L. Ginzton David L. Grove George Hatsopoulos Franklin A. Lindsay Paul W. McCracken Geoffrey H. Moore Michael H. Moskow James J. O'Leary

Robert T. Parry Peter G. Peterson Robert V. Roosa Richard N . Rosett Bert Seidman Eli Shapiro Donald S. Wasserman

Directors by University Appointment Charles H. Berry, Princeton James Duesenberry, Harvard Ann F. Friedlaender, Massachusetts Institute of Technology Jonathan Hughes, Northwestern J. C. LaForce, California, Los Angeles Marjorie McElroy, Duke Merton J. Peck, Yale

James L. Pierce, California, Berkeley Andrew Postlewaite, Pennsylvania Nathan Rosenberg, Stanford Harold Shapiro, Michigan James Simler, Minnesota William S. Vickrey, Columbia Burton A. Weisbrod, Wisconsin Arnold Zellner, Chicago

Directors by Appointment of Other Organizations Richard Easterlin, Economic History Association Edgar Fiedler, National Association of Business Economists Robert S. Hamada, American Finance Association Robert C. Holland, Committee for Economic Development James Houck, American Agricultural Economics Association David Kendrick, American Economic Association

Rudolph A. Oswald, American Federation of Labor and Congress of Industrial Organizations Douglas D. Purvis, Canadian Economics Association Albert T. Sommers, The Conference Board Dudley Wallace, American Statistical Association Charles A. Walworth, American Institute of Certified Public Accountants

Directors Emeriti Emilio G. Collado Solomon Fabricant Frank W Fetter

Thomas D. Flynn Gottfried Haberler

George B. Roberts Willard L. Thorp

Since this volume is a record of conference proceedings, it has been exempted from the rules governing critical review of manuscripts by the Board of Directors of the National Bureau (resolution adopted 8 June 1948, as revised 21 November 1949 and 20 April 1968).

Centre for European Policy Studies Officers Jean Frangois-Poncet, chairman Theo Peters, deputy chairman Martin Kallen, chairman of the finance committee

Peter Ludlow, director Johannes Krul, financial director

Programme Leaders (1986): Macroeconomic and Monetary Affairs Lars Calmfors, Institute of International Economics, Stockholm Jacques Dreze, Core, Universite Catholique de Louvain

Niels Thygesen, University of Copenhagen

Industry and Trade Henry Ergas, OECD, Paris Alexis Jacquemin, Universite Catholique de Louvain

Andre Sapir, Universite Libre de Bruxelles

Public Finance Luigi Spaventa, University of Rome Agriculture and Natural Resources Knud Jrirgen Munk, University of Aarhus Euro-South Relations Christopher Stevens, CEPS Research Advisory Committee (Economic and Social Affairs) Armin Gutowski, Chairman, HWWA, Institut fur Wirtschaftsforschung, Hamburg Richard Cooper, Center for International Affairs, Harvard University Paolo Fasella, Director-General, Directorate-General for Science, Research and Development, Commission of the European Communities Jacques Lesourne, Conservatoire Nationale des Arts et Metiers, Paris

Stephen Marris, Institute for International Economics, Washington Sylvia Ostry, Canadian Ambassador for Multilateral Trade Negotiations I. G. Patel, Director, London School of Economics and Political Science Jean Waelbroeck, Universite Libre de Bruxelles

Contents

Preface 1. An Introduction to the Issues and Analyses Robert E. Baldwin

xi 1

I. THE LEGAL FRAMEWORK

2. Legal Issues in US-EC Trade Policy: GATT Litigation 1960-1985

17

Robert E. Hudec Comment: Per Magnus Wijkman Comment: J. P. Hayes II. AGRICULTURE: TRADE AND PROTECTION

3. The Price and Welfare Implications of Current Conflicts between the Agricultural Policies of the United States and the European Community Dermot Hayes and Andrew Schmitz

67

Comment follows chapter 4 4. EC-US Agricultural Trade Confrontation Alexander H. Sarris Comment: Dieter Kirschke

vii

101

viii III.

Contents EMBARGOES AND STRATEGIC TRADE ISSUES

5. Strategic Trade, Embargoes, and Imperfect Competition Henryk Kierzkowski

135

Comment follows chapter 6 6. East-West Trade, Embargoes, and Expectations Alasdair Smith

153

Comment: L. A. Winters IV.

INDUSTRY: N E W PROTECTIONISM AND N E W COMPETITORS

7. The Steel Crisis in the United States and the European Community: Causes and Adjustments David G. Tarr

173

Comment follows chapter 8 8. Restrictiveness and International Transmission of the "New" Protectionism Carl B. Hamilton

199

Comment: Juergen B. Donges V. TRADE IN SERVICES

9. International Trade in Telecommunications Services Andre Sapir

231

Comments follow chapter 10 10. International Trade in Banking Services

245

C. R. Neu Comment: Frances Ruane Comment: Gary P. Sampson VI.

TRADE POLICY IN OLIGOPOLISTIC ENVIRONMENTS

11. High-Tech Trade Policy Kala Krishna Comments follow chapter 12

284

ix

Contents

12. Countervailing Duty Laws and Subsidies to Imperfectly Competitive Industries

313

Barbara J. Spencer Comment: Henry Ergas Comment: Harry Flam VII.

INTERACTION BETWEEN THE MACROECONOMIC ENVIRONMENT AND TRADE ISSUES

13. Macroeconomic Policy and Trade Performance: International Implications of U.S. Budget Deficits Rachel McCulloch

349

14. Economic Cooperation and Confrontation between Europe and the U.S.A.: A Game-Theoretic Approach to the Analysis of International Monetary and Trade Policies 369 Giorgio Basevi, Paolo Kind, and Giorgio Poli List of Contributors

389

Author Index

391

Subject Index

393

Preface

This volume is the result of a conference held jointly by the Centre for European Policy Studies (CEPS) and the National Bureau of Economic Research (NBER) in Brussels, Belgium, from 12 June to 14 June 1986. We wish to thank the staff of CEPS for their efforts in helping to make the conference successful. In presenting papers by North American and European trade specialists with theoretical, empirical, and institutional analyses of some of the major trade issues on which American and European Community officials disagree, the conference aimed to provide a better understanding of the economic causes of the disagreements and to consider various policy options that might assist in resolving these differences. The analyses are also informative for trade policy disputes generally, since many United States-European Community disputes are representative of disagreements within the entire trading system. We are grateful to the Ford Foundation for providing financial support for the conference, for a preconference authors' meeting, and for the publication of this volume. Sapir was largely responsible for organizing the conference and, together with Hamilton, assumed editorial responsibility for the papers by the European authors and comments on all the papers. Baldwin handled the North American papers. We wish to thank Annie Spillane and Mark Fitz-Patrick of the NBER for their efficient help in preparing the papers and comments for submission to the University of Chicago Press, and Donna Zerwitz of the NBER and Nancy Klatt of the University of Wisconsin for their editorial assistance. Tom Bayard of the Ford Foundation and Geoffrey Carliner of the NBER deserve special thanks for their valuable advice on the issues that should be analyzed and the format of the conference. We also wish to give particular thanks to Jean Waelbroeck for his support and advice xi

xii

Preface

throughout the project and for his traveling halfway around the world to present an overview of the conference. Robert E. Baldwin, Carl B. Hamilton, and Andre Sapir

An Introduction to the Issues and Analyses Robert E. Baldwin

1. The Importance of US-EC Trade Relations A minimum requirement for a viable trading system is the active support of both the United States and the member countries of the European Community (EC). This is not only because their trade makes up about one-half of total world trade but because the United States and members of the EC have been the main architects and supporters of the post-World War II international trading regime. Other major trading groups have generally been willing to accept the leadership of the United States and the EC in initiating multilateral negotiations aimed at reducing protection and modifying the rules of the General Agreement on Tariffs and Trade (GATT). In recent years a series of United States-European Community (USEC) disagreements have developed that threaten the degree of consensus between these two trading blocs that is necessary for the maintenance of a stable international trading order. They are on such diverse matters as the consistency with current GATT rules of particular actions taken by one of the parties, the need for new rules and for changes in existing rules to cover forms of trade not now subject to GATT discipline, the adequacy of present dispute settlement procedures, the proper agenda and procedures in new multilateral trade negotiations, and the relationship of trade policies to balance-of-trade deficits. As the some 90 members of the GATT embark on a new multilateral trade negotiation, the Uruguay Round, the United States and the EuRobert E. Baldwin is the Hilldale Professor of Economics at the University of WisconsinMadison, a Research Associate of the National Bureau of Economic Research, and director of the NBER's Trade Relations project.

1

2

Robert E. Baldwin

ropean Community face an historic challenge. They can either use the occasion to move toward the resolution of their disputes and thereby strengthen the trading system or they can adopt inflexible negotiating positions with the likely result that, as others follow suit, the negotiations become the occasion for a further weakening of the rules and arrangements of the world trading system. Fortunately, the Ministerial Declaration adopted in Punta del Este in September 1986 gives some promise that the first course will be followed. Besides agreeing upon a standstill and rollback of trade-restrictive measures inconsistent with the GATT, the participants included in the agenda as subjects for negotiation most of the issues on which US-EC disagreements have arisen. The purpose of this volume is to facilitate the resolution of the two blocs' present differences by analyzing some of the most important issues of disagreement and considering alternative policy options to reduce tensions and lessen the risks of a breakdown of the trading system. In carrying out this objective, emphasis is placed on utilizing appropriate combinations of historical, theoretical, and empirical analyses. Each general subject is analyzed, in most cases, from both an American and a European perspective. Of course, many US-EC disputes cover matters on which there is widespread disagreement within the entire trading system. Thus, the theoretical analysis in most of the papers, though not the institutional detail, is also relevant for studying trade policy in general, nonregional terms. 1.2 US-EC Litigation in the GATT Under the GATT, if one member considers that any benefits accruing under the Agreement are being nullified or impaired as a result of the actions of another member, that member can request consultations with the other party to resolve the problem. Should the dispute not be settled through this procedure, the complaining party can request that a panel of appointed experts make a report to the general membership with their judgment on whether the GATT rules have been violated. The general membership then decides whether or not to accept the judgment of the panel. An appropriate place to begin an analysis of US-EC disputes, therefore, is to examine the nature and frequency of cases between these two parties that have been brought before such panels. In part 1 of this volume, "The Legal Framework," Robert Hudec (chapter 2) examines the 80 GATT "lawsuits" filed between 1960, when the European Community became a full participant in GATT legal affairs, and 1985. Hudec finds that almost one-third of all the GATT lawsuits (26 of 80) during this period were between the United States and the Community. Furthermore, 45 of the remaining 54 cases in-

3

An Introduction to the Issues and Analyses

volved either the United States or the European Community as one of the parties. The volume of US-EC litigation has increased in recent years, but so too has the volume of GATT litigation in general. Most of the GATT lawsuits (43 of 80) have involved trade in agricultural products, and in most of these (25 of the 43), the EC has been the target of the complaint. Complaints about subsidies, both export and domestic production subsidies, and about tariffs, including the Community's variable levy on agricultural imports, dominate the list. Hudec concludes that the United States has litigated in the GATT mainly to satisfy certain domestic political imperatives, while the Community has litigated primarily for defensive purposes, without really believing in the process. Nevertheless, he believes the lawsuits between the two trading blocs have provided a peaceful alternative to real economic warfare. Although he thinks there is some reason to wonder whether GATT litigation can retain political credibility, he is generally optimistic that political leaders will continue to strengthen the dispute settlement procedures of the GATT. 1.3 Current Issues: Agriculture, Embargoes, and Declining Industries As Hudec's analysis demonstrates, the leading area of dispute between the United States and the European Community is agricultural policy. Much of the success of the Uruguay Round is likely to be judged on the extent to which these two parties resolve their differences on agricultural trade relations. In part 2, 'Agriculture: Trade and Protection," Dermot Hayes and Andrew Schmitz (chapter 3) and Alexander Sards (chapter 4) tackle this difficult issue and propose policies for dealing with the trade problems that have arisen. Both Hayes/Schmitz and Sarris agree that the Community's Common Agricultural Policy (CAP) entails heavy economic costs both to the United States and the Community and substantially distorts world agricultural markets. Hayes and Schmitz also show, however, that recently adopted U.S. farm legislation has long-run implications that are surprisingly similar to those of the CAP. Having described the Community's CAP and the U.S. agricultural policy, especially the Food Security Act of 1985, and shown how these policies have led to an international price war, Hayes and Schmitz propose specific policy changes they consider both politically feasible and welfare-enhancing. Under their proposal, all production would be sold at whatever price the market would yield, but through the use of a per-unit production subsidy, the government would ensure a certain reference income to those farmers whose operations are at the size that it most wants to support. Smaller, less efficient farmers would receive

4

Robert E. Baldwin

an income that is less than the reference income but more than from their market sales, while larger, more efficient units would receive at least the reference income, either from their market sales or the government. Hayes and Schmitz believe that this alternative to present agricultural policies would help alleviate the world oversupply situation in agriculture by shifting producers' emphasis from output-increasing technology to cost reduction and output-price enhancement. Sarris reviews US-EC agricultural policies from a European perspective. He outlines a simple model with random demand and supply shocks that is designed to capture the key economic features of the trade in grains between the two blocs and uses the model to estimate empirically the effects of actual and alternative US-EC grain policies. Sarris focuses in particular on three policy options available to the United States to offset some of its economic losses resulting from present EC policies. One is to take advantage of U.S. monopoly power by imposing an optimal export tax on grains; the second is to institute an optimal buffer stock scheme; and the third is to inflict a budget loss on the EC by introducing an export subsidy on U.S. grains. He concludes from his empirical analysis that an optimal export tax would more than compensate the United States for its CAP-induced losses but suggests that such a response is probably not politically feasible, since the U.S. Treasury rather than the U.S. farmer would be the big gainer. He also finds that an export subsidy is likely to prove too costly for the United States, since to inflict a $100 million annual budgetary loss on the Community would cost the U.S. Treasury $530 million. The best alternative, in his view, is the use of a government stockpiling policy. By buying for storage when world prices are low and selling when they are high, the U.S. Treasury is not hurt, producers benefit when they need it most, and consumers get more stable prices. While agricultural issues dominate the list of US-EC trade disputes, there are several areas of trade in manufactured goods and services where disagreements between the two trading blocs have arisen. Part 3, "Embargoes and Strategic Trade Issues," examines their differences over the use of embargoes as a means of inducing foreign countries to change a particular political action. A recent example is the U.S. embargo, introduced after the imposition of martial law in Poland in 1981, on sales by U.S.-based firms and their affiliates in foreign countries of equipment for use in building the Soviet gas pipeline from Siberia to Western Europe. Henryk Kierzkowski (chapter 5) and Alasdair Smith (chapter 6) point out that the difference in views between Western Europe and the United States on the wisdom of imposing embargoes has been evident on many other occasions. An imperfectly competitive framework is especially suitable for analyzing the embargo issue, and both Kierzkowski and Smith utilize this

5

An Introduction to the Issues and Analyses

approach. In doing so, both conclude that embargoes generally are not very effective in carrying out their intended purpose. Kierzkowski focuses on the following question: Can protection of a domestic industry considered to be strategic be justified when there is the possibility of an export embargo by a foreign producer? Since strategic industries often have a relatively small number of firms, to analyze this question he utilizes a model in which imperfect competition prevails. He demonstrates that, in the extreme case where open competition results in the product being produced in only one country, it could be advantageous for the importing country to produce the product for itself under import protection rather than risk the loss of the product because of a foreign embargo during the time needed to establish domestic production. Yet, when this extreme case is set aside and there are at least a domestic and a foreign firm supplying the domestic market, he finds that a foreign embargo cannot deal a devastating blow to the domestic economy. Strategic interdependence, as he terms the latter case, is, therefore, a far better state of affairs for a country than strategic dependence, as he terms the former situation. He argues, however, that strategic interdependence may be achieved without sacrificing efficiency by liberalizing foreign investment and providing for the freedom of establishment. As was demonstrated in the pipeline embargo case, a foreign monopoly operating within the frontiers of a country is less likely to deny goods and services to the host country even if ordered to do so by its home government. Smith presents a model of multilateral investment to focus on the embargo issue from a somewhat different viewpoint. How does the fact that the home-country government may impose an export embargo and thus reduce the profits of the multinational affect the company's decision to invest abroad? He assumes that it is also possible for a host-country firm to produce the good for its own market. Among the possible outcomes in this duopoly situation, two cases are of particular interest for the embargo issue. In one case, the multinational will choose not to invest abroad but instead to export its product to the foreign country if there is no threat of an export embargo by its government, but if this threat is strong enough, it will undertake foreign investment and thereby make the embargo ineffective. In the second case, the embargo threat is not strong enough to induce foreign investment by the multinational, but when the embargo is introduced, it becomes profitable for the host-country firm to undertake production and thereby make the embargo ineffective. As a practical policy issue, Smith concludes from his historical and theoretical analysis that, in most instances, embargoes are quite unlikely to succeed. In both the United States and the European Community there are a number of industries facing severe competition from third-country

6

Robert E. Baldwin

sources. The attempts to adjust to these new circumstances have led to US-EC disputes with these third countries and with each other. Part 4, "Industry: New Protectionism and New Competitors," examines certain aspects of protectionism and responses to this policy in two industries, steel and textiles. David Tarr (chapter 7) analyzes what he describes as the crisis that has arisen in the steel industries of the United States and the European Community as declining demand and the emergence of new lower-cost producers have reduced production and employment in both regions by more than one-third since 1974. The EC and the United States, he notes, have responded by adopting similar external policies, namely, greater import protection, but quite different domestic policies. At first, the Community attempted to maintain minimum prices for certain steel products with a system of voluntary production quotas; when this did not work, EC officials imposed mandatory production quotas. Shortly thereafter, a code on the subsidies provided by national governments to steel producers, aimed at reducing and finally eliminating such subsidies, was adopted. In contrast, the U.S. government has not intervened directly in the domestic market and has allowed losses suffered by domestic firms to be the guide in plant closings. These differences in domestic policy led to a major trade dispute in 1982 when U.S. producers filed charges of dumping and subsidization of steel exports against the EC. The U.S. Department of Commerce agreed that EC producers were being subsidized, some by substantial margins, and the International Trade Commission found that U.S. firms had suffered material injury. After the dispute reached the highest political level in both regions, however, and before countervailing duties were imposed, the Community agreed to a Voluntary Restraint Agreement (VRA) on steel. Tarr argues that a more viable, efficient EC steel industry would emerge if the Community eliminated its domestic controls on prices, output, and investment. These controls, he contends, create more distortions over time and make the adjustment problem worse. He also believes that the United States and the European Community should eliminate the nontariff barriers they have erected. His empirical investigations indicate that the costs of these trade barriers far exceed the adjustment costs they are designed to save. Carl Hamilton (chapter 8) considers two aspects of the protection of textiles that has been introduced by the United States and Europe by means of quantitative import restrictions: the levels of protection that these controls provide and the levels of rents earned by exporters of textiles because of these restrictions. The importance of these questions is reflected in the fact that trade in textiles and clothing makes up 9 percent of world trade in manufactures and 25 percent of the manufactured exports of the developing countries.

7

An Introduction to the Issues and Analyses

Using data on the prices of quota rights in Hong Kong and (for a short interval) in Taiwan, and an indirect method to calculate the degree of restrictiveness of quotas imposed on South Korean and Taiwanese textiles, Hamilton estimates the tariff equivalents of quotas plus tariffs on imports of textiles into the United States and the European Community from Hong Kong, Taiwan, and South Korea. He finds that the combined rate of protection from quotas and tariffs in the United States on textiles from these three suppliers ranges from about 45 percent to 65 percent. In contrast, the degree of restrictiveness on textile imports into the EC ranges from only about 25 percent to 35 percent. Rents derived by Hong Kong, Taiwan, and South Korea because of the quotas imposed by the United States and the EC are estimated by Hamilton at more than $500 million in 1983 alone. Some 80 percent of this amount is due to U.S. quantitative restrictions. While the United States and the European Community impose quantitative import restrictions on textiles from major developing-country exporters, they use only tariffs to restrict the flow of textiles between themselves. These policies, as Hamilton points out, can have the effect of mitigating and even nullifying the restrictive effects of tighter quantitative controls against the developing countries by stimulating increased textile trade between the two trading blocs. He notes in particular the prospect of increased textile exports to EC countries from Portugal and Spain when they become full members of the EC. Hamilton finds some empirical evidence that this kind of trade deflection has in fact occurred in textile and footwear trade. 1.4 New Issues: Services, High-Tech Products, and Strategic Trade Policy History will probably see as the most notable feature of the Uruguay Round of multilateral trade negotiations the fact that for the first time the member nations of GATT negotiated on trade in services as well as on trade in goods. When in 1982 the United States first proposed that services trade be covered in the next round of trade negotiations, the idea was rejected both by the developing countries and the European Community. The EC eventually agreed to support the proposal but the opposition of many developing countries led to the compromise that the services negotiations be formally separate from the negotiations on goods. In part 5, "Trade in Services," Andre Sapir (chapter 9) and Richard Neu (chapter 10) illustrate the kinds of problems and opportunities faced by negotiators in this field by examining two key areas of services trade, international telecommunications services (Sapir) and international trade in banking services (Neu). Sapir points out that negotiations on almost all forms of services are hampered by the absence of good data on the actual volume of trans-

8

Robert E. Baldwin

border trade and inadequate conceptual understanding of the differences between services and goods trade, especially whether the same economic principles can be used to evaluate the effects and benefits of services trade as have been traditionally used to appraise goods trade. His discussion of these problems indicates, however, that sufficient progress has been made on both issues to justify moving forward with a services negotiation. International telecommunications are, according to Sapir, an especially important service because they have not only enhanced the tradability of traditional services, such as banking services, but have increased the opportunities for international trade in new forms of information services, such as data processing and data-base services. Furthermore, a rapid series of innovations in information technologies and deregulation in the U.S. telecommunications markets are opening up new opportunities and challenges for the industry. There are, however, significant differences in views between U.S. and EC industry leaders over the extent to which trade in telecommunication services should be liberalized. European leaders generally favor the traditional view that national markets for telecommunication services should be organized on a monopolistic basis under government ownership or regulation to be economically efficient. The view in the United States is that deregulation increases efficiency. U.S. suppliers are now engaged in intense competition among themselves and they also want to compete in foreign markets. Interestingly, a new argument being put forward in Europe for restricting access to its markets is the need to have the opportunity to catch up with technological development in the United States, a development that may in part be a result of deregulation. Sapir notes that this type of disagreement is likely to arise in the negotiations on many forms of services. Richard Neu questions the official American position that the liberalization of international trade in banking services is a desirable trade policy objective. He supports his view with three main arguments. He first stresses the difficulties of trying to reconcile legitimate national needs to regulate banking with the demands for freer and fairer international trade. In the process, many important interests will be threatened and considerable time, energy, and political capital will have to be expended that, in his view, could be better used in negotiating on such issues as restrictions on foreign workers, rules for foreign direct investment in service industries, and international information flows. Neu also fears that too much liberalization may erode the safety and stability of the global banking system. The greater emphasis on foreign operations by the major banks may, for example, increase rather than reduce their riskiness. Similarly, an increased "interconnectedness" of the international financial system may make it easier for shocks to

9

An Introduction to the Issues and Analyses

move through the system. Competitive pressures on regulatory authorities to ease socially desirable requirements and the possibility of the banking system's being dominated by fewer, bigger banks are other concerns he expresses. Finally, Neu believes that technological advances in communications plus some changes in regulatory and establishment rules may enable local representative offices with good communications links to the bank's headquarters to handle all of the functions that now require the local presence of a foreign bank. In short, he believes that making liberalization of international trade in banking services a U.S. priority is to risk failure of the talks, and, more importantly, that success might bring about changes we may one day regret. Another notable feature of the agenda for the Uruguay Round is the explicit recognition of the growing importance of high-technology products in world trade. Community officials were skeptical about the appropriateness of including this subject on the agenda for a new trade round when U.S. officials first suggested doing so in 1982. While the Uruguay Ministerial Declaration mentions the importance of high-tech products, they are not specifically included in the subjects for negotiation. In part 6, "Trade Policy in Oligopolistic Environments," Kala Krishna (chapter 11) examines the industries producing high-technology products and asks whether such sectors require special consideration from a trade policy viewpoint. In the second paper in part 6, Barbara Spencer (chapter 12) raises another important trade policy issue of concern to U.S. and EC officials, given the fact that a large share of world trade is conducted by firms operating in an oligopolistic environment: Is there a need for the basic "unfair trade" rules of the GATT, such as the one covering countervailing duties, to be modified? Krishna points out that many high-tech industries are not only organized oligopolistically but often have a number of other special characteristics. One is the existence of network externalities, which exist when the usefulness, and thus willingness to pay for a good or service, increases with the number of other people who use the good or service. For example, the more people owning a phone, the greater the benefit each owner derives from the phone. Similarly, if the amount of software is related to the number of computers in use, an increase in the number of computers sold increases the available stock of software which, in turn, raises the amount consumers are willing to pay for the computer. As Krishna rigorously demonstrates, when a firm producing a product with network externalities competes with a foreign firm at home and abroad, the home government may be able to increase its country's economic welfare by subsidizing the domestic consumption of the firm's product. By increasing domestic sales at the expense of foreign sales in the home market, the subsidy will help the domestic firm in both

10

Robert E. Baldwin

the home and foreign market as consumers in both become more willing to purchase the product because of its greater use. In imperfectly competitive markets, this action could shift profits from foreign firms to domestic firms in both markets to an extent sufficient to raise national welfare. Of course, as she notes, this result does not mean that government subsidization could in fact increase national welfare. The government may not be well enough informed to identify a welfare-increasing policy; lobbying by interested pressure groups may bring about welfarereducing actions that were not expected by government bureaucrats; and the effects of foreign retaliation might more than offset the expected benefits. A quite different problem in an industry with network externalities is that a firm in the industry may choose to make its product incompatible with foreign competitors' products, thereby impeding competition and reducing national welfare. High-tech industries are also characterized by high research and development expenditures and significant experience effects. The first raises problems with international counterfeiting, while the second leads to pressures on the government to subsidize domestic firms so they can gain the advantages brought by experience. Aside from the existence of network externalities, when markets are imperfectly competitive government actions may be able to improve the strategic position of domestic firms relative to foreign firms and thereby raise national welfare at the expense of foreign welfare. Barbara Spencer, a pioneer in pointing out this possible effect of a government subsidy, asks whether countervailing duties levied by a government to offset the injury to its producers caused by foreign government subsidies to foreign firms will in fact just offset this injury when markets are imperfectly competitive. GATT rules (Article VI: 3) state that no countervailing duty shall be levied in excess of the amount of the estimated subsidy. Thus, if the subsidy is estimated to be 5 percent of the per unit value of a product, the countervailing duty cannot exceed 5 percent of the per unit value of the product. Spencer points out that whether countervailing duties levied under this rule just offset the harm done to domestic producers by foreign subsidization depends crucially on the nature of the subsidy. If the foreign subsidy takes the form of a direct subsidy per unit of exports, a countervailing duty equal to this amount will restore the production and profits of domestic producers to their pre-foreign-subsidy levels. But, as she shows, quite different outcomes occur when the subsidies apply to capital. If, as often is the case, the subsidy must be used by foreign firms to purchase new plant and equipment, a countervailing duty equal to the subsidy may be insufficient to restore the

11

An Introduction to the Issues and Analyses

production and profit levels of domestic firms. In contrast, if the subsidy is applied to existing capital equipment by foreign firms, the countervailing duty will result in a situation where foreign firms are worse off and domestic firms better off than before the subsidy. The implication of Spencer's analysis is that in applying the rules on countervailing duties in imperfectly competitive market structures, there is a need to examine more carefully the manner in which producer subsidies are utilized and to make the GATT rule more flexible to accomplish the intended purposes of these duties. It is important to resolve disputes such as those between the United States and the European Community over the proper way to measure the degree of subsidization, but even more fundamental issues relating to the GATT countervailing duty rule are in need of attention. 1.5 Trade Policy and the Trade Deficit The massive U.S. trade deficit with all major trading blocs has evoked a multitude of trade policy proposals to correct the imbalance. Not only is there considerable disagreement within the United States about the effects of these proposals but between foreign and U.S. political leaders. Thus, part 7 is devoted to the topic, "Interaction Between the Macroeconomic Environment and Trade Issues." Rachel McCulloch (chapter 13) explains why a country's current account balance is a macroeconomic phenomenon and describes how the massive U.S. budget deficit, coupled with a tight monetary policy, financial and industrial deregulation, enhanced fiscal incentives, capital inflows induced by actual and threatened increases in U.S. trade barriers, and liberalized restrictions on capital outflows from Japan combined to raise sharply the international value of the dollar and thus to increase the current account deficit significantly. She points out that reductions in the U.S. budget deficit and economic expansion abroad can help correct the U.S. external imbalance but cautions against overreliance on these measures. They will work, she notes, only if they reduce domestic absorption relative to domestic production and raise absorption abroad relative to foreign production. Trade policies that raise import barriers and provide for export subsidies in various forms are not likely to decrease the U.S. trade deficit, since they are unlikely to have much effect on the aggregate domestic production and absorption conditions that determine the state of the current account. Individual industries may benefit but their gains will be offset by import increases and export decreases in other sectors. Thus, while political pressures may bring about increased import protection and export subsidization and thereby worsen US-EC trade re-

12

Robert £ . Baldwin

lations, it is doubtful that these policies will alleviate the cause of these political pressures, the U.S. current account deficit with the European Community and other regions. The need for closer coordination by the major economic blocs of their macroeconomic policies to prevent conditions of excessive inflation, unemployment, or trade imbalances in the world economy is being increasingly stressed by both public and private leaders. Macroeconomists have responded by using game theory to investigate the economic consequences of governments' pursuing cooperative versus noncooperative policies. Giorgio Basevi, Paolo Kind, and Giorgio Poli (chapter 14) extend this line of research by developing a game-theoretic model that abandons the simple dichotomous approach to the issue—complete cooperation or no cooperation at all—and, instead, permits combinations of cooperation and noncooperation among countries on different macroeconomic policies. Its purpose is to provide insights into the problems of coordinating monetary and trade policies among countries within the European Community and between the EC or individual EC members and the United States. In response to an assumed exogenous shock that reduces output by 10 percent in all nations, the three countries in the model, representative of Germany, Italy, and the United States, each seek to minimize the deviations of consumer prices, the mark/dollar exchange rate, and the mark/lira exchange rate from their equilibrium values in the country. Each country uses changes in its money supply to achieve these goals. In addition, each country tries to minimize the deviations of aggregate output from its equilibrium value and uses import protection to stimulate national output. Postulating standard functional relationships for aggregate supply, aggregate demand, nominal wages, and a demand for money-based price levels, exchange rates, tariff rates, and interest rates, and assuming ''realistic" values for the various parameters, the authors simulate the effects of various combinations of cooperative and Cournot-Nash noncooperative strategies among the countries. Among the cases analyzed are (1) cooperative strategies by all three countries on both monetary and trade policy, (2) noncooperation among the three on these policies, (3) cooperation between Germany and Italy on monetary and trade policy but confrontation with the United States on these policies, (4) cooperation by all three on monetary policy but trade policy cooperation only between the two EC countries, and (5) cooperation on monetary policy only between the United States and Germany and on trade policy only between Germany and Italy. One important conclusion from the analysis is that cooperation between the Community and the United States in achieving their price

13

An Introduction to the Issues and Analyses

level and exchange-rate goals, coupled with a lack of cooperation on output objectives, leads to great fluctuations in output that, by inducing protection, result in worse output outcomes than from the original supply shock. The authors point out that this means that noncooperative solutions may be superior to partially cooperative ones. As might be expected, cooperation between the EC countries at the monetary and real levels is preferable collectively to each country's going it alone. But compensation payments by Italy to Germany and the United States would be required to induce the latter two countries to accept this strategy. An interesting result is that cooperation between Germany and the United States on monetary matters, coupled with no trade policy cooperation among the three countries, yields the preferred arrangement when trade policy cooperation between the European countries and the United States is ruled out. Consequently, as the authors note, dealing separately with cooperation at the monetary and the real levels leads to a situation in which European cooperation tends to fall apart because of the advantages to Germany and the United States of cooperating on monetary matters.

2

Legal Issues in US-EC Trade Policy: GATT Litigation 1960-1985 Robert E. Hudec

2.1 Introduction Recent years have witnessed a sharp growth of friction between the United States and the European Community in the realm of trade policy. In few areas has this friction been as apparent as in the growing volume of GATT litigation between the two parties. In one respect, the large number of GATT lawsuits can be viewed merely as a symptom of more fundamental substantive problems of the relationship. Increasingly, however, the litigation itself has come to be seen as a cause of conflict. The European Community has repeatedly accused the United States of misusing GATT legal procedures, charging that many U.S. lawsuits have no real legal foundation but are instead primarily political gestures made to satisfy domestic policy needs. The United States, in turn, has accused the Community of trying to subvert GATT litigation procedures—and indeed the integrity of GATT law itself—by resisting many of the suits brought by the United States. This paper examines US-EC GATT litigation from its inception in the early 1960s, until the end of 1985. The paper is divided into two main sections. Section 2.2 presents what is called a quantitative profile of US-EC litigation: Who has sued whom? How often? Over what issues? The US-EC litigation is compared with all other GATT litigation that took place during the same period. Section 2.3 then analyzes the US-EC litigation in its historical and political context. It seeks to identify the factors underlying the litigation Robert E. Hudec is Melvin C. Steen Professor of Law at the University of Minnesota Law School. N.B. GATT lawsuits are cited by their number in appendix A. 17

18

Robert £ . Hudec

behavior of the two parties, and offers a prediction of how these factors are likely to affect US-EC legal relations in the future. 2.2 A Quantitative Profile 2.2.1 The Sample The General Agreement on Tariffs and Trade (GATT) is an international agreement that establishes rules of behavior for governments in the area of international trade policy. The General Agreement also contains an adjudication procedure that permits member countries to bring "lawsuits" about violations of those rules. Since the GATT came into force in 1948, over 130 such GATT lawsuits have been initiated.1 This section of the paper examines the 80 GATT lawsuits filed from 1960 to the end of 1985—the time during which the European Community has been a full participant in GATT legal affairs.2 It focuses particularly on the 26 suits between the United States and the European Community during this period. (Appendix A presents a table listing all 80 lawsuits; Appendix B contains a separate list of the 26 US-EC lawsuits.) 2.2.2 The Parties Table 2.1 lists the number of appearances that each GATT member has made in GATT lawsuits from 1960 to 1985, listing separately the number of appearances as a defendant and as a plaintiff. Table 2.2 provides a detailed list of exactly which other parties were suing, and being sued by, the five most active GATT litigants—the EC, the United States, Canada, Japan, and Australia. The following data in tables 2.1 and 2.2 seem significant: 1. Almost one-third of the GATT lawsuits during this period (26 of 80) were lawsuits between the United States and the EC. 2. In addition to the 26 lawsuits between the U.S. and the EC, 45 of the remaining 54 lawsuits involved either the U.S. or the EC as one of the parties. Only 9 of the 80 lawsuits involved neither.3 3. The EC and the U.S. litigated more frequently with each other than with others. The United States accounted for 26 percent of the complaints filed against other GATT countries, but 53 percent of the complaints filed against the EC. The EC accounted for only 11 percent of the complaints filed against others, but 56 percent of the complaints filed against the U.S. The same disproportionate shares appear when one examines the activity of the U.S. and the EC as defendants. The EC was the target of 30 percent of the complaints filed by other GATT countries, but was the defendant in 57 percent of the U.S. complaints. The U.S. was the target of only 13 percent of the complaints filed by

19

US-EC GATT Litigation

Table 2.1

Appearances as Defendant and Plaintiff Defendants EC a U.S. Japan Canada U.K. Spain Greece Denmark Norway New Zealand Finland Switzerland Jamaica Brazil Chile

Plaintiffs 3 3

b

15c 9 7 3 2

Total 78 Plus one case with 10 defendants and one case submitted jointly by the 2 parties

U.S. 31 EC d 15 Canada 7 Australia 6 Brazil 3 Japan 2 Chile 2 Hong Kong 2 India 2 Uruguay Israel Korea Argentina Poland Nicaragua South Africa Finland Total 78 Plus one case with 15 plaintiffs, and one case submitted jointly by the 2 parties

includes complaints both against the EC and against member states: EC itself 27 France 3 Italy 1 Belgium 1 Netherlands 1 b 17by U.S. c 8 by EC. d There were no separate complaints by individual EC member states.

others, but 53 percent of the EC complaints were brought against the United States. 4. The US-EC litigation is not distinctive in terms of each party's role as plaintiff and defendant. The United States sued the EC twice as often as the EC sued in return (17 to 8), but the United States also sued other GATT countries twice as often as it was sued by them (14 to 7). The EC likewise played the same defendant-oriented role in GATT litigation with other countries, suing other GATT members less than half as often as it was sued by them (7 to 16). 5. With or without the EC, the United States was responsible for initiating a very large share of the GATT lawsuits during this period. It accounted for 40 percent of all complaints since 1961, with the EC

20

Robert E. Hudec

Table 2.2

Opposing Parties As Plaintiff, Sued:

1. EC (49a)

U.S. Canada Chile Finland Switzerland Japan

2. United States (47-)

EC Japan Canada Greece Denmark Jamaica Spain U.K. Brazil

3. Canada (14)

U.S. EC Japan

4. Japan (11)

U.S. EC

As Defendant, Sued by: 8 3 1 1 1 1

U.S. Australia Canada Chile Korea Brazil Hong Kong Argentina Japan (10 countries)

33

15 17 5 3 1 1 1 1 1

EC Canada Japan India Poland Nicaragua

31 3 3

EC Japan

3 3 7

U.S. Australia Canada India EC

~2 5. Australia (6)

8 3 1 1 1 1

15 U.S. EC South Africa

7 1 1

17 5 3 2

5 1 1 1

J.

9

5 6

a

Total includes one case submitted jointly by the U.S. and the EC.

a distant second at 21 percent. Excluding all lawsuits between the U.S. and the EC, the United States still accounted for a 26 percent of the rest, with Canada next at 13 percent. 6. With or without the U.S., the European Community has been the target of a very large share of the lawsuits during this period. The EC or a member state has been the defendant in 42 percent of the GATT lawsuits since 1961, with the United States a distant second at 20 percent. Excluding suits between the EC and U.S., the EC was the

21

US-EC GATT Litigation

defendant in 30 percent of the remaining cases, with Japan next at 17 percent. Summing up, it would seem that the United States and the European Community each had a rather pronounced legal tendency—legal aggressiveness in one case, and legal vulnerability in the other. It is perhaps no surprise, therefore, that the two parties litigated with disproportionate frequency when dealing with each other. 2.2.3 The Volume of Litigation over Time Table 2.3 presents a year-by-year breakdown of GATT lawsuit activity since 1960. The total number of all GATT lawsuits filed in the year is given in the first column of numbers. The following six columns give a breakdown of the litigation activity of the United States and European Community during each of those years. When GATT litigation is viewed over time, the following significant patterns emerge:

Table 2.3

Year

GATT Lawsuits, by Year All Cases

1960 1961 1962 1963 1964-1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985

none 2 3 la none 4

Total

80

a

4 4 2 1 4 3 6 4 6 5 15 7 6 3

All v. U.S.

U.S. v. All

U.S. v. EC

All v. EC

EC v. All

EC v. U.S.

— — —

—

—

—

—

a

—

— — —

— —

—

3

—

1

—

—

1

3 3

2 3

1

1

—

— —

2 3 1

— — — 1 1

—

3

1 2 1 1 2 1 1 9 1

2

2

— — — 1 1 6

2

— 3 1 3 1 2 3 9

2 2 1 4 1 2

— —

— — —

—

15

31

17

a

— — 1

— 1

— —

1

—

1 2 5 2 2

3 1 1

33

15

8

2

The 1963 case was the joint US-EC submission known as the "Chicken War.'

1

22

Robert £ . Hudec

1. The volume of US-EC litigation has been increasing, but so has the volume of GATT litigation in general. There were only 3 US-EC lawsuits in the 1960s, followed by 9 in the 1970s, and then 14 in the six years 1980-85. Other GATT litigation followed roughly the same rising curve—3 in the 1960s, 23 in the 1970s and 28 in the six years 1980-85. 2. The unequal distribution of plaintiff and defendant roles noted in the previous section occurred mainly in the period before 1980. In that period, the U.S. filed nine complaints against the EC while the EC filed only two in return. The same imbalance occurred, for both the U.S. and the EC, in their litigation against other GATT members during this period. In the years before 1980, the U.S. enjoyed a 10:1 plaintiff/ defendant ratio while the EC suffered a rather galling 1:8 ratio. 3. Since 1980, the plaintiff/defendant ratios of both the U.S. and the EC have become more balanced. The ratio in US-EC litigation in the 1980s has been only 8:6 in favor of the United States. In GATT litigation against other countries, the U.S. ratio has actually fallen into deficit (4:6), while the EC ratio has risen to an almost equal balance (6:8). 4. The single most striking change in the years after 1980 has been the emergence of the EC as a GATT plaintiff. Having filed only 3 GATT lawsuits in the years 1960-80, the EC filed 11 lawsuits in the years 1981-85. 2.2.4 Subject Matter Having examined which GATT countries were suing which other countries and when, we now turn to the question of what they were fighting about. Tables 2.4 and 2.5 present two sorts of data pertaining to the subject matter of these lawsuits. Table 2.4 presents a breakdown of disputes according to the product area affected by the trade policy measure in question.4 Disputes are divided into three broad categories: (1) complaints about measures that affect trade in agricultural and fishery products; (2) complaints about measures that affect trade in industrial and mining products; (3) complaints about trade measures of a general character that have, in the case at hand, no particular product focus or effect. The significant data in table 2.4 appear to be the following: 1. Of all the GATT lawsuits from 1960 to 1985, 54 percent of the complaints (43 of 80) involved trade in agricultural products. 2. The European Community has been the target of 58 percent of all lawsuits involving agricultural products (25 out of 43). 3. The percentage of agricultural complaints in litigation against the EC is twice as high as the percentage in litigation against other GATT countries. Litigation against other GATT countries involves agricultural trade measures only 38 percent of the time. Litigation against the EC,

23

US-EC GATT Litigation

Table 2.4 Grouping of Cases a

Subject Matter, by Product Area Agricultural & Fishery No.

%

Industrial & Mining

General

No.

%

No.

%

All Cases (80)

43

54

27

34

10

13

All v. EC (33) U.S. v. EC (17) Other v. EC (16)

24 13 11

73 76 69

5 1 4

15 6 25

4 3 1

12 18 6

All v. U.S. (15) EC v. U.S. (8) Other v. U.S. (7)

6 3 3

40 38 43

6 3 3

40 38 43

3 2 1

20 25 14

All v. Other (30) EC v. Other (7) U.S. v. Other (14) Other v. Other (9)

12 2 6 3

40 25 43 33

16 4 7 6

53 63 50 67

2 1 1 0

7 12 7 0

2

100

—

—

—

—

Unclassified (2) a

The term "Other" in the Grouping of Cases column means all other litigants except the EC and/or the U.S.

on the other hand, involves complaints about agricultural trade in 73 percent of the cases. 4. Lawsuits over agricultural trade measures are clearly the reason why the EC is the GATT's most frequent defendant. In disputes not involving agricultural products, the EC has been sued no more often than the U.S.—nine times each. 5. The volume of "agriculture" lawsuits against the EC is not due to any peculiarity of US-EC litigation. Other governments accounted for 44 percent of the agriculture lawsuits against the EC (11 of 25), and agriculture complaints represented an almost equally high percentage of their total complaints against the EC as did the agricultural complaints of the United States—69 percent to 77 percent. Everyone, it seems, had special problems with EC agriculture. In sum, table 2.4 points the finger for most of the EC's litigation problems at agricultural policy—meaning, of course, the EC's Common Agricultural Policy (CAP). The data also show that the litigationgenerating effect of the CAP is general, and not merely a peculiar U.S. reaction. Table 2.5 presents a breakdown according to the type of trade policy measure being complained about.5 Disputes are divided according to three basic types of policy measure: (1) complaints about subsidies,

24

Robert E . Hudec

including both export subsidies and domestic production subsidies; (2) complaints about tariffs, including EC variable levies; (3) complaints about nontariff barriers, including both border measures and internal measures, and both quantitative and tax-type measures. Table 2.5 also contains a final column that examines the extent to which discrimination has been a ground of complaint. Since trade barriers involving discrimination will already have been counted in the Tariff and Nontariff Barrier categories, the Discrimination category is separated from the other totals by being presented in parentheses in the column to the far right. The data in table 2.5 is less striking, but the following points may be noted: 1. Subsidies accounted for an important share of litigation against the EC (13 of 33 complaints—39 percent), but were only a negligible factor in the lawsuits against other GATT members during this period (3 of 45 complaints—7 percent). Of the 13 subsidy complaints against the EC, 10 involved agriculture.6 2. Subsidy complaints were an even larger share of U.S. complaints against the EC—8 of 17, or 47 percent. 3. The importance of subsidies led to further analysis of subject matter which revealed another phenomenon that cannot be shown on Table 2.5

Subject Matter, by Policy Measure Trade Barriers

Grouping of Cases

Subsidies

Tariff

Nontariff

Discriminato Element 3

No.

%

No.

%

No.

%

No.

%

All Cases (78)b

16

21

12

15

50

64

(18

29)

All v. EC (33) U.S. v. EC (17) Other v. EC (16)

13 8 5

39 47 31

3 1 2

9 6 13

17 8 9

51 47 56

( 6 ( 1 ( 5

30) 11) 45)

All v. U.S. (15) EC v. U.S. (8) Other v. U.S. (7)

2 2 0

13 25 0

3 2 1

20 25 14

10 4 6

67 50 86

( 5 ( 1 ( 4

38) 17) 57)

All v. Other (29) EC v. Other (29) U.S. v. Other (14) Other v. Other (9)

1 0 1 0

3 0 7 0

6 2 2 2

21 33 14 22

22 4 11 7

76 67 79 77

( 7

( 4 ( 3

25) 0) 30) 33)

—

—

1

100

—

—

—

—

Joint Submission a

(o

The number and percentage of "Trade Barrier" complaints containing a clear leagl attack on a discriminatory element. b Two cases (1, 68) could not be classified by type of measure.

25

US-EC GATT Litigation

table 2.5. The United States suffered almost exactly the same percentage of complaints as did the EC (6 of 16, or 38 percent) against what might be called antisubsidy measures. Four complaints against the United States involved legal objections to some aspect of the U.S. countervailing duty law.7 A fifth complaint involved a U.S. export subsidy on sales of wheat flour to Egypt, an act of retaliation against the EC wheat flour subsidy.8 And finally, a sixth complaint involved a U.S. corporate income tax law called DISC, an export subsidy justified in part as a response to alleged export subsidies built into the "territorial" income tax systems of other countries.9 It would seem, in short, that the U.S. response to subsidies caused as much irritation in other capitals as subsidies were causing in Washington. 4. Antisubsidy measures were the target of an even larger percentage of EC complaints against the U.S.— 4 of 8, or 50 percent. 5. The fact that 51 out of 64 trade barriers complained about during this period were nontariff trade barriers is not unexpected. After the 1967 Kennedy Round tariff cuts, tariff levels have fallen so low that they are rarely used anymore as instruments of trade policy. The relative proportion of tariff and nontariff litigation between the U.S. and the EC is not out of the ordinary. 6. The number of discrimination cases is perhaps a bit lower than one might have expected, but not much. Many observers, the author included, have remarked on the declining respect for the MFN principle over the past 20 years. Evidently, however, governments have not been overly preoccupied with the problem. The total of 18 out of 62 trade barrier complaints—a bit over one quarter—is not an insignificant share, but neither is it very large. 7. Discrimination is almost nonexistent as a factor in the US-EC litigation. In only 2 of the 26 lawsuits between them was discrimination a major issue.10 This is particularly surprising in view of the rather extensive discrimination practiced by the EC against the United States. Trade policy officials often call attention to the fact that, of all GATT countries, only the United States, Japan, Australia and New Zealand pay the full rates of the EC Common External Tariff anymore. Maybe so, but it does not seem to be bothering them very much. 8. Interestingly, 12 of the 17 discrimination complaints were made by "other" GATT countries. This tends to confirm the view often expressed in developed countries that discrimination is more dangerous for smaller countries than for the larger ones. At least smaller countries seem to react to it more vigorously when they end up on the wrong side. Of these various findings pertaining to the type of policy measure at issue, the only ones of any real significance to the US-EC legal relationship are the first four. The number one substantive issue in GATT litigation between the U.S. and the EC has been subsidies and measures

26

Robert £ . Hudec

responding to subsidies. These subsidy-related matters have been more prominent in US-EC litigation than in GATT litigation generally. 2.3 US-EC Litigation in Context 2.3.1 A Brief History of US-EC Litigation Several threads run through the history of US-EC litigation from 1960 to 1985. The timing of the U.S. lawsuits is quite closely related to events in Congress, with a majority of the lawsuits occurring either just before the Congress was to vote on new trade legislation, or seemingly in response to demands made by the Congress when passing such legislation. The timing of EC lawsuits, in turn, seems most closely related to the volume and vigor of U.S. lawsuits, with an increase in EC lawsuits following each major increase in U.S. litigation. The one substantive issue that seemed to dominate litigation strategy on all sides was the ever-present EC Common Agricultural Policy. Legal Actions in the 1960s The 1960s opened with a lawsuit by Uruguay that produced a significant challenge to the Common Agricultural Policy. The first phase of the lawsuit was directed to all developed countries, including the member states of the EC, attacking any and all trade restrictions against Uruguayan exports. Uruguay then added a second set of issues by asking for a legal ruling on the conformity of the EC's Common Agricultural Policy (CAP) with GATT, and on the conformity of the variable levy in particular. The GATT panel hearing the complaint twice declined to rule on this second set of issues, saying that the contracting parties had considered the variable levy before and had been unable to come to a decision.11 Had the United States wished to challenge the legality of the CAP in toto, this would have been a good opportunity. Apparently, however, the United States was not willing to do so. There is no evidence that the United States gave any support to Uruguay's request for a ruling. The first GATT legal complaints by the United States in the 1960s were a pair of 1962 complaints against France and Italy, attacking quantitative restrictions that were being maintained, without legal excuse, on products for which France and Italy had granted Dillon Round tariff concessions.12 The two complaints were the first GATT lawsuits filed by the U.S. in six years. They appear to have been intended to demonstrate the U.S. administration's resolve in enforcing trade agreement rights, for both werefiledjust at the time when the U.S. Congress was considering major trade legislation that became the Trade Expansion Act of 1962.

27

US-EC GATT Litigation

The two 1962 complaints also had the secondary purpose of warning the European Community against further expansion of its Common Agricultural Policy. The primary focus in both suits were restrictions affecting processed food such as canned fruits—products that had not been included in the original variable levy system and for which GATT tariff bindings were still in force. The United States apparently wanted to demonstrate that it had no intention of surrendering its GATT rights on such products. The complaint against Italy was settled at an early stage with a promise of liberalization on certain products. The French complaint could not be settled immediately, and so the United States asked for a panel decision ruling on its rights. A generally favorable ruling was made, but the panel asked the United States to defer its request for authority to retaliate pending another round of negotiations. The case was then settled with a promise by France of partial liberalization.13 The next major legal event of the 1960s was the celebrated "Chicken War" dispute.14 The European Community had withdrawn all GATT tariff bindings on variable levy products. This was legally permitted, but the Community was required to pay compensation, in the form of tariff concessions on other products, for the legal rights being taken back. If other governments did not regard the compensation offered as adequate, they were free to restore the balance in their own way by withdrawing an equivalent number of their own concessions. The Chicken War arose when the United States declined to accept the compensation offered for withdrawing the West German binding on poultry, and announced it would retaliate by withdrawing concessions of its own on $44 million worth of EC trade. The European Community contested the size of the retaliation, and this collateral dispute was submitted to a GATT panel. The United States then retaliated by withdrawing concessions of the amount determined by the panel—$26 million. The real purpose of the Chicken War retaliation was to give a dramatic expression of U.S. displeasure with the level of protection adopted in the CAP on poultry. One can never be certain what impact such retaliation has had, because one cannot know what would have happened without it. From all visible signs, however, the retaliation had no effect at all. The European Community went about finishing the CAP and setting its support prices much as before, with levels of protection uniformly higher than the United States thought appropriate. For the rest of the 1960s, US-EC legal relations remained quiet. This was part of a general lull in GATT legal affairs, for there were no GATT lawsuits of any kind during the years 1964-69. During these years, the United States directed its concerns about the CAP to the negotiating arena. It made a determined effort during the Kennedy Round to secure

28

Robert E. Hudec

some kind of legal ceiling on the level of protection in the CAP, repeatedly threatening to end the negotiations if such limits were not included. In the end, however, the United States did agree to go ahead with the industrial tariff cuts in the Kennedy Round, having received nothing more than a fig leaf in agriculture—the short-lived 1967 International Wheat Agreement. Legal Actions in the 1970s In the early 1970s, the United States brought a flurry of new GATT lawsuits. This new legal assault was intended partly as a foundation for another major piece of trade legislation—the Trade Act of 1974. But it was also a genuine effort to restore the effectiveness of GATT adjudication procedures by giving them work to do. U.S. legal relations with the European Community took a bad turn early in the 1970s with a rather long skirmish over EC tariff preferences on citrus products in favor of Mediterranean suppliers.15 The United States repeatedly protested the illegality of such preferences, but never formally demanded a legal ruling. The matter was eventually settled in 1973 with an agreement adjusting the seasonal tariff rates in a manner favorable to U.S. exporters. (It was about this time that U.S. and EC leaders reached a broad political agreement, known as the Casey-Soames agreement, in which the United States agreed to accept the general principle of EC preferential arrangements with the states of Africa and the Mediterranean.) The first formal legal complaints against the European Community were two lawsuits filed in 1972. One involved temporary " compensatory taxes" that the European Community had placed on imported agricultural products to adjust for the effects of the monetary disturbances of the time.16 The measures were acknowledged to violate GATT tariff bindings, and the Community promised prompt removal when conditions permitted. The United States repeatedly demanded a formal legal ruling, but the Community managed to delay GATT legal proceedings until the tax had been almost completely withdrawn, and the United States then agreed to drop the matter. The other 1972 lawsuit was a renewal of the 1962 complaint against France.17 The United States reopened the case byfilinga formal request for authorization to retaliate, charging that France had not yet removed all of the quantitative restrictions found in violation. The retaliation proceeding was averted at the eleventh hour by an undertaking to remove the remaining restrictions. The U.S. legal campaign evidently caused some irritation in Brussels, and it may also have raised some concern about the possibility of still further legal attacks upon EC policy. Prior to 1973, the European Community had responded to such "legalistic" forays by lecturing the United

29

US-EC GATT Litigation

States on the folly of seeking legal solutions to foreign trade problems. In 1973, however, the Community took up the legal sword itself. For the first time in its history, it filed a GATT lawsuit, charging that the newly enacted U.S. DISC law constituted a subsidy to U.S. exporters.18 While it is impossible to document the suspicion, most GATT observers believed that the Community had filed the lawsuit primarily because it believed that U.S. legal attacks were getting out of hand, and that the United States needed a lesson in legal humility. The United States was not chastened. It responded by filing three counterclaims against the Community—lawsuits against France, Belgium, and the Netherlands charging that the "territoriality" feature of their income tax systems had the same economic effects as DISC; if DISC was an export subsidy, said the United States, so were those tax laws.19 The United States insisted that the four complaints be adjudicated as an ensemble by a single GATT panel; it also demanded that the panel include outside tax experts skilled enough to understand the issues of tax theory raised by the U.S. legal position. Thus began a long and tortured proceeding that did not end until 1981. The United States took a fairly severe public relations beating throughout the history of the DISC cases. Everything the United States did was regarded as wrong by the majority of GATT countries. The U.S. legal position never had any support; DISC was a blatant subsidy, while territorial systems of income taxation were something that virtually every GATT government had employed since long before GATT existed.20 The techniques used by the United States to support its legal claim were likewise roundly criticized. In the view of most GATT countries, the U.S. insistence on linkage of the cases was totally improper. The insistence on special experts was also much criticized, and, though it was not the only cause of delay (the Community itself delayed for long periods), it certainly did add to the delay substantially. The U.S. isolation continued after the panel finally reported in 1976. Although the panel reports found all four defendants equally guilty, most other governments agreed that the findings against the EC countries were in error and that only the DISCfindingwas correct. Standing alone, the United States insisted that all four findings must rise or fall together and persisted in this position for five years. In 1981, the United Statesfinallyaccepted the majority view and agreed to a vaguely worded decision that was understood by everyone to set aside the adverse rulings on the French, Belgian, and Netherlands tax systems but not the ruling on DISC. In 1984, the United States repealed DISC and replaced it with a superficially more conforming law. If the purpose of the DISC case had been to show up the less-thanperfect quality of U.S. legal compliance, it could hardly have worked better. But, however embarrassing the DISC case proved to be, it did

30

Robert £ . Hudec

not slow the U.S. campaign to revive GATT dispute settlement procedures. Part of the reason may have been a new factor that entered the picture shortly after the DISC case began. In 1974, the United States executive branch went to the U.S. Congress to obtain new negotiating authority for the Tokyo Round trade negotiations. One of the conditions Congress added was a new procedure, known as section 301, that was designed to compel the U.S. executive branch to enforce U.S. legal rights more vigorously than it had in the past.21 Congress was no longer satisfied with a few demonstration lawsuits on the eve of each major trade law. It was now demanding a permanent, institutionalized permanent procedure that would create and maintain continual pressure for enforcement. Congress made clear that legislative approval of the Tokyo Round agreements would depend on a satisfactory record in the years ahead. The pressures generated by section 301 proceedings yielded seven GATT lawsuits by the United States during the years of the Tokyo Round negotiations (1975-79)—three against Japan, two against the European Community, and one each against Canada and Spain.22 Both complaints against the Community were filed in 1976. Both involved short-term trade measures in the agricultural sector, taken to alleviate surplus situations in product sectors benefitting from CAP price supports. One case involved a minimum-import-price regime for imported tomato products, and the other involved a mixing regulation requiring the use of surplus dairy products with imported animal feeds.23 Significantly, the imported products being restricted in both cases were once again products on which GATT tariff bindings had remained in effect. In both cases, the restrictive measures were withdrawn before the GATT panel proceedings could be completed, but in both cases the proceedings were carried out to a formal decision anyway, and both measures were found in violation. The GATT panel agreed with the U.S. position that a formal legal ruling of illegality was appropriate in order to deter similar "hit-and-run" measures in the future. In 1978, not long after the two 1976 complaints were decided, the European Community filed its second GATT complaint against the United States. It was a curious lawsuit. The United States had a quite strict countervailing duty law that had been partially suspended during the Tokyo Round negotiations in order not to prejudice negotiations for a new Subsidies Code. The suspension had expired before the end of the negotiations, and a crowded legislative calendar had delayed the U.S. Congress in passing a new law to extend it. As a consequence of the delay, several suspended investigations were suddenly in danger of triggering countervailing duties. In what appeared to be a warning of dire consequences if the suspension were not quickly renewed, the Communityfileda formal GATT complaint charging that any imposition

31

US-EC GATT Litigation

of countervailing duties would constitute a nullification and impairment of GATT benefits under Article XXIII.24 The legal grounds of the EC complaint were dubious because the U.S. law was pretty clearly covered by a reservation for pre-1947 mandatory legislation and had been used often before. The tone of the complaint was nonetheless quite demanding. The threatening tenor of the EC's Article XXIII action introduced a new type of Community legal response that was to be seen fairly frequently in the coming years—a quick, sharp, and rather belligerent threat of retaliation, often listing the actual products.25 Governments often make such threats for the purpose of providing tangible evidence of harm for friendly officials in the other government. On other occasions, such threats are understood as demonstrations to satisfy interests back home. But there is a thin line between such showcase threats and plain old-fashioned muscular diplomacy, and it was often difficult to tell which side of the line the EC was standing on. In either case, the Community was emerging from the defensive legal posture of its early years.26 Legal Actions in the 1980s In 1979, the U.S. Congress passed major legislation approving and implementing all the trade agreements negotiated in the Tokyo Round. The price of its approval was a promise from the Executive Branch that there would be even more vigorous enforcement of GATT legal rights in the future. To make sure that the promise was kept, Congress strengthened section 301 still further by increasing its scope, its retaliation authority, and its automaticity.27 Congress also made it clear that it now wished to become a permanent partner in the business of trade policy, with the relevant congressional committees exercising more or less constant oversight over day-to-day affairs. Needless to say, enforcement of GATT legal rights would be a central focus of that oversight. The message of the 1979 legislation amounted to a demand for some quick GATT legal victories to prove the value of the legal rights gained in the Tokyo Round. The climate within GATT had also become more conducive to litigation by this time. The pressure from the United States lawsuits in the early 1970s had brought about a general increase of government interest in GATT litigation. The Tokyo Round negotiations were followed up by the devoting of considerable time and effort to improving the GATT's litigation machinery. The results were significant—an extensive restatement of the Article XXIII panel procedure28 and a series of new and more rigorous panel procedures for most of the Tokyo Round "Codes." 29 Encouraged by all this attention, GATT litigation

32

Robert E. Hudec

began to increase noticeably after about 1977. Although most of the initial increase was litigation other than US-EC lawsuits, it was inevitable that these forces would eventually have an impact upon the USEC legal relationship as well. Nothing very exciting happened during the first few years following the Tokyo Round. The United States initiated a legal proceeding against the Community in 1980 to challenge a new United Kingdom regulation on poultry processing that required foreign suppliers to comply two years earlier than domestic suppliers.30 Although the new regulation seemed clearly in violation of both GATT Article III and the new Tokyo Round Standards Code, the United States eventually allowed the complaint to lapse when U.S. exporters reported that they had already learned to comply and were no longer concerned. The European Community responded with a 1981 complaint charging that the United States had impaired GATT tariff concessions on Vitamin B-12, because of the way in which it had converted the tariff to a new valuation basis.31 A GATT panel found that, although the U.S. mode of implementation had caused severe adverse trade effects, the method of tariff conversion had not been in violation of U.S. obligations. But the panel's report then went on to muddy the legal waters by suggesting that the United States should nonetheless modify its tariff anyway, voluntarily. The United States refused to comply with this suggestion, and the Community chose to regard U.S. inaction as a failure to abide by GATT obligations. The dispute still smolders.32 The relative legal quiet of these first years was deceptive because during this time a number of legal grievances against the European Community were working their way through U.S. internal procedures in Washington. Finally, in late 1981 the storm broke. Between December 1981 and July 1982, the United States filed seven GATT lawsuits against the European Community. Five concerned subsidies on agricultural products, one involved a renewal of the 1970-73 complaint about preferential tariffs on citrus products, and one involved an industrial trade problem—whether the EC's Value Added Tax (VAT) should be counted in calculating the price threshold for transactions subject to the new Tokyo Round Procurement Code. The seven complaints were: Export Subsidy on Wheat Flour (Subsidies Code) Export Subsidy on Pasta (Subsidies Code) Export Subsidy on Poultry (Subsidies Code) Production Subsidy on Canned Fruit and Raisins (Art. XXIII) Export Subsidy on Sugar (Subsidies Code) Preferential Tariff on Citrus Products (Art. XXIII) Treatment of VAT in Price Calculations (Procurement Code)33

33

US-EC GATT Litigation

The sugar complaint was not pursued, and the poultry complaint evolved into a slow-moving series of discussions between the U.S., the EC, and Brazil. The other five complaints were pressed quite hard, resulting in long and often bitterly contested proceedings that eventually produced legal rulings by a GATT panel. The legal claims in all five of the contested cases called for changes in EC policy that ranged from extremely difficult to politically impossible. The legal claims in the three agricultural subsidy cases attacked critical aspects of the CAP. The wheat flour case sought to establish legal limits on the amounts of CAP surplus production that could be disposed of via subsidized exports—limits that would have created impossible surplus disposal problems if the EC were to continue price supports, without production limits, at the high levels that had been considered politically necessary in the past. The pasta complaint sought to prevent, or at least seriously hinder, the Community from giving subsidies on exports of processed foods, a result that would have virtually destroyed the export markets of many EC food processors who, under the CAP, were being forced to pay extremely high prices for their raw materials. The canned fruit complaint would have caused similar effects on food processors selling within the EC market, for the complaint sought to bar subsidies to domestic producers for products whose EC tariff had been "bound" in previous GATT negotiations. This would have meant that local processors would have had to pay the full CAP prices for raw materials and then compete, without tariff protection, against imports made from raw materials at world prices. The legal claims in the citrus and VAT cases were of similar proportions. The citrus complaint attacked preferential tariff advantages that were a key economic incentive holding Mediterranean countries inside the European political orbit. The VAT complaint made demands on the EC's own procurement policy which the EC insisted were impossible to meet. All five of the panel decisions fell short of satisfactory resolution. The VAT decision was accepted but still appears to face great difficulty being implemented. In the other four cases, the panel reports were rejected by one of the parties and thus never became official GATT rulings.34 The wheat flour panel was unable to reach a decision on whether EC wheat flour exports had exceeded the "equitable share" standard of Subsidies Code Article 10. The United States attempted, unsuccessfully, to override the panel report by persuading the full Committee of Subsidies Code signatories to make a finding of violation itself, and when this failed the U.S. blocked adoption of the panel's no-decision report. The panel in the pasta case issued a 4-1 divided report in which the majority found that the export subsidy on pasta products was prohibited by Article 9 of the Subsidies Code. The EC,

34

Robert E. Hudec

supported by several other governments, refused to permit adoption of the majority report. Both the canned fruit and citrus panels rendered unanimous rulings that the EC measures in question constituted nonviolation nullification and impairment under GATT Article XXIII, but the EC blocked adoption of both these reports as well. At the close of 1985, the net outcome of the five U.S. complaints remained unclear. The VAT complaint was alive, unsettled, and under discussion. There were some more positive signs of movement in the four other cases. The legal complaint in the wheat flour complaint seemed quite dead, but the EC was heard to claim that it was limiting its wheat and wheat flour exports to a specified percentage of the world marked (14 percent)—an assertion that sounded a good deal like a claim of compliance with some new "equitable share" standard (defined by the EC).35 The canned fruit complaint appears to have been settled, without adoption of the panel report, on the basis of the EC's agreement to eliminate that part of the production subsidy which exceeded the difference between CAP prices and world prices for raw materials. The impasse of the citrus case caused the United States to retaliate unilaterally on pasta products (thereby snatching a bit of relief for the pasta case as well), but this action was then met by a counterretaliation from the EC.36 As the year ended, both parties claimed to be making progress toward a negotiated solution that would settle both the citrus and pasta cases at the same time. The years from 1981 to the end of 1985 found the United States occupied with carrying out the litigation started in 1981-82. These drawn out legal battles seem to have satisfied whatever political needs for GATT litigation the U.S. government might have had during this period. Following the 1981-82 complaints, the United States filed only one new GATT complaint up to the end of 1985, and that one was only a reopening of an earlier 1978 complaint against Japan.37 The European Community, however, was not too busy to respond. Between March 1983 and the end of 1985, the Community took seven legal actions against the United States—five conventional GATT lawsuits and two instances of retaliation. The five lawsuits were: Import Restrictions on Printed Matter—the "Manufacturing Clause" (Art. XXIII) Subsidy on Exports of Wheat Flour to Egypt (Subsidies Code) Tariff Reclassification of Machine-Threshed Tobacco (Art. XXIII) Ban on Steel Pipe and Tube Imports (Art. XXIII) Countervailing Duty Action on Wine (Subsidies Code)38 The two retaliation actions were a 1984 action exercising compensation rights under Article XIX in response to a U.S. "escape clause" action

35

US-EC GATT Litigation

on specialty steels,39 and the action mentioned above in response to unilateral retaliation by the United States in the Citrus case.40 Of the five GATT lawsuits filed by the European Community, three led to no decision. The pipe and tube case became moot, and the Egypt wheat flour and machine-threshed tobacco cases have not been prosecuted. The wine case also became moot when the law expired without having produced any new restrictions, but the EC nonetheless obtained a panel ruling that the law in question was inconsistent with the Subsidies Code. (Later, the United States would block acceptance of this ruling in retaliation for EC blockage of the several earlier panel rulings favorable to the U.S.). The Manufacturing Clause complaint also produced a panel ruling that the U.S. law was in violation of GATT, and here the Community followed up the decision by making a productspecific threat of retaliation if the U.S. Congress extended the law past its scheduled expiration in June 1986. (The law was later allowed to expire.) In addition to these specific legal actions, the European Community made what appeared to be a major change of legal policy in 1984 when it adopted a procedure modeled after the infamous (in Community terms, anyway) U.S. section 301 procedure.41 The new procedure, called the New Commercial Policy Instrument, permits private citizens to complain about GATT violations of other countries, and establishes a series of steps that could well force EC officials to respond with GATT complaints about such matters, in much the same fashion as section 301 does. As the curtain comes down on the first 25 years of US-EC litigation, the final scene shows neither party in retreat from the legal wars of the early 1980s. Both the European Community and the United States remain clad in full legal armor, both seemingly prepared and waiting for bigger and better GATT litigation in the years to come. The Community's New Commercial Policy Instrument is reportedly about to yield its first GATT complaint.42 Meanwhile, the U.S. Congress continues to grind out legislative proposals for stronger and stronger GATT enforcement procedures, and the executive branch has created a "strike force" looking for new GATT violations and a "war chest" to help induce discipline in subsidy matters.43 Both sides have agreed that the GATT dispute settlement is a problem that should be included on the agenda of any forthcoming round of GATT trade negotiations. From what has been seen so far, however, it is not at all certain that they have the same "problem" in mind. 2.3.2 The Underlying Reasons for US-EC Litigation What explanation can be given for the explosion of difficult and contentious GATT litigation in the early 1980s? More important, what

36

Robert E. Hudec

does the experience mean about the likely course of future legal relations between the United States and the European Community? Will the United States continue to follow the aggressive litigation practices it has followed in the past? And what legal policy will the European Community follow in the years ahead? The present section examines the factors behind the litigation just described, with an eye to answering these questions about the future. The Experience of the United States The one lesson that emerges most clearly from the history of the past 25 years is that GATT litigation has come to play an important role in maintaining political support for liberal trade policy in the United States. Throughout this period, the U.S. Congress has been willing to enact legislation authorizing trade negotiations and to resist most protectionist initiatives, but on each occasion the price for such liberal policies has been progressively more rigorous undertakings to enforce GATT obligations against other governments. The political importance of such enforcement mechanisms has reached the point where, in 1985, virtually every piece of new trade legislation introduced in Congress has contained some provision calling for more vigorous GATT enforcement.44 The political importance of enforcing GATT legal rights grows out of the public character of trade policy politics in the United States. Unlike arcane monetary policy issues, which are little understood, trade policy issues have a direct impact that is perceived by almost everyone. As a consequence, political leaders are held rather strictly accountable for what they do on trade policy matters. Leaders need little further justification for supporting protectionist positions, but they do need to justify support for liberal trade policies. The two concepts that work best as justifications in United States politics are "reciprocity" and "fair trade." Neither of these concepts ranks very high on the scale of economic rationality, but both have proved their effectiveness politically many times over. GATT legal enforcement happens to be an effective way for U.S. officials to demonstrate, in a politically credible manner, the existence of both reciprocity and fairness. The reciprocity justification is simple. To explain why foreign producers should be given trade opportunities in the U.S. market, political leaders must show that similar opportunities have been given to U.S. producers in foreign markets. The first step is to show that promises of equal access have been obtained from other governments. The second step is to show that such promises are actually being carried out. This second step has become a difficult hurdle in U.S. politics. Fact and myth have combined to create a fairly widespread belief that the U.S. executive branch fails to enforce most of the legal rights it receives

37

US-EC GATT Litigation

in trade negotiations—in order to maintain good political relations. GATT litigation is frequently used to answer these doubts. Although one or two lawsuits hardly proves anything about a government's attitude toward the hundreds of other enforcement issues that arise each year, they are nonetheless a useful symbol to back up declarations of good intentions. They give the executive branch something tangible to show to Congress and, perhaps more important, something tangible for Congress to show to its constituents. The importance of GATT lawsuits in demonstrating reciprocity has grown considerably in the past two decades due to the increasing importance of nontariff barriers in world trade. Nontariff barriers cannot be traded for one another the way that tariffs can. There is no common measure, for example, by which to trade the relaxing of an unduly restrictive safety standard on baseball bats for limiting the scope of an overly broad countervailing duty law. The only credible form of reciprocity in nontariff barrier negotiations is the reciprocity that comes from each country's adherence to the same set of legal obligations pertaining to each nontariff barrier. Such rule-compliance reciprocity is difficult to demonstrate, however, because most nontariff barrier rules consist of prohibitions, and it is always difficult to demonstrate that a government is not doing something. Lawsuits tend to meet this problem by providing a tangible sort of evidence that a rule-enforcement mechanism does exist, and that the mechanism is working—at least this once. In addition to providing a reciprocity justification, political leaders in the United States have also found it increasingly necessary to demonstrate that trade liberalization will take place only with respect to "fair" trade, and that U.S. producers will be protected from competition with various kinds of "unfair" trade. The primary source of such assurances has been the enactment of laws that stop unfair trade at the border, such as antidumping, antisubsidy, and similar anti-unfair trade remedies. In recent years, however, effective GATT regulation of subsidies has also been demanded, particularly as a political quid pro quo for curbing the excess rigor of the U.S. countervailing duty law. This is the reason for U.S. insistence on the Tokyo Round Subsidies Code, and for the fact, noted in section 2.2 of this paper, that subsidies have played such an important role in the history of GATT litigation to date. The political role served by GATT litigation in the United States explains the general litigiousness of the United States during this period. More must be said, however, about the large number of legal claims against the EC, and in particular the rash of legal claims that precipitated the litigation impasse in the early 1980s. The data in section 2.2 showed that the European Community was the most frequent defendant in U.S. lawsuits, and that measures af-

38

Robert E. Hudec

fecting agricultural trade were at issue in the great majority of these cases. In one respect, this was a perfectly understandable choice. Given a political need to litigate against trade barriers in order to demonstrate reciprocity, the EC's Common Agricultural Policy was an obvious target. Although the CAP was certainly not the only trade distortion in the world, it had three characteristics that made it politically the most visible: (1) it was large; (2) it involved agriculture;45 and (3) it was new. Novelty was perhaps the most important characteristic, for the loss of something currently possessed is always the loss most keenly felt. The CAP presented a new threat to a wide range of existing and potential U.S. export markets—both markets inside the EC and third-country markets where surpluses created by the CAP were likely to be disposed of. As politically irritating as the CAP was, however, it was not so completely damaging to U.S. interests that the U.S. government was prepared to risk a trade war over it. Many U.S. officials were persuaded by EC arguments that the CAP was the only cement that could hold the European Community together. For those who believed that the existence of a healthy European Community was of paramount importance to the long term security interests of the United States, acceptance of the CAP was thus a geopolitical necessity. Moreover, many sectors of the U.S. economy were benefitting from trade and investment opportunities created by the formation of the Community, and these sectors were also a political force of some importance opposed to risking trade war for the sake of lost agricultural exports. There were even some important agricultural interests on this side, for the CAP did not close all markets to exporters of U.S. agricultural products. United States agricultural exports to the Community actually grew, in round numbers, from $1 billion in 1960 to over $9 billion in 1980; these exports gave the United States a substantial export surplus in its agricultural trade with the Community throughout the period.46 The U.S. legal attacks on the CAP during this period can be understood only when viewed in terms of these opposing U.S. interests. The GATT lawsuits must be viewed, not as a simple attack on the CAP, but rather as an attempt to steer a middle course between the conflicting U.S. interests on this problem. The U.S. government was required to act vigorously in defending the agricultural interests injured by the CAP, for failure to do so would have been seen as a failure to protect U.S. reciprocity interests, and this would have caused a serious loss of political support for the liberal orientation of U.S. trade policy. If the U.S. government pressed too hard, however, it could very well damage the many other important economic and geopolitical interests benefitting from maintenance of the status quo. As the pressure mounted,

39

US-EC GATT Litigation

it would become clear that these latter interests were decidedly the stronger. GATT litigation provided the sort of compromise policy instrument that was needed in this situation. GATT lawsuits could achieve some positive results for the threatened U.S. interests. Legal pressure could in fact accomplish something: It could sometimes deter protectionist extensions of the CAP that were not essential to its basic operations. Indeed, on occasion GATT lawsuits might even be able to induce some self-restraint on larger issues by creating "events" that would focus pressure of all kinds on the problem in question. Most of the GATT litigation in question was in fact a good faith effort to make use of GATT legal forces in this manner. But—and this is the critical point—lawsuits were essentially a soft response. Despite their aggressive tenor, lawsuits themselves are just words; they merely threaten hostile action, putting off the actual implementation until the final decision is made, usually a year or two after the lawsuit is filed. Consequently, even on issues where the EC was unwilling to move at all, GATT lawsuits were a good way to buy time. They were aggressive enough to satisfy the political need for a reciprocity-protecting action, and yet did involve actual warfare. In other words, it would often make sense to file even a hopeless GATT lawsuit, because it would allow the U.S. government to avoid, for a while at least, the possibly damaging political reactions of those domestic interests who were going to be disappointed. This second part of the process may well have been the most valuable in terms of maintaining an open and liberal trade policy relationship between the United States and the EC. If a fully grown CAP, 1985 edition, had been presented to U.S. agricultural interests in 1961, the shock would probably have been too great to permit maintenance of normal trade relations. As it happened, of course, the CAP was revealed only in stages. Its economic effects could thus be accepted in smaller bites, each time with an assurance that this was as far as the United States would go. GATT lawsuits helped to convey such an assurance in a politically credible manner. In some instances, such delaying effects would solve the problem by themselves, for time alone can sometimes lead injured parties to accept their losses and to lose interest in further complaints. But even where the hurt persisted, the delays would serve to put off the day of reckoning for a while, thereby allowing normal relations to continue in the meanwhile. The U.S. GATT litigation during this period can be viewed in terms of this contain-and-retreat strategy. The first line of containment, once the U.S. government had accepted the basic design of CAP and its variable levy, was the effort during the early 1960s to persuade the

40

Robert E. Hudec

Community to limit the level of the variable levy, to keep support prices down, and to use production controls to manage supply. This was the objective of the Chicken War retaliation and of the Kennedy Round negotiating efforts. When this approach failed, U.S. officials drew two other "lines in the sand" to assure U.S. interests that the CAP would not be allowed to expand beyond the variable levy regimes protecting the European market. The second line of containment would be U.S. insistence upon preserving GATT bindings and other GATT disciplines on those products not covered by variable levy regimes (chiefly certain animal feeds and certain processed products). The third line would be insistence the effects of the CAP be limited to the EC market itself, so that they would not distort third-country export markets. The effort to hold the second line was evident in many of the GATT lawsuits challenging EC measures inconsistent with GATT bindings and obligations. A number of relatively small legal victories during the 1970s permitted U.S. officials to claim that the line was being held. In the case of animal feeds, this was generally true, for exports rose throughout the period. For processed products, there were problems. Given the very high primary product prices created by the CAP, it was unrealistic to expect that the EC would agree not to assist food processors disadvantaged by those abnormal raw material costs. Consequently, when the issue was finally forced directly, in the canned fruit and pasta lawsuits of 1982, the U.S. legal claims were almost guaranteed to produce an impasse. And, of course, they did. By the end of 1985, it looked as though the United States was prepared to make a partial retreat, at least to the extent of accepting EC assistance to food processors that reduced raw material costs to world price levels. The third line of containment, the defense of third-country markets against the subsidized export of EC surpluses, was initially to be held by GATT Article XVI:3 obligations on export subsidies. When EC export subsidies began to run wild in the early 1970s, the effort shifted to negotiating new and stronger legal prohibitions against export subsidies in the Tokyo Round Subsidies Code. Once again, of course, the idea of containing CAP surplus disposal without containing prices and production was wishful thinking. This became quite apparent in the late 1970s, when the EC indicated that it was unwilling to accept any new discipline in the Subsidies Code, and, indeed, was not even willing to sign the rather weak Subsidies Code it had helped to draft without first obtaining a private assurance of benign U.S. intentions (the notso-secret Strauss/Gundelach letter).47 The EC's unwillingness to limit its export subsidies was driven home in 1978 when the EC successfully "stonewalled" a pair of rather strong legal claims by Australia and Brazil challenging CAP export subsidies on sugar under the old Article XVI:3 rules.48

41

US-EC GATT Litigation

United States officials continued to claim, nevertheless, that the Subsidies Code had improved GATT discipline over the CAP. This somewhat dubious claim of success no doubt seemed justified by the political situation at the time, for without some show of "reciprocity" in the new Subsidies Code the U.S. Congress would not have been able to justify legislation implementing the many important trade agreements negotiated in the Tokyo Round, including much-needed reforms in the U.S. countervailing duty statute. Eventually, however, the claim of legal containment had to be demonstrated. The 1981-82 GATT lawsuits were the effort to provide that demonstration. Some U.S. officials may actually have believed that the Tokyo Round Subsidies Code had created workable new obligations. Others may have known better but were hoping that litigation could somehow achieve what the Tokyo Round had not. But even if there had been no hope at all, the lawsuits would have been necessary. In the actual context, the flurry of U.S. GATT lawsuits in 1981-82 can be viewed as a rather moderate action. The Tokyo Round had created expectations of changes in EC policy that had not occurred. The economic harms being caused by the CAP were growing, particularly in U.S. export markets in third countries. The situation had produced strong political pressure for some kind of vigorous response— something more than another round of tea-and-cookies diplomacy. GATT lawsuits were a response vigorous enough to satisfy the call for action, but still short of real economic warfare. They did not solve the problem, but they did buy more time. Unfortunately, by the end of 1985 the process of accommodating the conflicting U.S. attitudes towards the CAP had not yet played itself out, and so it is not possible to know whether the role played by GATT litigation will ultimately be a positive one. It may be that the promises of legal enforcement used to buy time in the short run will turn out to have exacerbated the reaction in the long run when it becomes clear that legal containment has not occurred. If so, the eventual reaction against the CAP may be more violent than it would otherwise have been. On the other hand, there are signs that the process of gradual accommodation may still be working. Even though most of the key lawsuits themselves have ended in legal impasse, they have produced changes in some aspects of CAP policy. So far, these changes have involved relatively small adjustments that in no way alter the basic course of the CAP. But the final settlements may turn out to be just meaningful enough to keep alive the idea that GATT law, used together with other instruments of diplomatic pressure, can impose limits worth having. Moreover, there is another round of negotiations in the offing, and these negotiations should produce at least something in the way

42

Robert £ . Hudec

of new agreements and new legal reforms to bolster claims of better enforcement in the future. It is still quite possible, therefore, that more lines-in-the-sand will be drawn, that a few more interests on the U.S. side will learn to live with disappointed expectations, and that the process of adjusting to the CAP will continue along much the same path as before. Up to this point, therefore, the balance sheet on US-EC GATT litigation does not appear to be as unfavorable as the world has been led to believe. The effects of any litigation have to be appraised against the background of realistic alternatives. Given a setting in which the EC's implementation of the CAP was causing major economic harms to U.S. export interests, it is difficult to avoid the conclusion that, until now, the U.S. lawsuits probably have bought more peace than war.49 The issue for the future will be whether the procedure can retain enough political credibility to continue performing this function. The Experience of the European Community Almost as remarkable as the United States' litigiousness during these years was the European Community's reluctance to become an active litigant. Throughout the 1970s, the Community seemed willing to submit to a rather considerable number of legal complaints without responding in kind, not only complaints from the United States but from other GATT governments as well. The DISC case was one instance of legal retaliation, but there were few others. Great powers do not usually accept such legal mistreatment so quietly. The Community explained its reluctance to litigate in terms of its basic approach to GATT and to international economic relations in general. Time and again the EC would lecture the United States and others on the theme that diplomacy was the best means to conduct trade relations. Lawsuits, the Community would argue, do not solve the economic and social problems at the root of most trade problems. The U.S. attempt to "legalize" GATT was, sad to say, merely the warped thinking of a naive and lawyer-ridden society. It happened, however, that this antilegalism also suited the EC's particular legal situation at the time. Although the EC's architects had done a reasonably good job of conforming to GATT rules, there were still several aspects of EC policy that did not fit comfortably within those rules. The Common Agricultural Policy had potential problems, both in the area of subsidies and with respect to the many consequential measures that might be needed to adjust for the impact of the CAP in peripheral product areas. In addition, a series of discriminatory arrangements with countries of Africa and the Mediterranean raised serious problems of compliance with GATT's MFN obligation. And finally, there were likely to be a number of corners to be cut as the EC expanded

43

US-EC GATT Litigation

to include new members, and also arranged economic relations with the countries of Europe who were not members. It was, in short, a time when a major realignment of European trade relationships was in process, and thus not a time when the Community wanted to encourage attention to legal obligations that defined the old order. If one accepts this analysis of the reasons for EC legal policy in the 1970s, the next question to ask is what has happened to that policy in the 1980s. Were the EC lawsuits of the 1980s just an outburst of irritation not likely to be repeated? Or were they a more deliberate strategy designed to dampen the litigation ardor of the United States and other GATT legalists? Or, perchance, has the EC policy changed in some more fundamental way? Is it possible that the EC has begun to believe in the efficacy of GATT law? For the present, it is probably wisest to remain rather skeptical about the possibility of any fundamental change in legal policy. Although the Community's early policy of antilegalism was certainly convenient, that does not mean it lacked conviction. To the contrary, it rested on longstanding traditions of economic diplomacy that will probably be very slow to change. In addition, although the European Community may no longer suffer from the exceptional number of GATT legal problems it once had,50 the Community still has more to fear from international legal obligations than do most other GATT members. The Community is a hothouse institution created by treaty rather than a sovereign nation-state. As a consequence, its internal powers are more exposed to charges of legal irregularity than is true of most other GATT members. International obligations tend to have greater legal effects internally, and once such obligations are found applicable it is much more difficult for Community organs to change them. The first few court decisions on GATT have indicated that at least some GATT norms do not have the kind of legal status that is needed to become fully binding obligations within the EC's internal law. But EC jurisprudence has not yet pronounced its final word on this subject. Some legal scholars are currently arguing that GATT law should be given a much higher place—that it should be considered binding on Community institutions as a way of making them conform to the original economic policy goals of the Rome Treaty.51 Anything that made GATT obligations more definitive might well encourage developments in this radical direction. Consequently, as long as the status of GATT within EC law is at all unsettled, it is to be expected that the political leadership of the Community will do nothing that would augment the legal status of GATT itself. If this is a correct analysis of Community legal objectives, what can be the reason for the Community's vigorous legal activities in the 1980s? The answer would seem to be that these legal activities have simply

44

Robert £ . Hudec

been a more vigorous form of defense against the more vigorous U.S. litigation policies of the early 1980s. In the terminology of the sport of boxing, the EC lawsuits would be called counterpunching. This reactive and defensive character seemed particularly clear in a recent US-EC exchange on legal policy in December 1985. Commissioner for External Relations Willy de Clerq presented U.S. Trade Representative Clayton Yeutter with a "very long and technical" list of U.S. trade practice that the EC Commission considered "unfair." According to news reports, the list had been prepared partially in response to a similar list of EC legal sins issued by the United States a few months earlier. A statement issued by Commissioner de Clerq explained that his own list was meant to show that "the United States does not have a prerogative on fair trade." News reports also quoted de Clerq as saying, "I underlined to Mr. Yeutter the Community view that it is necessary that the notion of fair trade be applied in the same manner on both sides of the Atlantic."52 It is doubtful that the EC's counterpunching strategy will reduce the volume of U.S. litigation very much. The analysis of U.S. policy in the previous section argues that U.S. GATT litigation is not the product of fuzzy legalistic thinking, but that it is rather a response to basic political needs in the conduct of U.S. trade policy. Counterpunching will not make those political needs go away. The most likely outcome in the near future, therefore, is that the United States will continue punching, and the Community will continue counterpunching, and together the two of them will continue to generate a substantial quantity of GATT litigation. Some Thoughts about the Future The analysis in this paper suggests that the United States and the EC are likely to continue to engage in frequent GATT litigation, but with neither side having its heart in quite the right place. The United States will often be litigating for the purpose of meeting certain political needs back home, and the Community will be litigating primarily for defensive purposes, without really believing in the process. The analysis suggests that these not-quite-real lawsuits may have been less irritating to US-EC trade relations than is commonly supposed. To the contrary, it appears that such lawsuits may well have been providing a peaceful alternative to real economic warfare. But there are clearly serious problems on the horizon. There is certainly reason to wonder whether, in view of the current state of legal impasse, GATT litigation can continue to retain its political credibility. It would be premature, however, to sign a death certificate. Despite the impasse to date, it may yet be possible to achieve certain meaningful results. In addition, GATT governments appear ready to

45

US-EC GATT Litigation

undertake further efforts to make GATT law work better, with a great deal of official optimism that improvements can be made. The optimism is significant. Political leaders' belief in the possibility of further GATT legal reform seems to be as durable as their belief in the value of "reciprocity"—and for the same reason. Whether or not the belief is true, believing leads to better outcomes than not believing. The other serious danger in the present situation is that the use (or misuse) of GATT legal procedures in these not-quite-real lawsuits will damage the long-term development of GATT law. Here again, however, the situation also has some brighter possibilities. The failures of USEC litigation during the past decade have already stimulated a number of procedural reforms that have made GATT litigation work better, in routine cases, than it did ten years ago. The need to keep GATT litigation credible will probably cause the strengthening process to continue. It is possible, therefore, that the well-functioning side of GATT law will be able to continue building on its already impressive record, and that one day governments will wake up to find that they have created a stronger legal institution in spite of themselves.

Appendix A GAIT Litigation,

1960-1985

There is no official classification of GATT "lawsuits" nor have scholars adopted any common definition. In this paper, the author uses the term to denote any GATT proceeding in which one GATT member has attempted to obtain an authoritative legal ruling that another member's action is either (1) in violation of GATT law or (2) has "impaired" the value of GATT rights (a special kind of GATT legal claim). The term is limited to legal claims against specified governments; it does not include requests for more general kinds of legal rulings. It includes all cases in which the complaining government began a lawsuit on the public record—i.e., took at least the first step in pursuit of such a legal ruling in a formal GATT proceeding or document. The list of lawsuits in this Appendix is based in part on data published in Hudec (1975, 227-96) and in part on the author's subsequent research, as yet unpublished except for a brief synopsis of the 1975-79 data in Hudec (1980, 200-203). This paper has reworked some of the previously published data, including a few cases not found in the earlier research and excluding several cases which, upon reexamination, could not be called a lawsuit under the criteria used here. Each entry in the Appendix presents a rather concise view of the lawsuit in question. The top line records the defendant, then a title

46

Robert £ . Hudec

describing the measure complained of, then the plaintiff, and finally the date of the first public complaint. The bottom line contains a pair of symbols describing the author's classification of the case according to two categories—the product sector affected by the trade measure complained of, and the nature of the trade measure itself. The symbol before the slash records the product sector: A = Agricultural and Fisheries I = Industrial G = General The symbol after the slash records the type of measure: T = Tariff T — D = Discriminatory Tariff NT = Nontariff Measure NT - D = Discriminatory Nontariff Measure S = Subsidy The classifications describe the primary focus of the case, and do not mean that other elements may not have been present.

Title

Complainant

Date

1. Fifteen developed countries and EC

"Recourse to Article XXIII" (effect of 562 restrictions on Uruguayan exports; request for ruling on variable levies) [A / not classified]

Uruguay

1961

2. U.K.

Tariff preference on bananas [A / T-D]

Brazil

1961

3. France

Residual BOP restrictions [A/NT]

U.S.

1962

4. Italy

Residual BOP restrictions [A/NT]

U.S.

1962

5. Canada

Antidumping duty on potatoes [A/NT]

U.S.

1962

6. [EC/US]

Joint submission: Trade value of U.S. withdrawal rights to compensate for EC withdrawal of poultry concessions [A/T]

[EC/US]

1963

7. Greece

Tariff preferences to USSR (on 30 industrial products) [I / T-D]

U.S.

1970

8. EC

Import restrictions on apples [A/NT]

Australia

1970

Defendant

47

US-EC GATT Litigation

Defendant

Title

Complainant

Date

Quota restrictions on grain [A/NT]

U.S.

1970

10. Jamaica

Excessive margins of preference [G / T-D]

U.S.

1970

11. EC

Compensatory taxes on imports [A/NT]

U.S.

1972

12. U.K.

Import restrictions on textiles [I/NT]

Israel

1972

13. France

Import restrictions (proposal for retaliation in complaint no. 3) [A/NT]

U.S.

1972

14. U.K.

Quotas on dollar area imports [A / NT-D]

U.S.

1972

15. U.S.

DISC tax legislation [G/S]

EC

1973

16. France

Income tax practices [G/S]

U.S.

1973

17. Belgium

Income tax practices [G/S]

U.S.

1973

18. Netherlands

Income tax practices [G/S]

U.S.

1973

19. EC

Adequacy of Article XXIV:6 compensation [A/T]

Canada

1974

20. Japan

Import restrictions of beef [A/NT]

Australia

1974

21. Canada

Import quotas on eggs [A/NT]

U.S.

1975

22. EC

Minimum import prices, etc. [A/NT]

U.S.

1976

23. EC

Minimum import prices, etc. (same as complaint no. 22)

Australia

1976

24. EC

Mixing requirement for imports of animal feed [A / NT]

U.S.

1976

25. Canada

Article XXVIII:3 retaliation [I/T]

EC

1976

26. U.S.

Ruling on definition of subsidy {Zenith case) [I / NT]

Japan

1977

9. Denmark

48

Robert E. Hudec

Defendant

Title

Complainant

Date

27. Japan

Import restrictions on thrown silk [I / NT-D]

U.S.

1977

28. EC

Export subsidies on malted barley [A/S]

Chile

1977

29. EC

Discriminatory import restrictions on TVs [I / NT-D]

Korea

1978

30. Norway

Discriminatory import restrictions on textiles [I / NT-D]

Hong Kong

1978

31. Japan

Import restrictions on leather [I/NT]

U.S.

1978

32. EC

Export subsidies on sugar [A/S]

Australia

1978

33. EC

Export subsidies on sugar [A/S]

Brazil

1978

34. U.S.

Countervailing duty actions [G/NT]

EC

1978

35. EC

Discriminatory import restrictions on apples [A / NT-D]

Chile

1979

36. Japan

Import restrictions on leather [I/NT]

Canada

1979

37. Spain

Restrictions on domestic sale of soybean oil [A/NT]

U.S.

1979

38. Japan

Restraints on import and sale of manufactured tobacco [I/NT]

U.S.

1979

39. U.S.

Import prohibition on tuna [I / NT-D]

Canada

1980

40. Spain

Discriminatory tariff treatment of coffee [A / T-D]

Brazil

1980

41. EC

Discriminatory import restrictions on beef [A / T-D]

Canada

1980

42. Japan

Import restrictions on leather [I/NT]

India

1980

43. EC

Internal restraints on imports of poultry (Spin Chill case) [A/NT]

U.S.

1980

49

US-EC GATT Litigation

Defendant

Title

44. U.S.

Discriminatory application of injury test in CVD cases [I / NT-D]

India

1980

45. U.S.

Import duty on vitamin B-12 [I / T]

EC

1981

46. EC

Production subsidies on canned fruit [A/S]

Australia

1981

47. U.S.

§337 restraints on imports of automobile spring assemblies [I/NT]

Canada

1981

48. EC

Export subsidy on wheat flour (Subsidies Code complaint) [A/S]

U.S.

1981

49. EC

Import restrictions on watches, radios, etc. [I / NT-D]

Hong Kong

1981

50. EC

Export subsidy on pasta (Subsidies Code complaint) [A/S]

U.S.

1982

51. EC

Export subsidy on poultry (Subsidies Code complaint) [A/S]

U.S.

1982

52. EC

Production subsidy on canned fruit and raisins [A/S]

U.S.

1982

53. Canada

Restrictions on imports in Foreign Investment Review Act [I/NT]

U.S.

1982

54. EC

Export subsidy on sugar (Subsidies Code complaint) [A/S]

U.S.

1982

55. EC

Sugar regime (export subsidy) [A/S]

Argentina Australia Brazil Colombia Cuba Dominican Republic India Nicaragua Peru Philippines

1982

56. EC

Falklands War embargo [G / NT-D]

Argentina

1982

Complainant

Date

50

Robert E. Hudec

Defendant

Title

Complainant

Date

57. EC

Preferential tariff on citrus [A /T-D]

U.S.

1982

58. EC

Treatment of VAT in price calculations (Procurement Code complaint) [I/NT]

U.S.

1982

59. Brazil

Export subsidy on poultry (Subsidies Code complaint) [A/S]

U.S.

1982

60. Japan

Internal regulations on softball bats (Standards Code complaint) [I/NT]

U.S.

1982

61. Finland

Internal regulations on footwear [I/NT]

EC

1982

62. Switzerland

Article XIX measures on table grapes [A/T]

EC

1982

63. U.S.

Denial of MFN tariff treatment [G / T-D]

Poland

1982

64. EC

Restraints on imports of VTRs [I/NT]

Japan

1982

65. Japan

Import restrictions on leather (revival of complaint no. 31) [I/NT]

U.S.

1983

66. U.S.

Import restrictions on printed matter ("Manufacturing Clause") [I/NT]

EC

1983

67. U.S.

Export subsidy on wheat flour sales to Egypt (Subsidies Code complaint) [A/S]

EC

1983

68. Japan

Nullification and impairment of benefits in general [G / not classified]

EC

1983

69. U.S.

Discriminatory import restrictions on sugar [A / NT-D]

Nicaragua

1983

70. U.S.

Tariff reclassification of machine-threshed tobacco [A/T]

EC

1983

51

US-EC GATT Litigation

Defendant

Title

Complainant

Date

71. Canada

Antidumping investigation of electrical generators from Italy (Antidumping Code complaint) [I/NT]

EC

1983

72. EC

Import restrictions on newsprint [I/NT]

Canada

1984

73. Chile

Measures on dairy imports [A/NT]

EC

1984

74. Canada

Internal tax on gold coins [I/NT]

South Africa

1984

75. New Zealand

Antidumping duties on electric transformers [I/NT]

Finland

1984

76. EC

Operation of beef and veal regime [A/NT]

Australia

1984

77. U.S.

Ban on steel pipe and tube imports [I / NT-D]

EC

1984

78. Canada

Internal regulations on alcoholic beverages [I / NT]

EC

1985

79. U.S.

Import restrictions on products containing sugar [A/NT]

Canada

1985

80. U.S.

Countervailing duty action on wine (Subsidies Code complaint) [A / NT]

EC

1985

52

Robert £ . Hudec

Appendix B US-EC GATT Litigation, Defendant

1960-1985

Title

Complainant

Date

3. France

Residual BOP restrictions [A/NT]

U.S.

1962

4. Italy

Residual BOP restrictions [A/NT]

U.S.

1962

6. [EC/US]

Joint submission: Trade value of U.S. withdrawal rights to compensate for EC withdrawal of poultry concession [A/T]

[EC/US]

1963

11. EC

Compensatory taxes on imports [A/NT]

U.S.

1972

13. France

Import restrictions (proposal for retaliation in complaint no. 3) [A/NT]

U.S.

1972

15. U.S.

DISC tax legislation [G/S]

EC

1973

16. France

Income tax practices [G/S]

U.S.

1973

17. Belgium

Income tax practices [G/S]

U.S.

1973

18. Netherlands

Income tax practices [G/S]

U.S.

1973

22. EC

Minimum import prices, etc. [A/NT]

U.S.

1976

24. EC

Mixing requirement for imports of animal feed [A/NT]

U.S.

1976

34. U.S.

Countervailing duty actions [G/NT]

EC

1978

43. EC

Internal restraints on imports of poultry ("Spin Chill" case) [A/NT]

U.S.

1980

45. U.S.

Import duty on vitamin B-12 [I/T]

EC

1981

48. EC

Export subsidy on wheat flour (Subsidies Code complaint) [A/S]

U.S.

1981

50. EC

Export subsidy on pasta (Subsidies Code complaint) [A / S]

U.S.

1982

53

US-EC GATT Litigation

Defendant

Title

Complainant

Date

51. EC

Export subsidy on poultry (Subsidies Code complaint) [A/S]

U.S.

1982

52. EC

Production subsidy on canned fruit and raisins [A/S]

U.S.

1982

54. EC

Export subsidy on sugar (Subsidies Code complaint) [A/S]

U.S.

1982

57. EC

Preferential tariff on citrus [A / T-D]

U.S.

1982

58. EC

Treatment of VAT in price calculations (Procurement Code complaint) [I/NT]

U.S.

1982

66. U.S.

Import restrictions on printed matter ("Manufacturing Clause") [I/NT]

EC

1983

67. U.S.

Export subsidy on wheat flour sales to Egypt (Subsidies Code complaint) [A/S]

EC

1983

70. U.S.

Tariff reclassification of machine-threshed tobacco [A/T]

EC

1983

77. U.S.

Ban on steel pipe and tube imports [I / NT-D]

EC

1984

80. U.S.

Countervailing duty action on wine (Subsidies Code complaint) [A/NT]

EC

1985

Notes 1. In an earlier study using somewhat more inclusive criteria, Hudec (1975, 278-90), the author counted 54 lawsuits from 1948 through 1959. Appendix A of the present study, based partly on the earlier work and partly on unpublished research, counts 80 lawsuits from 1960 to the end of 1985. The introduction to Appendix A explains the author's definition of a GATT lawsuit. 2. The EC was created by the Treaty of Rome in 1957, but began to function only gradually. The key event in the EC's assumption of GATT legal responsibility was the 1960-61 Dillon Round trade negotiations in which the GATT

54

Robert E. Hudec

bindings of EC member country tariffs were replaced by bindings of the EC Common External Tariff. The year 1960 provides a convenient breaking point for there were no GATT lawsuits that year. 3. The nine cases not involving either the U.S. or the EC were complaints no. 2, 12, 20, 30, 36, 40, 42, 74, and 75. (Note that, here and throughout the study, suits against EC member states are treated as EC litigation only when the defendant was a member state at the time of the suit.) 4. It must be emphasized that this classification is based on the author's judgment as to the main product interest in the dispute, where such a product or product sector can be identified; it does not mean that no other products were involved. 5. Once again it must be emphasized that the classification is based on the author's judgment as to the primary subject of the complaint and does not mean that no other types of trade barriers were involved. With respect to discrimination in particular, it should be noted that cases were counted only if the complainant made a major legal issue of the fact; several cases involving barriers with elements of discrimination were not counted because not enough was made of that aspect. 6. The ten agricultural subsidy cases were complaints no. 28, 32, 33, 46, 48, 50, 51, 52, 54, and 55. The three other subsidy complaints against the EC were complaints no. 16, 17, and 18—the U.S. complaints about the subsidy effects of the "territoriality" principle in French, Belgian, and Netherlands income tax law. These three complaints were counterclaims in response to the EC's DISC complaint, discussed at notes 18-19 infra. 7. The complaints involving the U.S. countervailing duty law were complaints no. 26, 34, 44, and 80. 8. Complaint no. 67. 9. Complaint no. 15. 10. Complaints no. 57 and 77. 11. GATT, BISD, 11th Supp., pp. 100-101 (1963); id., 13th Supp., pp. 4 8 49 (1965). 12. Complaints no. 3 and 4. 13. The complaint returned to the GATT agenda ten years later. See complaint no. 13, discussed at note 17 infra. 14. Complaint no. 6. 15. The controversy is not listed as a lawsuit in Appendix A. It is described in Hudec (1975, 232). 16. Complaint no. 11. 17. Complaint no. 13. 18. Complaint no. 15. (DISC: Domestic International Sales Corporation) 19. Complaints no. 16, 17, and 18. Roughly speaking, the theory was that territorial systems permitted the shifting of some export income to "tax havens" and that DISC was merely an equivalent tax haven within the United States. 20. The United States itself had once permitted taxpayers to shift many kinds of income to off-shore (in other words, "territorial") tax havens, before outlawing this practice in its so-called Subchapter S reforms in the 1960s. 21. Section 301, Trade Act of 1974, as amended 19 USC 2411 et seq. 22. The seven complaints were complaints no, 21, 22, 24, 27, 31, 37, and 38. All but the last were the direct result of private companies filed under section 301. Since 1979, section 301 proceedings have yielded only four more new GATT complaints, all against the EC. They were the wheat flour, pasta, canned fruit,

55

US-EC GATT Litigation

and citrus cases, complaints no. 48, 50, 52, and 57, discussed at notes 33-36 infra. One other GATT lawsuit was filed pursuant to section 301 after 1979, but it was a 1983 complaint reopening an earlier lawsuit against Japan on leather goods (complaint no. 65); it was then extended in 1985 to some other leather products. (The 1985 extension is not listed as a separate complaint in Appendix A). 23. Complaints no. 22 and 24. 24. Complaint no. 34. 25. Instances of actual retaliation include: the Article XIX retaliation in response to Australia's escape clause measures on autos and shoes (see GATT DOCS. C/M/154, - 155, - 156 [1982] ); the Article XIX retaliation in response to the U.S. escape clause measures on specialty steels (see GATT DOC. L/5524/Add.l5 [1984]); and the "lemons and walnuts" counterretaliation in response to the U.S. unilateral retaliation on pasta products in the citrus case (see note 36 infra). Notable threats of retaliation included those directed at the 1982 U.S. countervailing duty proceeding on carbon steel, at the 1984 U.S. Wine Equity Act, and at the 1985-86 proposed legislation to extend the U.S. Manufacturing Clause law. 26. It is difficult for a U.S. observer to judge whether the Community's posture in these cases was really more belligerent than, say, the average U.S. reaction to such situations. It seems so to the author, but U.S. citizens tend not to take their own government's rhetoric very seriously. 27. Trade Agreements Act of 1979, Title IX, codified as 190 USC 2411 et seq. 28. "Understanding Regarding Notification, Consultation Dispute Settlement, and Surveillance," printed in GATT, BISD, 26th Supp., pp. 210-18 (1980). 29. The new panel procedures are described in detail in Hudec (1980). 30. Complaint no. 43. 31. Complaint no. 45. 32. For example, the B-12 problem is still listed as one of the unfair U.S. trade practices in the de Clerq memorandum of December 1985, described at note 52 infra. 33. The seven complaints are, respectively, complaints no. 48, 50, 51, 52, 54, 57, and 58. 34. The only GATT body with power to make legal rulings is the plenary assembly of GATT members, called the Contracting Parties, or its agent, the somewhat smaller executive assembly called the GATT Council. Typically, both bodies rule by "accepting" the reports and rulings prepared by the GATT panel that hears the case. By tradition, however, decisions of the Contracting Parties or the Council are taken by consensus, which means that any country, including the defendant, can block the decision to accept a panel report. This power to defeat legal rulings is held in check by another tradition that GATT bodies should not relitigate cases decided by panels, but should accept all but the most clearly erroneous decisions. Prior to 1976, this second tradition had been quite strong, with only one panel report ever having been rejected. Since then, however, there have been at least 12 instances where acceptance of a panel report has been blocked—in whole or in part—by one of the parties. All the rulings in the 4 DISC cases (complaints no. 15, 16, 17, and 18) were blocked by the losing party—the first for about five years and the other three permanently. After that, blocking actions by either the defendant or a losing plaintiff occurred in complaints no. 37, 47, 48, 50, 52, 57, 74, and 80. 35. See, for example, Brown (1985, 177).

56

Robert E. Hudec

36. The two retaliations went into effect in a complex series of halting, linked steps that stretched over a good part of 1985. The United States raised duties to prohibitive levels against pasta (ameliorated somewhat by exchange rate behavior), and the Community responded in kind on walnuts and lemons, items that represented an equivalent amount of trade. See International Trade Reporter, vol. 2 (Washington, D.C.: Bureau of National Affairs, 1985), pp. 835, 898, 1131, 1389-90. 37. Complaint no. 65. Another leather product was added to the complaint in 1985. 38. The five complaints are, respectively, complaints no. 66, 67, 70, 77, and 80. 39. See note 25 supra. 40. See note 36 supra. 41. Council Regulation (EEC) No. 2641/84, of 17 September 1984, O.J. (1984) L252/1. For one recent discussion of the new instrument, see Van Bael and Bellis (1985, 197-216). 42. The subject of the complaint is the United States' section 337 remedy which permits restrictions on imports accused of patent infringement and other "unfair" characteristics. 43. The strike force was created in September 1985 as part of an executive branch campaign to counter a tidal wave of protectionist sentiment in the Congress—another case, by the way, of how promises of legal enforcement are used to justify support for liberal trade policy positions. For an amusing account of the first three months of strike force activities, see "Trade Strike Force, Hamstrung by Turf Battles, May Prove Political Liability to Administration," Wall St. Journal, 7 January 1986, p. 60. For two legislative proposals to fund a war chest to match foreign subsidies, see H.R. 3296 and H.R. 3667, 99th Cong., 1st Sess. (1985). 44. See, for example, the list of pending bills on this one subject alone in International Trade Reporter, vol. 3 (Washington, D.C.: Bureau of National Affairs, 1986) pp. 75-76. 45. It is a commonplace among trade policy experts that agricultural interests have political clout. This political clout is usually observed in the context of rent-seeking at home, where agricultural interests appear to have better-thanaverage success in persuading their governments to grant benefits such as price or income supports, protection against import competition, or export subsidies. But that same clout can also be expected to earn better-than-average results when agricultural interests believe their export markets are being damaged by foreign actions like the CAP, and ask their government to do something about it. Clout is clout, whatever the issue. 46. For a discussion of these trade figures, see Talbot (1985, 39) and also Brown (1985, 176). The $9 billion figure was a peak, but U.S. exports in the following four years averaged well above $7 billion. 47. The author has never seen the letter, but he has heard descriptions from both EC and U.S. officials whom he has interviewed. As might be expected, each side has a somewhat different description of the contents. One side views the letter as a U.S. undertaking not to challenge the level of CAP exports as they stood in 1979; the other side views it as a more general assurance that the U.S. was not seeking to destroy the CAP as such. Given the EC's knowledge that the letter was not to be made public, the EC would not have been justified in treating it, whatever its contents, as a meaningful commitment. The letter belongs more to the category of symbolic gestures salvaged by the losing side on an issue. Its significance here is simply that it was needed at all—that after

57

US-EC GATT Litigation

having insisted that the 1979 Subsidies Code be limited to a rather tame repetition and elaboration of existing GATT obligations, the EC was still unable to give an unqualified commitment to observe it. 48. Complaints no. 32 and 33. It is quite interesting to note that the United States, usually an active voice in most GATT legal proceedings, sat rather conspicuously on its hands during the litigation phase of this case—up to the end of 1980. It would appear that the highest priority for the U.S. delegation during this time was to obtain the EC's signature on the Subsidies Code, an objective that required tender treatment. The United States finally joined the fray in the working party consultations that followed the litigation, but by then the game had been lost. Indeed, a good case can be made that the subsequent wheat flour case was lost on these playing fields as well. 49. The use of GATT litigation as a peace-keeping device should be interesting to social scientists interested in primitive legal societies. One cannot help being struck by the similarity between the legal behavior of the GATT governments described in this paper and the behavior of primitive Anglo-Saxon clans described by Malone (1970, 1): The primordial seed from which [legal actions of] both crime and tort were to germinate was the blood feud that was characteristic of any barbaric society organized along lines of blood kinship. The defense of the honor of the clan by resort to warfare against the harm-inflicting outsider and his entire kin was a traditional practice with roots deep in the need for survival of the family unit. The outrage that cried for revenge lay not so much in the desire to enforce atonement for the bodily harm inflicted upon the wounded family member as in the humiliation that was suffered by his entire kin group. The primary object of law was to provide a substitute for the feud; and, as would be expected, the remedy that eventually emerged was of a character to offer balm where the hurt was deepest—in the clan's profound sense of indignity. 50. The problem of discriminatory favors for developing countries has been largely taken off the legal agenda by the GATT's adoption of the Enabling Clause in 1979 at the end of the Tokyo Round, and by the considerable amount of ad hoc discrimination now practiced by most other developed countries— for example, the current United States and Canadian policies of granting special trade favors to certain Caribbean nations. As for other legal problems that appeared significant in the 1970s, most have by now been addressed and debated, usually inconclusively, leaving most complainants with a sense that their complaints have achieved as much as possible. In addition, the EC practices in question are beginning to acquire the legitimacy of existence over time. Time, of course, is not a full guarantee of legal security, as was clearly shown by the 1982 U.S. complaint about long-standing EC citrus preferences (no. 57). 51. For a recent article reviewing the status of GATT obligations in the Community's internal law and arguing for a more prominent role, see Petersmann (1985). 52. The exchange is reported in International Trade Reporter, vol. 2 (Washington, D.C.: Bureau of National Affairs, 1985) pp. 1567-68; id., vol. 3 (1986), p. 50. Another fact which tends to confirm the Ec's less-than-complete legal commitment to GATT litigation is that, to the author's knowledge, all GATT litigation continues to be handled, not by the EC Commission's legal staff, but by policy officials in the Directorates General responsible for diplomatic representation of the EC on commercial policy matters.

58

Robert £ . Hudec

References Brown, Susan. 1985. Review. In Confrontation or negotiation: United States policy and European agriculture, ed. Randall B. Purcell. Millwood, N.Y.: Assoc. Faculty Press. Hudec, Robert E. 1975. The GATT legal system and world trade diplomacy. New York: Praeger. . 1980. GATT dispute settlement after the Tokyo round: An unfinished business. Cornell International Law Journal 13: 145-203. Malone, Wex S. 1970. Ruminations on the role of fault in the history of the common law of torts. Louisiana Law Review 31: 1-39. Petersmann, Ernst-Ulrich. 1985. International and European foreign trade law: GATT dispute settlement against the EEC. Common Market Law Review 22: 441-87. Talbot, Ross B. 1985. The foundations of the CAP and the development of USEC Agricultural Trade Relations. In Confrontation or negotiation: United States policy and European agriculture, ed. Randall B. Purcell. Millwood, N.Y.: Assoc. Faculty Press. Von Bael, Ivo, and Jean-Francois Bellis. 1985. International trade law and practice of the European Community. Bicester, Oxfordshire: CCH Editions Ltd.

Comment

Per Magnus Wijkman

Papers on GATT reform seem to fall into one of three interrelated categories. Most debate the need to change the rules of the trade game. In this category fall arcane papers on reform of the safeguard clause, the intricacies of "graduation," and the sophistry of conditional mostfavored-nation treatment. Others are concerned with the need to change the players in the game in order to restore a liberalizing momentum to the trade regime. Their authors claim that this momentum is exhausted because the existing distribution of bargaining power has deadlocked negotiations. Introducing new issues—or reviving old ones—could attract new players to the game and change the balance of negotiating power. In this category fall discussions in the corridors of power on whether to include services, high-tech goods, agriculture, and apparel in a new round of negotiations. Finally, we have a few papers discussing whether to change the referee in the game. Professor Hudec's paper falls in this exclusive category dealing with dispute settlement procedures. Its takeoff is his valuable statistical compilation that shows that half of the GATT lawsuits between 1960 and 1985 occurred during the

Per Magnus Wijkman is head of the Research Secretariat at the National Board of Trade, Stockholm, Sweden.

59

US-EC GATT Litigation

1980s. Robert Hudec's contribution is to explain why litigation has become so fashionable today. In the tradition of good detective novels, the author first presents the most litigious parties and quickly identifies the most frequent plaintiff—the United States—and the most frequent defendant—the EC. One third of GATT lawsuits were between these two protagonists. Thereafter, Hudec presents the most common subject matter of disputes: agriculture and nontariff barriers (NTBs). He shows that the EC has been sued primarily for its use of subsidies and NTBs to protect agriculture while the United States has been sued primarily because of its application of countervailing duties. Finally, he explains the motives and the behavior of these two major parties with the master detective's psychological insight. Hudec presents three explanations for the American administration's recourse to litigation, of which I find the third the most interesting. First, he suggests that the intention of American administrations is to achieve reciprocity in the term's new American sense. However, litigation based on this misinterpretation of reciprocity is doomed to be ineffectual. GATT's nondiscrimination principle ensures the United States of equal access to a foreign market only in the sense that U.S. goods need not pay higher tariffs than other foreign goods pay. It does not provide the United States with equal access in the sense that tariff levels on American goods exported to, e.g., Brazil, cannot be higher on average than tariffs on Brazilian exports to the United States. It is this latter discrepancy in terms of bilateral market access that has concerned the Americans recently. If the American objective is to "level the playing field," a policy of litigation is flawed. Tariff negotiations are the appropriate and classical remedy to this problem. Consequently and secondly, Hudec explains increased litigation by the spread of NTBs. This explanation is based on his explicit assumption that elimination of NTBs cannot be negotiated through reciprocally balanced concessions as can tariffs.1 Instead, general rules concerning their use must be formulated, after which litigation can be used to enforce rule observance. This observation is correct for some types of NTBs. However, it does not apply to quantitative restrictions. Grayzone measures, such as "voluntary" export restraints (VERs), normally reflect an out-of-court settlement, by which a country agrees to limit its exports to a particular market in exchange for part of the scarcity rents generated by restricting trade. The proliferation of VERs is not an explanation of litigation but an alternative to it. Hudec's second explanation thus only holds for other NTBs such as subsidies and public procurement, and these account for but half of GATT lawsuits. 1. For a different view see L. Alan Winters, "Negotiating the Abolition of Non-Tariff Barriers," Oxford Economic Papers 39, no. 3 (1987):465-80.

60

Robert £ . Hudec

Hudec's third explanation of the increased frequency of litigation is that American administrations have used it as a way to accommodate Congress and to placate domestic pressure groups. In particular, litigation over the common agricultural policy (CAP) has been a means to accustom constituencies to unpleasant commercial policy realities rather than a way to change these realities. This explanation dominates the first two and makes them superfluous. If administrations have been aware of their ineffectiveness in modifying the CAP, they have been conducting a policy to deceive pressure groups. If unaware, they have been engaged in a policy of self-deception. Hudec seems inclined to the former, Machiavellian view. To be successful for any length of time such a policy requires tacit collusion between governments, since each government must fool its own constituency but not its adversary. The policy assumes that pressure groups will not tire of ineffective policies—whether ineffectiveness is intended or not—and lobby the administration for more effective action but instead resign themselves to political realities. The U.S. propensity to go to court in order to placate domestic interests has increased international trade tensions. Hudec shows how the rash of litigation in the 1980s results from increased U.S. aggressiveness and the EC's adoption of a counterpunching strategy. The "trade hostilities" in 1986 occasioned by Spain's and Portugal's membership in the EC support this thesis. Towards the end, Hudec's paper portrays the two commercial superpowers standing "clad in full legal armor" anticipating "bigger and better GATT litigation." This prospect fills an observer from a small country with fear of being crushed under the heavy armor of falling giants. It is often claimed that small countries need the protection of legal institutions more than large countries do. This observation implies that small countries can be expected to use litigation more often than large countries do. Hudec's statistics do not bear this out. Big countries sue more often than small countries. Big countries sue other big countries. When small countries sue they also sue big countries. This reflects the obvious fact that countries litigate when their bargaining power is too small to negotiate a satisfactory out-of-court settlement. But it also suggests that large countries are more confident than small countries that litigation will provide satisfaction. This is a worrisome conclusion. A first cause for concern is that the strategy attributed to the United States by Hudec risks overloading GATT's dispute settlement capacity. This capacity is already insufficient and would become even less accessible to small countries whose needs for dispute settlement may be greater, though less dramatic, than the two superpowers' needs. A second cause for concern is that this strategy risks discrediting the dispute settlement mechanism. Its basic purpose is to settle dis-

61

US-EC GATT Litigation

putes, not to appear to be settling disputes. Dispute settlement has taken on the characteristics of a Norman jousting tournament in which the battles are for the benefit of the spectators rather than for the spoils of war. While recognizing that these ritual jousts may help avoid general trade wars, one still may ask whether it is not an abuse of the dispute settlement system. What will happen when a real dispute arises and needs to be settled? This prospect worries small countries, which lack negotiating strength to fall back on in bilateral dispute settlement. It should also worry the EC and the U.S. in case either should need conciliation to settle a major dispute. Therefore, let me conclude with some reflections on the real culprit: a flawed dispute settlement mechanism. The tendency to litigate for the sake of litigation rather than to seek a remedy cannot be forbidden. However, it can be discouraged by measures which increase the costs of litigation. Displeasure with an adversary's action should not be sufficient cause to go to court. In addition, the outcome of a lawsuit should be unpredictable. If predictable, both parties have strong incentives to reach an out-of-court settlement provided that the dispute settlement mechanism is credible. The concurrence of a large number of lawsuits and a high predictability of outcome indicates that the mechanism is being used for other purposes than dispute settlement. This will reduce its credibility in the long run. How can one increase the risks to litigants of litigation? One way is to reduce members' ability to influence the outcome. Their influence is both indirect via the selection of panel members and direct by vetoing the Council's adoption of a panel's report. Members cannot only "pack the jury" but also "hang the judge" and large countries tend to exercise this power more often than other GATT members. Having a number of standing panels, sitting for four-year periods, would increase the independence and continuity of the panel mechanism. Not allowing parties to a dispute to vote on Council adoption of a report (the "GATTminus-two" formula) would diminish the predictability of the outcome. These are minor changes. Major changes are not possible in the GATT system. The GATT contract depends on self-enforcement for compliance with its basic principles.

Comment

J. P. Hayes

I would like to pick up Dr. Wijkman's remark, that "some trade disputes are serious." It struck me that one of the most serious disputes, J. P. Hayes is a senior fellow at the Trade Policy Research Center, London.

62

Robert £ . Hudec

over the possibility of U.S. countervailing duties on EC steel in 1982, does not qualify for inclusion under Professor Hudec's definition of GATT lawsuits (although the Community raised the matter in the GATT Committee on Subsidies and Countervailing Duties).1 Is it the case that the GATT dispute settlement procedure leads to concentration on the particular points at issue, and does not allow wider trade-offs? Whatever the answer, US-EC trade disputes typically have to be seen in a broader context of transatlantic relations. For example, on the Washington view of the 1982 steel dispute, Levine reports: At a time when harmony was needed, US-EC relations were bad and seemingly getting worse. The Soviet Union was in Afghanistan, Poland was in turmoil, and the proposed stationing of additional U.S. missiles in Europe and President Reagan's vocal anti-communism were testing the alliance. . . . Washington was also concerned about undermining the EC. Steel was, along with the Common Agricultural Policy, one of the underpinnings of the Community.2 This case suggests the question: Why do some trade disputes lead to GATT "lawsuits," under the definition, and others not? Professor Hudec suggests that the Gatt dispute settlement procedures perform a useful political function in the U.S. On the Community side, there are feelings of irritation which I think are by no means entirely synthetic. Thus, in the Commission one hears complaints that the U.S. had been "testing GATT dispute settlement to destruction." One complaint on the Community side is that, given the separation of powers between the executive and legislative branches, the U.S. cannot be relied on to comply with GATT findings. (Reference is made to the U.S. manufacturing clauses and the Wine Equity Act; also to the delay in replacing DISC.) The Community may feel vulnerable not only because it has policies which are subject to attack but also because the strength of the executives in Community countries removes a plausible excuse for noncompliance. In fact, I suggest that we have a curious game here, in which each side wishes to use the dispute settlement mechanism to restrain the other from doing certain things, but wishes to avoid being restrained itself. (The question why governments wish to follow some of the policies to which they are attached raises wider issues, which I will just touch on later.) 1. For the issues as seen from the Community's side, see Frank Benyon and Jacques Bourgeois, "The European Community-United States Steel Arrangements" Common Market Law Review 21, no. 2 (June 1984): 305-54. 2. Michael K. Levine, Inside International Trade Policy Formulation: A History of the 1982 US-EC Steel Arrangements (New York, etc.: Praeger, 1985), 37-38.

63

US-EC GATT Litigation

Another aspect of the Community's attitude is dislike of the idea that GATT panels might in effect make law by establishing precedents.3 In this view, rules for international trade should be established by negotiation, and the GATT dispute settlement procedures can only lead to firm conclusions when the relevant rules have already been agreed upon. The problem then is that in important areas—subsidies and countervailing duties being a notable example—the rules have not been agreed upon. Professor Hudec's paper might be read as suggesting that the arrangements are reasonably satisfactory as they are; but in some quarters, at least, there seems to be a felt need for improvement. The proposals I have seen might be roughly grouped in five categories. First, there are what might be roughly called administrative improvements—for example, time limits and increase of the professionalism of GATT panels. However, such changes might still leave the outcome subject to negotiation between the parties. A more far-reaching change, which perhaps goes beyond the boundaries of this category, is that the last word on a dispute should rest with the panel rather than with the GATT Council. This (like the proposal that the parties to the dispute should abstain in the final vote on the adoption of the panel report) appears to move dispute settlement nearer to a judicial process. I am not clear whether this would be acceptable, at any rate to the Community; and in any case the problem of securing compliance would remain. The second category is strengthening of the machinery for monitoring compliance. The problem of sanctions would remain. I understand that the GATT has been very reluctant to authorize retaliation against countries found to be in breach of GATT rules: there is a clear danger that the "remedy" could be as bad as, or worse than, the disease. Thus, even with strengthened monitoring, dispute settlement might continue to depend on moral suasion. Third, it has been suggested that, where a case turns on a disputed interpretation of the GATT, the panel should make recommendations for the revision or interpretation of the relevant provision. Because any such revision or interpretation might alter the previously negotiated balance of concessions, the issue should be sent to the appropriate committee or a special working party for negotiation.4 A possible prob3. The present dispute settlement process has been described by de Lacharriere: "The common purpose of all the procedures for the settlement of disputes is not, strictly speaking, to ensure compliance with the law but to arrive at settlements acceptable to the parties concerned. Hence it is not a matter of sanctioning a breach of a rule, still less of punishing it, but of restoring the balance of advantage in trade between the parties" (Guy Ladreit de Lacharriere, "Case for a Tribunal to Assist in Settling Trade Disputes," The World Economy, 8, no. 4, [December 1985] 340). 4. Gary Clyde Hufbauer and Jeffrey J. Schott, Trading for Growth, Policy Analyses in International Economics no 11 (Washington, D.C.: Institute for International Economics, 1985); cited in Gardner Patterson and Eliza Patterson, "Importance of a GATT Review in the New Negotiations," The World Economy 9, no. 2, (June 1986): 153-69.

64

Robert E . Hudec

lem with this is that issues would be dealt with one at a time, thus giving little or no opportunity for trades. It may be better to continue the practice of storing up disputed issues to be negotiated simultaneously in a GATT round. Fourth, there are suggestions that GATT rules should be incorporated into national law, so that private parties could invoke them in the national courts. I am inclined to doubt whether this is in the area of practical politics.5 Fifth (and going further afield), it might be said to be a function of the GATT to prevent governments from doing things which they ought not to want to do in any case. Public education has a part to play here. In the United Kingdom, cost of protection studies seem to be having a certain effect. (In France, on the other hand, a senior official told me that no such studies have been made, and that there is no demand for them). In any case, we should not be overoptimistic about the influence of economists. It might be more effective to build up countervailing power by strengthening the machinery for consulting consumer, user, and importer interests. However, governments may be reluctant. At present, they can secure political benefit by conferring advantages on producer groups, and those who bear the costs do not appear to be particularly aggrieved. With greater and more balanced consultation, someone would be aggrieved in the end—the producers if the upshot were to withold protection, and the consumers, other users, or importers if protection were granted over their objections. Nevertheless, more extensive consultations on matters of trade policy would be highly desirable in the general interest.6 I live in hope that someone may have proposals that will solve all the problems. 5. "I would propose that the main international effort should be directed to securing more perfect national justiciability of the personal rights which it is the ultimate function of international agreements to protect. Adjudication procedures at the international level— determinations by GATT panels, for example—are procedures of the executive branch which any truly independent national judiciary should have the full right to ignore" (Jan Tumlir, "Conceptions of the International Economic and Legal Order," [book review article], The World Economy, 8, no. 1 (March 1985): 87). 6. I must confess that I had not, and still have not, thought out the implications of the idea that individuals might have access to the GATT dispute settlement proceedings; nor of the different suggestion that third parties should be encouraged and helped to complain {Trade Policies for a Better Future: Proposals for Action, Report of an Independent Group under the Chairmanship of Fritz Leutwiler [Geneva: GATT Secretariat, 1985], 46-47; cited in Patterson and Patterson, op. cit.). Also relevant to parts of this discussion may be: Ernst-Ulrich Petersmann, "International and European Foreign Trade Law: GATT Dispute Settlement against the EEC," Common Market Law Review, 22 (1985): 441-87.

3

The Price and Welfare Implications of Current Conflicts Between the Agricultural Policies of the United States and the European Community Dermot Hayes and Andrew Schmitz

3.1 Introduction The agricultural policies currently in place in the United States and the European Community are similar in one respect: both are designed to increase the export market share without regard to the prices or revenues these products earn. A price war is currently being waged between the two trading communities in which, as with any price war, there is a welfare transfer from the combatants to the consumer (in this case, food-importing countries). The wealth transfers are enormous. When the 1985 Farm Bill takes effect in the United States, both areas will be exporting food at well below the cost of production.1 This avoidable situation is the result of a misguided attempt by the European Community (EC) to increase the income of its agricultural producers, followed by a somewhat more justifiable attempt by the U.S. to counter the policies of the EC. It is a situation that, if unchecked, will cause most of the finances allocated for agricultural income support to be used eventually as a food subsidy for the industrial competitors of both communities. Using a rather broad definition of a "farmer," there are approximately 6 million farmers in the EC-10 and 2 million in the U.S. "Roughly speaking, in 1983 the governments of the United States and the European Community had farm policies that cost the taxpayers and consumers a minimum of $80 billion" (Johnson 1984), or $10,000 per farmer. This figure approximately equaled the average net farm income in the Dermot Hayes is an assistant professor in the Department of Economics at Iowa State University, Ames, Iowa. Andrew Schmitz is a professor of agricultural economics in the Department of Economics at the University of California at Berkeley, California. This work was initiated while Hayes was a graduate student at Berkeley.

67

68

Dermot Hayes/Andrew Schmitz

U.S. in that year (Petit 1985) and was well above that which prevailed in the EC. It will become clear later in this paper that this cost to benefit ratio has deteriorated since then. Moreover, despite this enormous wealth transfer, farmers are in poor financial condition in both communities. We begin our discussion in section 3.2 with an analysis of the common agricultural policy (CAP) of the EC. We argue that the CAP's pricing mechanism has acted as an incentive for individuals to overproduce, while the institutional design of the EC has caused national governments to encourage such overproduction. We then discuss how the technologies that have evolved in response to these incentives have shifted the supply curve for individual commodities to the right, making a reduction to the new equilibrium price level politically infeasible. We show that in the long run the effect of these above-equilibrium prices will necessitate an ever increasing reliance on export subsidies. We then review the performance of the CAP in achieving the objectives of an efficient use of resources, stabilized markets, and fairness. Unsurprisingly the single policy instrument of the CAP (price fixing) has not achieved its multiple objectives. In section 3.3 we discuss U.S. agricultural policy as it relates to USEC trade relations. Particular attention is paid to the Food Security Act of 1985, which we view as the U.S. response to the excesses of the CAP. The long-run implications of this policy are surprisingly similar to those of the CAP, a fact which allowed us to avoid a tedious repetition of the theoretical analysis. In section 3.4 we discuss developments in this price war as they relate to international comparative advantage. This is followed by a discussion of the direction in which the US-EC dispute is heading. We conclude in section 3.5 with a proposal to ease the current tensions. It consists of a specific set of policy changes which were designed to be politically feasible yet welfare enhancing. Although this policy alternative was designed with specific reference to CAP reform, it would achieve maximum benefit if used as the basis for a bilateral policy alignment. 3.2 The CAP 3.2.1 European Behavior under the CAP The long-term trend in the increase in the volume of agricultural production in the Community has been 1.5 percent to 2 percent per year, although internal demand has increased by only 0.5 percent per year. This spectacular surge in agricultural production in Europe will continue and could well gather momentum in coming years (Commission of the European Community 1985a).

69

US-EC Agricultural Policies

The European Community is already the largest exporter of meat and dairy products in the world. It has been a net exporter of wheat since 1974 and has recently become a net exporter of coarse grains (Meilke and de Gorter 1985). While the EC-10 remains a major importer of oilseeds, production of sunflower seeds in the 1985-86 crop year is expected to be more than 300 percent above the 1979-84 average. For rapeseed the increase is 167 percent (U.S. Department of Agriculture 1986). U.S. exports of soybeans to Europe have fallen 25 percent since 1983 (Rosson 1985). If quotas are imposed on cereal output, it seems inevitable that this # trend towards self-sufficiency in oilseeds will continue. Exports of European wines to the U.S. have recently been the subject of GATT litigation because of countervailing duty action by the United States (see chapter 2). Furthermore, when asked if he felt the U.S. Wine Equity Act (a measure that permits import retaliation and export subsidies if other countries do not reduce their import barriers on wine) could lead to a trade war, Mr. Bruno Julian, the Community's agricultural representative said, "We would certainly move against the import trade in soybean and soybean meal" (Wines and Vines, March 1984). There seems to be no end in sight to this increase in EC agricultural production. The Commission has estimated using current trends that by 1991 cereal "stocks will amount to around 80 million tons which is almost half the Community production and more than three times the quantity which the Community was able to sell on the world market when conditions were most favorable" (Commission of the European Community 1985c). To put this figure in perspective, for the period 1978-82 the U.S. produced on average 64 million metric tons of wheat per year, while total world wheat exports, including intra-EC trade, were 89.7 million metric tons. Between 1975 and 1984 the U.S. share of the world's grain trade has declined from 53 percent to 43 percent. In March 1985 Mr. Julian stated that "the EC believes it is entitled to a fair share of the world market and it will be aggressive in trying to obtain this share" (Rosson 1985). In figure 3.1 the value of import levies collected by the EC is expressed as a percentage of expenditures on export refunds. The absolute values of these numbers are presented in table 3.1. To dispose of surplus production the Community has increasingly depended on export refunds, which are, in effect, a food subsidy for importing countries. It is widely felt that U.S. farmers are more efficient than their European counterparts (Petit 1985). The question then arises as to the source of the increase in EC exports. In table 3.2 the relative rates of protection in the EC, Japan, and the United States are presented. It can be seen that this increase in exports was not in response to world

70

Dermot Hayes/Andrew Schmitz

1984 Year Fig. 3.1

E.C. Import levies as percentage of export subsidies (19741984)

Table 3.1

Import Levies Collected and Export Subsidies Paid by the European Community, 1974-1984 (million of European currency units) Year

Export Subsidies

Import Levies

1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984

590.6 968.9 1666.2 2703.9 3749.6 4981.8 5695.0 5208.5 5053.3 5559.7 6717.5

225.0 510.0 1040.1 1816.9 1872.7 1678.6 1535.4 1264.9 1522.0 1475.4 1946.6

prices. For the commodities listed, the protection rates in the EC were, with only one exception, significantly higher even than Japan's. This, however, is not the sole source of the problem. In figure 3.2 wheat support prices in West Germany, France, and the United States are shown. These prices, which represent the minimum amount eligible farmers could expect to receive at harvest, increased to a similar degree on both sides of the Atlantic. In 1973 and 1974 actual wheat prices

71

U S - E C Agricultural Policies

Table 3.2

Nominal and Effective Rates of Protection for Processed Agricultural Commodities in the European Community, Japan, and United States in the mid-1970s (percentage by value)3 Japan

EC Nominal

Effective

Nominal

U.S.

Effective

Nominal

Effective

Meat products

33

165

18

69

6

10

Fruit and vegetables

45

75

19

49

15

37

Dairy products Cheese Butter

59 77

276 1,328

35 45

175 418

12 10

35 47

22

82

26

69

4

0

49

95

24

75

11

35

148

148

25

268

23

253

Grain Products Corn Flour and cereal preparations Soybean oil

Source: A. J. Yeats, "Agricultural Protectionism: An Analysis of Its International Economic Effects and Options for Institutional Reform," Trade and Development, no. 3 (1981). a The nominal rate is the percentage by which the domestic product price exceeds the world price. The effective rate is the percentage by which protected value added exceeds value added without protection.

CH-DW.G.

o — o France • - • U . S . TGT U.S. LN.

70

72

74

76

78 Year

Fig. 3.2

W h e a t s u p p o r t prices (1970-1984)

82

84

72

Dermot Hayes/Andrew Schmitz

received in the U.S. were higher in dollar terms than the average of those received in the Community. Figure 3.3 shows the wheat yields per harvested hectare in some of the main producing countries. Some enlightening comparisons concerning the effect of the CAP can be made. In the late 1950s when the Community was formed, yields in the United States were about 70 percent of those achieved in France. By 1982, after the CAP had been in operation for several years, this figure had fallen to 40 percent. In addition, wheat yields in the U.K. showed only a moderate increase up to 1973; they had been surpassed for the first time ever by those in France in the previous year. After the U.K. joined the EC in 1973, these relative trends were reversed, with British yields increasing faster than those of the French. This postmembership performance cannot all be explained by relative prices, however (figure 3.3). This is borne out by the behavior of yields in the United States. If movements along the supply curve were responsible for all of the increase in European yields, a more marked increase in American yields could have been expected in response to high prices in 1973 and 1974. It is obvious from these figures that certain aspects of the structural design of the CAP act to shift out the supply curve. There are two possible reasons for this. First, there is no downside price risk for commodities that come under the CAP. Sandmo (1971) has shown that when output prices are volatile, planned output

1980 Fig. 3.3

1983

Wheat yield per harvested hectare for various countries

73

US-EC Agricultural Policies

decreases because risk-averse producers defer investments and avoid yield-increasing technologies to reduce input costs and exposure to risk. In the United States prior to the 1985 Farm Bill, this was the case as it still is in Australia and Canada. Prices in these countries depend on volatile world market conditions. Before 1985 American farmers could purchase downside price protection by removing 20 percent of their wheat base from production. By doing so, however, they exposed themselves further to changes in world prices via the opportunity cost of this unused land when market prices rose above costs of production. Until recently only a minority of U.S. producers (about 30 percent) chose this option. The method of agricultural price-support funding in the EC is the second potential cause of the outward shift in the EC supply curve of agricultural products. Individual European governments are aware that they receive the full benefits of output-increasing research, while bearing only a proportion of the costs involved in disposing of excess production. This may be one of the reasons why national research expenditures have increased substantially in recent years (see table 3.3). National public expenditure in favor of agriculture actually increased in the late 1970s despite a plan for shifting toward common funding (table 3.4). It is a commonly expressed sentiment in the Community that the solution to excess production is a return to "more market-oriented prices" (Commission of the European Community 1985a). The inelastic demand curve for food items ensures, however, that most individual producers could not survive at these prices, as the equilibrium revenues would be even lower than they were prior to the CAP. 3.2.2 Long-Run Supply Response in Agriculture Any attempt to measure the price elasticity of supply in agriculture is faced by many problems that preclude accurate measurement. Farmers take many years to respond to aggregate price changes. While it may be possible to estimate a quantity response caused by producers switching between enterprises, to estimate the elasticity of supply when all prices are simultaneously increased to levels above equilibrium is much more difficult. On balance, higher prices encourage the development and adoption of output-increasing technologies. However, the mastery by producers of the more complicated methods of production requires a considerable learning period. At the same time the stabilization of agricultural prices in the EC encourages agricultural producers to move outward on their marginal cost curves, i.e., adopt higher-cost methods of production. Separating the effects of these two factors operating on supply is impossible, consequently, in figure 3.4, it is assumed that the supply-increasing effects of stabilized prices is included in the overall price index.

2,241.0 2,770.6 2,950.5 2,239.7 2,515.4 2,731.6

1,589.4 1,513.4 1,568.2 1,670.5 1,670.4 1,636.5

1975 1976 1977 1978 1979 1980

2,595.9 1,810.5 1,942.4 2,067.1 2,164.8 2,882.2

Italy

Belgium 101.3 115.6 144.9 197.5 236.2 229.7

200.5 236.3 258.5 288.5 307.5 330.0 b

14.4 21.9 28.9 16.3 18.5

Luxembourg 1,493.7 1,206.1 931.9 685.9 855.4 1.075.5

United Kingdom 176.7 215.1 239.2 297.3 281.3 360.9

Ireland

134.0 158.2 177.1 224.7 277.2 273.9

Denmark

8,546.9 8.047.7 8,241.6 7,687.7 8,326.7 9,520.3C

European Economic Community

Source: Commission of the European Community, Perspectives for the Common Agricultural Policy: The Green Paper of the Commission. Brussels: Agricultural Information Service of the Directorate-General, July 1985. includes basic research but excludes Social Security expenses for farmers. These were of the order of 17 billion European currency units in 1980, i.e., 143 percent of the European Agricultural Guidance and Guarantee Fund expenditure of 175 percent of the national expenditure in favor of agriculture. b No data available. Luxembourg not included.

France

Year

The Netherlands 3

National Public Expenditure in Favor of Agriculture, 1975-1980 (million European currency units)

West Germany 3

Table 3.3

649.9 929.9 1,315.7 2,441.2 2,464.9 2,596.3

Year

1975 1976 1977 1978 1979 1980

Italy 961.3 1,091.2 1,000.1 1,195.9 1,694.8 1,930.0

France

1,219.4 1,453.5 1,631.7 1,511.4 2,380.5 2,963.1 543.9 771.0 907.2 1,111.2 1,402.3 1,569.7

The Netherlands Luxembourg 6.0 8.5 10.2 25.3 13.9 12.6

Belgium 187.1 348.3 435.2 574.5 769.7 596.4 631.9 511.7 416.9 1,193.9 992.6 991.1

United Kingdom

246.7 234.4 602.5 358.1 484.2 609.7

Ireland

European Agricultural Guidance and Guarantee Fund Expenditure, 1975-1980 (million European currency units)

318.1 438.9 639.3 583.8 644.3 640.4

Denmark

4,764.3 5,787.4 6,958.8 8,995.3 10,847.2 11,909.3

European Economic Community

Source: Commission of the European Community, European Agricultural Guidance and Guarantee Fund: Annual Report, various years.

West Germany

Table 3.4

76

Dermot Hayes/Andrew Schmitz

>/MC

00

_ ro J

>^i

P

l

03

X

i | i1

a?

o

1 1

i

i >

1

I

i

i i

!


a2 is the new value of a2 needed to keep p2 constant, and p is given in (32). If a2 = a2, and —e2 + fi Pi < -ex + fx p, in other words, if the EC average net exports are smaller than the average U.S. net exports, as is the case between the U.S. and EC grain exports, it is clear by comparing (53) and (52) that (54)

A (GG2) | < | A (GG{)

in other words, the U.S. treasury loss is larger than the loss inflicted on the EC. This last point has been supported empirically. Paarlberg and Sharpies (1984) found that a budget cost to the United States of $1.9 billion would be required to increase the wheat subsidy costs of the EC by $100 million. Similarly, Anderson and Tyers (1983) estimated an annual U.S. Treasury cost for wheat of between $800 million and $1 billion to increase the annual EC budget by $130-200 million. As recent literature on the political economy of protection has illustrated (Baldwin 1984), governments do not usually adopt policies on

120

Alexander H. Sams

total welfare grounds; the differential political power of the affected groups determines the final outcome. In the U.S. grain trade, there are four major affected groups: producers, consumers, the Treasury, and the international grain-trading industry. Of the three policies mentioned above, the first, export taxes, will hurt producers and the grain traders (by decreasing exports), and benefit consumers, and the Treasury. The second, buffer stocks, will tend to benefit producers and hurt consumers, the Treasury, and grain traders, although the average effects over the cycle will be small. The main effect of this policy will be a tendency toward stabilization of receipts or expenditures of the main groups. Finally, export subsidies will benefit producers and grain traders and hurt consumers and the Treasury. The relative magnitudes of the gains and losses, of course, would be different under the three policies. Furthermore, over the market cycle the effects of the policies will not be the same. While export taxes and subsidies will generate the same effects on the interest groups under both tight and surplus conditions, buffer stocks will tend to generate the opposite effects in the two situations. The U.S. reliance mostly on buffer stocks and partially on export subsidies (guised as food aid) might be an indication of the relatively balanced strengths of the various private interests and a weakness of the Treasury. 4.6 Some Empirical Estimates for Cereals The theory outlined in the previous sections illustrated the potential effects of the US-EC agricultural trade conflict. In this section, the magnitudes involved are quantified in the case of wheat and coarse grain trade. No detailed econometric estimates of elasticity parameters are derived, and the numbers should be considered as reasonable approximations. The basic data in the computations and the results are shown in table 4.3. While all the quantity and price data have been estimated empirically by the author, the elasticity figures come from several published estimates, including those of the GOL model (Liu and Roningen 1985). As simulated, only EC policies are represented, while for the United States it is assumed that an approximately free and open market prevails. This is quite realistic for 1981-1983, as the numbers in table 4.1 indicate. The results show that both the United States and the EC lose substantial amounts on average because of the EC grain-support and stabilization policies. The total average annual U.S. welfare loss for all grains is $1,401 million; for the EC, $404 million. The U.S. and the EC would gain these amounts if both switched to free trade. It is interesting that the EC loses in the long run and because of fluctuations on all

121 Table 4.3

EC-US Agricultural Trade Confrontation Data Parameter Values and Effects of Wheat and Coarse Grain Policies (Quantity figures are in thousand metric tons (nit)., prices in US$/mt, GL, and GSi in million US$) Coarse Grains b

Wheat

Average production Average consumption 3 Average net exports 3 Short-run supply price elasticity Long-run supply price elasticity Short-run demand price elasticity Long-run demand price elasticity Value of a, Value of p j Value of a, Value of pi Long-run gain GLf Short-run gain GS,C

U.S.

EC

72,305 29,410 42,895 .17

57,828 45,953 11,875 .05

.17

.40

-.14

-.14

-.14

-.28

1.0 1.0 3,367 170.7 -989 5.5

1.3 .28 1,932 221.8 -309 -9.9

i

Rest of World

— —

-54,774

— —

U.S.

EC

211,873 153,488 58,385 .07

68,166 71,792 -3,626 .05

Rest of World

— —

-54,759

— —

.07

.25

-.le

-.80

-.11

-.18e

-.3e

-.80

-.31

-.28e

1.0 1.0 17,853 129.4d -422 .4

1.3 .3 6,023 168.2 -95 -4.9

— —

— —

3,129

— — —

3,724

— — —

Sources: FAO Production and Trade Yearbook (various years), Sarris (1985), and own estimates based on several published studies. GLt and GSi computed. a The figures are averages for the 1981-83 years. includes maize, rye, barley, sorghum, oats, millet and mixed grains. c For the definitions of GLt and GS/, see equations (22) and (23) in the text. d For coarse grains, the maize price is taken as the representative price. e For the rest of the world, the elasticities refer to the excess demand.

products, while the U.S. gains because of the changes in the probability distributions, but not enough to overcome the huge losses caused by the average price depression. If the EC were to switch to free trade, that is, if we set a2 = p2 = 1> the model predicts that the average world wheat price would increase by 10.8 percent and the average world coarse-grain price would increase by 5.3 percent, while the standard deviation of the world wheat price would decline by 18 percent and by 3.8 percent for coarse grain. The estimates for wheat are close to those of Sarris and Freebairn (1983). The gains the United States can obtain by applying an optimal export tax in every period, see equation (38), are substantial. For wheat, they average $3,116 million, and for coarse grains, $3,335 million annually. The reason for this large potential gain is that the U.S. export tax substantially raises the average world price (by an average 56 percent

122

Alexander H. Sarris

for wheat and 90 percent for coarse grains). Domestic U.S. prices are heavily depressed, however, and the large gainer is the U.S. Treasury. Notice that the estimated gains are larger than the CAP-inflicted losses. But such a policy is politically untenable unless the Treasury proceeds are redistributed to farmers, which is unlikely to happen. The net gains to the United States from instituting an optimal bufferstock policy, which is wholly financed by the U.S. Treasury, are small but positive. For wheat, they average $14.6 million a year, and for coarse grains, $5.8 million annually (assuming a value of 7 equal to .01, which corresponds to $10 per metric ton annual storage cost when total excess government stocks are 1 million tons). The maximum gains from the buffer stock, that is, when 7 = 0, are $56.8 million for wheat and $142.2 million annually for coarse grains. The interesting result here is that an optimal U.S. government stock policy does not burden the Treasury on average and is very beneficial for farmers, since prices for their products are substantially stabilized. Finally, we can estimate how much the U.S. Treasury would lose by inflicting a budget loss on the EC by means of export subsidies. Using the expressions derived earlier and the results of table 4.3, it can be shown that to inflict a $100 million annual loss on the EC budget in the wheat sector, the average U.S. wheat export subsidy would need to be $12.30 per metric ton and the average annual U.S. budget cost would be $530 million, or about five times as much as the EC's loss. This roughly five-to-one ratio of U.S. losses to EC losses is close to what Anderson and Tyers (1983) also found, using a more complex empirical model. The recent weakness of the U.S. dollar implies that a given percentage increase in the EC export subsidy would necessitate an even larger U.S. budget outlay. The fact that the United States is currently (mid-1986) pursuing this policy, despite the low value of the dollar and a budget-conscious government, might be an indication of the enormous pressures from producers and the grain industry, strengthened by electoral politics. 4.7 Summary and Conclusions The Common Agricultural Policy of the EC entails heavy economic costs to both the United States and the EC, and it distorts the world markets substantially. Since CAP reform is a slow and tedious process, given the structure of decision making within the EC, we have explored some options by which the U.S. might defend itself from the detrimental effects of CAP policies. While it was shown both theoretically and empirically that an optimal U.S. export tax, a measure that exploits U.S. monopoly power, would more than compensate the United States for the CAP-induced losses, it is probably politically infeasible. A U.S.

123

EC-US Agricultural Trade Confrontation

export subsidy designed to strain the EC budget would be too costly for the U.S., since for every dollar of damage inflicted on the EC budget, about five U.S. budget dollars would have to be spent. The only policy that seems to alleviate the political strains induced by the U.S. farmer is a government stockholding policy. By buying for storage when world prices are low and selling when world prices are high, the U.S. Treasury is not hurt, the producers benefit when they need it most, and the consumers get more stable prices. U.S. grain carryover figures from the last two decades show that the United States has indeed been manipulating stocks. Whether this has been done in an optimal way is a subject for further empirical research. The results of this paper suggest that the U.S. ought to consider a more deliberate long-term storage policy, rather than be forced to stockpile as a last resort by the inevitable swings of the market. Finally, it appears that a combination of a small export subsidy and a stockholding policy might allow the United States to recoup the welfare losses inflicted on it by the EC without hurting U.S. farmers in periods of market downturns.

References Abbott, P. C. 1979. Modeling international grain trade with governmentcontrolled markets. American Journal of Agricultural Economics, 61, no. 1 (February). Anderson, K., and R. Tyers. 1983. European Community's grain and meat policies and U.S. retaliation: Effects on international prices, trade, and welfare. October, Research School of Pacific Studies, Australian National University, Canberra. Mimeo. Baldwin, R. E. 1984. Trade policies in developed countries. In Handbook of International Economics, vol. 1, ed. R. W. Jones and P. B. Kenen. Amsterdam: North-Holland. Butler, N. 1983. The ploughshares war between Europe and America. Foreign Affairs (Fall). Chambers, R. G., and R. E. Just. 1981. Effects of exchange-rate changes on U.S. agriculture: A dynamic analysis. American Journal of Agricultural Economics 63, no. 1 (February). de Gorter, H., and K. D. Meilke. 1985. The EEC's Common Agriculturel Policy and international wheat prices. July, International Trade Policy Division, Agriculture Canada. Mimeo. Eldor, R. 1984. On the risk-adjusted effective protection rate. Review of Economics and Statistics 66, no. 2 (May). Food and Agriculture Organization (FAO) of the United Nations. 1975. Agricultural protection and stabilization policies: A framework for measurement in the context of agricultural adjustment. C75/LIM/2 (October), Rome. 1983. "International Agricultural Adjustment." Fourth Progress Report, C83/21 (August), Rome. 1985. "International Agricultural Adjustment." Fifth Progress Report, C85/21 (August), Rome.

124

Alexander H. Sarris

Green Europe. 1985a. EEC-USA relations: The commission's view, No. 31. 1985b. Perspectives for the Common Agricultural Policy: The Green Paper of the Commission, no. 33 (July). Koester, U. 1982. Policy options for the grain economy of the European Community: Implications for developing countries. International Food Policy Research Institute, Research Report no. 35 (November). Liu, K., and V. O. Roningen. 1985. The world grain-oilseeds-livestock (GOL) model: A simplified version. U.S. Department of Agriculture, Economic Research Service. ERS Staff Report no. AGES 850128 (February). Paarlberg, P. L. and J. A. Sharpies. 1984. Japanese and European Community agricultural trade policies: Some U.S. strategies. U.S. Department of Agriculture, Foreign Agriculture Economic Report no. 204. Washington, D.C. Petit, M. 1985. Determinants of agricultural policies in the United States and the European Community. International Food Policy Research Institute, Research Report 51 (November). Sampson, G. P., and R. H. Snape. 1980. Effects of EEC's variable import levies. Journal of Political Economy 88, no. 5 (October). Sampson, G. P., and A. J. Yeats. 1977. An evaluation of the Common Agricultural Policy as a barrier facing agricultural exports to the European Community. American Journal of Agricultural Economics 59: 99-106. Sarris, A. H. 1982. Export taxes versus buffer stocks as optimal export policies under uncertainty. Journal of Development Economics 11, no. 2 (October). 1985. Domestic price policies and international distortions: The cases of wheat and rice. April, Food and Agriculture Organization of the United Nations, Global Perspective Studies Unit. Mimeo. Sarris, A. H., and J. Freebairn. 1983. Endogenous price policies and international wheat prices. American Journal of Agricultural Economics 65, no. 2 (May). Schuh, G. E. 1974. The exchange rate and U.S. agriculture. American Journal of Agricultural Economics 56: 1-13. Sheehy, S. 1984. The agrimonetary system: A reinterpretation. European Review of Agricultural Economics, vol. 11-2. Stanton, B. F. 1986. Production costs for cereals in the European Community: Comparison with the United States, 1977-1984. Cornell University, Agricultural Experiment Station, A.E. Res. 86-2 (March). Valdes, A., and J. Zietz. 1980. Agricultural Protection in OECD countries: Its cost to less-developed countries. International Food Policy Research Institute, Research Report no. 21 (December).

Comment

Dieter Kirschke

The papers presented by Hayes/Schmitz and Sarris* are interesting and animating. It has been a pleasure reading them and commenting on Dieter Kirschke has been a country economist at the General-Directorate for Development of the EC Commission and is now professor of agricultural policy at the Technical University of Berlin. These comments reflect the author's personal view and are not an official EC position. *Chapters 3 and 4 in this volume.

125

EC-US Agricultural Trade Confrontation

them. I want to discuss the papers together and put them in the framework of the current US-EC confrontation in agricultural trade. There are basically two problems to be dealt with: (1) Does the EC's Common Agricultural Policy (CAP) have a negative impact on the United States? and (2) Is the present U.S. reaction justified? Evaluation of the CAP National agricultural policies are orientated towards domestic problems and possible trade effects principally have to be considered as mere by-products. This is widely agreed upon and has clearly been stated in the papers. It is in particular true for the CAP, but the 1985 Farm Bill seems to indicate that it is also true for the United States. The basic feature of the CAP is to support farmers' income by price policy and, thus, protectionism. The shortcoming of such a policy approach have been extensively discussed and the main criticism is summarized in the Hayes/Schmitz paper. There is no need to discuss these arguments further. From an economic point of view, the CAP certainly cannot be considered a first- or second-best policy. To put it in other words: the objective function that is maximized by the CAP is difficult to imagine, let alone to describe. CAP Impact on the United States Price Level on World Markets There is no doubt that EC agricultural protectionism tends to decrease the price level on several world markets. Hayes/Schmitz and Sarris give an overview of a multitude of studies that all demonstrate this effect, but differ in their empirical estimates. It is obvious that a price decrease on world markets is harmful to U.S. exports such as wheat. The price level effect is less certain, however, if dynamic aspects are taken into account. In this regard the papers are somewhat contradictory and confusing. The general question to be answered refers to the link of protectionist policies and structural change in agriculture. Hayes/Schmitz argue that the CAP tends to shift out the supply curve in European agriculture. I can follow their argument that a possible risk reduction may increase output; the rough comparison of the yield developments in the EC and the United States, however, is not sufficient to examine the thesis empirically. Sarris argues exactly the opposite. According to him high agricultural prices would leave marginal farms in production and would thus hamper structural change. This sounds reasonable and the argument has in fact been discussed extensively in the German agricultural economics literature (Koester 1977). It may indeed be described as a reversed infant

126

Alexander H. Sarris

industry argument. In any case—a comprehensive analysis on the link between price support policies in agriculture and structural change, integrating the arguments presented in the papers, still has to be given. Price Instability on World Markets The traditional argument has been introduced by Johnson (1975) who stated that the CAP would tend to fix internal prices and thus would make world market prices more volatile. Sarris confirms the recent analyses (Schmitz and Koester, 1984) that such a statement only reflects part of the problem and may indeed be wrong. The CAP may also help to stabilize world market prices, for instance, by stabilizing expectations and thus production in European agriculture, and possibly, by shifting world production to less risky production areas. The actual impact of the CAP on world market price instability therefore can only be judged case by case and must be derived by means of comprehensive empirical analysis. Sarris's Model In this context a brief discussion of Sarris's model is justified. Sarris seems to be restrained in his comments on the "standard partial equilibrium trade liberalization exercise" as he calls it—and then proceeds himself in a similar way. The genuine feature of his model, of course, is its stochastic formulation. The model underlines the fact that policy analysis under uncertainty will seldom yield definite, but rather conditional results which require a concise discussion of relevant parameters. Politicians will not like this kind of result, though complexity is not a shortcoming of the model, but rather of reality. The importance of stochastic modeling for policy analysis is therefore emphasized and, in particular, the reader's attention is drawn to some interesting particularities of Sarris's model, notably the simultaneous consideration of short- and long-term effects and the modeling of currency effects under uncertainty. Some questions concerning Sarris's model should nevertheless be asked. First, both the agricultural policy of the United States and the EC have been modeled by introducing a variable price differentiation between the domestic and the world market price. The CAP, however, can rather be described as a price-fixing protectionist policy, and I wonder why Sarris did not model it this way (Kirschke 1985). Second, the analysis of, for instance, a U.S. export tax is based on a state contingent policy concept which could hardly be implemented under real world conditions. A preferable way might be to model policies on the basis of a quasi-deterministic concept according to which policy interventions are based on expected supply and demand functions. Third, Sarris derives explicit formulas for expectations and higher mo-

127

EC-US Agricultural Trade Confrontation

ments, which is welcome as a precise analytical exercise, but practically restricts the analysis to linear functions and additive uncertainty. It is therefore suggested to equally take into account approximation formulas in stochastic calculus and stochastic simulation. In this context I agree with Sards who says that "not only price levels but also the higher moments of the price distributions should be computed," but I have got to contradict him when he states that "this has not been done to any extent yet." In fact, several analyses have been done in recent years though these have not always been published in Anglo-Saxon journals.1 Aggregation of CAP Effects Summarizing the discussion it is difficult to state that the United States has actually and undoubtedly suffered from the CAP. This is only obvious if the classical static price level effect is considered. It is less obvious if dynamic effects and CAP effects on world market instability are taken into account. It finally becomes doubtful if indirect effects are to be included in the analysis, such as an increased EC demand for cereal substitutes due to the CAP. The papers by Hayes/ Schmitz and Sarris concentrate on partial market effects and do not really deal with such additional effects. This is not a shortcoming of the analyses, but simply points to the complexity of the problem to which there is no clear-cut answer available. U.S. Policy Options Let us now stop complicating the discussion and take a simplified view of the world. Consider a single commodity, say wheat, that is exported by the United States and that is protected by the EC's protectionist policy, and suppose that only the classical static price level effect of the CAP on the world market is taken into account. Under these restrictive conditions the CAP's negative impact on the U.S. is evident. From an economic point of view, the interest of the United States would then be to pursue policies that could help to raise the world market price. What kind of policy options would the U.S., in fact, have? Inducement of a CAP Change The U.S. could obviously plead for a CAP change in order to reduce EC surplus production and thus increase the world market price level. The selfish way would be to enforce EC surplus reduction by any

1. Kirschke (1987) gives an overview of such studies and analizes the CAP's pricefixing protectionist policy under uncertainty by means of different stochastic methods.

128

Alexander H. Sarris

possible CAP change, no matter how EC economic welfare is affected. The altruistic way would, however, mean to help the EC to implement a first-best CAP from an economic point of view, which could be beneficial for both sides. The perspective for such a CAP change is widely agreed upon and would consist of a separation of efficiency and distribution objectives in European agriculture. This would mean the abolition of price supports and protectionism and the establishment of direct income subsidy schemes. The proposal for a CAP change presented by Hayes/Schmitz goes in this direction, but unfortunately not very far. They suggest to support farmers' income by differentiated production subsidies which, in my view, is a combination of a partial price differentiation and a partial quota system. It very much resembles the former French approach for a CAP reform and might also be considered as a new variation of the "small is beautiful'' concept in agricultural policy formulation. Even if the nondebatable argument of political feasibility is admitted I cannot see that their proposed system will be easier to implement and administer than some direct income support scheme. Therefore I wonder why Hayes/Schmitz did not go some step further towards a first-best CAP. Adjustment to the CAP In a more passive way the United States could also take the CAP for granted and adjust its policies accordingly. A first option would then be to use market power on the world market and thus increase export revenues. Sarris has clearly pointed out that such a policy could be realized by an optimal export tax which equates the domestic price in the United States to the marginal export revenues. In fact, this is the well-known terms-of-trade argument for trade interventions in order to increase national economic welfare. Sarris has also stated that a buffer stock policy could equally help to enhance U.S. economic welfare even if it were to be financed by the U.S. alone. A final U.S. reaction to the CAP might be the demand for compensation for CAPinduced losses. This alternative has not really been discussed by the authors. The Actual U.S. Behavior Increasing Complaints The actual U.S. behaviour in the present agricultural trade confrontation does not really reflect the stated options. In the context of the simplified scenario developed here, U.S. complaints about the CAP are certainly justified, but they increase at a time when this could not be expected. First, EC protectionism has rather diminished than in-

129

EC-US Agricultural Trade Confrontation

creased for several agricultural commodities in recent years. This is first of all a currency phenomenon due to changes in the dollar's value, but also indicates a change to a more modest price policy in the EC. Second, the EC has introduced and discusses the introduction of further specific policy measures to cut surplus production such as producer coresponsibility levies and quotas for milk production, which may be questioned from an EC welfare point of view, but, from a U.S. point of view, lead in the right direction. It should, finally, be kept in mind that an overall assessment of CAP impacts on the United States cannot really be given. Furthermore, the U.S. itself carries a major responsibility for the current GATT regulations which have allowed the establishment of the CAP as it is. This has clearly been stated by Hayes/ Schmitz and Sards. Uneconomic Reactions Much of the present U.S. complaint is based on the market share argument. It is true that EC protectionism tends to reduce U.S. shares in the world markets, but it is also true that there is no monocausal relationship between the CAP and export market shares. Following the discussions above, in particular, the drop from 53 percent to 43 percent in the U.S. share of the world's grain trade between 1975 and 1984, as documented by Hayes/Schmitz, cannot simply be attributed to the CAP, but may rather reflect the change in the U.S. monetary policy. Hence, thinking in terms of a simple market share does not really consider the economics involved. Another astonishing feature discussed by Hayes/Schmitz is the recent U.S. agricultural policy change. The United States seems to be willing to repeat the same mistakes the EC has made under the CAP. Direct price support schemes, for instance, are strengthened under the 1985 Farm Bill. The U.S. has introduced an enormous export subsidy program with the interesting abbreviation "bicep." This is directly opposed to the optimal export tax argument and Hayes/Schmitz convincingly argue that this program directly wastes resources and transfers wealth to cereals-importing countries, many of which are centrally planned economies. From a classical economic point of view, indeed, the U.S. behaviour cannot be described as anything else but irrational. Political Economics in Agricultural Trade The U.S. reactions may, however, be rationalized if the purely economic view of the present agricultural trade conflict is broadened arid some aspects of political economy are taken into account. In view of the evident difficulties for a fundamental CAP reform the U.S. behavior may simply be explained as a calculated attempt to change the CAP at all and, thus, reflect strategic behavior. It may also be a response to a

130

Alexander H. Sarris

change in public opinion or increased lobbyist activities. There is no need to say that internal policy restrictions often dominate rational policies in a economic sense which may, indeed, be illustrated by the CAP itself. Finally, some persons argue that the U.S. reactions should most of all be considered as an attempt to turn away from negative domestic policy impacts on agriculture. Apart from this, a seemingly institutional shortcoming in economic policy formation is worth being considered. According to Harvey's Abilene Paradox people in committees agree on decisions that, as individuals, they know are stupid (Dixon 1986). In other words: economic decision making or economic negotiations often end with policies that cause damage to all parties involved. A possible explanation to this phenomenon is given by Fisher and Ury (1981), who describe position bargaining as a central shortcoming of negotiations. This is the defence of a position once taken by whatever reason which does not reflect one's real interest. This is rather a psychological than an economic phenomenon. The current U.S. market-share thinking in the confrontation with the EC may be interpreted as a typical case in point. It is a position that may not be economically reasonable, but which has been taken and will be defended. Some more examples for unfruitful position bargaining in international economic negotiations and especially in the context of the CAP could easily be identified. In general one is tempted to state a "Law of Economic Negotiations" that could read as follows: If there are economic negotiations there will be a position no matter how uneconomic it is. The avoidance of an open agricultural trade war between the United States and the EC in the summer of 1986 raises some hopes that both sides are willing to reduce the conflict and come back to the economics involved. The excellent papers presented by Hayes/Schmitz and Sarris and the discussion during this conference may further help to strengthen economic reasoning in the current US-EC confrontation. References Dixon, M. 1986. How innovators work, and what stops them. Financial Times. 6 March 1986. Fisher, R., and W. Ury. 1987. Getting to YES: Negotiating agreement without giving in. Boston: Houghton-Mifflin. Johnson, D. G. 1975. World agriculture, commodity policy, and price variability. American Journal of Agricultural Economics 57: 823-28. Kirschke, D. 1985. Trade uncertainty in a price-fixing protective system. Zeitschriftfur die gesamte Staatswissenschaft (Journal of Institutional and Theoretical Economics) 141: 269-81.

131

EC-US Agricultural Trade Confrontation

1987. Agrarmarktpolitik bei Unsicherheit. Berlin: Duncker & Humblot. Koester, U. 1977. EG—Agrarpolitik in der Sackgasse. Baden-Baden: Nomos. Schmitz, P. M., and U. Koester. 1984. The sugar market policy of the European community and the stability of world market prices for sugar. In International agricultural trade, ed. G. G. Storey, A. Schmitz, and A. H. Sarris, 235-59. Boulder and London: Westview.

Strategic Trade, Embargoes, and Imperfect Competition Henryk Kierzkowski

5.1 Introduction "Strategic trade" is one of those catchwords frequently used in public debates and policy pronouncements which through repetition seems to have acquired an impression of clarity. The concept is, however, ambiguous and I therefore propose to begin my paper by discussing alternative usages of the term. Thomas C. Schelling (1984) tells the following episode in his lovely book Choice and Consequences: "I was invited to talk about strategic aspects of social problems. I asked what that meant. I was told it meant whatever I meant. So, with the uneasy feeling that it had just been used on me, I set out to characterize the 'strategic approach.' " As for myself, I have decided to follow a different strategy—here is this word again! I set out to see what others meant by strategic trade. This is attempted in section 5.2 which looks at the evolution of the concept in the context of U.S.-Western European trade relations. The reader must be forewarned that he will not find there a precise definition of strategic trade. Nevertheless it is interesting to see how the emphases on different aspects of the term shifts over time. Section 5.3 of this paper presents a model of strategic trade which lends itself to policy analysis. The question I ask is the following: Can protection of an industry considered to be strategic be justified when a possibility of embargo exists. It seems that a model of monopolistic competition is best suited to answer this question because it naturally Henryk Kierzkowski is professor of economics at the Graduate Institute of International Studies in Geneva. I wish to thank Professors K. Abbott, G. Hufbauer, and P. Mathieu for helpful comments and suggestions.

135

136

Henryk Kierzkowski

leads to concentration of production in the hands of one or a few producers. But for an embargo to be really painful there must be costs associated with replacing foreign supplies denied by it. The time necessary to build a new or reestablish an industry constitutes such a cost. I also argue in this paper that other policy measures can be taken to minimize costs of embargo. What is really necessary is that a country move from a position of strategic dependence to that of strategic interdependence. It would, however, be costly, unwise, and almost impossible to achieve a position of total strategic independence. 5.2 Historical Background One might expect that the Articles of the General Agreement on Tariffs and Trade would contain a definition of strategic trade. Alas, this term does not exist for the organization dealing with international trade. Only Article XXI can be construed to refer to strategic trade, albeit indirectly, as it states that the Articles of GATT need not apply in cases where essential security interests are involved. Unfortunately, the concept of "essential security interests" is even more ambiguous than "strategic trade." It is up to individual contracting parties to the GATT to decide when security interests are, or are not, at stake. It is worth recalling that the countries invoking the security exception under GATT's Article XXI were hardly ever, if at all, challenged in the past. The country which, in the aftermath of World War II, has been preoccupied with strategic trade and remains so is the United States. This concern is to some extent shared also by the Western European countries. It is revealing therefore to take a quick look at the history of U.S.-European relations in this field. By the late 1940s the United States sought to regulate, through domestic legislation and international collaboration, exports of goods with military and strategic importance.1 The Export Control Act of 1949 aimed at imposing limits on exports of military and strategic goods to countries with which the United States was not actually at war. Previously, similar acts had only been enacted during periods of military confrontation, wars, or in the face of extraordinary emergency. In addition to domestic legislation limiting strategic trade, the United States attempted to forge international cooperation in this field. In 1950 the United States set up jointly with the United Kingdom, France, Italy, and the Benelux countries the so-called Coordinating Committee for East-West Trade (COCOM) with the aim of compiling and enforcing lists of strategic goods of which exports to communist countries were embargoed. COCOM continues to exist and at present consists of all the NATO countries, with the exception of Iceland and Japan.2 The COCOM

137

Strategic Trade, Embargoes, and Imperfect Competition

embargo list is revised and updated every two to three years. The embargoed goods fall under three different categories: (1) military weapons, (2) nuclear goods, and (3) strategic goods with military and civilian uses. The last group is of particular interest to the economist. Unfortunately, no detailed information is publicly available as to what constitutes a strategic good; it seems, however, that 80 percent are in the category of electronics. The enactment of the Mutual Defense Assistance Control Act in 1951 resulted in another list of embargoed goods. The "Battle Act," as the Mutual Defense Assistance Control Act is known, called for denial of U.S. aid—military, economic and financial—to countries which exported embargoed goods. As long as Europe shared U.S concerns visa-vis communist countries and accepted U.S. political leadership, the Battle Act helped to develop a joint approach to the problem at hand. It would seem that in the late '40s and the early '50s the United States had a very decisive influence in determining what constituted trade in strategic goods and in enforcing restrictions on exports of those goods. Since then, however, time and changing conditions have weakened the strength of the U.S. position. One can attribute this development to several factors. First of all, the United States ceased to be the dominant supplier of strategic goods, particularly when the term is used only with regard to military hardware and equipment. It becomes much more difficult to set and enforce a global embargo policy when the number of independent suppliers increases. Furthermore, the potential threat of "using the stick" in the form of the Battle Act became ineffective when Western Europe grew independent of U.S. economic assistance, that is, roughly speaking, at the end of the Marshall Plan. It also has to be said that the maintenance of a joint U.S.-European policy towards exports of strategic goods to third countries required not only a good deal of political cooperation and cohesion among respective governments but also a strong backing on the part of concerned industries. The European business community eventually came to see great commercial opportunities in East-West trade and has become determined to exploit them.3 The American business community, while not exactly following suit and openly advocating liberalization of trade restrictions with regard to strategic trade with the Soviet Union, certainly became aware of lost opportunities. When President Nixon made a dramatic shift in the U.S. policy toward the People's Republic of China, American industry (and the general public) by and large favored liberalization of restrictions on strategic trade. In discussing the post-World War II history of strategic trade one must not overlook atomic weapons, equipment, and materials. Many countries joined efforts in developing rules governing trade in strategic

138

Henryk Kierzkowski

goods related to nuclear technology. In an attempt to check the spread of the nuclear threat, the Nuclear Nonproliferation Treaty was put into effect in 1968. The Nuclear Suppliers' Guidelines which followed the treaty requires the exporting countries to obtain assurances from the importers of nuclear material and equipment that the transferred goods and processes be used only for legitimate purposes. Once again, while sharing the basic objectives of the Nuclear Nonproliferation Treaty, conflicts and tensions have occasionally developed between the United States and Western Europe. Most recently, a nuclear reactor sale by West Germany to Brazil provoked strong objections from the U.S. To sum up, strategic trade during the first two decades after the end of World War II concerned mainly goods with direct and indirect military use. The rules of the game were primarily defined by political considerations. Gradually, though, it emerged that the United States and Western Europe, in spite of being close political allies, did not always share the same views on the matter. As one American expert put it: "Since World War II, Canada, France, and Britain, and other important trading partners have vigorously opposed the extraterritorial reach of American export restrictions through diplomatic and legal means. Even the extraterritorial aspects of American controls on exports to the Soviet Union of militarily useful goods and technology— controls accepted in principle and administered cooperatively by our allies—have led to sustained irritation in Europe".4 As time went by and the political cohesion of the Western allies weakened, tensions and differences of opinion between the United States and Western Europe mounted with regard to strategic trade with third countries. The most recent incident has involved the Siberian pipeline. The main facts of the pipeline episode are still fresh in the memory of the general public, but are well worth repeating. From the beginning the Reagan administration saw the 3,700 mile pipeline from Siberia to Western Europe as exposing the receiving countries to excessive dependence on a Russian source of energy supply. Even though, according to various projections, the pipeline would never satisfy more than 5 percent of Western Europe's energy demand, the U.S. administration considered the willingness of its allies to go ahead with the deal as imprudent at best. Additional concern was provoked by the fact that the Soviet Union stood to earn $10-12 billion a year in hard currencies from gas exports. These earnings could be used to purchase Western goods and technology to increase Russian military strength. Following the imposition of martial law in Poland on 13 December 1981, President Reagan ordered sanctions against the Soviet Union because of its "heavy and direct responsibility for the repression in Poland." Trade sanctions included severe limitations on exports of equipment and engineering know-how related to oil and gas exploration, production, and transmission. Following President Reagan's ac-

139

Strategic Trade, Embargoes, and Imperfect Competition

tion, exports and reexports of U.S. oil and gas equipment and associated technical data fell under very strict and specific licensing restrictions although more liberal and general licensing had formerly been allowed. Furthermore, outstanding validated licenses could be suspended or revoked. It appears that the extraterritorial effects of the December 1981 regulations brought the United States into open conflict with certain Western European countries. U.S. firms were ordered to cease exports of pipeline-related equipment and technology to Western European buyers if they knew that the material would be reexported to the Soviet Union. Furthermore, the embargo of certain exports to the Soviet Union also applied to U.S. owned or controlled foreign firms. It was immaterial whether those firms made any use or not of U.S. materials or technology in their exports to the Russians. While Western European countries were initially supportive of President Reagan's action, they were not prepared to follow the U.S. suit. Their support was limited to making public gestures of approval. When in June 1982 President Reagan decided to harden his stance against trade with the Russians, Western Europe came out strongly against the American embargo. Both the British and the French governments moved to block restrictions imposed by President Reagan on U.S. owned or controlled firms operating in England and France. One of the major outside suppliers for the Siberian gas pipeline was the French subsidiary of the American company Dresser Industries based in Dallas, Texas. Its contract with the Soviet Union, signed prior to the imposition of the restrictions, called for delivery of 21 compressors worth $18-20 million. When the American parent company of Dresser France asked its subsidiary to obey the U.S. government sanctions, the French Minister of Industry answered by instructing the company to ignore the embargo and proceed with the contract.5 In response to the French government action, the U.S. Commerce Department decided to apply penalties by issuing a denial order. The order barred Dresser France from receiving any technology, services, and equipment from its parent or any other American firm. In the end, the U.S. government was able to bring Dresser France to a halt by ordering its parent company in Texas to interrupt the flow of crucial data provided on a continuous basis through telephone and satellite channels. The pipeline episode, and especially the Dresser case, shows the extraordinary vulnerability that modern technology imposes on trading partners. Consequently, the concept of strategic trade must be considerably widened for it goes far beyond the confines of military uses. The concept of strategic trade must also cover various types of services. What seems interesting is that certain services such as design, engineering, and data processing have acquired strategic characteristics

140

Henryk Kierzkowski

in the production process—there are no substitutes for them. While it used to be that production of thege services and production of respective goods had to coincide geographically, modern technology allows these two stages to be increasingly separated. This separation could involve a high degree of strategic dependence. Thus Dresser France could be run on information provided by its parent company from Dallas or it could stop functioning if some vital services provided longdistance were withheld. In analyzing U.S.-Western European conflicts over strategic trade, legal scholars tend to focus on the question of the extraterritorial application of American law as being the main source of the problem. I think the conflict is more fundamental. The basic reason for U.S.Western European frictions stems from differences in the hierarchical importance attached to commercial policy in relation to foreign policy. In the view of American policy makers, foreign trade is part and parcel of foreign policy, and commercial policy should be subjugated to U.S. foreign political interests. Europeans, on the other hand, accept much less willingly the dominance of foreign policy over trade policy. Much more interdependence is seen in Europe in this respect, and occasionally foreign commercial interests seem to influence heavily, if not dictate, foreign policy. I hasten to add that this is particularly true of France and England, the two countries which still have global or at least regional strategies and whose political influence continues to count. Mrs. Thatcher frankly admits to be batting for England when she goes to Arab Gulf countries and triggers off an avalanche of commercial contracts. It is clear even to a casual observer that the French attitude towards the Iran-Iraq war is primarily determined by economic interests. It would seem, therefore, that there are rather strict limits to which Western European countries are prepared to subjugate their commercial interests to their own foreign policy, and certainly they are much less willing to abdicate their trade policy to the requirements of U.S. foreign policy. Today, there are also strong doubts about the efficacy of economic measures the United States proposes to take; the most recent example was provided by U.S. proposals for a trade embargo against Libya and Nicaragua. Also, there is often a feeling of unfairness in U.S. demands for sacrifices from its European allies. During the pipeline episode Michel Jobert, the French Foreign Trade minister, put it very blatantly: "[I]f the United States wants to respect its oil embargo, let it start by not delivering eight million tons of grain." It is my opinion that in spite of friendship and mutually shared longterm objectives, U.S.-Western European confrontations over issues of strategic trade will continue in the future. These conflicts will tend to follow the same pattern: A political or even a military confrontation with a third country will provoke the United States to take drastic

141

Strategic Trade, Embargoes, and Imperfect Competition

economic measures, among them the barring of certain types of trade. U.S.-USSR tension will probably continue to be a frequent starting point of these crises.6 There may well be other initial causes for embargoes. Libya, the Middle East, and Central America are geographic areas where a major crisis could easily erupt and force the United States to take decisive action. It can be almost taken for granted, for reasons indicated above, that Europe will behave in a much more restrained way than the United States, and it will certainly be less willing to use economic sanctions or join a U.S. initiated embargo. Confrontations with a third country will occasionally also lead to U.S.-Western Europe conflict. To ensure maximum efficiency of embargo measures, the United States may be forced to restrict exports of certain goods, technology, or services to its allies as well, or at least to U.S. firms operating in Western Europe. It is even probable that a denial of certain types of technology and services would be more likely and a more punishing action than the refusal of goods. What kind of costs and advantages are associated with strategic trade and associated embargoes? What policies should the government follow? I turn to these two issues in the next section. 5.3 Modeling Strategic Trade I now wish to develop a model of strategic trade which would incorporate at least some parameters affecting U.S.-Western European trade relations, either bilaterally or vis-a-vis third countries. Let me first address the question of the degree of market competition and strategic trade. Government policies towards this type of trade crucially depend on market structure. 5.3.1 The Standard Model In the standard trade model, perfect competition is assumed to prevail in the markets for goods and services. Trading equilibrium usually involves nonspecialization when countries have similar factor endowments. In equilibrium, exportables and importables are produced by a large number of firms. Firms can expand their output at will. Furthermore, reallocation of resources between sectors involves no adjustment costs. The model of international trade under perfect competition seems rather ill-suited to analyze the question at hand. For, in the absence of specialization, an interruption of U.S. exports to Europe for one political reason or another could not produce disastrous effects. A lack of goods previously produced in the United States could be immediately replaced by their perfect substitutes produced in Europe. If Western

142

Henryk Kierzkowski

Europe decided to take no action of its own and continue to export to the United States, it Would start running a trade account surplus. Of course, Europe could retaliate by banning its exports to the United States. If worst came to worst, an escalation of restrictions on exports could eliminate trade altogether. Both the United States and Western Europe would suffer welfare losses. In the limiting case one would move from free trade before embargo to autarchy after the imposition of embargo. Welfare losses would be exactly equal to the gains from free trade, and they would fall on both sides. What is important, however, is that the autarchy position could be reached immediately and at no extra cost. If, however, in addition to the United States other countries could supply Europe with products it needs, worst would not have to come to worst. Trade diversion would occur without Europe's being pushed back into the position of autarchy. To be sure, trade diversion would involve welfare losses as certain very efficient producers in the U.S would have to be replaced by less efficient producers elsewhere. Thus, the existence of alternative sources of supply could substantially reduce welfare losses inflicted on Western Europe by the United States. Enough has been said already to make clear that the standard trade model could be extended to the analysis of embargo, but it does not produce interesting results. The welfare cost of embargo can at most equal the gains associated with free trade. In a multicountry trading system, these losses must be even smaller. What are the policy implications of embargo analysis in the competitive set-up? In particular, should Western Europe reduce its dependence on trade with the United States in "sensitive" areas in anticipation of possible trade disruptions? The answer is an emphatic no. By taking preventive (and protective) action Europe would deny itself the benefits of free trade without achieving anything in return. Our conclusion follows from the proposition that free trade some of the time is better than autarchy all the time. However, this conclusion hinges on specific assumptions associated with the standard trade model. 5.3.2 Imperfect Competition Models Let us now move to the opposite case, namely that of monopoly. Suppose that we deal with an industry in which the existence of increasing returns to scale leads to one producer's taking over the whole market. We can think of the aircraft (or even better the satellite) industry as being one such example. To set the stage for the analysis I first wish to consider equilibrium in the United States and Western Europe prior to trade. Western Europe's airplane is called the Airbus; the American counterpart goes under the name of Boeing. Both aircraft have very similar

143

Strategic Trade, Embargoes, and Imperfect Competition

technical features—size, carrying capacity, speed, fuel efficiency, and so on. Once produced they can operate exactly n years provided that proper maintenance service is applied. In the Ricardian vein, one may assume that labor is the only factor of production. The amount of labor required to produce one Airbus is a, while it takes a* units of labor to manufacture one Boeing. (From now on variables with asterisks will refer to the United States.) In addition to its use in the production of airplanes, labor needs to be expanded to service them. The respective input-requirement coefficients per unit of time are s and s*, respectively. The nature of technology is such that only the producer has the know-how for servicing its airplane. Thus when final users buy airplanes they also commit themselves to buying a future stream of services. Suppose that, after a period of autarchy, trade in airplanes and associated services becomes possible between the United States and Western Europe. For the sake of simplicity let us assume that both economies are stationary. Thus, the United States and Western Europe will build their respective airplane fleets (of either Airbuses or Boeings), and then they will try to maintain a constant number of airplanes in operation. Which company will build the entire stock of planes for the American and European users depends on their relative competitiveness. Note that the concept of relative competitiveness now embraces production of goods as well as services.7 The concept could also cover their cost of financing. With regard to the latter factor, if Boeings and Airbuses were bought under credit conditions arranged by the producers or their governments, the sale of every airplane would carry an interest rate charge which would depend on the price of the airplane and the respective interest rates, i and /*. I do not wish to go into the discussion of what will be the actual price of the airplane charged by the dominant producer when the new equilibrium emerges. In order to determine who will prevail in the market it is rather more important to ask what is the very minimum price that each producer could charge without losing money.8 The cost of keeping an airplane in operation consists of three elements in this model: First, the depreciation cost which is the price of the airplane divided by the length of its useful life. The costs of producing an airplane are aw and *. The lowest depreciation cost could thus be awln and a*w*/n. Second, in addition to production costs there are also servicing costs. These costs are sw and S*H>*. Finally, there are also interest rate charges. (It is assumed, quite realistically, that the airplane producers also provide financing.) It is quite plausible that there exist economies of scale in servicing and in that case s = s(X + Z*) and s* = s*(X + Z*) with the properties

144

Henryk Kierzkowski

that s' < 0, s*'0. X + X* denotes the total production of airplanes for the North American and the European markets in the steady-state equilibrium. In order to capture economies of scale, servicing is done in one location, say, Seattle, Washington; London, England; or Toulouse in the south of France. The assumption of the existence of economies of scale in servicing of airplanes, ships, or oil rigs contains a large dose of realism. ^ We are now in a position to ask which producer is likely to capture the combined Europe-North American market. The winner will be the producer who will be able to offer a lower joint price of the airplane (or rather depreciation), maintenance servicing, and cost of financing. These prices are given by equations (1) and (2) and can be best thought of as the minimum prices the two producers would charge for leasing a plane. (1)

P = s(X + X*)w + awln + iaw.

(2)

P* = s*(X + X*)w* + a*w*/n + i*a*w*.

Equations (1) and (2) clearly show that one's comparative advantage may stem from one or more possible sources. With production and servicing technologies assumed identical, the country with the lower wage rate and the lower interest rate would capture the whole market. The latter element is very interesting because it is usually absent in trade models. Lower interest rates in the United States, with other conditions being the same, would allow the Boeing producer to sweep the whole market regardless of its size. This case is illustrated in figure 5.1a. The PP curve shows the minimum cost curve for the European producer and the P*/5* curve represents the same thing for the U.S. producer. Interestingly enough, comparative advantage in "financial packaging" can lead to domination of a market for goods and services. This advantage may be real or policy-induced. Figure 5.1a shows how governments can give domestic firms a competitive advantage by providing them with interest-rate subsidies. What seems of particular importance is that the "break" need not be big to lead to market dominance. Another interesting case of competitive advantage is depicted in figure 5.1b. Here the only difference between the Boeing and Airbus producers is that w* > w, and s*(X + X*) < s(X + X*), i.e., the U.S. wages are higher but the U.S. is more efficient in servicing at any level of output. The net result of this case is that when the market is relatively small, the low-wage country tends to dominate it; however, for a sufficiently large market, the country more efficient in producing services becomes the sole supplier. While the foregoing analysis suggests which producer may capture the market, it does not tell us what the equilibrium price will be. Surely,

145 P.

Strategic Trade, Embargoes, and Imperfect Competition P*

Pp

P

*

p* P

p

p*

Hg.

Fig. 5.1

a

X + X*

~

Fig.

b

X + X*

Minimum prices and market size

that price will not be equal to P or P* but it will also contain some pure monopoly profit. How much higher the actual price will be above the minimum cost of production is irrelevant for our analysis. It can be said, however, that the size of pure monopoly profit will depend on the PIP* ratio and the threat of potential reentry. It will also depend on the commercial policy pursued by the importing country. When monopoly profits are present, some of them can be recaptured by the importing country through tariffs, taxes, and so on. The trading equilibrium price will be lower than the autarchy price. Consequently, the steady state output of airplanes for each market should be expected to be higher than before the opening to trade. Finally, the welfare level reached by each trading partner will be higher than under autarchy. I now turn to the consequences of an embargo. Suppose that Europe ends up being the importer of airplanes and associated services. One would expect that Europe would switch to Boeings only gradually rather than all at once. In any case, the transition should not take more than n periods after trade is allowed. Assume that as of time t0 Europe's entire fleet already consists of Boeings. Once it switches to Boeings, Europe could be dealt a very severe blow by being denied, not so much American airplanes, but rather maintenance services. This is easily accomplished in this model because servicing of Boeings owned by Europe is done, because of economies of scale, in Seattle, Washington.

146

Henryk Kierzkowski

The U.S. government could turn off the flow of maintenance services at will without worrying about the problem of extraterritorial application of the U.S. law. Suppose that indeed an embargo is imposed at time tk. The stock of Boeings held by Western Europe becomes completely useless under the extreme assumption that Europe is unable to service American airplanes. This is the major cost of the embargo. If the cut-off of services was perceived permanent by Western Europe, a new fleet based on Airbuses would have to be built from scratch.9 This surely could not be accomplished instantaneously. Suppose that it would take T periods to get the production running again and then the new steady-state position could be reached at once. The fact that adjustment takes time (it may involve other costs as well) means that from tk to tk+T Europe would be deprived of using airplanes. At the end of the transition period, Europe would return back to the autarchy position. Note that a permanent embargo would lead to autarchy even if Europe could service the existing fleet of Boeings. Europe's servicing capability could only reduce the adjustment cost. In order to evaluate the welfare effects of the embargo, one needs to introduce Europe's welfare function. Suppose it is of the following form: U - U(x,y), where x is the flow of consumer services provided by the existing stock of airplanes and y stands for consumption of all other goods (for simplicity's sake "all other goods" are nondurable). The utility function often has a property that U(0,y) = U(x,0) = 0, i.e., both goods are indispensable. In that case the utility level achieved by Europe at different points in time can be shown in figure 5.2. The level of utility achieved between t0 and tk corresponds to the trading equilibrium and it can be called U(T). The utility level enjoyed for t > tk+T represents the autarchy level and can be denoted U(A). By our previous argument, it must be true that U(A) < U(T). We can now evaluate the present value of the stream of utility generated between t0 < t < oo. The rate of time preference is assumed to be r. (3)

^ , = /[/(. . .)e~«dt + } U(. . .)e~«dt. t0

tk + T

If Western Europe never switched to Boeing its welfare would be (4)

M > 2 = / U{. . .)e~«dt. o Comparison of equations (3) and (4) clearly shows that reliance on American airplanes can increase or reduce Western Europe's welfare depending on whether

(5)

?(t/(. . .) - U(. . .))e-«dt § 'Yu(. . .)e~«dt . to

*k

147

Strategic Trade, Embargoes, and Imperfect Competition

U(x. y)

to Fig. 5.2

tk

tk+T

TimQ

Welfare over time

The above condition says that trade some of the time may not be superior to autarchy all the time. The reason is that gains from trade achieved between t0 and tk may not be big enough to compensate for subsequent welfare losses. It should be stressed that this result does not depend on the assumed utility function. If the utility function had the property that U(xfi) > 0 and U(09y) > 0, the utility level achieved between time tk and tk+T would be positive but necessarily smaller than U(A) because some resources would have to be engaged in rebuilding Europe's airfleet. Our analysis so far has not allowed for uncertainty of an embargo's being imposed, but this problem could easily be taken care of. One need only introduce the probability of an embargo occurring over a time interval. Suppose that, if no embargo had occurred by time t, the probability of this happening over the time interval (t, t + di) is given by e~zt. Now the probability that an embargo will be imposed during oo

the interval (0,) is Je~ztdt. Given this probability one could reforo

mulate our problem in terms of the expected present value of welfare. Recent literature on international trade under monopolistic competition suggests that there are cases where restricted trade is superior

148

Henryk Kierzkowski

to free trade. In the case of an embargo, one needs to know whether no trade is better than restricted trade. Equation (5) holds the answer to the question. It is clearly possible to envisage a situation where a threat of embargo would cause a country to eliminate its dependence on foreign trade. Having constructed such a case we want to stress its special character and point out that other policy measures should be considered before a country retreats into autarchy. Strategic dependence of Western Europe on American airplanes or rather maintenance services provided by the Boeing producer could be eliminated altogether if maintenance could be done by Europeans themselves. That may or may not be possible depending on a particular product and the market structure. This dependence could at least be reduced if servicing was done by the Boeing producer, but in Europe rather than in Seattle, Washington. In this situation, the ability of the U.S. government to enforce an embargo could be substantially reduced. (Yet one should not forget the Dresser France case discussed earlier.) The last conclusion should be good news for those who are in favor of liberalization of trade in services and freedom of establishment. Of course, freedom of setting up servicing stations in Europe does not imply that the Boeing producer would want to do it. In fact, the assumed existence of economies of scale in servicing pushes for a central location of servicing activities in any case. To have a servicing network in Western Europe in addition to the one in North America would reduce benefits from scale economies. However, Western Europe could well insist on such a solution before switching to Boeing. Some efficiency could be foregone for greater security. In the model developed above, price bidding would result in a monopolistic market structure. The producer who is potentially able to offer a lower price takes over the whole market. It is interesting to consider now the case of duopoly and see how our analysis and policy conclusions change. Our basic model can be readily modified so that it leads to a duopolistic market structure. It suffices to assume that when trade is allowed to take place, the two producers follow the Cournot strategy in determining their supplies for each market. Such a model has been recently applied by Brander (1981), Brander and Krugman (1983), and others, so there is no need to dwell on details. It turns out that we can even assume constant marginal costs in production of airplanes and their servicing without affecting the results in a substantial way. The general line of the argument goes as follows. Under the Cournot strategy, the Boeing producer sets the optimal levels of output for the American and Western European markets assuming that the Airbus producer will keep his respective level of production unchanged. The Cournot model is symmetrical, hence the

149

Strategic Trade, Embargoes, and Imperfect Competition

European producer follows the same logic. Solving the model results in four reaction functions, two for each producer. One pair of such reaction functions is shown in figure 5.3 with regard to the European market. The vertical axis measures the steady state number of Boeings sold every period in the European market. The number of Airbuses sold in the same market is measured along the horizontal axis. The European reaction function, EE, is steeper than the American reaction function, AA. The fact that the two curves intersect means that both airplanes will be used in Europe. A similar solution holds for the U.S. market. In figure 5.3 the U.S. duopolist is shown to have a more than 50 percent share of the market. This would happen if either a* < a, or c* < c, or both inequalities happen to hold. In the duopoly model, simultaneous presence of the two producers in the Western European market implies that an embargo on sales of Boeings or denial of maintenance services cannot deal a devastating blow to Europe. While it is still true that Western Europe's stock of Boeings can be rendered useless, there are European planes in operation and, even more important, there is a continuing stream of new

(s + t)XH, while a sufficient condition for direct investment to be superior is that G < (s + i)XE. 6.3.3 The Multinational as Duopolist If there is potential entry from a single domestic producer, the model now takes the form of a multistage game. In the initial stages, the multinational decides whether to sell, and, if so, whether to export or to invest, while the domestic firm decides whether to enter. For the moment, let us not identify the order in which the different firms' entry decisions are made. At the final stage of the game the firms compete, if both have entered, and output levels get determined. I assume that the decisions at earlier stages of the game are made in the knowledge of the equilibria of later stages, so the equilibrium is perfect. The equilibrium is solved by working backwards. In the event that the domestic firm does not enter, we know from the previous section what are the payoffs to the alternative actions open to the multinational. If both firms enter the market, I assume that they act as Cournot duopolists, which is to say that each firm in equilibrium takes the output

165

East-West Trade, Embargoes, and Expectations

of the other firm as given in choosing its own output level. It should be noted that an implication of the analysis of Fudenberg and Tirole (1984) is that the results will not be robust, at least in detail, to changes in this assumption about the nature of the competitive interaction at this stage of the game. If the multinational supplies the market by exporting, then, indicating the host country firm by the subscript 1 and the multinational by the subscript 2, the respective objectives of the firms are independent of the now-sunk costs and can be written max P(XX + X2)XX - cXx ,

(2)

max P(X{ + X2)X2 - (c + s + i)X2 ,

while, when the multinational has established a plant, the objectives are the same but with s + t = 0. In figure 6.2, the interaction between the two firms in the post-entry game is represented by the two reaction curves which describe the optimal output of each firm as a function of its rival's output. If we suppose only that the demand function P(XX + X2) is sufficiently well behaved so that the reaction curves slope and cross in the way shown in figure 6.2, then it is easily shown that an increase in s +1 moves the

X EH Fig. 6.2

XEXHH

XH

Duopoly equilibria

X2

166

Alasdair Smith

multinational's reaction curve in to the left, and the equilibria in both situations are as illustrated in figure 6.2. The presence of transport costs and tariffs drives the equilibrium away from the symmetric duopoly solution (XHH,XHH), to a lower level of sales XEH for the exporting firm, and a higher level of output XHE for the host country firm. It is easily seen also that the variable profits of the host country firm are higher and of the multinational lower when the multinational is an exporter than when it is a direct investor. It is also obvious that a firm will make greater profits as a monopolist than as a duopolist. Note how the choice between exporting and investing is different in duopoly from in monopoly. A monopolist simply compares the fixed cost of investment with the transport and tax costs of exporting. A duopolist, whose output level, other things equal, is less than that of a monopolist, might be expected to be more likely to choose the export option. But this is to ignore the strategic value of making the fixed investment which lowers variable unit cost by s +1, and thereby, with Cournot competition, expands the multinational's output and profit while contracting its rival's output. 6.3.4 Embargo Threats and Entry Decisions In applying this model to the analysis of U.S.-European trade, I make the apparently eccentric assumption that Europe, both East and West, is to be treated as one country, with the U.S. as the other country. The justification for this structure is that it is simple but also sufficient to illustrate the essential points about the effects of multinationality on the effectiveness of embargoes. I assume that an embargo can only be effective if the multinational has chosen to export; that, in other words, the actions of foreign subsidiaries are beyond the reach of the multinational's home government. This is a very considerable narrowing of the focus of the model, and restricting of the potential effectiveness of an embargo, and this feature of the model is justified also precisely in order to focus attention on the way that multinationality may reduce the effectiveness of embargoes. The assumption that the multinational can make positive profits as an exporter, so that it will never choose to stay out of the host-country market, means that, in the absence of an embargo, the multinational must choose between exporting and investing, while the host country chooses between entering and not entering. However, an embargo could exclude the multinational from the market. The payoffs to the various decisions derive from the Cournot equilibria of the post-entry games discussed in the previous section and are set out in table 6.2, where H is the host country firm, M the (potential) multinational, and in each payoff pair, M's payoff is given first.

167

East-West Trade, Embargoes, and Expectations

Table 6.2

The Entry Game Not Enter M

Export Invest Excluded

A, O B-G, O O, O

H

Enter

C£,C„-F-G D-G, D - F - G O, B - F - G

The payoffs A, B, CE, CH, and D are given by A = P{XE)XE - (c + s + t)XE B = P(XH)XH - cXH CE = P(XEH + XHE)XEH — (c + s 4- t)XEH = CH P\XEH + XHE)XHE — cXHE D = P{XHH + XHH)XHH — cXHH We have assumed that CE > 0 and we know that B >A > CE and that B > CH > D > CE. An embargo will, by assumption, have no effect if the multinational has chosen to invest. If, however, it has chosen to export, then an embargo will move it down to the "excluded" row in table 6.2. But insofar as an embargo is anticipated, it changes the payoffs to investment by the multinational, reducing the expected values of A and of CE towards zero. It also raises the expected value of CH towards D, but the host country firm can always enter after the imposition of an embargo, so it need only concern itself with the effect of the anticipation of an embargo on the behavior of the multinational. The effect of an embargo will depend on which equilibrium is chosen and unfortunately the above inequalities are not sufficient to narrow down greatly the range of possible equilibria. I shall here consider only some of the possibilities. It is natural to model firms' entry decisions as being sequential, and it is important, in doing this, to observe that only a decision to sink entry costs—investment by M, entry by H—is an irreversible commitment. A decision by M to export or by H not to enter is a decision which could be reversed after the rivalfirmhas made its entry decision, and as we have noted above, a decision by H not to enter could be reversed after an embargo is imposed on exports by M. For more formal analysis of sequential decision making in this model, see Smith (1987). Strategic Foreign Direct Investment Suppose that M makes its entry decision first, and suppose that (B>)CH>F+G>D so that H will enter only if M is an exporter or is

168

Alasdair Smith

excluded. Suppose also that CE>D-G and A>B-G, so that simple comparison of the costs of exporting with the costs of foreign direct investment would indicate that M should export, whatever action is taken by H. In this case, it is actually optimal for M to invest, because by doing so it will ensure that H chooses not to enter. This example illustrates the potential strategic role of foreign direct investment as an entry-deterrent in this kind of model, to add to its role in strengthening oligopolistic power discussed at the end of section 6.3.3. Thus, a multinational may have strategic incentives to undertake foreign direct investment even when cost comparisons do not seem to justify such a policy, and in this event it puts itself beyond the reach of an embargo independently of the embargo threat itself. Embargo-Induced Foreign Direct Investment In the event that F+G>CH>D, H will choose not to enter if M is not excluded, whether or not it makes its entry decision before M. If B — G>A, then M will undertake foreign direct investment, and the embargo has no effect. If, on the other hand, A>B-G in the absence of embargo threats, M would choose to be an exporter. The threat of an embargo, however, may reduce the expected value of A below B-G, in which case M will be induced by the threat of embargo to invest, and thereby make the embargo ineffective. Entry in Response to Embargo Continuing with the previous case, suppose that the embargo threat is not perceived by M to be sufficiently strong to reduce A below B-G, so that M is a monopoly exporter. It does not follow that an embargo will be effective, for once M is excluded from the market, H faces returns of B- F—G from entry, and it will enter if this is positive. An Effective Embargo It follows, then, that an embargo will be effective only if the fixed costs F + G of entry by H are higher than the profits B that it can make as a monopolist, if the fixed costs G of foreign direct investment by M are sufficiently high so that the returns A to exporting exceed the net profits B-G of foreign direct investment, and if embargo threats are not perceived to be serious enough to reduce the expected value of A below B-G and induce foreign direct investment. 6.4 Conclusions The cases analyzed briefly in the previous section are not exhaustive, but they do illustrate various theoretical possibilities, and in particular illustrate in a formal model some of the issues which were raised earlier

169

East-West Trade, Embargoes, and Expectations

in the paper. The requirements for a successful embargo, in the absence of effective extraterritorial control of subsidiaries and licensees are quite demanding. Embargoes seem unlikely to succeed if there has been diffusion of technological knowledge and independent production capability through multinationals either in response to market pressures or in response to perceived embargo threats; or if the nation contemplating the embargo has a reputation for the political use of trade restrictions; or, finally, if the acquisition of the requisite knowledge and capability by independent foreign firms is not prohibitively costly. The skepticism of the Western European countries and of the Office of Technology Assessment may be well founded.

References Bayard, Thomas, O., Joseph Pelzman, and Jorge Perez-Lopez. 1983. Stakes and risks in economic sanctions. The World Economy 61: 73-87. Becker, Abraham S., ed. 1983. Economic relations with the USSR. Lexington, Mass.: Lexington Books. Cooper, Richard N. 1987. Trade policy as foreign policy. U.S. trade policies in a changing world economy, ed. Robert M. Stern, chap. 6. Cambridge, Mass.: M.I.T. Press. Daoudi, M. S., and M. S. Dajani. 1983. Economic sanctions: Ideals and experience. London: Routledge and Kegan Paul. Dunning, John H. 1979. Exploring changing patterns of international investment: In defence of the eclectic theory. Oxford Bulletin of Economics and Statistics 41: 269-96. Ethier, W. J. 1983. Modern international economics. New York: Norton, chap. 7. Fudenberg, Drew, and Jean Tirole. 1984. The fat-cat effect, the puppy-dog ploy, and the lean and hungry look. American Economic Review (papers and proceedings issue) 74: 361-66. Guillaume, Jean-Marie. 1983. A European view of East-West trade in the 1980s. Chap. 6 of Becker (1983). Hanson, Philip. 1983. The role of trade and technology transfer in the Soviet economy. Chap. 3 of Becker (1983). Hirsch, Seev. 1976. An international trade and investment theory of the firm. Oxford Economic Papers 28: 258-70. Horstmann, Ignatius, and James R. Markusen. 1987. Strategic investments and the development of multinationals. International Economic Review 28: 10921. Office of Technology Assessment. 1983. Technology and East-West trade: An update. Washington, D.C.: United States Congress. Rode, Reinhard, and Hanns-D. Jacobsen, eds. 1985. Economic warfare or detente: An assessment of East-West economic relations in the 1980s. Boulder: Westview Press. Smith, Alasdair. 1987. Strategic investment, multinational corporations, and trade policy. European Economic Review 31: 89-96.

170

Alasdair Smith

Stent, Angela. 1981. From embargo to ostpolitik: The political economy of West German-Soviet relations, 1955-1980. Cambridge: Cambridge University Press. Vernon, Raymond. 1971. Sovereignty at bay: The multinational spread of"U.S. enterprises. New York: Basic Books.

Comment

L. A. Winters

These papers* are interesting and informative attempts on a difficult subject. Each comprises a historical-cum-political analysis followed by a simple model of the sort with which economists are used to working. While there are many similarities between the historical sections of the papers, the models are nicely complementary: Kierzkowski deals with the potentially embargoed country, while Smith deals mainly with the embargoing country and its firms. It is no disrespect to either author to note that their models have a very long way to go to match the subtlety of their historical analyses. Nonetheless most of my comments concern the former rather than the latter. In U.S.-European trade relations two aspects of embargoes are of interest: first the partners' different attitudes towards embargoes on third countries, and second the possibilities of embargoes between the two. The former revolves around the concepts of linkage and leverage discussed by Smith. The distinction between these concepts is undoubtedly interesting and fruitful, but I am not sure that it lies entirely in the realms of frustrated vs. fulfilled expectations. Both linkage and leverage involve the target country—the USSR—paying a price, i.e., doing something it would not otherwise do. Both require the initial expectation that the West continue to deliver on its side of the bargain. Both require that Western promises/threats to continue/curtail delivery are believed. Both are defined over the same pair of outcomes: either both the USSR and the West "deliver," or neither does. It seems to me that the crucial distinction lies not in expectations but more in the process by which the "price" is declared. Under linkage this is done quietly and implicitly, under leverage overtly and explicitly. At least in the Jackson-Vanik case, leverage entailed the USSR's paying a higher price—it added national humiliation to the price of Jewish emigration. Not only was this unpalatable to the highest Soviet authorities, but it may also have upset subtle coalitions within the Soviet hierarchy, bringL. A. Winters is professor of economics at the University College of North Wales, Bangor, U.K., and a research fellow of the Centre for Economic Policy Research, London. ^Chapters 5 and 6 in this volume.

171

East-West Trade, Embargoes, and Expectations

ing into the game new players who were excluded while it was being played on purely commercial grounds.1 Turning to the possibilities of U.S.-European embargoes, Smith's conclusion that firms are likely to prefer foreign direct investment over exporting as a means of evading embargoes seems correct. It probably does not generalize to more extreme political situations, however. Both the possibility of retaliatory nationalization of foreign assets, and the possibility that the embargo or other pressure would bring the target country's economy to its knees reduces the relative attractiveness of foreign direct investment. Thus, for example, I suspect that firms considering supplying South Africa today would favor exporting, with its lower initial commitment, to direct investment. Kierzkowski argues convincingly that the case for restrictive trade policy based on the threat of future embargo is rather slight. Even so, I believe his model tends to exaggerate it. First, the case for policy requires that the perceived private costs of embargo fall short of the social costs. With the exception that governments might have better information about future foreign policy shocks, this seems unlikely. Second, his example of airliner sales, in which purchase involves a necessary commitment to foreign servicing, provides precisely the sort of circumstances in which leasing is an attractive alternative to outright purchase. Under leasing arrangements the purchaser at least avoids losing his capital expenditure in the event of an embargo. Third, the nonstorability of the embargoed good is not completely generalizable. Storage may be a cheaper response to the threat of embargo than is autarchy; certainly strategic stocks of several goods do exist. Fourth, the utility function in which C/(0,y) = 0 means that no insurance against embargo is worthwhile. If higher ^-consumption did compensate for the loss of JC, countries might try to insure themselves on foreign insurance markets, although it is not guaranteed that such contracts would be honoured. More to the point, however, if y were worth something during the embargo, there would be more incentive to maximize the level of y-capacity prior to any embargo; that is, less incentive to sacrifice y-output for jc-output by a policy of autarchy. 1. See H. Raiffa, The Art and Science of Negotiation (Cambridge: Harvard University Press, 1982), on the negotiating penalties of trying to humiliate one's opponent.

7

The Steel Crisis in the United States and the European Community: Causes and Adjustments David G. Tarr

7.1 Introduction The years since 1974 have been very difficult ones for the steel industries of the United States and the European Community (EC). Production in both regions has dropped by more than one-third and employment has fallen even more. In recent years there have been either large losses or small profits.1 Data on these trends are presented in table 7.1. In response to these developments the U.S. government and the Commission of the European Community have adopted rather similar external policy measures but quite different domestic measures. How well are these measures suited to meet the problems affecting the industry and how might they be modified to deal more effectively with these problems? In order to answer these questions, the fundamental causes of the industry's problems are discussed in section 7.2. The policy responses and their effects in the U.S. and EC are described in section 7.3. The last section of the paper evaluates these policy responses. 7.2 The Causes of the Crisis 7.2.1 New Entrants in the International Steel Market In the past 30 years there has been a dramatic shift in the pattern of steel production, exports, and imports around the world. The basic David Tarr wrote this paper while visiting at the International Economic Research Division of the Development Research Department of the World Bank. He is a senior economist with the Federal Trade Commission. The author would like to acknowledge the helpful comments of Robert Baldwin, Andrzej Olechowski, Hans van der Ven, and Theo Dage. The views expressed are those of the author and do not necessarily reflect those of the World Bank, Federal Trade Commission or those acknowledged.

173

1977

113,700 126,121 38,984 22,094 23,335 4,923 11,256 4,329 20,467 47 686

1974

132,195 159,881 53,232 27,021 23,804 5,873 16,227 6,447 26,667 110 536

1974 na 155,526 53,232 27,020 23,789 5,840 16,225 6,448 22,318 110 535

1984 83,941 119,316 39,389 19,000 24,061 5,739 11,305 3,987 15,121 166 548

101,455 127,732 43,838 23,176 26,501 5,264 12,322 4,618 11,277 2 734 160,000 200,869 67,701 33,282 34,168 8,230 19,151 8,200 28,870 91 1,177

1977 153,700 202,536 66,924 32,512 39,385 8,512 19,659 6,380 27,954 90 1,120

1980

1984 135,300 172,851 51,556 28,829 37,271 7,965 15,664 6,380 23,991 345 850

Capacity (crude) (thousands of metric tons)

1980

Crude Steel Production (thousands of metric tons)

512,395 794,770 232,037 157,833 95,656 25,077 63,738 23,503 196,926 na na

1974

Steel Production, Capacity, and Labor Employed in the U.S., EC, and Member Countries

452,388 721,619 209,465 142,992 96,593 23,293 49,752 17,437 178,874 704 2,509

1977

398,829 597,873 197,406 104,940 99,528 21,047 45,220 14,904 112,120 527 2,181

1980

Employment (thousands of workers)

236,002 445,843 152,467 85,064 75,611 18,748 37,184 12,713 61,856 657 1,543

1984

Sources: International Iron and Steel Institute, Annual Statistical Report, various years for production data; Eurostat, Iron and Steel Yearbook, various years, for EC capacity and employment data; and American Iron and Steel Institute, Annual Statistical Report, various years for U.S. capacity and employment figures.

United States EC9 FRG France Italy Netherlands Belgium Luxembourg United Kingdom Ireland Denmark

Table 7.1

175

The Steel Crisis in the U.S. and EC

trends are summarized in table 7.2. In 1950, the United States produced almost half of the world's steel output; by 1984, its share was less than 12 percent of this output. The U.S. became a net importer of steel in 1959 and has imported over 20 percent of its steel throughout most of the 1980s. The European Community tripled its steel production between 1950 and 1970. Its output has been reduced from the peak of the early 1970s, but the EC remains a net exporter of steel products. Japan emerged as a major player in the international steel market during the 1960s and early 1970s, with an increase in production of over 100 million tons between 1960 and 1973 (the peak year for Japanese production). In fact, Japan's increased production capability is underestimated by the production figures because Japan is estimated to have been operating at only about 65 percent of capacity during the last decade (Bradford 1986, 82). As will be shown in the next section, in the last 10 to 15 years some developing nations, especially Brazil and Korea, have also emerged to become important players in the international steel market. 7.2.2 Decline in Demand In the past 12 years, growth in overall demand for steel has been extremely sluggish; it can essentially be characterized as a period of zero growth. In 1973-74 steel production peaked in the noncentrally planned economies at about 519 million metric tons. After a recession in 1975, steel demand began to recover until it reached its 1973 level in 1979-80. Since 1980, however, steel production has remained at or below 500 million metric tons in the noncentrally planned economies with 1982 and 1983 being especially disastrous years for steel demand. The broad trends are summarized in table 7.3. The years 1973-74 were very encouraging ones for the future of the world steel industry. Most forecasters of steel demand predicted significant growth in steel demand over the 1975-85 period.2 As a result, a number of nations undertook major expansion projects for their steel industries. By the early 1980s, however, the optimism of 1974 had turned to pessimism as many companies, especially those in the U.S. and the EC, experienced significant losses. Overall trends mask an even more ominous trend as far as the United States, the EC, and other industrialized countries (ICs) are concerned, From 1973 to 1984 there was a significant decline in steel consumption in the industrialized countries, from 416 to 322 million metric tons, while consumption in the developing countries (DCs) increased from 100 to 181 million metric tons, a total of 81 percent. Over this period demand in the United States fell by 36 million metric tons (crude steel equivalent) to 113 million tons and by 33 million metric tons (crude steel equivalent) in the EC to 95 million tons. Thus, while steel demand

Percent of World Total 69.2% 49.4 33.7 25.6 29.6 23.8 17.8 16.9

Output 3 7.5 20.3 27.8 53.2 107.8 151.6 140.1 132.5

Net Exports

-0.7 0.9 2.2 1.6 -0.2 -6.3 -12.4 -22.9

16.2% 34.2 59.8 48.4 27.6 21.6 14.1 11.7

1.8 14.6 49.2 96.8 99.3 131.5 111.8 91.5

European Community

Percent of World Total

b

Output a

United States

1.2 5.0 8.0 9.0 9.7 7.4 15.0 11.3

Net Exports

b

0.9 5.3 24.4 102.9 122.8 116.4

—

Output 3

The Changing International Positions of the U.S., EC, and Japanese Steel Industries

1.1% 2.6 6.7 16.1 15.5 14.9

—

Percent of World Total

Japan

a

.4 2.5 22.3 34.7 31.8

— —

b

Net Exports

Source: Adams and Mueller (1986). In million net (or short) tons of raw steel. b Exports minus imports, in million net tons of steel products. (One product ton is roughly equivalent to 1.25 tons of raw steel.)

1870 1900 1920 1950 1960 1970 1980 1984

Table 7.2

10.8 41.1 82.3 200.0 360.3 637.8 792.2 783.0

Output 3

World

177

The Steel Crisis in the U.S. and EC World Steel Production Trends

Table 7.3

Levels (millions of metric tons of crude steel equivalent)

ICs DCs a Subtotal CPEs Total

Growth (% per annum)

1960

1973

1980

1982

1984

1960-73

1973-80

1980-84

233 20 253 87 340

463 57 520 179 699

407 94 501 215 716

338 97 435 210 645

376 113 489 221 710

5.4 8.4 5.7 5.7 5.7

-1.8 7.4 -0.5 2.7 0.3

-2.0 4.7 -0.6 0.7 0.0

Source: Compiled from data available in International Iron and Steel Institute, Annual Report, various issues. a China is included in the DCs.

Statistical

in the developing world has been increasing, the opposite trend is apparent for the industrialized world. These trends are summarized in table 7.4. Most forecasts of demand for the 1985-95 decade predict continued slow aggregate growth of about 1 percent annually for the noncentrally planned economies. Again, however, there is a sharp contrast in the expected growth patterns of the industrialized countries and the developing world. For the next decade, the International Iron and Steel Institute (IISI) forecasts annual growth of 0.1 percent in the industrialized countries, but 3 percent in the developing countries. The growth of steel demand in the developing world and not in the industrialized countries is explained by the IISI's intensity-of-steeldemand curve. The curve is based on data that this organization has collected showing that per capita consumption of steel increases (at a decreasing rate) with national per capita income up to a maximum and then begins to decline,3 There are a number of reasons why the steel intensity curve has the observed shape. First, infrastructure expendiTable 7.4

World Steel Consumption Trends Levels (millions of metric tons of crude steel equivalent) 1960 1980 1982 1984 1973

ICs DCs a Subtotal CPEs Total

218 39 257 89 346

427 87 514 181 695

360 147 507 214 721

298 144 442 209 651

339 163 502 218 720

Growth annum) 1980-84 1973-80

(% per 1960-73 5.3 6.4 5.5 5.6 5.5

-2.4 7.8 -0.2 0.0 0.5

-1.5 2.6 -0.2 0.5 0.0

Source: Same as Table 7.3 except for 1960 data, which are taken from the World Bank data base.

178

David G. Tarr

tures tend to be significantly reduced after a given level of development is reached. Second, the share of service industries has been rising in most Organization for Economic Cooperation and Development (OECD) countries since 1974. As economies shift into banking, financial services, and insurance, and away from traditional smokestack industries, there is a decline in the demand for steel. Third, there is a saturation level for some consumer durables, such as refrigerators; once reached, this results in slower growth. A fourth reason is that manufacturers of products such as automobiles and cans have substituted alternative materials, and fifth, technological advances have reduced the demand for steel. For example, it is estimated by IISI that continuous casting reduces the demand for raw steel by 15 percent. In the decade from 1974 to 1984, the industrialized countries' production of steel by continuous casting went from 15 percent to 64 percent, while that of the developing nations went from 15 percent to 36 percent. Also, the development of stronger, thinner gauge steel has reduced the demand for final steel products.4

Table 7.5

Unit Cost for Inputs: U.S. and Japan (dollars per metric ton of steel produced) Total

Labor

Iron ore

Scrap

Coking coal

Year

U.S.

Japan

U.S

Japan

U.S.

Japan

U.S.

Japan

U.S.

Japan

1976 1975 1974 1973 1972 1971 1970 1969 1968 1967 1966 1965 1964 1963 1962 1961 1960 1959 1958 1957 1956

294.65 270.27 215.55 161.21 155.11 145.98 137.23 125.25 119.40 117.70 113.21 112.99 114.97 116.01 118.74 122.50 120.18 113.98 122.18 110.00 110.84

161.93 159.26 147.30 100.97 83.56 81.28 78.05 69.93 67.78 69.53 71.86 76.38 75.20 79.03 81.56 91.59 85.08 90.04 98.65 133.21 119.83

143.55 132.87 100.91 87.31 89.52 85.03 80.81 75.18 70.35 69.88 65.93 65.06 67.00 69.62 71.36 72.36 71.83 66.67 70.09 60.24 54.67

49.64 49.93 42.60 35.32 31.97 27.98 23.22 21.20 20.83 19.93 20.68 22.11 20.97 23.76 24.10 21.94 23.01 25.02 30.12 26.79 26.66

44.51 37.58 29.66 24.42 23.84 22.85 21.54 20.34 20.65 20.10 19.95 19.92 20.41 19.60 19.93 20.58 19.47 17.25 19.75 18.17 17.51

26.87 27.85 21.65 17.62 16.97 19.43 17.47 16.66 16.99 16.68 18.14 18.63 16.73 17.80 18.97 18.54 17.91 18.08 21.20 31.55 25.78

21.82 18.98 34.10 17.08 11.26 8.53 10.05 8.60 6.71 6.73 7.72 8.56 8.25 7.39 6.83 9.45 8.24 10.87 9.94 10.95 17.78

22.72 17.23 33.65 23.38 12.04 9.06 16.05 14.00 12.16 15.73 14.88 16.75 19.27 18.12 17.43 30.09 23.16 24.59 19.37 37.98 35.15

53.73 52.40 29.20 17.44 16.08 15.15 12.80 10.29 10.69 10.83 9.99 9.78 9.74 9.16 10.17 10.21 11.48 10.93 13.09 12.73 12.15

41.38 43.18 29.84 15.18 14.65 16.76 14.65 11.72 10.91 10.27 10.84 10.94 10.05 10.99 12.33 11.85 11.50 13.03 16.75 23.03 20.01

Source: Federal Trade Commission Staff Report (1977, pp. 113-14).

u w c CO 0>

0> to

£

O

U >

I ^

1s

CD

3

C CD

5

S^

S

CD

X CD ^ CD . 2 to

iS

OX) I S *->

I D 0> CJO cd

CD

13

M

1*

CD

CD

13 CD

^

V

cd

o ^ x

CD ON H J

CD ' ^

c5

CD o •£ .2 o c CD c U CD C L cd X ) •C iS a CD -4-» CD cd ps

for 0 = qes

The first expression states that for Europe-to-U.S. trade to take place the domestic U.S. price equals the combination of the domestic European price, the U.S. tariff, and the transport costs. If not, equation (8) holds; that is, there will be no trade. From (3) the equations for domestic prices in Europe and the United States can be written as: (9)

pe = px (1 + txe){\ + uxe)

and (10)

ps = px{\ + txs){\ + uxs),

where px is the domestic price in the rest of the world. Assuming that Europe-to-U.S. trade takes place, i.e., (7) holds, and inserting (9) and (10) into (7), we can then write: (11)

px{\ + txe){\ + uxe){\ + tes){\ + zes) = px(\ + txs)(l + uxs).

If one ignores existing tariff preferences, and assumes that Europe and the rest of the world face the same tariff in the United States, this expression reduces to (12). The assumption of identical tariffs faced by European and rest-of-the-world suppliers in the United States is likely to hold for commodities subject to NTBs. Typically, tariff preferences on "sensitive" commodities are minimal in the United States and Europe. (12)

1 + uxs = (1 + uxe){\ + txe) (1 + zes).

Equation (12) states that where there is Europe-to-U.S. trade, an increase in the restrictiveness of U.S. NTBs will cause prices to rise in Europe. Suppose import quotas are reduced, or rules of origin for countries in the rest of the world are changed in order to restrict imports. If introduced by the United States, such policy changes would increase the U.S. MTE rate, uxs, and raise the domestic U.S. price. This in turn would lead to increased imports from unrestricted European producers. Since in case 1 imports into Europe are assumed to befixedbecause of European NTBs, the increased exports from Europe to the United States would reduce overall supply in Europe, and consequently increase the domestic price in Europe. To achieve a given percentage increase in the U.S. price by increasing the restrictiveness of U.S. NTBs, U.S. imports from the rest of the world must be reduced by a larger volume than would be the case without interdependence between the United States' NTB policy and European exports to the United States. In other words, the more restrictive U.S. NTB must reduce the overall supply in the combined NADC market to raise the domestic price in the United States.

210

Carl B. Hamilton

Case 1 is illustrated in figure 8.1, which depicts the supply and demand curves of the two countries, "Europe" and "U.S.A.," and the European export supply curve, Ses. The world market price is px, and is assumed to be constant (this is discussed later). The domestic European price is above the world market price at /?5 because imports from the non-NADCs—"the rest of the world"—are restricted to a0au so that domestic supply in Europe is 0a0. U.S. producers sell at home at the price of /?§, which in the figure is assumed to be above the European domestic price even if European tariff and transport costs are zero. At price /?§, the domestic U.S. supply is 0b0 and imports are b0bi. Assuming that the higher U.S. domestic price is not high enough to cover the tariff and transport costs that Europeans face when selling in the U.S. market, all imports originate in the rest of the world. When the U.S. government decides to reduce imports from the rest of the world, the U.S. domestic price increases frompg to ps2 and U.S. production increases from 0b0 to 0b3. This has no effect on trade with Europe. However, suppose the U.S. government reduces imports from the rest of the world even further, to b2b5. Trade deflection is prevented in the real world by rules of origin. Consequently, such rules are assumed here to prevent European importers from buying from the rest of the world and reselling such goods in the United States. In this case, the U.S. domestic price reaches the level at which European producers find it profitable to sell in the United States, in other words, at the price /?g(l + tes)(l + zes). When this price is reached, the U.S. market supply curve kinks at €. Beyond € the combined U.S.-European supply curve is Ss+es. The U.S. price increases topj, U.S. domestic production increases to 0b4, and European production sold in the U.S. is b4b2. European production and price increase to 0a4 and /?f, respectively, and European consumption is reduced to 0a3. The European exports to the United States are the result of both an increase in European production and a decrease in European consumption. Case 2: Only one of the two NADC partners maintains an NTB against the rest of the world on the commodity in question. Suppose Europe protects itself only with an ad valorem tariff and that the European tariff is lower than the combined tariff equivalent of the U.S. tariff and the U.S. NTB. (Otherwise the trade flow would be reversed.) The European domestic price is /?§ = px(\ + txe), and again making the simplifying assumption that tes = txs, (12) can be rewritten as, (13)

(1 + txe){\ + zes) = (1 + if").

Equation (13) states that, when trade takes place, there is an upper limit on the restrictiveness of the U.S. NTB that is determined by the European tariff rate against the rest of the world and by cross-Atlantic

Fig. 8.1

pS(i+t e b )0 + z e b )

Case 1: Two policy intervals

EUROPE

.s+es

b5

b 0 b3 b 4

b2

i

M !

b,

!

fp—i—PN

m

USA

Q

DS

212

Carl B. Hamilton

transport costs. In the extreme case, European producers will export their entire output to the United States while the European market will be served exclusively by imports from the rest of the world. If the United States (or in the reverse case, Europe) wants to prevent its NADC partner from increasing exports as it tightens NTBs against the rest of the world, it must also introduce an NTB on imports on crossAtlantic trade. Figure 8.2 illustrates case 2. At the world market price of px and a tariff equal to pg px, Europe imports a^ax from the rest of the world. Suppose again that the U.S. government decides to reduce imports from the rest of the world. Considering a U.S.-to-Europe transmission of protection, one can distinguish among three policy intervals. Interval I: No policy transmission. As long as the U.S. domestic price increase resulting from the tighter NTB is kept below the European supply price, /?g(l 4- tes){\ + zes), the increase in U.S. domestic supply and the effect on U.S. consumers and producers is determined by the U.S. supply curve only (along kt). Exporting from Europe does not take place, that is, qes = 0, and ps < pg(l + tes){\ + zes). Interval II: Complete policy transmission. Suppose U.S. imports from the rest of the world are reduced further. The U.S. domestic price reaches the supply price of imports from Europe, so that the supply curve kinks at € and is horizontal up to m. Imports to Europe from the rest of the world increase by the same amount as European exports to the United States, an amount shown by the segment €m on Ss+es. U.S. production remains unchanged at 0b2. Within Europe, domestic producers now supply less than 0a0, the difference between 0a0 and the Europeans' supply of European home demand being exported to the U.S.A. In this situation, qes < gg, where gg is the initial level of European production (0a0 in the figure), andps = /?g(l + tes)(\ + zes). Interval III: Modified policy transmission. If U.S. imports from the rest of the world are reduced even further, the point m is reached where Ss+es ki n k s upward and intersects with the U.S. domestic demand curve at n. The U.S. price increases to p\\ U.S. consumption falls; U.S. production increases to 0b3; and imports from Europe increase beyond the initial European production volume of 0a0 to 0a2, since the price of European output sold in the U.S. increases top\. In Europe a wedge is driven between the export price received by producers and the import price /?g paid by consumers. Since there is no European production for home demand, the European government could, in principle, dismantle its protection without affecting the price received by European producers, i.e., allow its consumers to enjoy the world market price px while its producers sell at the higher U.S. domestic price. If Europe imports its entire domestic demand from the rest of the world, would this drive up the world market price? No, because the

Fig. 8.2

Pod+t es )(1+z es )|

Case 2: Three policy intervals

EUROPE

p|(l+tes)(l+zes)f

b0

b2 b3

USA

214

Carl B. Hamilton

suppliers in the rest of the world can redirect the supply originally intended for the United States to Europe, and the combined U.S. and European demand from the rest of the world will remain unchanged (along €ra) or fall (beyond m). If anything, the world market price will fall. In this policy situation, qeQ < qes and /?g(l + tes) (1 + zes) < p5. It is sometimes said that developed-country tariffs are no longer important. However, as cases 1 and 2 indicate, this view overlooks the (potential) interplay between one importing country's NTBs and another importing country's tariff levels. More specifically, it is not necessarily one country's NTB combined with its tariff that determines its combined trade barrier on a product. Suppose both NADC partners initially maintain NTBs, but then one (say, Europe) liberalizes imports. Case 2 represents this situation. If Europe lowers its nontariff trade barriers, this stimulates European producers to look towards the United States, since that market becomes relatively more lucrative. Consequently, the U.S. domestic price is reduced. The EC-EFTA countries' liberalization of textile and clothing imports from Spain and Portugal through the enlargement of the EC illustrates how reduced protection may also be transmitted to another country, namely, the United States. Portugal has a well-established, competitive textile and clothing industry, but presently faces quantitative restrictions (VERs) on its exports to the EC and EFTA countries.23 Wage rates in Portuguese textile and clothing production are shown in table 8.1. They are not only below those of the United States and northern European countries, but are close to the levels of Hong Kong, South Korea, and Taiwan. These wage levels should give Portugal a competitive edge in Europe when the EC-EFTA VERs are removed in 1990. If, as would be expected, the European price on textiles and clothing then declined, other European suppliers would export to the United States. The case 2 mechanism would be put in motion, reducing the U.S. domestic price and production. No U.S. rules of origin that discriminate against textile and clothing exports from Portugal could prevent such an indirect trade deflection. 8.3.2 Empirical Evidence Is it possible to observe such transmission effects on intra-NADC trade because of NTBs on "sensitive" commodities? Can NTBs be said to have caused larger cross-Atlantic trade flows than otherwise would have occurred? Ideally one would like to compare two sets of commodities: those subject to NADCs' NTBs and those not subject to such NTBs, with both sets displaying the same shift in comparative advantage away from NADC producers to producers in the rest of the world. In practice,

215

The "New" Protectionism

the problem with such a comparison is that almost all developingcountry manufacturing exports in which important shifts in comparative advantage have taken place are subject to NTBs. Thus, the set of commodities displaying the same shifts in location of production as the commodities where the bulk of NTBs are found is too limited to be useful as a "control" group. What I will use instead to indicate the effect of NTBs on trade flows is a comparison of changes in suppliers' shares of home demand (apparent consumption), specifically: (a) domestic, (b) partner NADC, and (c) rest-of-the-world suppliers. Home demand of commodity j is defined as domestic production, Pj9 plus imports, Mj, minus exports, Ej. The supply of home demand is then divided into the following three sources of origin (dropping the subscript j ) : a. Domestic share of home demand (D): D = (P - E)/C b. NADC countries' share of home demand (N): N = MNADCIC c. Rest-of-the-world's share of home demand (X ): X = MXIC In any one year, these shares sum to unity, with the change in the combined shares between any two years summing to zero. In the traditional literature on customs unions, AN > 0 implies gross trade creation. However, the measure of gross trade creation does not tell whether AN > 0 has been at the expense of domestic supply (a shift from highto low-cost suppliers) or rest-of-the-world supply (a shift from low- to higher-cost partner suppliers). To ascertain this, net trade creation is defined as AD < 0. Furthermore, A X < 0 is defined as trade diversion. If it is assumed that the NADCs have lost comparative advantage in a commodity, one would expect to find that, in the absence of NTBs against the rest of the world, there would be net trade creation, in other words, AD < 0, and an increase in the share of imports in home demand would come mainly from the rest-of-the-world suppliers. However, since it is known that NTBs in some instances have covered only part of a commodity category—for example, only certain types of footwear—and have not always covered all exporters, one should allow for the possibility of NTBs' being "leaky" rather than "watertight." Thus, the following hypotheses are formulated: In the absence of NTBs, and given the lower costs of the rest-of-the-world suppliers, one would expect AD < 0, AN < 0, and A X > 0. With NTBs one would expect AN > 0, and, if the NTBs were "leaky," one would expect that A X > AN, but still with AN > 0. Summarizing the hypotheses: In the absence of. NTBs: AD < 0, AN < 0, and 0 < A X With "watertight" NTBs: AD < 0 or 0 < AD, 0 < AN, and A X < 0 With "leaky" NTBs: AD < 0, 0 < AN, and A i V < A I

216

Carl B . Hamilton

Five NADCs—the United States, West Germany, France, the United Kingdom, and Sweden—are studied, first separately and then as a combined group. Their shares in 1975 and 1983 are compared, and then also are compared to the U.S. shares for 1981 and 1983. The latter comparison is of interest because the U.S. dollar appreciated significantly between those two years (40 percent against the European Currency Unit, ECU). Commodities are classified on the basis of the three-digit ISIC code. For the United States, four NTB-restricted commodity groups are studied: textiles (ISIC 321), clothing (322), footwear (324), and basic iron and steel (371). This last group differs from the others in that there are NTBs between Europe and the United States (as well as subsidies, minimum-price rules, production quotas, etc.). However, the traderestricting actions taken by the NADCs against each other in steel were designed to be in line with historical trade patterns. That is, the NTBs in steel tend to take the form of discrimination against the newly industrialized countries. Table 8.4 indicates the share of U.S. domestic demand supplied by domestic producers, by exporters in other NADCs, and by the rest of Table 8.4

Shares of Apparent Consumption: the United States (percent)8

Commodity

1975

1981

1983

Change 1975-83

Change 1981-83

Textiles: Domestic NADCs ROW

96.3 1.2 2.5

94.5 1.6 4.0

94.7 1.5 3.8

-1.6 + 0.3 + 1.3

+ 0.2 -0.1 -0.2

Clothing: Domestic NADCs ROW

90.2 1.2 8.6

81.6 1.2 17.2

79.7 1.5 18.8

-10.5 + 0.3 + 10.2

-1.9 + 0.3 + 1.6

Footwear: Domestic NADCs ROW

76.2 9.4 14.4

67.8 8.0 24.2

56.7 9.6 33.8

-19.5 + 0.2 + 19.3

-11.1 + 1.5 + 9.6

Steel: Domestic NADCs ROW

90.4 3.9 5.6

86.1 6.3 7.5

87.1 5.4 7.5

-3.3 + 1.5 + 1.8

+ 1.0 -0.9 -0.1

Source: Constructed from "The OECD Compatible Trade and Production Data Base 1970-1980," OECD Department of Economics and Statistics, Working Paper no. 31, March 1986, Paris. a NADC is North Atlantic developed countries and ROW is rest of the world. Figures have been rounded off.

217

The "New" Protectionism

the world between 1975 and 1983. For textiles, the changes in shares of home demand are small, and the hypothesis of "leaky" NTBs appears to hold for the 1975-83 period. With clothing, there was much more net trade creation. With the rest of the world taking almost all of the reduced domestic share of home demand, the hypothesis again holds that NTBs were "leaky" yet produced some trade deflection. The net trade creation in footwear was almost twice as large as in clothing—particularly so after 1981, when the U.S. restrictions on nonrubber footwear imports from Taiwan and South Korea were lifted. The "leaky"-NTBs hypothesis holds both for 1975-83 and for 198183. Steel imports behaved similarly to textiles. Thus, over the period 1975-83, there was net trade creation. NADC partners' shares of U.S. home demand increased, but less than the increases from the rest of the world. U.S. NTBs were "leaky," and permitted import penetration by the rest of the world, especially in clothing and footwear. The increases in NADC partners' shares of U.S. home demand likely slowed down the observed increases in the rest of the world's import penetration into the United States. Table 8.5 shows developments in France, West Germany, the United Kingdom (all of them members of the Community), and Sweden (a member of EFTA) in clothing and footwear trade. For these four European countries, the NADC group includes partners in EC and EFTA that have tariff-free access. There is net trade creation in all four countries in both commodities. Apart from West Germany, the NADCs' shares of home demand in clothing increased more than the shares of suppliers in the rest of the world. Consequently, the "leaky"-NTBs hypothesis does not hold for France, Sweden, and the United Kingdom. Sweden stands out as a country in which the partner share increased much more than the share of the rest of the world. With regard to footwear, all European countries experienced much faster growth in import penetration of NADC partners than in import share of the rest of the world. Overall, the net trade creation in Europe benefited NADC suppliers much more than in the United States, but domestic shares of home demand in clothing were lower in Europe than in the United States. Finally, in table 8.6 all NADCs are treated as a group (the United States, Canada, Belgium-Luxembourg, Finland, France, West Germany, Italy, Netherlands, Norway, Sweden, and the United Kingdom). In addition to the four commodity groups mentioned above, a group of "all other manufactures" is included.24 Over the period, there was net trade creation in the "all other manufactures" category of minus 3 percent; the rest of the world (which includes Japan25) accounted for just over half of that net trade creation.

218

Carl B. Hamilton

Table 8.5

France: Domestic NADCs ROW

Shares of Apparent Consumption: France, West Germany, the United Kingdom, and Sweden (percent)a 1975

1983b

Change 1975-83 c

83.4 11.3 5.2

66.9 19.9 13.3

-16.5 + 8.5 + 8.0

83.2 11.8 5.0

66.5 23.0 10.5

-16.7 + 11.2 + 5.5

55.1 24.8 20.1

26.8 35.1 38.1

-28.3 + 10.3 + 18.0

56.7 34.6 8.7

38.9 45.2 15.9

-17.8 + 10.6 + 7.2

72.5 11.5 16.0

60.0 18.2 21.8

-12.5 + 6.7 + 5.8

80.5 10.4 9.1

64.0 22.1 13.9

-16.5 + 11.7 + 4.8

38.0 42.4 19.7

14.0 55.9 30.0

-24.0 + 13.6 + 10.4

36.1 52.3 11.6

20.9 63.1 16.1

-15.3 + 10.8 + 4.4

clothing

footwear Domestic NADCs ROW Germany: Domestic NADCs ROW

clothing

footwear Domestic NADCs ROW U.K.: Domestic NADCs ROW

clothing

footwear Domestic NADCs ROW Sweden: Domestic NADCs ROW

clothing

footwear Domestic NADCs ROW

Source: Same as table 8.4. NADCs is North Atlantic developed countries and ROW is rest of the world. Figures have been rounded off. b For Sweden 1981. c For Sweden 1975 to 1981. a

Textiles display about the same changes as "all other manufactures." Clothing and footwear, in contrast, had more net trade creation and the rest-of-the-world suppliers increased their shares of clothing and footwear relatively more than of textiles. In steel, there was little net trade creation and the increased import penetration was due entirely

219

The "New" Protectionism

Table 8.6

Commodity Textiles

Clothing

Footwear

Steel

Shares of Apparent Consumption: North Atlantic Developed Countries (NADCs) Combined (percent)8 Source

1975

1983

Change 1975-83

Domestic NADCs ROW Domestic NADCs ROW Domestic NADCs ROW Domestic NADCs ROW

83.9 14.8 1.3 78.6 12.0 9.4 73.1 18.5 8.4 85.0 13.8 1.2

80.6 16.0 3.4 69.3 13.0 17.7 56.1 23.9 20.0 84.5 14.4 1.1

-3.3 + 1.2 + 2.1 -9.3 + 1.0 + 8.3 -17.0 + 5.4 + 11.6 -0.5 + 0.6 -0.1

Domestic NADCs ROW

82.9 13.5 3.6

79.9 14.9 5.2

-3.0 + 1.4 + 1.6

All other manufactures

Source: Same as table 8.4 ROW is rest of the world. Figures have been rounded off.

a

to NADC suppliers. There was even a slight trade-diversion effect against the rest of the world. In summary, net trade creation seems to have been much stronger in clothing and in footwear than in textiles, "all other manufactures," and steel. In clothing and footwear, the NADCs' partner penetration of home demand was larger than with "all other manufactures," but smaller than the penetration of the rest-of-the-world suppliers. One interpretation of this is that, while NTB protection against the rest of the world has not been "watertight," it has still generated increased intra-NADC trade in clothing, footwear, and steel.26 Considering just textiles and clothing under the MFA, there are indications that there has been significant trade deflection since 1983. No doubt the United States became more restrictive in its bilateral agreements with developing MFA suppliers after 1982. The United States also changed its rules of origin for textiles and clothing so as to restrict imports. (Hong Kong and China were primarily affected by this.) In combination with the strong demand for imports in the United States, this policy is likely to be the major factor behind the widening gap between the growth of imports from restrained developing-country producers and from unrestrained West European suppliers.

220

Carl B . Hamilton

From 1982 to 1985, imports of textiles and clothing processed from MFA fibers, originating in developing MFA suppliers, increased very slowly compared to imports from Western Europe. In 1985, such imports increased by a mere 4 percent, while imports from Western Europe rose by 24 percent (see table 8.7). During the same period, the share of developing-country MFA suppliers in the United States' imports of textiles and clothing processed from MFA fibers declined by 4 percent, whereas the share of imports from Western Europe increased by 6 percent. But this is only part of the story: The United States' policy also had the effect of boosting the import supply of clothing made of fibers not covered by the MFA (linen, ramie, silk, and jute). Thus, non-MFA fiber clothing products increased from 80 million square yard equivalents (SYE, a volume measure used by the United States government) in 1983 to over 500 million SYE in 1985, or by almost 600 percent. They now account for 10 percent of MFA clothing imports. These developments illustrate not only the trade-pattern effect of NTBs but also the effect on commodity composition. Because of NTBs, Western European producers have an incentive to supply the United States with certain types of restricted textiles and clothing, even though they may not have a comparative advantage in these commodity groups. (The same applies to their supply of the domestic MFA-protected European market, of course.) Table 8.7

United States Imports of Textiles and Clothing Processed from MFA Fibres from All Sources, from the Developing MFA Suppliers, and from Western Europe 1982

Total imports Developing MFA suppliers Western Europe

Total imports Developing MFA suppliers Western Europe

Total imports Developing MFA suppliers Western Europe

6.11 3.73 0.55

1983

1984

Billions of Square Yard Equivalent 7.73 10.15 4.55 5.98 0.76 1.34

1985 10.83 6.22 1.66

Percentage Increase over Previous Years 27 32 7 22 32 4 38 76 24

100 61.0 9.0

Percentage Share in Total Imports 100 100 100 58.9 58.9 57.4 9.8 13.2 15.3

Sources: U.S. Department of Commerce, International Trade Administration TQ 2010, TQ 2210, TQ 2310 and Major Shippers Report. Table taken from International Textiles and Clothing Bureau, "Textile and Clothing: Developments in Industry, Technology, and Trade during MFA III", Geneva, 1986.

221

The "New" Protectionism

In the previous section, neither case 1 nor case 2 applies fully to the North Atlantic developed countries' trade during the periods studied. On both sides of the Atlantic, the systems of VERs have "leaked" and have allowed increased import penetration, partly because VERs cannot cover new suppliers and new products, and partly because VERs were designed to permit some increases in imports. 8.4 Summary and Conclusions The "new" forms of protectionism of the North Atlantic developed countries are frequently pointed to as a serious and growing problem for the world economy. However, surprisingly little is known about the extent and impact of that new protectionism on trade flows, prices, and welfare, especially in the exporting countries. The first part of this paper estimated the restrictiveness of, and rent income generated by, voluntary export restraints (VERs) on clothing exported in the early 1980s to the NADCs from the Big Three. Most VERs today are on clothing and textiles, and these two commodity groups account for almost 10 percent of world trade in manufactures and around 25 percent of developing-country manufactured exports to the NADCs. Exports to the NADCs are dominated by the Big Three, which make up almost two-thirds of the U.S. import market and onethird of the European import market. Restrictiveness was measured directly, and also indirectly by using estimates of the restrictiveness of one exporting country's VERs to derive the restrictiveness of another exporting country's VERs. The main findings were: (1) the combined tariffs and VERs of the United States in the early 1980s were higher than Europe's; (2) the importtariff equivalent (MTE) of the VERs varied with the business cycle in the importing NADCs; (3) the lower-bound MTE estimates for Taiwan and South Korea were lower than the rates of Hong Kong; and (4) the rent income transferred to the Big Three was substantial. It averaged at least half a billion U.S. dollars over the two-year period 1982-83, with around 80 percent of this total being transferred from the United States, and approximately two-thirds of the total rent going to Hong Kong. The second part of the paper investigated the implications of the fact that the NADCs do not apply nontariff barriers to imports from each other (with the notable exception of steel), whereas they do apply such barriers against the rest of the world. This has an overlooked implication for the transmission of protectionist policies among the NADCs through indirect trade deflection. This mechanism operates to modify, or even nullify, the effects on the domestic economy of increased protectionism against the rest of the world. It cannot be counteracted

222

Carl B. Hamilton

through rules of origin. Thus, the enlargement of the EC to include Portugal and Spain could have an important impact on the U.S. government's ability to protect U.S. producers of clothing. The final section of the paper presented empirical evidence on the possible effect of nontariff barriers to trade on cross-Atlantic trade flows. Although rest-of-the-world suppliers of "sensitive" commodities, like footwear and clothing, increased their shares of the NADCs' home demand between 1975 and 1983, the shares of partners on the other side of the Atlantic also increased, pointing to the conclusion that NTBs have been "leaky."

Notes 1. A recent paper on the types and frequency of NTBs is by Nogues, Olechowski, and Winters (1986). 2. Of course, the question of evaluation ought to be a problem for the policymakers who constructed and introduced the measures, provided that the politicians' objective has been to restrict imports. However, this objective is not self-evident. See Yoffie (1983). 3. The agreements violate the Community's competition laws in addition to the article of the Rome Treaty disallowing new national trade restrictions. 4. Langhammer (1982) and Hamilton (1986a) and (1986b). 5. Choi, Chung, and Marian (1985) and Yoffie (1983). 6. The developed countries' protection regarding textiles and clothing is formally legitimized within GATT through the Multifiber Arrangement (MFA). See, e.g., Wolf (1983) and Choi, Chung, and Marian (1985) for descriptions of the MFA. 7. GATT (1985). 8. GATT (1984). 9. Publication of Quota Weekly ceased in mid-1984 after 18 months. 10. One feature of the quota market in Taiwan is that there is trade both in rights to export that are valid for the present year only ("temporary quotas") and trade in rights to export that are valid for all future periods ("permanent quotas"). Unfortunately, the registered prices for "permanent quotas" either were almost constant over time—the data from the TTF, or contained too few observations to be useful—the data from Quota Weekly. For a theoretical paper on intertemporal aspects of trade in quotas, see Anderson (1987). 11. It should be noted that, owing to the "upgrading effect" of VERs, there is a tendency within each commodity category for an exporting country to concentrate on higher-quality and more costly grades (see Falvey 1979 and Feenstra 1984). This upgrading mechanism works toward a narrower range of quality differences within each commodity category than would have been the case, or could be expected to be the case for unrestrained exports. 12. Readers interested in the variations over time in import-tariff equivalent rates are referred to Hamilton (1986a). Export trade statistics are not available on the MFA classifications. In table 8.2, the United States' export-tax equivalents are converted into import-tariff equivalents by exploiting the fact that ordinary U.S. import statistics—but not the MFA trade statistics—are given

223

The " N e w " Protectionism

on both a c.i.f. and an f.o.b. basis. The conversion was made by taking the ratio of c.i.f. and f.o.b. values on clothing imported from Hong Kong, which turned out to be 1.07. On precise references to statistical sources, and on the use of disaggregated unit values, see also Hamilton (1986a). 13. I tried in vain to collect quota prices for South Korea, also through Korean colleagues and research institutes. An indirect method seems to be the only one possible since trade in quotas is strictly forbidden, and this law is enforced. See also Breitenacher, Galli, and Grefermann (1986, 229). Also, apart from different intensities of policing, controls in South Korea are more effective than they probably can ever be in Taiwan or in Hong Kong, as the bulk of Korea's clothing trade is concentrated in a few large corporations, compared to a very large number of small firms in Taiwan and in Hong Kong. 14. Werner International Management Consultants (1986). 15. Breitenacher, Galli, and Grefermann (1986, 87). 16. Cable and Baker (1983, 54). 17. United States International Trade Commission (1985, 29-30). 18. Hong (1981, table 8.19). Also, "[the] expansion of exports of textiles, wearing apparel, and miscellaneous manufactures can be attributed more to basic comparative advantage of Korea (i.e., low wages) than to subsidized interest rates" (Hong 1981, 382). It is clear from Hong's analysis that, if anything, the labor-intensive parts of South Korean industry were discriminated against when it came to the allocation of capital. 19. The difference between U.S. f.o.b. and c.i.f. clothing prices from Hong Kong is around 7 percent. 20. Hufbauer and Schott (1985) and Hufbauer, Berliner, and Kimberly (1986). 21. An exception, in addition to steel, is the less important Canadian global import quota on footwear. 22. Because of tariff preferences like the General System of Preferences (GSP), the tariff on goods originating in Europe is not necessarily the same as the tariff on goods originating in LDCs. 23. Portugal has been a member of EFTA since its formation on 4 January 1960. The VERs on Portugal imposed by other EFTA members was against Article 11 of the EFTA charter, which forbids all quantitative trade restrictions after 31 December 1961. 24. ISIC groups 311, 313, 314, 323, 331, 332, 372, 381, 382, 383, 384, 385, and 390 combined. 25. The United States restricts imports of textiles and clothing from Japan. 26. It would be desirable in future research to identify precisely the restricted categories. As of today, there are no production data on a tariff line-commodity classification.

References Anderson, J. E. 1987. Quotas as options: Optimality and quota license pricing under uncertainty. Journal of International Economics 23, no. 1/2 (August). Breitenacher, M., A. Galli, and K. Grefermann. 1986. Perspektiven des Welttextilhandels. IFO Studien zur Industriewirtschaft, no. 30. Munich. Cable, V., and B. Baker. 1983. World textile trade and production trends. The Economist Intelligence Unit, Special Report no. 152. London.

Choi, Y-R, H. S. Chung, and N. Marian. 1985. The multi-fibre arrangement in theory and practice. London and Dover, N.H.: Francis Pinter. Falvey, R. 1979. The composition of trade within import-restricted import categories. Journal of Political Economy 87: 1105-14. Feenstra, R. 1984. Voluntary export restraint in U.S. autos, 1980-1981: Quality, employment, and welfare effects. In The structure and evolution of recent U.S. trade policy, ed. R. Baldwin and Anne O. Krueger. Chicago and London: University of Chicago Press. GATT. 1984. Textiles and clothing in the world economy. Geneva: GATT. 1985. International trade 1984-1985. Geneva: GATT. Hamilton, C. B. 1986a. An assessment of voluntary restraints on Hong Kong exports to Europe and the U.S.A. Economica 53 (August). 1986b. The rise and fall of footwear protectionism. Institute for International Economic Studies, Stockholm. Mimeo. Hong, W. 1981. Export promotion and employment growth in South Korea. In Trade and Employment in Developing Countries, Part 1, Individual Studies, ed. A. O. Krueger, H. B. Lary, T. Monson, and A. Narongchai. Chicago and London: The University of Chicago Press. Hufbauer, G. C. and J. J. Schott. 1985. Trading for growth; The next round of trade negotiations. Institute for International Economics, Washington, D.C. Hufbauer, G. C , D. T. Berliner, and A. E. Kimberly. 1986. Trade protection in the United States, 31 case studies. Institute for International Economics, Washington, D.C. Langhammar, R. 1982. ASEAN manufactured exports in the EEC markets: An empirical assessment of common and national tariffs and nontariff barriers confronting them. In ASEAN-EC economic relations, trade, and investment, ed. A. Narongchai and H. C. Rieger. Institute of South-East Asian Studies, Singapore. Nogues, J., A. Olechowski, and A. Winters. 1986. The extent of nontariff barriers to industrial countries' imports. World Bank Economic Review, 1, no. 1. United States International Trade Commission (USITC). 1985. Emerging textile exporting countries. Washington, D . C : USITC. Werner International Management Consultants. 1986. Comments on the hourly labour costs in the primary textile industry, winter 1985/86. Brussels and New York. Mimeo. Wolf, M. 1983. Managed trade in practice. In Trade Policy in the 1980s, ed. W R. Cline. London: MIT/Cambridge Univ. Press. Yoffie, D. 1983. Power and protectionism, New York: Columbia Univ. Press.

Comment

Juergen B. Donges

Hamilton has developed an interesting method for estimating tariff equivalents of VERs. Direct international price comparisons may have been preferred, but they are difficult to make because of lack of data Juergen B. Donges is vice-president of the Kiel Institute of World Economics and Hon. professor of political economy at the University of Kiel.

225

The " N e w " Protectionism

in most cases; moreover, there is always the tricky question of how to purge consumer prices, if they are available, from transport costs, handling margins, taxes, and the like. Therefore, any attempt to improve on reduced-information methods in this field is useful and stimulating. The order of magnitude of restrictiveness that Hamilton finds for VERs on clothing is quite plausible, though for the EC his implicit tariffs are somewhat lower than those reported elsewhere (including several studies which have been done at the Kiel Institute of World Economics). The estimates on the recent income accruing to the exporting countries by means of VERs also look sensible. For the sake of comparison, these rents should be expressed as a percentage of GDP or export value. I submit that they are important in relative terms, which is one explanation for the fact that exporting countries do not forcefully oppose VERs. I would like to add that in the case of the EC, various member states have requested time and again (temporary) exclusion from the common external tariff (invoking Article 115 of the Treaty of Rome). Textiles and clothing account for the lion's share of all exemptions granted. This could mean that the tariff-cum-NTB degree of restrictiveness in individual countries might be higher than measured in this paper. Another fascinating topic is the international transmission of protectionist policies. Hamilton puts his finger on a key attribute of the present protectionist system, namely that individual countries do not protect, but rather groups of countries protect. Leaving aside only agriculture, EC protection exists in a number of tiers: one against the developing countries; one against Japan, the United States, and a few other industrial countries; one against the centrally planned economies; and one tier encompassing a free trade area with EFTA. Thus Hamilton correctly perceives the results of raising nontariff barriers against some countries only as indirect trade deflection. This could tend to mitigate the effects of new trade barriers after a change in comparative advantage within such a quasi-customs union. For instance, the costs of protection associated with the U.S.-Japanese VER in automobiles should be lower than suggested by standard theory for just this reason: U.S. small car imports can be diverted from Japan to Europe. Be that as it may, further research on the transmission of protectionism among trading partners is extremely welcome. In particular, I wonder how sensitive the findings are to assumptions on supply elasticities in the new lost-cost producers. It should also be interesting to disentangle trade flows across the North Atlantic trade at least with regard to clothing and textiles in order to capture more fully the trade effects of the MFA within the EC.

226

Carl B. Hamilton

It is true that indirect trade deflection illustrates that VERs (as other NTBs) are not watertight. This may reduce their restrictiveness in particular cases. And yet, they create a sort of policy-induced uncertainty. This worries me because it may easily become a source of export pessimism in developing countries and it tempts them to embark upon, or continue with, an excessive import substitution (in spite of the substantial distortions and the notably growth-retarding effects which this strategy usually generates). As to Tarr's paper,* it persuasively shows what other studies also have portrayed, namely that government intervention in favor of a particular sector does not resolve perennial structural adjustment problems of declining industries. It is just an illusion to believe that protection by itself transforms underlying comparative cost disadvantages into advantages. On the contrary, the so-called "breathing spaces" reduce the incentive to adjust in an efficient manner, especially if we deal with an industry such as the iron and steel industry, which the general public may want to see shielded from imports at any rate; survival is perceived as necessary also for noneconomic reasons (e.g., for national defense, as a symbol of strength). The futility of sectorspecific protectionist measures explains why so many NTBs that were at one time announced as temporary have endured over time, and not only in steel. I share Tarr's sceptical view on the EC's steel policy (initiated in 1977). Incidentally, three years before, the Kiel Institute had already published a comprehensive study that showed that steel production in traditional locations (such as the United States, EC, Germany) was bound to come under increasing adjustment pressures from Japan and several NICs (Wolter 1974). In addition to the reasons given in Tarr's paper I would like to add two observations: First, subsidies to inefficient steel mills are of little help even if these mills use them to modernize their production capacity and to diversify activities, because these options are also open to the more efficient steelmakers (which are in a much better financial position to carry out the investment). Second, as the decision on the allocation of production quotas across the Community has become a political issue (much influenced by pressure groups from structurally weak regions in which steel production is concentrated), the quotas are typically set up too high. This makes the process of cutting down capacity even more troublesome, particularly so if it affects capacity which was created through modernization (nobody wants to write off costly investment immediately after it is undertaken). Incidentally, there is an incentive for firms to invest today in excessive capacities in order to get a higher quota tomorrow. *Chapter 7 in this volume.

227

The " N e w " Protectionism

As to the EC's code on national subsidies of 1981, it is worth recalling that it blatantly violates Article 4 of the ECCS founding Treaty of Paris which explicitly forbids all kinds of national subsidies in member states and that it runs counter to Articles 2 and 3 of that treaty which call for efficiency in steel production and openness in trade among member states and vis-a-vis third countries as well. The language used in that code to define the conditions under which a steel mill qualifies for subsidies at first glance sounds sensible and progressive. But, in fact, a strange set of rules has been set up which allows governments to go ahead with distorting subsidization. The subsidy was to expire at the end of 1985. In December 1985 the Council of Ministers decided to extend the code until 1990. Any similarity of European steel policy with the history of the Common Agricultural Policy, or with that of the Multifibre Arrangement, is purely coincidental!

References Wolter, Frank, 1974. Strukturelle Anpassungsprobleme der westdeutschen Stahlindustrie—Zur Standortfrage der Stahlindustrie in hochindustrialisierten Landern. Tubingen: J. C. BL Mohr.

International Trade in Telecommunications Services Andre Sapir

9.1 Introduction In recent years, there has been growing American sentiment toward promoting U.S. service industries and enhancing their international competitiveness. At the 1982 ministerial meeting of the General Agreement on Tariffs and Trade (GATT), the United States trade representative (USTR) proposed for the first time that service transactions be added to the agenda for the next round of multilateral trade negotiations. However, the developing countries, led by Brazil and India, rejected this proposal. These countries have been reluctant to enter international negotiations on services for two reasons. First, U.S. negotiators unintentionally conveyed the message that the liberalization of trade in services would be a zero-sum game. Their eagerness to dismantle barriers to trade in services is perceived by many as simply serving the self-interest of large U.S. service corporations. Second, many GATT signatories are not enthusiastic about trade liberalization in this area. They fear that an international system of rules for trade in services will interfere with their national policy objectives. Although the European Community (EC) shared many of these apprehensions, it was instrumental in finding the compromise adopted at the 1982 meeting: contracting parties with an interest in services would undertake national studies of problems in services trade. Most industrialized nations submitted reports for the November 1984 GATT session, and a working group was established to improve information about services. Andre Sapir is professor of economics at the Free University of Brussels. I am grateful for detailed comments and suggestions from Robert Baldwin. I have also benefited from comments by several participants at the conference.

231

232

Andre Sapir

During the following two years, the EC again played a crucial role in reconciling the strong American stand on including services in the new round with the insistence of the developing countries that services remain outside the scope of GATT. Under the compromise reached at the September 1986 ministerial meeting which launched the Uruguay Round, the multilateral trade negotiations are to proceed along two parallel tracks: one for goods and one for services. The services negotiations are aimed at establishing an international set of rules that might eventually be incorporated into the GATT system. One sector of special concern to U.S. officials that illustrates the problems of negotiating in the services area is telecommunications services. U.S. interest in this field goes considerably beyond the perception that American firms have a competitive advantage in supplying telecommunications services. Telecommunications play a central role for almost all forms of services by providing an infrastructure for international trade in services. The United States has devoted particular attention in recent years to issues of telecommunications trade with the EC. The European telecommunications market—the world's second largest—tends to be much less open than the American market. This is a result of the difference in institutional arrangements on the two sides of the Atlantic. In European countries (as in most of the world), telecommunications services are provided largely by government monopolies known as PTTs (Posts, Telegraph, and Telephones). In contrast, telecommunications services in the United States are supplied by private firms which operate in an increasingly competitive environment since deregulation was launched over a decade ago. U.S. interest in telecommunications services also stems from the changes in this sector caused by an ongoing technological revolution. New information technologies are much faster, and it now costs less to process, store, retrieve, manipulate, and transmit data. Information technology now also encompasses a wide array of convergent and linked information-goods and information-services activities. The information-goods industry includes computers, data recognition equipment, telecommunications equipment, and other related hardware, while the information-services industry consists of computer services, information storage and retrieval, and telecommunications services. Information technology has led to the merger of data-processing and telecommunications activities into telematics, which involves both information-goods and information-services industries. As a result, the traditional dividing line between goods and services has become blurred. The extension of telematics internationally has given rise to transborder data flows that are similar to trade in information services and can be defined as the electronic international movement of computer-readable

233

Telecommunications Services

data across telecommunications networks. This paper analyzes the trade policy issues of transborder data flows, with particular emphasis on US-EC trade relations. Although there is a general trend toward information intensity throughout the economy, it varies by industry. Thus, the impact of transborder data flows on international transactions will vary across industries. In particular, the possibilities for international trade in services have been greatly enhanced by transborder data flows because of their "high information-technology content in both product and process" (Porter and Millar 1985, 154). However, the exact nature of the impact of transborder data flows will depend on the nature of services. I will argue in this paper that information flows enhance mrrafirm trade in some service industries and interfirm trade in others. Both intrafirm and arm's-length information flows depend upon the efficient operation of telecommunications networks. How open should telecommunications services be to competitive forces? The United States and the European PTTs have two opposite views on this matter, with Europe favoring as wide a monopoly as is possible. This paper examines the implications of alternative regulatory environments on telecommunications trade. The plan of the paper is as follows: section 9.2 analyzes the impact of information flows on the structure of international trade in services, while section 9.3 examines the policy issues involved in the organization of international telecommunications. The last section offers some perspectives on the potential for future trade negotiations in the telecommunications field. 9.2 The Impact of Information Flows on Service Trade Information technology, by increasing the speed and efficiency of transmitting information around the world, is rapidly changing the previous landscape of international service transactions. It has created new services with substantial scope for international trade flows, namely information services. In addition, information technology has greatly enhanced the tradability of traditional services. Information services have as a primary function the collection, processing, and/or transmission of information in an electronic form. They include data-processing and data-base services. These services are essentially long-distance, linking a computer facility to a remote user via a communications system. Therefore, they are highly tradable. In their case, transborder data flows tend to be arm's-length transactions between the provider of the service and an unrelated user. Complementary innovations in data processing and telecommunications are opening up the way toward a greater international division

234

Andre Sapir

of labor in services in the same way that the industrial revolution affected agricultural and manufactured goods. "[A] key factor in the growth of world trade [was] a sequence of transportation innovations that opened up continental hinterlands, reduced the cost of transoceanic shipping, and made possible the preservation of perishable food products during extensive voyages over land and sea. These innovations included the rapid growth of the railroads after 1830, the expanding role of the iron steamship after 1850, and the introduction of refrigeration on both freight cars and steamships beginning in the 1870s. . . . With these complementary innovations, there began to emerge, by the end of the nineteenth century, a truly worldwide agricultural division of labor" (Rosenberg 1982, 58 and 251). The increased tradability of services is conceptually equivalent to a reduction in transport costs and has two consequences for the pattern of trade. On the one hand, reduced transportation costs make it feasible for service industries to reallocate certain activities to least-cost locations and to export their products to other locations. On the other hand, reductions in transport costs make possible the greater exploitation of economies of scale. In a world of low transport costs, the size of domestic markets plays less of a role in shaping trade patterns, and production can be concentrated in fewer locations. Several trade economists have recently drawn on the work by Hill (1977), who emphasizes the nonstorable nature of services: production and consumption must generally occur in the same location and at the same time. This characteristic provides the basis for the fact that services are generally not traded in the papers by Bhagwati (1985), Deardorff (1984), and Sampson and Snape (1985). Bhagwati (1985) and Sampson and Snape (1985) divide services into two categories: those that require the physical proximity of the producer and the consumer and those that do not; the latter are referred to as "separated" services by Sampson and Snape. Within the group of services for which physical proximity is essential, a further distinction is drawn between those that necessitate the movement of the producer, the consumer, or both. The distinction carries major policy implications. For "separated" services, trade liberalization is similar to trade liberalization in goods. But for the majority of services, freedom of international transactions would require freedom of movement of either producers or consumers. It is precisely the latter issue which has clouded the prospect for negotiations on service transactions. In particular, many countries are reluctant to open the Pandora's box of rules governing foreign investment (involving so-called "rights of establishment" and "national treatment" regulations) and foreign labor. It has been argued that the advent of information technologies has largely eliminated these issues because it now "makes little difference

235

Telecommunications Services

where in the world the buyer and the seller or user and provider are located, as long as their computers are linked together through modern communications systems" (Feketekuty and Hauser 1985, 7). My own view is that this argument needs to be qualified. The enhanced tradability of services is partly offset by a quality-uncertainty problem that arises because of the intangible nature of services. Contrary to most goods where buyers can rely on physical attributes to judge the quality of a product, with services it is reputation that plays an overriding role in the selection of suppliers. To be sure, uncertainty about product quality is a feature of markets for many goods, but services are "[virtually all . . . impossible to evaluate until they are used" (Shapiro 1982, 20). There is a continuum of services from highly tangible ones (almost goods-like) to highly intangible ones. For instance, most insurance services are highly intangible, while routine transportation services tend to be very tangible. The more intangible the service, the more its market structure tends to be characterized by nonprice competition. The reputation of producing a quality product is also the major intangible asset of service firms. As discussed in the vast literature on multinational enterprises (for recent surveys, see Caves 1982; Dunning and Rugman 1985; and Teece 1985), the possession of intangible assets creates an incentive to sell in foreign markets in order to maximize the rent that can be gained. Although this is equally true for both goods and services, there is a fundamental difference between the two. In the case of most goods, intangible assets owned by producers are embodied in the physical attributes of the product. However, for service activities the product itself is intangible. Therefore, goods do not generally require consumers to be close to producers for ascertaining quality and can be exported to foreign markets. In contrast, services generally require such a close interaction between producers and consumers that they must often be produced in the markets where they are consumed. Several authors have noted that service firms often become multinational in order to follow their customers (see Caves 1982 and references cited therein). Service firms tend to acquire an intangible asset: a quasi-contractual relation with their customers "based on trust that lowers the cost of contracting and the risks of opportunistic behavior. If the service firm has such a quasi-contractual relation with a parent MNE (Multinational Enterprise), it enjoys a transactional advantage for supplying the same service to the MNE's foreign subsidiaries" (Caves 1982, 11). The more intangible a service, the more difficult it is to export, especially to an unrelated party. This proposition has two implications. First, the principle of comparative advantage might not apply to highly intangible services in the sense that the country with the lowest pro-

236

Andre Sapir

duction cost might not be able to export a service for which it lacks reputation. Second, the degree of tangibility of a service determines the channel for possible trade flows. The less intangible it is, the more likely it is that international trade will consist of arm's-length transactions between unrelated parties. Conversely, the more intangible a service, the more trade will tend to flow through intrafirm channels. In other words, the more intangible the service, the more transborder data flows can be expected to take place outside the international market place and within multinational firms. An important policy implication is that, contrary to what might have been expected, transborder data flows have not eliminated the issues of "rights of establishment" and "national treatment," but they have changed the nature of the service establishment that is required to conduct business in a foreign country. 9.3 Telecommunications Services 9.3.1. Telecommunications Services as an Infrastructure To what extent do telecommunications services serve as an intermediate input for other industries in the economy? In principle, the answer could be obtained from input-output tables that provide interindustry linkages. However, in most countries (particularly in Europe), telephone and postal services are jointly operated so their activities cannot be distinguished in input-output tables. The United States, where the businesses of telephone and postal services are totally unrelated, is an exception. Consequently, one can use the 1977 input-output table of the U.S. economy (disaggregated into 537 industries) to estimate the role of telecommunications services as an infrastructure input in the United States. Table 9.1 shows that the telecommunications sector sells 44 percent of its output to intermediate users. The bulk of this demand comes from service industries: "wholesale and retail trade," "finance and insurance," "business services," and "health, education, etc.," account for over 50 percent of the output sold to industrial users. Table 9.2 examines the "telecommunications intensity" of various industries—that is, the direct requirements of telecommunications services per dollar of industry output—of the various industries. The table reports only on industries with requirements of one percent or more. Service industries are obviously the largest relative users of telecommunications inputs. In the case of "business services" and "finance and insurance," inputs of telecommunications services are about 2 percent of output and over 5 percent of intermediate inputs.

237

Telecommunications Services

Table 9.1

Distribution of the Telecommunications Sector's Output, 1977 (millions of dollars) Total demand Total intermediate demand Wholesale and retail trade Finance and insurance: Banking Credit agencies Brokers Insurance carriers Insurance agents Business services: Computer services Consulting services Legal services Other business services Health, education, etc. Total final demand

52,868 23,404 5,584 2,523 933 356 311 531 392 2,619 272 287 374 1,686 1,868 29,464

Source: U.S. Department of Commerce, Bureau of Economic Analysis, Detailed InputOutput Structure of the U.S. Ecomony, 1977, (Washington, D.C.: Government Printing Office, 1984). Table 9.2

Telecommunications Service Inputs as Share of Total and Intermediate Inputs in Selected Sectors (in percent)

Sector Amusements Automobile services Printing and publishing Health, education, etc. Hotels Wholesale and retail trade Communications services Business services: Computer services Consulting services Legal services Other business services Finance and insurance: Banking Credit agencies Brokers Insurance carriers Insurance agents

Share of Total Inputs

Share of Intermediate Inputs

1.00 1.10 1.20 1.20 1.40 1.50 1.80 1.90 (1.83) (2.09) (1.82) (1.97) 2.00 (2.07) (3.81) (3.51) (1.12) (2.11)

2.07 2.33 2.26 3.30 3.90 5.26 9.00 7.20 (8.02) (6.19) (9.14) (7.19) 5.08 (6.84) (6.28) (9.76) (2.10) (12.56)

Source: U.S. Department of Commerce, Bureau of Economic Analysis, Detailed InputOutput Structure of the U.S. Economy, 1977, (Washington, D.C.: Government Printing Office, 1984).

238

Andre Sapir

It appears from these tables that telecommunications services, although required by nearly every industry, are used primarily by service industries. They are not only among the main customers of telecommunications services, but also exhibit the highest degree of telecommunications intensity. This conclusion applies to international as well as to domestic activities. International transactions in services primarily involve transborder data flows, either as a service activity per se (usually as interfirm flows), or as an infrastructure input for other services (generally as intrafirm flows). Therefore the expansion of international service transactions depends crucially on the existence of an efficient telecommunications network. 9.3.2. International Telecommunications Services; Institutional Framework Until recently, there has been a consensus among countries that telecommunications services are provided most efficiently within a country by a monopoly. The only difference in viewpoints was over what form of organization would ensure that the monopolist behaved in society's best interest: either a regulated private firm (as in the United States and a handful of other countries) or a government-owned organization (as the PTTs in Europe and most other countries). The key to the belief that a monopoly was needed was that each country shared the same technology, and that this technology was evolving at a relatively slow pace. The monopoly principle was also extended from domestic to international markets. In effect, the international market for telecommunications services became, and largely still is, a cartel of national monopolies that share the same technology and offer the same services. As one observer has noted, this market structure has prevented telecommunications services from being traded. "Far from encouraging countries to exchange different services, it instead has ensured that all countries produce the same services. The capability of producing these services has then been shared among all countries, not traded" (Reid 1985, 18). In recent years, this traditional concept of the international telecommunications network has been disturbed by a rapid succession of innovations in information technologies. The resulting challenge to the traditional view has come almost entirely from the United States, where the innovations have led to a series of changes in the regulatory environment. As a result, the structure of the U.S. telecommunications market has shifted during the past decade from a regulated monopoly toward one with intense interfirm competition. The more rapid changes in technology and regulatory policy in the United States have created

239

Telecommunications Services

several imbalances in the international network, with consequences for trade in telecommunications equipment and services. 9.3.3. Trade Policy Issues in Telecommunications Services The traditional network offered homogeneous public telecommunications services accessible to all those with the requisite hardware. However, the development of new information technologies and computerto-computer data communications created specific needs for which the public network was ill-suited. In response to a need for transborder data flows, countries agreed to provide private leased lines that offered certain services not available on public networks (for example more reliable and faster services—see Ergas 1984). As indicated by Antonelli (1984), private leased lines are particularly effective channels for intrafirm transborder data flows that rely on standardized or compatible hardware and software."Moreover, leased lines guarantee greater appropriability and thus greater security for regular ongoing flows of information among [MNE] headquarters and affiliates" (Antonelli 1984, 337). In contrast, for transborder data flows between unrelated parties (relying on heterogenous hardware and software), the use of public networks is more appropriate, provided that they can supply sufficiently high-quality data communications. Outside of the United States, private leased lines are provided under conditions that preserve the monopoly character of telecommunications. They must be denied at any time if they infringe on the services available on the public network. In particular, the use of private leased lines is limited to subscribers, who are then prohibited from reselling telecommunications services. The rule requiring that leased lines be used solely for the subscriber's internal communications has been challenged by the emergence of new services that combine computer processing and telecommunications. In order to preserve the spirit of the rule, a distinction was drawn between services that primarily involve computer processing—dataprocessing and data-base services—and those that engage mostly in data transmission, referred to as enhanced or value-added services, such as electronic mail, videotext, and protocol conversion. Under the interpretation adopted in most countries, the rule forbids the use of private leased lines for providing enhanced services. The situation is quite different in the United States, where present regulatory conditions distinguish between two types of communications services: basic services, which only transmit information through the telecommunications network, and enhanced services, which also modify that information. Basic services are provided by monopolies subject to regulations; enhanced services are provided by suppliers in com-

240

Andre Sapir

petition with each other who have unrestricted access to private leased lines and who can resell their unused transmission capacity. Therefore the previous consensus among countries has given way to a disagreement about the extent to which telecommunications services should be competitive. The United States and the European PTTs hold opposite views on this matter, the latter arguing for a monopoly that would supply both basic and enhanced services. This lack of international consensus is hampering the emergence of trade in (enhanced) telecommunications services. As Reid (1985) concludes, "the development of trade is a two-step process; first a variety of services must be allowed to emerge within individual countries, then these services must be permitted to be sold internationally" (p. 35). An illustration of the type of dispute that has arisen because of the lack of a consensus in telecommunications is the disagreement between the United States and Germany on Bundespost restrictions on the use of private leased lines for transborder dataflows.The Bundespost forces firms to link their private leased lines to public data networks, requires that data processing occur in Germany before such data is transmitted across international leased lines, and charges usage-sensitive rates, in addition to fixed charges, on some international leased lines. The German position is that the Bundespost must maintain its monopoly over all telecommunications services in order to provide them most efficiently (the natural monopoly argument) and protect its revenues (the cross-subsidization or social equity argument). On the other hand, the United States maintains that the German policy encourages inefficiency and violates the 1985 Declaration on Transborder Data Flows by which OECD (Organization for Economic Cooperation and Development) members pledged to promote international data flows. The natural monopoly issue has raised considerable debate in Europe concerning U.S. deregulatory and divestiture experience. Some (mostly PTT officials) view this experience with skepticism because of regulatory confusion and the deterioration of universal service. Others (many officials of the EC Commission and the British government) would like to follow the U.S. lead toward increased competition on the grounds of better and cheaper services and the stimulation of innovation. But, so far, there is no serious study of the production characteristics of the European telecommunications industry that could help reveal whether or not the natural monopoly argument is appropriate. However, there are studies of the rate structure in Europe. A recent analysis shows that substantial welfare losses result from the practice by the Bundespost of setting prices for local and long distance calls at 10 and 140 percent, respectively, above marginal costs. The huge profits arising from these telecommunications services are used to cover def-

241

Telecommunications Services

icits from postal activities and as revenue sources for the general government budget (see Neumann, Schweizer, and von Weizsacker 1985). Particular trade disputes between Europe and America on telecommunications services are part of a wider issue. The unanimous international consensus in favor of monopoly organizations for providing national telecommunications services has been shattered by rapid changes in information technology. The challenge facing the two trading blocs will be to develop a new consensus on the international provision of telecommunications services, taking account of the new technological environment. 9.4 Conclusions This paper has argued that telecommunications services are the cornerstone of a new information technology that can transform the scope and nature of international trade in services. Depending upon whether services are more or less intangible, their enhanced tradability is likely to involve intrafirm or arm's-length information flows. In either case, an efficient telecommunications network needs to be established, drawing lessons from the U.S. experience that allows more competition in the provision of both services and equipment. The alternative for Europe and other areas is to be left behind, not only in telecommunications services but also in many other service activities. One conclusion of this study is that domestic policy changes—breaking up domestic monopolies in services and equipment—rather than new international rules, hold the key to international competition and trade in telecommunications services. This has important implications for the multilateral trade negotiations on services launched at Punta del Este in September 1986. In almost all countries, services are more regulated by the government than most other activities. It is generally thought that the very nature of services necessitates regulating entry into the industry in order to achieve optimal economic efficiency or other national objectives. There are two reasons why free entry might result in wasteful allocations of resources. First, the intangible nature of many services (banking, engineering, insurance, etc.) requires the imposition of minimum standards in order to prevent welfare losses to consumers arising from low-quality services. Second, the production characteristics of some services (communications, transport, etc.) implies that excessive competition might be detrimental to welfare because of an inefficient scale of operation. Although government regulations of services are imposed for purely domestic reasons, they may also have repercussions on international

242

Andre Sapir

transactions. Rules aimed at protecting consumers make it very difficult, if not impossible, for foreign corporations to obtain "rights of establishment" or to be granted "national treatment." In principle, all matters involving discriminatory practices could be dealt with by extending existing GATT rules to services. But the main problem is the rights-of-establishment issue, which is not covered by these rules. Another instance is when national regulatory practices—although they might be nondiscriminatory—diverge sufficiently to prevent international transactions from occurring, as in telecommunications services. Here the issues are primarily domestic matters involving regulatory philosophies. This point has been recognized by the ministerial declaration of Punta del Este which indicates that rules for trade in services must respect the policy objectives of national laws and regulations. However, it remains to be seen how much progress can be achieved along such lines by the 92 GATT signatories. It might be, instead, that negotiations involving regulatory issues would be better conducted, for the moment, among countries that share common economic situations and principles. The OECD, therefore, might be a more appropriate forum for such negotiations.

References Antonelli, Cristiano. 1984. Multinational firms, international trade, and international telecommunications. Information Economics and Policy 1:333-43. Bhagwati, Jagdish. 1985. Trade in services and developing countries 10th Annual Geneva Lecture delivered 28 November at the London School of Economics. Caves, Richard E. 1982. Multinational enterprise and economic analysis. Cambridge: Cambridge University Press. Deardorff, Alan V. 1984. Comparative advantage and international trade and investment in services. Paper presented at a conference on Current Issues in Trade and Investment in Service Industries: U.S.-Canadian Bilateral and Multilateral Perspectives, 19-20 October, Ann Arbor, Michigan. Dunning, John H., and Alan M. Rugman. 1985. The influence of Hymer's dissertation on the theory of foreign direct investment. American Economic Review 75: 228-32. Ergas, Henry. 1984. Monopoly and competition in the provision of telecommunications services. In Changing Market Structures in Telecommunications, ed. H. Ergas and J. Okayama. Amsterdam: North-Holland. Feketekuty, Geza, and Kathryn Hauser. 1985. The impact of information technology on trade in services. Office of the U.S. Trade Representative. Mimeo. Hill, T. P. 1977. On goods and services. The Review of Income and Wealth 23: 315-38. Neumann, Karl-Heinz, Urs Schweizer, and C. Christian von Weizsacker. 1985. Welfare analysis of telecommunications tariffs in Germany. In Public Sector Economics, ed. J. Finsinger. New York: St. Martin's Press.

243

Telecommunications Services

Porter, Michael E., and Victor E. Millar. 1985. How information gives you competitive advantage. Harvard Business Review, 63: 149-60. Reid, Ann. 1985. Trade in telecommunications services: The current institutional framework and the potential for change. Committee for Information, Computer and Communications Policy, ICCP(85)12, OECD, Paris. Rosenberg, Nathan. 1982. Inside the black box: Technology and economics. Cambridge: Cambridge University Press. Sampson, Gary P., and Richard H. Snape. 1985. Identifying the issues in trade in services. The World Economy 8: 171-81. Shapiro, Carl. 1982. Consumer information, product quality, and seller reputation. Bell Journal of Economics 13: 20-35. Teece, David J. 1985. Multinational enterprise, internal governance, and industrial organization. American Economic Review 75: 233-38.

10

International Trade in Banking Services C. R. Neu

10.1 Introduction The international or "tradable" nature of banking services (and of financial services more generally) has been well established. It is difficult to identify any class of services for which an international demand or a capacity for international supply has been more clearly demonstrated. Foreign banking institutions are prominent in most financial centers of the developed world. "International banking" has become almost a cliche. But despite the apparent ease with which banking operations have crossed national boundaries in recent years, there remain important obstacles to efforts by banks to serve customers in foreign markets. Increasingly, these obstacles are becoming a focus for international debate and dispute. Obstacles to international trade in banking services arise for the most part because of the special nature of banking services and the importance that all nations place on the regulation of banking operations. In every country, banking operations are subject to special regulations and restrictions. These regulations and restrictions are almost always intended to ensure the stability of national banking systems, to provide national authorities with effective instruments for economic management, or generally to encourage thrift and other social virtues. That these policies sometimes hinder the establishment of foreign banking operations or restrict the scope of such operations once they have been

C. R. Neu is a senior economist in the Economics and Statistics Department of the RAND Corporation. The views expressed in this paper are those of the author and do not necessarily reflect the views of the RAND Corporation. 245

246

C. R. Neu

established is usually an unintended (although not always deeply regretted by domestic banking institutions) side effect. Having grown up in response to particular national circumstances, policies toward banking operations vary widely from one nation to another. There is no international consensus about what are and are not legitimate activities for banks, about the mechanisms required for adequate supervision of banks, or about the roles that banks are supposed to play in the larger economy. Under the circumstances, it is not surprising that there is no consensus about the rules under which banks from different countries should be allowed to compete with each other within national markets. International trade in services seems to have been identified as the next frontier for trade liberalization—especially by U.S. policymakers—and trade in banking services is likely to be an important topic for discussion whenever a new round of trade talks gets started. In this paper, I consider several aspects of trade in banking services and note the possibilities for and the problems facing expanded international trade in banking services. To the extent that this paper has a unifying theme, it is the question of whether trade in banking services really should be made a priority item in a larger effort to expand opportunities for trade in services. My tentative conclusion is that—at least for the major industrialized countries—the game is just not worth the candle. To identify progress in this area as an essential requirement for the upcoming round of trade negotiations is to risk the failure of the talks. But much more important, to succeed in facilitating expanded trade in banking services may be to accomplish something we may one day regret. In part, this conclusion arises from consideration of the obstacles to be overcome. Painstaking negotiations will be required to reconcile legitimate national needs to regulate banking with the demands of freer and fairer international trade. In the process, many important interests will be threatened and a great deal of political capital will have to be expended—capital that might be more fruitfully employed in seeking improved trading opportunities in other areas. These problems are, I think, well understood and do not need elaboration here. My reluctance to make a big issue out of trade in banking services stems also from some uncertainty about whether more international trade in banking services is unambiguously a "good thing." Raising a question like this among economists may seem a bit odd. Our instinct is to urge freer trade in all areas. But as I began to think about trade in banking services, I found myself becoming less sure that my instincts were right in this case. None of us believes, after all, in completely "free" banking at the national level. We may argue about what constitutes an appropriate

247

Banking Services

degree of bank regulation or what regulatory mechanisms are most effective, but we all recognize a need for some regulation. If we do not want "free" banking at the national level, maybe we should not want "freer" banking at the international level. Pertinent in this regard is the current debate over whether and under what conditions to allow increased interstate banking in the United States. To be sure, many of the arguments against interstate banking reflect little more than a desire in some quarters for m/ranational protectionism. But behind all of this smoke, there do lie some serious questions about the consequences of more interstate banking. It is difficult, for example, not to harbor at least the suspicion that the Continental Illinois debacle might have been considerably harder to deal with if the banks involved had been truly national. If there is a case for limiting interstate banking, maybe there is also a case for limiting international banking. Obviously, there are no simple answers to these questions. In this paper, I try to lay out the potential benefits of increased international trade in banking services and to note on what conditions the existence of these benefits seems to depend. I also try to describe some of the difficulties and potential dangers that may lie in the international supply of banking services. While I hope that my analysis is correct, I will be satisfied if others' efforts to correct my mistakes bring forth some new thinking on this subject. 10.2 The Issues My first task is to establish some reasonably concrete notion of what is at issue when we are talking about increased international trade in banking services. I should stress that I am not offering a judgment here on which aspects of trade in banking services most urgently require international discussion. Rather, I am simply noting which aspects of trade in banking services seem to be on the agenda for the upcoming trade round. The first step in this discussion is distinctly positive, not normative. (For a broader survey of important policy issues relevant to international banking—mostly issues that will not be addressed in the course of trade talks—see Pecchioli 1983.) For the most part, we are not talking about lending to, deposit-taking from, or foreign exchange services for large corporations, sophisticated managers of large portfolios, state enterprises, or national entities. These consumers of banking services generally have access to services from any large bank in the world. They can avail themselves of these services in the home country headquarters of whatever bank they may choose. The most important barriers that banks face in this kind of business are those imposed on them by their home country regulatory authorities. It is evident that banks in many countries have developed

248

C. R. Neu

successful strategies (too successful in some eyes) for attracting and carrying on this sort of business. Barriers to trade in these banking services do not seem to have generated much concern. To the extent that there is a public policy problem here, it is probably the other way around: how to limit these activities to levels considered safe. Neither, more generally, are we talking about those services that are often classified as constituting truly "international" banking—that is, services supplied to foreign residents from offices in a bank's home country. Sometimes these operations can be hindered by restrictions on the customer's ability to do business with a foreign bank. If a firm is required, for example, to buy its foreign exchange from or to hold its foreign currency balances at its central bank, it may have that much less business to transact at some foreign bank. As a practical matter, though, these kinds of restraints do not seem to have become a serious issue for debate. The range and volume of services affected do not seem to be large, and one hears few complaints about these restraints. These sorts of restrictions are sometimes a reflection of economic or balance-of-payments problems, and a country imposing them may not be the most attractive place for banks to look for new customers anyway. Moreover, attempts to limit the opportunities for potential customers to deal with foreign banks sometimes foster a whole new demand for foreign banking services as the restricted customer goes through complicated maneuvers to keep his assets abroad and beyond the reach of the authorities of his country. What people really seem to mean when they are talking about increased trade in banking services is what is often called "multinational" banking—services that require a local presence by a foreign bank: lending to small and medium-sized firms, mortgage lending, retail deposit-taking, consumer finance, and a host of so-called non-assetbased services such as securities underwriting, local-currency bond trading, foreign exchange services for small and medium-sized firms, brokering, custody services, cash letters, lockboxes, and funds collection and disbursal services. These are not services that can easily be supplied to, say, German residents from a banking office in New York. To offer these services, a bank needs to have an operation (as bankers and generals like to say) "on the ground" in the market to be served. (For a general discussion of this issue, see U.S. Government 1984. For a sampling of the debate over how best to gain access for U.S. banks to foreign markets, see U.S. Senate 1984.) 10.3 Obstacles to Trade in Banking Services The requirement for a local presence brings a foreign bank face to face with the domestic banking policies of the host country, and this

249

Banking Services

is where the trouble begins. In the past, national policies have explicitly prohibited the establishment of foreign-controlled banking institutions. As recently as 1984, for example, establishment of foreign banks' branches and subsidiaries was either prohibited by law or not permitted by prevailing policy in Australia, New Zealand, Iceland, and Sweden (OECD 1984). These absolute bars to entry have been relaxed in the past few years, however, and today only Iceland among the OECD (Organization for Economic Cooperation and Development) countries continues to prohibit all foreign-controlled banking operations. (A number of OECD countries require that foreign-controlled banking entities take on a particular legal form, usually that of a locally incorporated subsidiary.) The "right of establishment" (to use the current jargon) remains a thorny issue in negotiations between industrialized and developing countries (U.S. Government 1984). The latter sometimes see (probably quite correctly) financial services as a growth industry. Further, they see (also probably correctly) a network of financial institutions as an important part of the infrastructure necessary for economic development. As a result, some developing nations have adopted classic infantindustry strategies with respect to financial services, excluding foreign banks from their domestic markets in hopes of encouraging the development of the local industry. Among industrialized and financially sophisticated countries, though, the simple right of establishment is usually not a major problem. The real problems in relations among the industrialized countries arise from policies that have the effect—intentionally or otherwise— of discriminating against operations of foreign-controlled banking institutions. This discrimination can take many forms. Foreign-controlled banks may, for example, be prohibited from engaging in certain lines of business or serving particular customers. The most common restrictions of this sort are regulations preventing foreign banks from underwriting bond issues or managing or participating in securities issues and measures that deny foreign banks access to government deposits. Foreign banks are sometimes limited to a specified share of the national banking market. In Canada, for example, foreign-controlled banks may not hold more than 16 percent of all domestic bank assets. Branches of foreign banks must in some instances meet special capital-adequacy criteria. Foreign-controlled banking entities are also faced with a variety of "nuisance" obstacles to efficient operations. Sometimes these take the form of immigration policies that hinder the use of homecountry nationals or requirements that host-country nationals be included in the institution's management or on its board. Sometimes foreign banks do not enjoy the same access to government-controlled communications and information transfer services as do domestic banks.

250

C. R. Neil

Restrictions are sometimes placed on the advertising of foreigncontrolled entities. In the U.K., for example, advertising by banks based outside the European Community must include a warning about the absence of deposit insurance and of possible transfer and exchange risks. Even the names of foreign-controlled institutions are sometimes subject to regulation. The names of foreign banks operating in Switzerland must not give the impression that the bank is Swiss, and in the Netherlands certain conditions must be met before an institution based outside the European Community can use the word "bank" in its name. (These examples of restrictions on foreign banking operations are drawn from OECD 1984; Walter 1985; U.S. Government 1984; and U.S. Treasury 1984.) In the industrialized andfinanciallysophisticated world, foreign banks are generally permitted to establish representative offices as they please. At most, permission of the host government is required, and this permission is generally granted. These representative offices can be active in marketing a bank's services and in prospecting for customers. They cannot actually conduct business—cannot make loans, cannot take deposits. A full accounting of all the restrictions facing foreign banks is beyond the scope of this paper. In 1984, the OECD compiled a summary of regulations affecting the establishment and operations of foreigncontrolled banks within the OECD area (OECD 1984). Even this summary required 84 pages. Moreover, any list of such regulations will be out of date almost before it is printed. Since publication of the OECD volume, Japan, New Zealand, Australia, Sweden, Germany, the Netherlands, France, and the U.K. have all altered policies to reduce obstacles to foreign banking operations. (Personal communication, U.S. Treasury Department). Despite these movements toward liberalization, there seems little reason to alter the conclusion reached by Ingo Walter in his recent study of barriers to trade in banking and financial service: "Foreign-based financial institutions operate in one of the most restrictive environments in international trade" (Walter 1985). Any agreement on principles for trade in banking services will almost certainly include the requirement that "national treatment" be extended to foreign banking entities, making them subject to the same rules as domestic banks. The only alternative basis for agreement, "reciprocity"—country A allows banks headquartered in country B to carry on only those operations that country B permits to banks headquartered in country A—would produce a regulatory nightmare with foreign banks in a major financial center subject to a variety of different limitations. Neither would reciprocity do much to advance trade in banking services. Banks operating in foreign markets would be restricted to offering only those services allowed to domestic banks in

251

Banking Services

both home and host countries, very likely fewer services than could be offered by domestic banks in either country. The general stance of the U.S. government in international negotiations on trade in banking services is to seek national rather than reciprocal treatment of U.S. banks abroad (U.S. Senate 1984). But in some cases, national treatment can be as much of an obstacle to trade in banking services as it is a principle for allowing freer trade. Canadian banking rules provide a particularly neat example of how national treatment can place a foreign bank at a disadvantage. In order that all banking institutions may be unambiguously subject to Canadian law and to facilitate oversight of banks by supervisory authorities, the Canadian Bank Act requires that all banking entities operating in Canada be locally incorporated. This effectively prohibits the establishment of branches in Canada by foreign banks. Foreign banks may operate in Canada, however, through locally incorporated subsidiaries. These subsidiaries, so-called schedule B or 4'closely-held'' banks, are subject to the same regulations as domestic schedule B banks. These subsidiaries are invariably smaller than major Canadian domestic banks, and they have no legal claim on the resources of the foreign parent bank beyond their initial capitalization (although a "comfort letter" from the foreign parent is required). As a result, they are generally seen as poorer credit risks than large Canadian banks and must therefore pay higher rates for deposits than do large Canadian banks. Before the Bank Act came into force in 1980, foreign banks conducted business in Canada through special, nonbank financial companies, operating under a different set of regulations than did domestic banks. There were many disadvantages to this arrangement as far as the foreign banks were concerned, but they were able to fund their operations by issuing commercial paper in the name of their parent banks. After 1980, however, the new schedule B subsidiaries had to issue certificates of deposit in their own names, and subsidiaries of the larger U.S. banks found themselves paying about 25 basic points more for funds than previously. In the view of some foreign bankers, the extension of national treatment to their Canadian operations did more harm than good. 10.4 Why Trade in Banking Services Has Become an Issue Foreign banks have been well represented for a number of years in most (although by no means all) major financial centers of the industrialized world. In the past, foreign banks usually did not compete with domestic banks for domestic business. The foreign banks traditionally concentrated their efforts on serving the multinational companies headquartered in their home countries that they had "followed" abroad. Often (as with U.S. banks in London in the 1970s) they took advantage

252

C. R. Neil

of less burdensome regulations abroad, offering services to their traditional customers at home or in third countries through a foreign office where banking regulations allowed types of operations prohibited at home. In an important sense, the establishment of a London branch of a U.S. bank during the mid-1970s probably represented more an export of British banking services to the United States than an export of U.S. banking services to the U.K. The London branch of the U.S. bank was, after all, staffed mostly by Englishmen whose salaries were paid largely by U.S. corporations using the branch to gain access to the Eurodollar market. When foreign banks did serve local customers, it was usually in some transaction that involved the bank's home country or its financial markets—matters in which the foreign bank would be recognized as having particular expertise. Domestic and foreign banks usually found themselves in direct competition only in providing banking services for customers in third countries. This specialization by foreign banks reflected in part their previous customer relationships. Banking, it is widely said, is a personal business, and foreign banks naturally concentrated their attention on the customers with whom they had had the closest contact. It also reflected the fact that, even in the absence of overt discrimination, foreign banks were generally at a disadvantage in serving local markets. Without the branch networks, the established customer relationships, and the familiarity with local markets that domestic banks enjoyed, foreign banks found themselves without secure local currency deposit bases. To fund their local-currency business, foreign banks had to rely primarily on the interbank market. In general, it was more expensive for foreign banks to borrow local-currency funds from domestic banks than it was for the domestic banks to raise funds directly through deposit taking. Thus, foreign banks found themselves at a disadvantage vis-a-vis domestic banks. This dependence on the interbank market for funding also encouraged a certain competitive restraint on the part of the foreign banks: If you try too aggressively to win the clients that are the bread and butter of a domestic bank, you may find your access to funds at its interbank window suddenly terminated. In recent years, though, this situation has begun to change. As foreign banks have gained confidence and experience in dealing with hostcountry firms, they have gradually begun to see opportunities for attracting more local business. Stronger local deposit bases have improved their competitive position in the local market; and to attract these deposits, the foreign banks have sought more aggressively to serve local customers. At the same time, changing economic conditions have brought new opportunities for banks to offer non-asset-based services. The booming demand for foreign exchange services in the

253

Banking Services

late 1970s and early 1980s, for example, opened a major new market in which a foreign bank could serve just as well as a domestic one. Local-currency funding was much less important in these new lines of business, and therefore the disadvantage faced by foreign banks was less. A worldwide trend toward financial market deregulation and financial market innovation has also produced a variety of new possibilities for banks, and subsequently a variety of new financial instruments and services. New markets have opened in which domestic banks have no long-standing position. Indeed, in some cases a foreign bank with experience at offering a particular service in its home country might find itself with an advantage over domestic banks in a market where such activity is emerging for the first time. A growing "internationalization" of all business activity also has contributed to eroding barriers to operations by foreign banks. Firms in all countries are making increasing use of foreign financial markets. It is no longer unusual, for example, for a U.S. firm to borrow Swiss francs and convert them to dollars to fund a U.S. investment program, knowingly assuming a foreign exchange risk in return for a lower interest rate. As firms become more comfortable with the idea of borrowing foreign currencies, foreign banks without local-currency funding will find their positions strengthened. Another result of deregulation has been an increased emphasis on retail banking as a growth strategy. Services to retail depositors and borrowers have traditionally been among the most heavily regulated services provided by banks, and deregulation has therefore opened a particularly rich set of opportunities in this area. In the United States, and in other industrialized countries as well, retail customers are becoming more aggressive, demanding a wider variety of banking services at finer prices. Many see retail financial services as a major growth industry, and the ability of foreign banks to provide these kinds of services is particularly at issue. Finally, technological progress in telecommunications and data processing may have brought new possibilities for economies of scale. This may be encouraging efforts by banking institutions to grow, and foreign markets may in some cases provide the most attractive opportunities for growth. The last few years, then, have seen an erosion in the "natural" barriers to foreign bank operations. As these barriers have declined, what are perceived as "unnatural" barriers—barriers arising from government policies—have loomed relatively larger. This has happened at a time when the incentives for expansion by banks into foreign markets are also growing. The not very surprising result is intense new interest

254

C. R. Neii

in how national policies do or do not hinder trade in banking services. (For an indication of interest within the banking community in improving access of U.S. banks to foreign markets, see U.S. Senatel981.) 10.5 What Really Is Being Exported What is at issue, then, in the debate over international trade in banking services is the right of foreign banks to compete for customers in the domestic market of the host country. To what extent, though, does the provision of these services really constitute an export? A visit to the foreign branch of any major U.S. bank will reveal an office staffed overwhelmingly by natives of the host country. Sometimes this staffing pattern is required by local immigration laws; in many OECD countries, obtaining work permits for foreign personnel can be difficult. (Restrictions on foreign workers can form very effective barriers to foreign competition in many service industries, and these rules will almost certainly become an important focus for discussion whenever talks on service trade are begun in earnest. It may turn out that the real key to agreement on rules for trade in services will be agreement on policies towards foreign workers.) But the use of native staff is also simple good business. Language, cultural affinity, and direct experience with prevailing local conditions are important elements in successful banking relationships, and it is in native staff that these qualities are often most easily found. Funding, the other principal input to the production of banking services, is generally procured locally as well. Foreign banking operations generally rely only minimally on funding from the parent bank, turning instead to the local interbank market and, when possible, to local deposits for raising funds, particularly local-currency funds. The other inputs to this production process—facilities, communications, data processing, and so on—are also generally purchased locally. Banking services are not like automobiles or textiles. Export either of the latter, and you employ more workers and raise incomes in the exporting country. To the extent that other inputs also are produced in the exporting country, you increase employment and income that much more. Export banking services in the sense described above, though, and you mostly generate jobs for foreigners, pay interest to foreign depositors, and pay foreign suppliers for most of your other inputs. What the home country gets is some remitted profits and a modestly increased demand for supervisory and headquarters personnel. For those who see increased exports of services as an important way of increasing employment and income, banking services may turn out to be a disappointment.

255

Banking Services

The other side of the coin is that the importation of banking services is unlikely to have a major impact on employment of importing country nationals in the banking sector. If a foreign bank captures a significant share of the domestic market, workers displaced from a domestic bank are likely to find opportunities to move across the street, going to work for the newly arrived foreign bank. Indeed, if some local workers do not move across the street, the chances of a foreign bank's gaining a share of the local market are probably small. Neither will the capture of some share of the market by foreign banks affect the demand for local deposits or for other inputs to the banking process. If anything, the arrival of a foreign bank is likely to increase demand for these inputs. A fear that increased imports will leave some local resources underutilized (a reasonable concern, at least in political terms, in the case of merchandise imports) is probably not justified in the case of banking services. Those who see international trade in banking services as a potential threat to local employment or local income are probably worrying unduly. What are really being traded in the case of banking services, it would seem, are managerial services, technical know-how, networks of contacts, certain analytical and information services, and the name of the parent institution. The only local resource that may be left underutilized by an increase in foreign banking activity is local bank management. (Similarly, the home-country resource most likely to be more in demand as a result of increased trade in banking services is bank management. A cynic might see in this an explanation for the vehemence with which senior bank executives sometimes approach questions concerning international trade in banking services.) What is being exported and imported are not really banking services at all, but the technical and managerial structure necessary for the local production of banking services and the capital necessary to establish a branch or subsidiary to deliver banking services. This leads to my first general observation. Discussions of rules for international trade in banking services would seem to have much more in common with discussions about rules for international direct investment than with discussions about merchandise trade. Perhaps the inclusion of banking services in a round of negotiations aimed at liberalizing other, more traditional forms of trade is not the best idea. This line of argument also raises a suggestion for empirical work. As far as I can tell, no one has a good idea about how much "trade" there really is in banking services. We do not have a clear picture of the value of services provided to banking customers by foreign banks. Neither do we have any idea what fraction of the value of these services is accounted for by returns to factors of production located in the

256

C. R. Neii

foreign bank's home country. I suspect that the truly "traded" component of banking services is small, that most of what may appear as the provision of banking services by a foreign bank is really local production. I suspect, therefore, that there is much less to trade in banking services than may at first meet the eye. It would be interesting to see some careful empirical work in this area. 10.6 The Benefits of Freer Trade in Banking Services To argue that the direct employment and income effects of trade in banking services are small is not to argue that trade in banking services is an unimportant or an uninteresting issue. The centrality of the banking function to the operation of the wider economy is such that questions about how the banking system operates and by whom it is operated are always important. There is also a particular analytic attraction in the question of trade in banking service. Precisely because the direct employment implications of such trade are relatively minor, some of the more emotional arguments for and against freer trade are minimized. This gives more leeway for the more fundamental analysis of the type that economists prefer. I now turn to more traditional arguments about the benefits of trade in banking services. Traditionally, two sorts of gains arise from freer international trade: 1. Increased opportunities for trade can promote more intense competition within national markets and can lead to more efficient production of the traded good or service. 2. Each trading nation gains by specializing in the production of the goods or services for which it enjoys a comparative advantage. Dealing with the first of these benefits is relatively simple. Lack of competition in the banking sector, as in any other sector, can allow inefficient providers to survive, reducing the range of services available to customers, and raise the prices of services that are provided. The experience of countries that have deregulated their banking sectors so as to allow more competition provides ample proof of these points. Increased competition has forced some banks out of business and forced survivors to offer a broader range of customer services at lower prices. This is undoubtedly a benefit to customers. What is not clear is whether the presence of foreign banks is necessary to foster more intense competition. Most obstacles to competition in banking services arise from local regulations that prohibit entry into the banking sector, fix interest rates, or otherwise restrict competition. It seems likely that important gains in this direction could be achieved simply by adjusting national policies. Indeed, if these policies

257

Banking Services

are not adjusted, allowing foreign banks to operate in domestic markets will probably do little to encourage competition; newly arrived foreign banks will sink into the same comfortable noncompetitive life enjoyed by domestic banks. Where restrictive regulations have been relaxed, more intense competition seems to have arisen even without a surge of activity by foreign banks. Thus, it would seem that allowing international trade in banking services is neither necessary nor sufficient to bring about the undoubted benefits arising from more competition in the banking sector. The issue of comparative advantage in banking services is more difficult. If banks in one country can produce banking services more efficiently, in terms of the other goods and services produced in that economy, then can banks in another country, and if this efficiency can be "exported"—that is, if it persists when a bank from the first country offers services in the domestic market of the second—there will be welfare gains to be had by allowing banks from the first country to supply banking services in the second. If there is no substantial difference in the relative costs of producing banking services in the two countries, or if differences cannot be transferred to another market, then there is little to be gained from trade. We have, then, two kinds of questions to answer: 1. What factors can bestow on a bank (or banks based in a particular country) greater relative efficiency in providing banking services? 2. Can this efficiency be "exported"? Let us consider a number of possible sources for banking efficiency. (In the rest of this discussion, I will sometimes make points in terms of absolute rather than strictly comparative advantage. I do this for convenience of presentation. All of these points could be made—more clumsily—in terms of comparative advantage without altering the nature of the results. For a survey of studies on what constitutes advantage in the international supply of banking services, see Aliber 1984. For a more general discussion of comparative advantage in services trade, see Deardorff, 1984.) The availability of sufficient numbers of well trained and disciplined clerical workers at a relatively low wage will undoubtedly give a country a strong advantage in producing banking services. Such workers are important inputs to the production of banking services, and a plentiful supply of them will reduce the cost of producing such services. But this workforce is not exportable. As we have already noted, clerical staff in the branches and subsidiaries of foreign banks are generally hired in local labor markets. Thus, being headquartered in a country endowed with the kind of workforce necessary for banking is not likely

258

C. R. Neil

to bestow any particular international advantage on a bank. Thus, there seems little reason to expect gains from trade to arise because of differing labor force characteristics in different countries. The efficiency of bank operations can vary, of course, with the quality of bank management. At any given time, there will be a group of banks, not necessarily all in the same countries, that will be able to offer better or cheaper services than can other banks by virtue of their simply being better run than other banks. To the degree that this managerial talent reflects expertise in dealing in particular financial markets, it will not be transferable to markets in other countries. But to the extent that it is embodied in personnel policies, operating procedures, data-processing operations, or other techniques not unique to particular markets, it may be "exportable." Freer international banking may give rise to welfare gains if customers are able to avail themselves of higher-quality services when from time to time they are offered by foreign banks. One might wonder, though, whether gains of this sort are likely to be very significant. Little banking technology is proprietary. Bank officers with detailed knowledge of banking products and internal operating procedures frequently move from bank to bank, and bank personnel frequently go abroad to learn the latest techniques employed by foreign correspondent banks. It seems unlikely that any bank has unique procedures or management secrets that are not known to other banks. Further, most large banks have the size and capital resources necessary to implement good ideas they pick up from the outside. (Among U.S. banks at least, the ability to copy new ideas quickly is striking.) If differences in product lines or management practices persist for extended periods, they are likely to reflect some adaptation to local conditions. When foreign products and management techniques differ systematically from local ones, the advantage must be presumed to lie with local practice. In these circumstances, the products and practices of the foreign bank are unlikely to be exportable, and the gains from freer trade in banking services are again likely to be minimal. A more likely source of advantage in the provision of banking services lies in differing abilities to attract deposits and to raise capital. These abilities will in turn be strongly influenced by national policies toward bank regulation. Banks in one country might, for example, be perceived as being somewhat safer than banks in other countries. Perhaps this is because they are supervised by a regulatory authority seen as particularly wise, careful, or strong. Perhaps it is because prevailing national banking regulations guarantee particularly prudent behavior. Or perhaps banks headquartered in that country have access to a particularly reliable lender of last resort. Whatever the reason, these banks will be able to raise capital and to attract deposits more cheaply than

259

Banking Services

will banks less favorably situated. As a result, they will have a cost advantage over banks in other countries and should be able to offer loans at lower rates and nonasset services at lower prices. These banks will enjoy a true advantage; and if this advantage can be maintained in an "export" market, there will be gains to allowing such banks to provide services in foreign markets. Banks' competitive positions can also be affected by reserve requirements and regulatory attitudes about capital adequacy. Banks required to hold large amounts of non-interest-earning reserves are at a disadvantage vis-a-vis banks not required to do so. Banks permitted a low capital/asset ratio should be able to offer finer spreads than a bank forced to operate with a lower gearing ratio. Obviously, competitive advantage in these circumstances is bought with some increase in risk, but other policies—such as strong government guarantees or particularly close supervision—may compensate for these higher risks. In terms of usual trade theory, if two countries have differing endowments of the factors necessary for efficient banking (such as confidence in the banking system or a set of regulatory policies that accommodate banking), and if these endowments are reflected in cross-border operations, then there may be significant gains from trade. Whether these advantages are exportable will depend on the mechanism by which export is accomplished. If the exporting bank establishes a branch to serve a foreign market, many of the characteristics of the parent bank are exported. The bank's home-country regulations apply to the operation of the branch, home-country authorities oversee its operation, and all the resources of the parent bank stand behind it. To the extent that these characteristics inspire confidence and allow the branch to attract deposits or capital easily, the advantages enjoyed by the parent bank are exportable. Also exportable, in general, are home-country reserve requirements and definitions of capital adequacy. If these are the basis of a competitive advantage, they too can be transferred. What is not entirely clear is the extent to which the foreign branch has access to the lender-of-last-resort facility of the home country central bank. The Concordat concluded among major central banks in 1975 suggests that the central bank of the parent should be responsible for foreign branches. (The text of the Concordat has been published in IMF 1981.) Some doubt remains, however, about whether foreign branches can really count on support of their home-country central banks. Fortunately, these guarantees have never been put to a test. Even if the support of the home country central bank were certain, the export of confidence in the parent bank could never be complete; depositors are understandably hesitant to rely on promises by a foreign central bank that might have to be mediated through a foreign legal system.

260

C. R. Neil

If, on the other hand, the mechanism for export of banking services is a foreign subsidiary, export of the parent bank's advantages is more problematic. The subsidiary has no legal recourse to the resources of the parent beyond its initial capitalization. Neither do regulatory authorities in the home country of the parent bank have any direct control or supervision of the subsidiary. Reserve requirements and capital adequacy tests of the host country are operative for the subsidiary. And the central bank Concordat makes the host-country central bank the lender of last resort for the subsidiary. (That home-country central banks can feel little responsibility for foreign subsidiaries was dramatized in 1982 when, in the wake of the Banco Ambrosiano affair, the Bank of Italy refused to extend guarantees to deposits in Banco Ambrosiano's Luxembourg subsidiaries.) I believe that the only things that are exported when a foreign subsidiary is established are the parent bank's ability to raise capital and the parent bank's upper-level management structures. It is hard to see how the establishment of a subsidiary can provide an effective mechanism for the export of other competitive advantages. This leads to my second general observation: If we are seeking welfare gains from trade in banking services, we probably must allow cross-border branching; subsidiaries will probably not do the trick. While establishing foreign branches may be the only way effectively to export the important qualities of a bank to foreign markets, branches can pose serious difficulties for host countries. The host country can generally exercise no control over and only limited surveillance of the parent bank, and thus has little control over the soundness of a banking institution providing services to its residents. This can be overcome by requiring the formation of a subsidiary clearly separate from the parent bank. But to do so is sometimes to get the worst of everything. By insisting that foreign banks operate as subsidiaries, for example, Canada has gotten American bank managers who may or may not fully understand the Canadian financial markets in which they operate, without any recourse to the substantial assets and reserves of the parent American banks should these managers get into trouble. A fundamental dilemma is posed: The benefits of trade in banking services can be realized only if host-country authorities are willing to relinquish regulatory control and oversight of some banking institutions operating in their territory. Insulation from the dangers of foreign banking practices can be bought only at the price of insulation from the benefits of these practices as well. 10.7 The Dangers of International Trade in Banking Services Many banks apparently feel that the difficulties of exporting operating efficiencies across national boundaries can be overcome and that

261

Banking Services

successful operation in foreign markets is possible. Why else would banks show such interest in establishing a presence in foreign markets? If banks think they can succeed in foreign markets, why should we not allow them to try their luck? Indeed, since some gains may arise from increased trade in banking services, why should we not do everything possible to remove barriers to such trade? The principal reason for caution in encouraging more trade in banking services is that this trade may erode the safety and stability of the global banking system or, depending on prevailing regulatory policies, the safety and stability of either host- or parent-country domestic banking systems. In every country, regulatory authorities and central banks bear at least some of the risks inherent in banking operations. The decision to commence banking operations in a foreign market may increase the risks borne by regulatory authorities in either the home or the parent country without necessarily imposing any additional costs on the bank wishing to engage in multinational business. If, for example, deposits are fully guaranteed by national authorities, depositors have no reason to demand higher interest rates even if banks take on riskier operations. Similarly, if regulatory authorities do not increase the costs of deposit insurance as banks take on riskier operations (and how are regulators realistically to measure risks?), banks do not face the full extra costs of engaging in riskier business. Thus, moves that may appear profitable to an individual bank, may impose extra risks on the wider society and may result in a hidden loss of overall welfare. In the usual jargon, moral hazard may result in negative externalities when a bank decides to engage in cross-border operations. Does the entry of a bank into a foreign market increase the riskiness of its operations? Some have argued that by allowing banks to broaden the scope of their operations, by allowing them to enter new lines of business or to extend their operations geographically, we allow banks to diversify their operations and thereby reduce their risks. This argument makes sense only if the quality of bank management is independent of or is positively correlated with the scope of banking activity. I think that it is not unduly uncharitable to say that recent experience has, at the very least, raised some questions about the validity of this proposition. The diversification of bank portfolios away from traditional forms of lending into international lending and a variety of new services does not seem to have brought about any marked reduction in riskiness of major banks. Neither has it made these banks noticeably better able to retain their balance when traditional loans (in, for example, the energy or agricultural sector) go bad. For some time now, the managements of big banks have faced incentives to grow—often without taking full account of the risks thus incurred. Recent difficulties have complicated the process of growth, but they have probably not altered the underlying incentives. It is still

262

C. R. Neu

a better and better paid job to run a bigger bank. Incentives to growth may also have been strengthened by recent problems with bad loans: One way to deal with a fixed volume of bad loans is to enlarge the bank enough to absorb them, and foreign markets may appear to be a fruitful field for expansion. The trick, of course, is to avoid taking on more bad loans because the bank is growing too fast. Some have argued that the rush by major banks to establish a presence in new markets either at home or abroad has distracted management attention from the basic business of credit assessment and risk control. The result is at least a perception that major banks are in a more precarious position today than at any time since the Great Depression. The expansion of banking activities into new markets also places new demands on regulatory authorities. As new financial instruments appear, for example, regulators cannot always clearly identify what constitutes credit exposure for a bank. When a bank moves into a foreign market, domestic regulators may not fully understand the circumstances prevailing in foreign markets or may not have access to information essential for assessing the creditworthiness of foreign borrowers. The presence of foreign banks in national financial markets may also increase somewhat the "interconnectedness" of the international financial system, making it easier for shocks (and there will inevitably be shocks) to propagate through the system. Perhaps one should not make too much of this point. The Penn Square/Continental Illinois fiasco made it clear that the degree of interconnectedness is already such that the failure of a rather small Oklahoma bank could lead to severe difficulties for a number of foreign banks that had deposited large amounts with Continental. (For a readable and entertaining account of the Penn Square debacle, see Singer 1985. A more traditional account is available in U.S. Senate 1982.) The situation might not have been appreciably worse if large numbers of foreign retail customers had also been at risk. If there is any difference between the situation we have now and the situation that might prevail if there were a significant increase in trade in banking services, it might lie in the political implications of a bank crisis originating abroad. The non-U.S. banks that found themselves in a difficult spot because of Continental's troubles were large enough and sophisticated enough that they might reasonably have been expected to make informed judgments for themselves about the soundness of the U.S. banks they dealt with. U.S. regulatory authorities guaranteed the large deposits of these banks (a step not required by U.S. banking law) in order to prevent a bad situation from becoming worse, not because they felt any moral or political obligation to protect these foreign banks. Financial authorities outside the United States also might

263

Banking Services

not have felt much obligation to make whole these banks. As long as the losses incurred did not threaten the stability of national banking systems, there was no reason to act to minimize the losses suffered by banks that should have known better than to deposit too much at Continental. The situation might have been quite different, though, if Continental had had a host of foreign retail customers. These customers could not reasonably have been expected to assess the relative safety of the various banks seeking their deposits. If a bank is allowed to take retail deposits, it is not unreasonable to expect local authorities to make sure it is sound. The burden of assessing the soundness of foreign banks in this situation falls more heavily on local regulatory authorities—who may or may not be up to the job. Perhaps life would be simpler all around if local authorities were responsible for local banks only and were not expected to take on the daunting task of investigating the soundness of every foreign bank that seeks to open up shop in their area of responsibility. To the extent that one is willing to rely on the terms of the central bank Concordat, which calls for the monetary authorities of the parent country to stand behind deposits in the foreign branches of banks headquartered in its territory, one might come to the mildly ironic conclusion that the most sensible national policy toward trade in banking services would be to seek to attract branches of foreign banks into local markets while discouraging branching of domestic banks into foreign markets. This, of course, is just the opposite of what is generally sought in discussions of trade in banking services. It may be that one of the few things actually being exported when a bank establishes a branch in a foreign country is the guarantee of the home-country regulatory authority to make whole a foreign depositor who might lose money as a result of problems within the multinational bank. The regulatory authority in question may or may not receive compensation for this extension of its contingent liabilities. If it does not, or if these costs are borne by home-country interests, the taxpayers, banking customers, or bank shareholders, country A may be providing deposit insurance for the residents of country B at little or no cost to B. Another danger inherent in wider trade in banking services is the pressure that might be brought to bear on national regulatory authorities to make banks under their jurisdiction internationally competitive. We have become accustomed to pleas to governments to ease this or that requirement for pollution control or worker safety in order to make some domestic manufacturing industry competitive in world markets. Regulatory competition of this sort is often regrettable, the result being more injured workers or dirtier air. But at least where manufacturing industries are involved, the suffering that results from unwisely waived

264

C. R. Neu

regulations is usually limited to the constituents of the government that knuckled under to pressure. Imagine what might happen if trade in banking services became a major item in some countries' balance of payments. Pressures would arise from the banking lobby that local reserve requirements were too high to permit effective competition with foreign banks regulated by more "reasonable" authorities or that local reporting requirements were too burdensome and imposed costs that were pricing domestic banks out of foreign markets. I am not being an alarmist; pressure of this sort is already being brought to bear. In January 1986, the U.S. Federal Reserve proposed the eminently sensible idea of tying bank capital requirements to the riskiness of bank asset portfolios. An official of the American Bankers Association opposed such a step on the grounds that it would "make the industry less competitive with financial institutions that wouldn't have to meet such standards" (Nash 1986). By giving in to these pressures, regulatory authorities would endanger not only the assets of their own citizens but perhaps the stability of the entire global banking system. One might imagine the result of an international competition among national bank regulators, each trying to make banks in his own country more competitive. The regulators might seek to achieve this end by insisting on extremely prudent behavior, enforced by strict regulation. Alternatively, regulators might seek competitiveness through lax regulation, low reserve requirements, and relaxed attitudes towards capital adequacy. Badly regulated banks might drive out well-regulated ones if the near-term cost advantage of being badly regulated were sufficient to allow the former to take markets from the latter. Given the difficulties already facing bank regulators, perhaps it would be better not to allow considerations of international competitiveness to influence regulatory policy. Perhaps it would be better if banking remained a national industry. My third observation, then, is that banks contemplating an expansion into foreign markets may not face the full costs of such an expansion. By deciding to proceed with their plans, they may impose excessive costs on all the rest of us. 10.8 International Trade in Banking Services and the Optimal Size of Banks Another reason for caution in encouraging increased trade in banking services is that it might lead to banking markets being dominated by fewer, bigger banks. There is no international bank regulator. As bank operations sprawl across the globe, who is to guarantee prudent behavior? The only possible answer seems to be that we must rely on

265

Banking Services

market forces to supply the necessary restraint. Before the Continental affair, U.S. regulators had announced what seemed an appealing policy. They clearly stated their intention to limit deposit insurance to the $100,000 per depositor per bank prescribed by the policies of the Federal Deposit Insurance Corporation. Depositors of larger amounts were presumed to have the resources and the sophistication to form their own judgments about how prudently the banks they placed deposits with operated. By placing these larger depositors at risk, the authorities hoped to impose some market discipline on U.S. banks. Banks with suspect loan portfolios or less than entirely reliable funding arrangements would, the thinking went, find themselves paying a premium for their deposits and thus have an incentive to clean up their acts. For a policy like this to work, large depositors must believe that they really can lose their money. The catch, as it turned out, was that Continental was just too big to be allowed to fail; U.S. authorities eventually guaranteed all deposits fully. This is not to suggest that U.S. authorities did the wrong thing in the Continental case; they probably had no other choice. (At least at the time they thought that they had no other choice.) But once they had guaranteed all deposits, the previously announced policy was seen to have been a bluff. It seems unlikely that in the next few years any national banking authority will be able credibly to threaten the large depositors of any major bank with the possible loss of funds. This suggests that if we are trying to design the global banking system for tomorrow, we ought to be seeking ways to produce banking institutions that are big enough to capture the substantial economies of scale that mark the banking business, but are not so large that they cannot fail from time to time. Continental was thought to be too big to be allowed to fail, and there are today plenty of banks bigger than Continental was. If we are going to have to rely on the market to encourage prudent banking behavior in the future (and where else will we have to turn?), we might want to consider how to reduce the size of some of the larger banks, or at least how to slow their growth sufficiently to let growth of the world financial system reduce their relative size. As the costs of information transfer and data processing continue to fall, it may also be that the minimum efficient size for a bank is decreasing. One wonders if there will be a continuing need for the very large financial institutions that we have today. I do not know of any serious efforts to estimate costs functions for the production of banking services. Work of this sort might yield some interesting insights into the optimal size of banking institutions. What does this have to do with freer trade in banking services? One possible consequence of freer trade in such services is an increasing

266

C. R. Neil

concentration in the global banking industry. Little is known about market dynamics in the banking industry. I do not really know whether large banks have important competitive advantages over small banks. It is certainly possible, though, that only large banks will have the resources to jump into new markets. And if such banks succeed in capturing a share of foreign markets, it may well be at the expense of smaller banks in the host country. My fourth observation is that freer trade may result in bigger banks controlling more of the total market for banking services—bigger banks whose failure is likely to be less thinkable than was Continental's. Restricting the geographical spread of banks is almost certainly not the ideal way to slow their growth. It may, however, be the most natural and politically feasible way to do so.

10.9 What Kinds of Agreements Are Necessary? There is no denying the fact that perceived obstacles to trade in banking services are a source of tension today among the developed, financially sophisticated countries of North America, Western Europe, and the Pacific rim. Banks in all of these countries frequently want to do more in foreign markets than host-country policies will allow, and these banks put pressure on their governments to negotiate them a better deal. I wonder, though, how long-lived this kind of pressure will be. We have seen that the principal issues are the right of establishment in foreign markets and the terms under which foreign banks are to be allowed to operate once they are established. We have also seen that for some kinds of business, local establishment seems to be essential. Will this continue to be the case? I have no doubt that personal representation in a particular market will always be required. The need for bankers and customers to meet face to face seems unlikely to go away. Almost all developed countries allow the free establishment and unhindered operation of representative offices of foreign banks. These offices can manage customer contacts, negotiate terms of transactions, and generally market the foreign bank's services. They cannot actually book business. If the activities of a foreign representative office are to bear fruit, a foreign customer must ultimately deal with the headquarters of the parent bank or one of its branches in a third country. There are three kinds of obstacles to these cross-border transactions. The first is regulatory: one of the countries involved in the transaction may have capital controls or currency restrictions that prohibit the external transaction contemplated. The second is practical: communications difficulties, time-zone differences, and so on may hinder the

267

Banking Services

smooth execution of the transaction. The third is legal: a customer with a grievance will have to deal with a foreign legal system. All indications are that obstacles of the second sort are being rapidly eliminated. Communication and electronic funds transfers are becoming easier and cheaper. Financial market hours are being extended and financial institutions are becoming increasingly willing and able to transact business around the clock. Customers also are becoming increasingly comfortable with the prospect of electronic financial transactions. When a customer uses an automatic teller machine, transfers funds by wire, or takes advantage of automatic funds collection or disbursement services, does it really matter whether the computer handling the transaction is in his own country or not? Indeed, we are told that what seem to be purely domestic transactions will soon actually be international, as, for example, U.S. banks place "back office" data processing functions abroad in low-wage areas. Obstacles of the first and third types may also be reduced as countries seek to adjust their policies and their laws to encourage foreigners to bring their banking business to institutions headquartered there. It seems possible that before very long all of the functions that now require a local presence by a foreign bank may be adequately handled by a combination of a local representative office and a very good communications link to the bank's headquarters. If this style of operation should become the norm, much of the difficult debate over appropriate rules for foreign bank establishment, what constitutes truly national treatment for foreign banks, and so on will have been unnecessary. In this possible world of electronic banking, the real issues will be what kinds of national controls on capital movements and foreign exchange transactions are suitable, how international information flows (potentially involving very sensitive information) are to be regulated, who will be responsible for the smooth operation of international interbank clearing systems, and what level of foreign exposure is appropriate for banks in the eyes of their homecountry regulators. There is no reason to think that these issues will be any easier to settle than those involved in setting up foreign branches and subsidiaries. My own view is that rules for operating this truly international (as opposed to multinational) banking system will be much more difficult to devise than the rules that would govern simple establishment and operation of foreign bank branches and subsidiaries. But we will have to face these problems anyway. Expending time, energy, and political capital to facilitate foreign establishment of branches and subsidiaries, a style of doing business that already appears somewhat outmoded, may represent a diversion of resources from more important work. My fifth and final observation is that we may be getting ready to negotiate an unnecessary set of agreements.

268

C. R. Neii

10.10 Conclusions My conclusions are really rather simple. For a variety of reasons, I suspect that the next round of trade talks will have a better chance of being productive if the thorny questions of how foreign banks should be allowed to operate in national markets are downplayed. These issues will be contentious, and there is no guarantee that success in making multinational banking more prevalent will make the world a better place. Much better to concentrate attention on questions of restrictions on foreign workers (critical to many service industries), rules for foreign direct investment in service industries, international information flows, and international payments restrictions. The overall tone of this paper may appear to be anti-free trade. This is certainly not my general view of the world. We have for a long time, though, recognized that banking is not just another industry, and we should at least consider the possibility that our usual approaches to international trade may not be appropriate to the banking industry. More important, I am troubled that economists have not played a larger part in the debate over all aspects of international financial activity, leaving this debate largely to "men of affairs." If nothing else, I hope this paper may provoke some further consideration of this subject by economists, if only to rebut my positions. It might even be that by thinking harder about the place of banking in international trade, we may develop a clearer idea of banking's role in the domestic economy.

References Aliber, R. Z. 1984. International banking: A survey. Journal of Money, Credit, and Banking 16, no. 4 (November) part 2: 661-78. Deardorff, A. V. 1984. Comparative advantage and international trade and investment in services. Unpublished manuscript. International Monetary Fund. 1981. International capital markets. Occasional Paper, no. 7 (August). Washington, D.C.: IMF. Nash, N. C. 1986. U.S. regulators want banks to cover risky loans better. New York Times, 16 January, p. A l . Organization for Economic Cooperation and Development. 1984. International trade in services: banking. Paris: OECD. Pecchioli, R. M. 1983. The internationalization of banking: The policy issues. Paris: OECD. Singer, M. 1985. Funny money. New York: Knopf. United States Government. 1984. U.S. national study on trade in services: A submission by the United States Government to the General Agreement on Tariffs and Trade. Washington, D.C.: Government Printing Office, 1984-464694/20595. United States Senate. 1981. Service industries development act. Committee on Commerce, Science, and Transportation, Subcommittee on Business, Trade,

269

Banking Services

and Tourism. Hearings to consider Senate Bill 1233, 20-21 October. Committee serial number 97-88. United States Senate. 1982. Failure of the Penn Square Bank. Committee on Banking, Housing, and Urban Affairs. Hearings 10 December. United States Senate. 1984. National treatment of banks, Committee on Banking, Housing, and Urban Affairs. Hearings to consider Senate Bill 2193, 26 September. United States Treasury Department. 1984. Update of Report to Congress on foreign government treatment of U.S. commercial banking organizations. Originally published in 1979. Walter, I. 1985. Barriers to trade in banking and financial services. Thames Essay no. 41, Trade Policy Research Centre, London.

Comment

Frances Ruane

Before responding to the papers by Andre Sapir and Richard Neu,* I would like to begin by making some general remarks about trade in services. This subject has become very topical among trade policy makers in recent years because of both the growth of international trade in services and the repeated suggestions that trade in services may be incorporated within the GATT.1 It has also merited some discussion among international trade theorists who have been concerned primarily with the conceptual issue of whether the framework which they use to analyze trade in commodities is appropriate for trade in services.2 If such a framework is appropriate, then the results obtained for trade in commodities and the policy implications for trade negotiations carry through to trade in services. From a trade theory perspective, one's immediate instinct is to presume that, at least as far as economic aspects are concerned, trade in services is similar to trade in commodities, whatever the legal, institutional, or payment differences between them. This contrasts with the approach taken by Hill who argues strongly that services have two specific characteristics which goods do not have and consequently "goods and services belong in different logical categories" (1977, 336). The two specific characteristics identified and discussed at length by Hill are nonstorability and physical proximity, which suggest the reDr. Frances Ruane is a fellow of Trinity College, Dublin, and a visiting associate professor at Queen's University, Kingston, Ontario. *Chapters 9 and 10 in this volume. 1. For an overview of the history of trade negotiations in services, see Malmgren (1985). 2. See, for example, Hindley and Smith (1984), Deardorff (1985), Jones (1985), Markusen (1986a, 1986b), Melvin (1985, 1986a, 1986b), Sampson and Snape (1985), and Sapir (1985).

270

C. R. Neii

quirement that production and consumption of each service must coincide in time and in space, respectively. While these characteristics are common of many services, they are not common to all,3 and I, for one, am not convinced of the merits of searching for characteristics which are common to all services. What is more important is that we have a framework which allows us to analyze the different channels through which trade in services takes place, especially at a time when we are contemplating negotiations which will affect these channels.4 Focussing on channels rather than basic characteristics, I have found it helpful to look at services conceptually in terms of whether they are actually tradable or nontradable, and whether or not, for any given economy, foreign-owned factors are required in the production of services. Figure C9-10.1 illustrates. Box 1 contains those services which are produced using domestic factors only and which are sold only on the domestic market. This category includes what one would consider traditional services, such as restaurants, general retailing, etc. Box 2 contains services that are produced by domestic resources only but that are traded internationally. Examples of such services are life insurance, engineering designs, etc. Box 3 contains services which are nontraded but which require foreign as well as domestic factors in production. Examples of such services are certain types of international banking, and certain types of consultancy. Finally box 4 contains those services which are both traded internationally and which use foreign as well as domestic factors. An example of such a service would be a U.S. bank in Brussels which services loans to Italy. While in each case I have given examples of services which may fall into a particular box, I wish to stress that I do not believe that one should necessarily attempt to classify whole sections of the service sector in this way; rather in looking at a particular service sector activity, it might be helpful to identify it in terms of this classification. In the context of the issues of this conference and both Sapir's and Neu's papers, I found this framework useful in that it allows one to identify, according to the way in which a service is classified, the focus for policies and negotiations. For example, services in box 1 are strictly domestic and are indistinguishable from what trade theorists have called nontraded goods. The prices of such services are determined by the parameters of the production process, the availability of domestic factors of production, domestic consumer tastes and the degree of competitiveness in that sector. Clearly services of this type should not fall within the arena of trade negotiations. By contrast, for policy purposes, 3. Exceptions have been noted, for example, in Bhagwati (1985). 4. This approach is similar to that taken in Melvin (1986a, 1986b).

271

Banking Services

TYPE OF FACTORS DOMESTIC

DOMESTIC & FOREIGN

NONTRADED

1

3

TRADED

2

4

TYPE OF SERVICES

Fig. C9-10.1

Analysis of services with respect to the tradability of outputs and the ownership of production factors.

services in box 2 should be treated in like manner to commodities, where instead of tariff and quota negotiations, intergovernmental negotiations would take the form of negotiations on barriers to entry, such as domestic regulation. Furthermore, it must be recognized that differences in the methods of payment and the legal status of the services will have implications for such negotiations. In the case of services in box 3, the service itself is not traded internationally but is produced in the country in which it is to be consumed. Furthermore, there is no necessity in many instances for the foreign-owned factor to physically cross national boundaries; the only requirement is that it combine with local factors to produce the service. Of course in many instances it will move physically but it is the foreign ownership that is the essential component.5 This feature of service markets is similar analytically to the case of international factor mobility where the foreign factor is used in the nontraded sector; see for example, Jones, Neary, and Ruane (1983). The characteristics of services in box 3 indicate clearly that negotiations in this area are about the use and payment of internationally-owned factors and the nature and extent of competition in the domestic market. GATT negotiators in such areas would find themselves breaking new ground, as the whole area of international factor mobility, use and payment typically lies outside the GATT framework. I will return to this issue below. In the case of box 4, we have the combination of trade at the output and factor level, so that policies appropriate in both box 2 and box 3 may be relevant here. Turning now to the papers presented in this session, Richard Neu provides a very detailed qualitative account of the various activities which fall under the heading of international banking. (The difficulties involved in quantitatively measuring the scale of trade in banking services are discussed in St.-Hilaire and Whalley [1986] who provide some 5. This point is particularly important in the case of services. See Deardorff (1985).

272

C. R. Neu

estimates of trade in banking services for the United States.) Because he looks at banking in isolation, Neu's paper tends to emphasize the differences rather than the similarities between banking and other economic activities. In fact, as Neu implicitly recognizes in his first general observation, international banking typically involves the use of domestic and foreign-owned factors to produce a nontraded service; in other words, in terms of the framework presented above, international banking services fall primarily into box 3, with some activities in boxes 2 and 4. Apart from some initial capital and a few key personnel, the major foreign-owned factors involved are management services and information, both of which can be combined with domestic factors without leaving the parent headquarters. Once one recognizes this, international banking seems less complex as an activity for analysis and the details in the paper can be seen in a more structured way. Not surprisingly, a major issue discussed in the paper is whether or not there are welfare gains from increased international banking. Formally, if we accept that international services are primarily in box 3, this question is analogous to the question posed in the international trade literature of whether or not there are gains from increased mobility of sector-specific factors when the sector involved is both nontraded and noncompetitive. This question is difficult to answer since we are quite obviously in a second-best world. However, we can make some general observations about the elements which would determine whether the answer is likely to be positive or negative. In particular, the source of the noncompetitiveness in the domestic market is crucial. For example, if the source of the distortion is government regulations, and the foreign banks are offering no additional services, then it is quite possible that increased foreign banking will reduce welfare. To the extent that new services are offered, there will be an offsetting welfare gain. On the other hand, to the extent that the source of the noncompetitiveness is due to cartelization and that this is reduced or abolished by increased international banking, then there is likely to be a welfare gain. Looking at the experience in the United Kingdom (Llewellyn 1986) and Ireland (McGowan 1986), for example, it seems clear that a considerable portion of the increase in competition in the banking sector in recent years and the pressure for deregulation is due to the growing presence of foreign banks. A further element in determining whether there are welfare gains from international banking is the treatment of taxation of the returns to the internationally-mobile factors. The issue of taxation of factor rewards and profits in international banking is a somewhat surprising omission from Neu's paper. In his paper, Neu analyzes the welfare gains of increased international banking by reference to banking as a traded service; in other words, he analyzes banking services as if they are box 2-type activities,

273

Banking Services

although he subsequently describes them in terms of box 3 activities.6 He argues that if foreign banks can produce banking services more efficiently than domestic banks, then there is a potential gain from international trade in banking services; in his opinion, a crucial determinant of whether or not this gain will be transferred to domestic consumers is the organization of the international banking system. In particular, he states, in his second observation, that "we must probably allow cross-border branching; subsidiaries will probably not do the trick." His argument is that unless there is branch banking, the characteristics of the parent bank are not exported, and hence the gains from international banking are not realized. I found his argument unconvincing, as both subsidiaries and branches have equal access to the major foreign-owned factors employed in banking, referred to above. Furthermore, I do not know of any evidence to support the view that subsidiaries have to pay higher interest rates than branches to secure funds, which might support Neu's contention that branches can transfer benefits to domestic consumers that are not transferred by subsidiaries.7 However, the issue of whether or not international banking takes places through subsidiaries or branches is extremely important in the context of the optimal size of banks from a global perspective. As Neu points out, one of the major concerns with increased international banking is that the economies of scale involved will most likely result in a few very large banks which governments may find themselves having to bail out if there is a major crisis. Since it is hard to imagine any agency which could effectively or sensibly regulate banks internationally, the only effective method of regulating international banks is good local regulation, and, for this to operate, international banking must operate through subsidiaries. Good regulation would ensure that neither Neu's third observation, that foreign banks may not bear the full costs of their expansions, nor his fourth observation, that international banks may become undesirably large, would be realized in practice.8 Good regulation benefits both home and host countries as well as the international bank itself, by reducing its monitoring costs. Unlike Richard Neu, Andre Sapir attempts to imbed his analysis of telecommunications in the general context of trade in services. In fact, 6. This view of trade in banking services is supported by St.-Hilaire and Whalley who conclude, on the basis of their empirical analysis, that "[s]ervice trade, as it relates to banking, may be an important investment issue, but does not appear to be that important a trade issue" (St.-Hilaire and Whalley 1986, 11). 7. In Ireland, for example, it would appear that foreign banks are able to compete on equal terms with domestic banks (see McGowan 1986). 8. There is also no evidence so far of an international cartel emerging as a result of economies of scale in international banking. In fact, as Andre Sapir suggests in his paper, it seems likely that lower technology costs will reduce the relative importance of economies of scale in future decades.

274

C. R. N e u

using the framework outlined above, it is possible to pinpoint the role which the development of information technology and telecommunications has played in the recent internationalization of the services sector. In effect, many of the services which would typically have fallen into box 1, i.e., strictly local services, have through the development of telecommunications and information technology during the past fifteen years moved into boxes 2, 3, or 4. The analogy here with the effect of the reduced cost of transportation on trade in commodities, as noted by Sapir, is very strong. Furthermore, future developments in telecommunications may well lead to more services moving between boxes. To take an example from the banking area, developments in telecommunications and computers may result in some banking services which are currently undertaken by local subsidiaries and branches being transferred to the bank's headquarters, involving the movement of activities from box 3 to box 2. Such developments will be influenced not only by developments in telecommunications but also by international and national regulatory activities, which may reinforce or counteract the pressures created by technological changes. In his general discussion of services, in addition to the two characteristics already noted by Hill (1977), Andre Sapir notes that intangibility is a further characteristic of services, where the intangibility arises from the difficulty of evaluating services ex ante.9 As a result of this intangibility, Sapir argues, "it is the principles of reputation and qualification that play an overriding role in the selection of suppliers." He goes on to argue, and I agree with him fully, that services characterized by such intangibility are likely to be traded more intra- rather than mterfirm, as the greater the degree of intangibility, the more difficult it is to trade with an "unrelated party." However, he subsequently suggests that if a country with the lowest production cost of a particular service is unable to export that service because it lacks reputation, the principle of comparative advantage might not apply. I have to disagree with his interpretation of this occurrence as a possible violation of comparative advantage, because he is not in essence comparing two identical services. Either one can look at this issue formally in terms of comparative advantage under uncertainty or one can interpret the reputation of a service as an attribute of the service (essentially internalizing the uncertainty) which means that the two services are not identical. Hence, while I agree with Sapir's point that low-cost centres may initially have a problem establishing a reputation for certain services (as indeed Japan and many of the NICs had with manufactured

9. Sapir develops the notion of intangibility from earlier work by Caves (1982) and Shapiro (1982).

275

Banking Services

products in the past three decades), I would not agree that this should be interpreted as a violation of comparative advantage. Returning to the crucial role that telecommunications may play in the expansion of international services, I should like to add two related points here in the context of Sapir's paper. First, and this applies to the telecommunications sector as well as other service sectors, changes in technology may quickly render regulatory negotiations obsolete, and regulatory mechanisms may themselves lead to particular technological developments. To view technological developments as random in this context could be dangerously naive, just as presuming that nontariff barriers might not emerge where once there were tariff barriers has proved to be in goods markets. Second, in the telecommunications field technological developments raise very interesting issues about our traditional concepts of natural monopolies. Where once it was common to think of certain industries as being, in some fundamental sense, natural monopolies, we realize that this description may not be appropriate in the longer term. Thus, while the basic infrastructure of telecommunications fits comfortably with our idea of a natural monopoly, the market for the range of services which such a system can facilitate may be more appropriately described as competitive. It seems to me that the phenomenon which Andre Sapir describes in the telecommunications sector may be much more widespread. In conclusion, my strong impression from these two papers is that GATT negotiators should be very wary about entering the arena of trade in services. I would, therefore, support the cautionery stance taken in Uruquay, as referred to in Andre Sapir's introduction. In almost every section of services, negotiators would have to deal, not only with international factor usage, where the factors may or may not move across national boundaries, but also, in many instances with domestic regulation. Rather than attempting to negotiate trade-offs between different countries' discriminatory regulations, time might be better spent attempting to convince governments of the benefits of sensible domestic regulation.

References Bhagwati, Jagdish N. 1985. Trade in services and developing countries. 10th Annual Geneva Lecture delivered 28 November at the London School of Economics. Caves, Richard E. 1982. Multinational enterprises and economic analysis. Cambridge: Cambridge University Press. Deardorff, A. 1985. Comparative advantage and international trade and investment in services. In Trade and investment in services: CanadalU.S.

276

C. R. Neu

perspectives, ed. Robert M. Stern, 39-71. Toronto: Ontario Economic Council. Hill, T. P. 1977. On goods and services. The Review of Income and Wealth 23: 315-38. Hindley, Brian, and Alastair Smith. 1984. Comparative advantage and trade in services. The World Economy 7: 369-89. Jones, R. W. 1985. Comment on Deardorff. In Trade and investment in services: Canada!U.S. perspectives, ed. Robert M. Stern, 72-76. Toronto: Ontario Economic Council. Jones, Ronald W., J. Peter Neary, and Frances P. Ruane. 1983. Two-way capital flows: Cross-hauling in a model of foreign investment. Journal of International Economics 14: 357-66. Llewellyn, David T. 1986. Financial deregulation—the British experience. Irish Banking Review (Autumn): 55-65. McGowan, P. 1986. Competition in Irish banking. Irish Banking Review (Autumn): 27-40. Malmgren, Harold B. 1985. Negotiating international rules for trade in services. The World Economy 8: 11-26. Markusen, James R. 1986a. Trade in producer services: Issues involving returns to scale and the international division of labour. October, University of Western Ontario. Mimeo. 1986b. Intra-firm service trade by the multinational enterprise. October, University of Western Ontario. Mimeo. Melvin, James R. 1985. Comment on Deardorff. In Trade and investment in services: CanadalU.S. perspectives, ed. Robert M. Stern, 77-82. Toronto: Ontario Economic Council. 1986a. Dimensionality and intermediation in economic analysis. October, University of Western Ontario. Mimeo. 1986b. Trade in services: A Heckscher-Ohlin approach. October, University of Western Ontario. Mimeo. St.-Hilaire, France, and John Whalley. 1986. Some estimates of trade flows in banking services. Working Paper. University of Western Ontario, Centre for the Study of International Economic Relations. Sampson, Gary P., and Richard H. Snape. 1985. Identifying the issues in trade in services. The World Economy 8: 171-81. Sapir, Andre. 1985. North-South issues on trade in services. The World Economy 8: 27-42. Shapiro, Carl. 1982. Consumer information, product quality and seller reputation. Bell Journal of Economics 13: 20-35.

Comment

Gary P. Sampson

While international service transactions (e.g., shipping, telecommunications, patents, finance) have long been the subject of intensive mulGary P. Sampson was employed by the United Nations Conference on Trade and Development at the time of writing. He is now Senior Counsellor, General Agreement on Tariffs and Trade (GATT), Geneva. The views in this paper are his own and not necessarily those of either institution.

277

Banking Services

tilateral deliberation, in recent years interest in this area has grown. At least one of the reasons for this is mentioned in both papers*: Richard Neu and Andre Sapir indicate that some—especially policy makers in the United States—consider international service transactions to be the next frontier for trade liberalization. Progress toward an agenda for negotiation, however, has been particularly slow, and both papers contain useful material in helping to understand why. For the purposes of exposition, it is possible to group some of the reasons under two headings. First, various initiatives (notably the United States' initiative at the November 1982 Ministerial Meeting in GATT) have attempted to deal with services as a sector rather than as individual activities. Second, the services sector has been raised in the context of "trade liberalization." Services as a Sector Dealing with services as a sector throws into sharp relief the problems of dealing with an extremely heterogeneous group of economic activities in a systematic manner. While some services are functionally related (e.g., telecommunications is an input into banking), others such as neurosurgery and the provision of party rental facilities have little in common. Maritime transport and fast food franchising are very different economic activities with very different production functions. Treating services as a sector led to a search for characteristics that would permit what are essentially heterogeneous activities to be dealt with in a common analytical framework—something that is not possible by simply grouping them as "invisible" or "residual" items as has been done on past occasions. Various characteristics have been identified and Andre Sapir, for example, pays particular attention to the nonstorable nature of services. Because of this, the production and consumption of services must generally take place in the same location at the same time. This leads to an important conclusion—the production and consumption of services frequently requires mobility of the factors producing the services or of the consumers consuming them. In this sense, services and goods are different. A further extension of the conclusion is that controls on international services transactions (or "barriers to trade in services") come from restricting the movement of the factors producing the services and/or the consumers who receive the services. Sapir also addresses a related question: namely, whether the principle of comparative advantage holds for this diverse group of activities in the same way as it holds for goods. After a brief review of the literature Sapir concludes that in this respect there is no difference between services and merchandise trade. *Chapters 9 and 10 in this volume.

278

C. R. Neu

When looking at services (individually or as a sector) it is clear that further complications arise as there is a dearth of systematic quantitative information on international services transactions. The authors make this plain in their case studies of the banking and telecommunication industries. With respect to measuring the value of transactions, Richard Neu raises the very important question of what is in fact being traded when an international service transaction takes place. If the sale of the banking service is done via an "on-ground" presence in the importing country (e.g., through branch banking), there are important conceptual and practical difficulties in measuring the value of the service provided. We do not know, for example, what share of value added (or "retained value") is accounted for by returns to factors of production engaged in the importing and exporting country. At the same time, Neu throws some cold water on the hopes that an international expansion of banking services will create jobs in the exporting country; in his view, exporting banking services may mean that jobs are created by the foreign bank mostly for foreigners. Liberalizing Trade in Services The second group of reasons explaining the protracted nature of discussions is that they are taking place in the context of trade liberalization of services activities. For tariff liberalization in goods—as carried out in the postwar GATT rounds of multilateral trade negotiations—life was relatively uncomplicated. Ad valorem tariffs can be compared across goods. An important complication arises with the services sector when discussing "trade liberalization" in the traditional GATT sense of an exchange of "concessions." Should such concessions be exchanged across activities (e.g., market access in banking against market access in telecommunications) or should they be within the same service activities? As Richard Neu notes, the idea of exchanging preferential market access for different foreign banks operating in the same market is something that would prove to be unmanagable. In addition, unlike services, goods pass through customs houses, and tariff barriers are very apparent. As mentioned above, however, international services transactions frequently require the movement of the factors of production and receivers of the services; trade liberalization in services involves changing national rules and procedures. Services include a number of "sensitive" activities with national laws relating to their production and distribution—banking and telecommunications are two examples. For activities of national importance, intergovernmental discussions are sure to be intense and protracted. Andre Sapir, for example, notes that the governments of the EC and the United States have very different perceptions as to how the tele-

279

Banking Services

communications industry should be structured. The EC assigns a special importance to telecommunication activities, and national regulation has resulted in a state monopoly. The United States government has a different perception and private firms operate in an increasingly competitive market. The U.S. government promotes competition and recent technological developments are such that competition will be further promoted in the United States. The interesting question which Sapir poses is whether with these technological developments it will be possible to contain the monopoly in the EC. Richard Neu also pays particular attention to the sensitive nature of banking and, as he says, since countries have legitimate national needs (in regulating banking) and perceptions about what is in their national interests—painstaking negotiations would be necessary to reconcile these legitimate national needs. Whether sensitive activities or not, liberalization of services trade involves difficulties. In looking at these, it is useful to consider three groups of services. First, there are services that do not, because of their nature, cross national boundaries. These must be consumed in the importing country by the receiver of the service. Andre Sapir mentions the PTT in Europe; here for example, local mail delivery services can only be traded internationally if there is international movement of factors of production. Other examples include the construction of roads and other public works such as the provision of municipal sewerage facilities. Fields (the receiver of the service) cannot be shipped abroad to be ploughed. But there can be international transactions in these activities if there is movement in the factors of production; labor or capital must cross the border for the service to be produced in the country of consumption, and must face the national laws relating to foreign investment and the granting of work visas. A further example is the provision of branch-banking services. Richard Neu gives some interesting insights into what he calls multinational banking (as compared with international banking), which requires an4 'on-the-ground'' presence for the local market to be served. Here, national policies have in some countries explicitly prohibited the establishment of foreign-controlled banking institutions; rights of establishment have not been granted to foreign banks. Neu notes, however, that since the establishment rules in most industrialized countries have been liberalized, the real problems in these countries now arise from policies that have the effect of discriminating against the operation of foreign-controlled banking institutions. This is the problem of national treatment for foreign banks operating in the domestic market. Richard Neu gives the example of banking services which could potentially be supplied to foreign residents from offices in the bank's home country if there were not national rules which discriminate spe-

280

C. R. Neu

cifically against foreign operations. Firms in the foreign country may be prohibited from carrying out foreign exchange business, for example, as their potential clients have to conform to national rules and hold and exchange foreign currency with their own central bank. Thus, where factor movement is important for international service activities, "right of establishment" and "national treatment" emerge as major issues. Restrictions on factor movement are imposed by means of the issuance of work permits and constraints on the behavior of foreign capital. Foreign Investment Review Boards and Departments of Immigration are notoriously sensitive institutions when it comes to deregulation, and progress in this direction can therefore be expected to be slow. A second class of services includes those that can be produced with or without factor movement. Presumably, a bridge to be constructed in Australia can be equally well designed in Sydney or San Francisco, but the life of the consulting engineer may be considerably simplified if he can visit the site of construction when necessary or perhaps even install an office. So, too, for computer programmers wishing to talk to their clients, and insurance companies hoping to sell insurance to theirs. There are many banking and telecommunication activities that are facilitated through a local presence in the importing country. Thus, in such cases, if there are restrictions on the movement of factors, the provision of the service may not be prohibited—it is just rendered more difficult. Finally, for an international transaction to take place in some services, it is necessary to have mobility of the receiver of the service. The Taj Mahal cannot be shipped to California to be viewed by Americans any more than Disneyland can be shipped to India. There must be mobility of the receivers. So, too, for university students who study in foreign countries, and aircraft which are shipped internationally for specialized repairs. Here the constraints on trade are again national regulations—tourist visas granted by the government of the country exporting the service and foreign exchange controls for residents of the country of import. Both papers could be improved if they contained better quantitative material. As noted above, data are hard to find, but they are not nonexistent. There are no statistics in the paper by Richard Neu and some of the quantitative material in the paper by Andre Sapir could be improved. The market access question for telecommunication products would have been more informative if the author had presented more detailed information on the nature of the goods traded internationally, the direction of trade flows, and the tariff and nontariff barriers confronting such trade. Such information would assist in forming an opinion as to whether trade liberalization in these goods should be a priority matter for any forthcoming trade negotiations.

281

Banking Services

The authors conclude by offering suggestions as to what all this means for any future trade-liberalizing negotiations in banking and telecommunications. Richard Neu concludes that the next round of trade talks will have a better chance of being productive if the "thorny" question of how foreign banks should be allowed to operate in national markets is downplayed. Based on the evidence, this seems to be a very sensible statement and one wonders to what extent this conclusion applies to a variety of other thorny services activities. Andre Sapir's conclusions appear no less reasonable. He notes that the focus of future international negotiations will be productive if they are directed toward establishing a greater degree of understanding between governments with regard to their "regulatory philosophies" and toward a "harmonization of national attitudes." He is of the view that such negotiations are best carried out between governments with common economic situations and principles; as far as the EC and the United States are concerned, the appropriate forum is the OECD. There is of course merit in this, but decisions that are taken on the nature of international telecommunication activities in future years will have effects that transcend these economies. The interests of producers and consumers of services deserve to be represented in such negotiations, even if their countries are not members of the OECD.

11

High-Tech Trade Policy Kala Krishna

11.1 Introduction With some 90 nations about to embark on a new round of multilateral trade negotiations under GATT (General Agreement on Tariffs and Trade), there is concern that rules for "high-tech" industries be high on the agenda. The United States has been the leader in expressing this view because it perceives that the products of its high-tech industries will have (or could have) a long-run comparative advantage. Aho and Aronson (1985, 44) point out that "[t]he United States pushed for work on high technology at the 1982 GATT Ministerial Meeting, but did not convince other countries that high-technology industries should be handled any differently from other industries." In fact, they say that "the initiative was so poorly defined that LDC (less-developed countries) representatives asked how high-technology discussions could be related to transfer of technology, which is a legitimate question but not what the United States had in mind." More recently, in March 1984, the U.S. government officially called for new GATT negotiations in the area of high-tech goods. The United States informed the GATT Council meeting at that time that it would begin bilateral trade talks on hightech goods with interested delegations. These were to be the basis for future council discussions. Other countries, however, questioned the urgency of these negotiations. Why is it useful to think of high-tech trade separately, and why has the interest in trade policy for these industries been on the upswing? The two questions are closely related. There are two reasons for this Kala Krishna is assistant professor of economics at Harvard University and a faculty research fellow of the National Bureau of Economic Research.

285

286

Kala Krishna

increase in interest. First, trade in these products has been a growing part of total trade. Second, a number of high-tech industries possess certain characteristics that provide opportunities for currently legal (under GATT rules) and profitable unilateral trade policy that could be harmful to all parties in the international arena if practiced by all parties. In other words, these characteristics make the industries "fall between the cracks" of trading arrangements that are meant to prevent such suboptimal situations from arising. The importance of high-tech trade and the sector's special characteristics indicate that policy for this sector should be thought of separately and its special characteristics taken into account. US-EC trade currently totals about $108 billion a year. A breakdown of the leading items exported to and imported from the European Community can be found in the tables of the USITC publication "Operations of the Trade Agreements Program," 37th Report, reproduced at the end of this paper as tables 11.1 and 11.2. It remains to identify the characteristics that make some high-tech industries special. The first characteristic is that network externalities play a significant role in determining the demand for the products of these industries. Network externalities are said to exist when the utility that a user derives from consuming a good and the user's willingness to pay for a good increase with the number of people who also consume the good or are expected to do so. These externalities arise in a number of ways, both directly and indirectly. In communications equipment, such as telephones, they arise directly. People derive a greater benefit from a phone if all the people they wish to communicate with also possess a phone. They may arise indirectly, for example, if the amount of software produced is related to the number of computers in use, so that increases in the number of computers would increase the available stock of software which in turn would raise peoples' willingness to pay for a computer. Such network externalities arise for many products that are not "hightech"—for example, they arise for many durable goods since the availability of servicing for any durable good is likely to be related to the number of units of the good already sold in the market.1 The network externalities are especially important, however, in such key high-tech industries as computers and telecommunications systems. In part, this is because these and a number of other high-tech goods are informationrelated, and goods with this characteristic tend to have greater network externalities. For industries in which network externalities are important, expectations about the size of the network, i.e., the number of units of the good sold, are a major determinant of the demand for such goods.2

287

High-Tech Trade Policy

This paper analyzes how network externalities and expectations about their size can provide a special role for trade policy. The conditions for such a role to exist are shown to be quite severe. Other characteristics of high-tech industries that may also serve as the basis for government intervention in international trade are briefly mentioned at the end of the paper. Recent work in international trade theory has led both academics and policy makers to a better understanding of trade policy in imperfectly competitive markets.3 It is now understood that to the extent that national interests do not include the well-being of foreigners, in particular of foreign firms, there may be a case for trying to draw away the profits of foreign firms. This can be done directly, or if this is illegal, indirectly, by altering the behavior of domestic firms in order to improve their strategic position. Attention has focused on the use of taxes and subsidies for such purposes. This point was first made in Brander and Spencer (1984) who showed that in a particular strategic setting government subsidization of a domestic firm competing with a foreign firm in a foreign market would improve the domestic firm's strategic position if no other governments attempted this as well. As might be expected, this conclusion aroused a good deal of interest in many circles. It has become clear through recent work4 that the appropriateness of this kind of policy depends very much on the nature of the strategic setting. Loosely speaking, if the firms5 choose their actions on the basis of incorrect beliefs about the actions of other firms and the government understands how these are incorrect and can precommit to tax/subsidy schemes, there is a role for government policy to correct the distortion arising from the incorrect beliefs of the firms. If, for example, a domestic firm expects its foreign rival to keep sales constant while it increases its sales, and in fact its foreign competitors tend to reduce their sales whenever this happens, a wedge exists that can be exploited by a knowledgeable government. In this case the government would give the domestic firm a strategic advantage by subsidizing it. In contrast, if the foreign competitor increased its sales with those of the domestic firm, the optimal policy would be imposition of a tax. The existence of consumer expectations about network size creates another possible role for government policy. In industries where expectations of network externalities are important, firms may believe that they are unable to influence expectations. This may be an incorrect assumption. If, for example, firms believe that an increase in their output will not increase the expected network size on which consumers base their purchasing decisions, while consumers really do adjust their expectations about network size in response to output changes, then

288

Kala Krishna

there is a role for a government that knows the true relationship to offset the incorrect beliefs of firms with a subsidy that would induce the firms to produce more. Another possibility, that in view of existing GATT rules is perhaps more disturbing, relates to the existence of strategic multimarket interactions.6 Such interactions are bound to be important in those hightech industries where expected network size affects demand in all markets, domestic and foreign. There is a possible role for the government in trying to shift profits from foreign to domestic firms by an appropriate policy for domestic firms exporting to foreign markets. Under these conditions a purely domestic policy, such as a tax/subsidy on consumption for the domestic market only will have repercussions for the firm's behavior in foreign markets. The basic idea can be understood quite easily with a simple example. Consider a situation where the individual firm takes expectations about network size as given. The firm is assumed to compete in both foreign and domestic markets. Network expectations adjust to network size so that in equilibrium the expected network size is the actual one. If the government subsidizes only national sales of the product, the firm will produce more.7 This, in turn, will raise expectations about the network size of the product at home and abroad, and therefore affect the product's international competitiveness. This natural interdependence of markets via network effects could be used strategically by a government to affect a firm's behavior in international markets in order to promote the national interest.8 This possibility is particularly disturbing as trade laws are designed to deal with direct government subsidization of exported goods, but are not framed in a way to deal with purely domestic policies that have indirect international effects. Under GATT rules, countervailing duties are allowed if there is a foreign subsidy on a product and imports of the product cause material injury or retardation of growth to the domestic industry producing the same good. Thinly disguised subsidies to exports are prevented under this clause. An example is the "x radial steel-belted tires" case in which Canada subsidized a tire company to build a new factory in an area with high unemployment. Since virtually all of the product was to be exported and thus likely to cause material injury to U.S. competitors, it was found to be countervailable. The countervailing duty provisions clause cannot be used, however, in the case of "purely domestic" policies, i.e., where the subsidy applies only to goods consumed domestically. This leaves only the last resort, the escape clauses, as the means to deal with such policies. These provisions can be used if increased imports cause or threaten serious injury to domestic producers or if increased imports threaten national security. The requirements for obtaining protection by these two routes

289

High-Tech Trade Policy

are quite stringent, however. Thus, "purely domestic" policies with favorable effects on international competitiveness could look quite attractive as a means of promoting exports. The next sections present a simple model that helps isolate and interpret the effects of network externalities and expectations on optimal trade policy. Two versions of the model are considered in the next two sections. Section 11.2 examines the implications for profitshifting possibilities of the two previously mentioned characteristics of high-tech industries, network externalities and expectations about them. Section 11.3 contains an analysis of the nature of multimarket interactions that occur because of these characteristics and the implications of these interactions for trade policy. The last section discusses some other characteristics of high-tech industries and the problems they may and do cause for trade policy. 11.2 Expectations, Network Externalities, and Profit Shifting Recent work on the role of government policy when oligopolistic firms operate in international markets has focused on the possibility of profit shifting from foreign to domestic firms and on the trade-off between the gains to domestic firms from such profit shifting and the losses to domestic consumers. A number of recent papers have developed this idea. One of the earliest papers is by Brander and Spencer (1984), who show in a model with one home firm and one foreign firm acting as Cournot duopolists and competing in a third market, that the optimal government policy is to subsidize exports. Dixit (1984) extends this result to cases with many firms and shows that the same result holds as long as the number of domestic firms is not too great. Eaton and Grossman (1986), in an insightful paper, show that the Brander and Spencer result can be interpreted as a special case of a more general policy. Their basic interpretation is in terms of a government acting in response to differences between conjectured and actual responses of foreign firms to changes in the domestic firms' output when they are competing in a third market.9 They also develop a number of extensions, including allowing for many firms and domestic consumption. As argued previously, demand for a firm's product depends on the expected size of the "network," that is, the expected number of units of compatible goods sold. These expectations are very important in industries with network externalities. Government intervention may be useful in influencing these expectations. This idea is explored in this section using a model that embeds the Eaton and Grossman model in it.10 Use is made of the simplest version of the model that makes the point, namely, the case where a domestic and foreign firm compete in a third market and there is no consumption of the product in either the

290

Kala Krishna

home or foreign country. Extensions along the lines of allowing more firms or domestic consumption are possible, and the interested reader can use Eaton and Grossman (1986) as a guide to do so. The next section analyzes multimarket interactions with firms competing in both the domestic and foreign markets. When network externalities exist, a consumer's willingness to pay for a product depends on the expected network size for the product. The products of the two firms could be compatible or incompatible with each other, and are substitutes for each other. If they are compatible, expected network size is the expected total output of both firms; if incompatible, the network is the expected output of each firm. The inverse demand functions for the home and foreign firms are given by P(x,x*,N^) and P*(x,x*,N*E)y where x,x* and NE, N*E are the outputs and expected network sizes of the home and foreign firms, respectively. The superscript E denotes that these are the expected levels of these variables. If the products are compatible, NE = N*E = xE + x*E. If they are incompatible, NE = xE and N*E = x*E. Of course, price, P, falls as x or x* rises since the products are substitutes, and rises with increases in the expected network size. Similarly, P* rises as N*E rises and falls as x or x* rises.11 Consumers are assumed to base their demand for a firm's product on the expectations they hold about a firm's network size. Firms take these expectations as dependent on their output. They assume that a unit change in x(x*) creates an e(e*) change in the expected domestic (foreign) output with consequent effects on expected network size. The special case where expectations about network size do not change as the firm's output changes arises when e and €* equal zero. However, in general, expectations are perceived as being affected by a firm's output. In addition, firms have conjectures about how their opponent will react to changes in their output. 7 and 7* denote the conjectural variations parameters of the domestic and foreign firms. Finally, expectations about network size must fulfill a consistency condition that allows them to be tied down. A natural condition is that expectations about the network sizes of the firms are fulfilled in equilibrium. This defines a "fulfilled expectations equilibrium." It is useful expositionally to be slightly more abstract at this time. An (e,e*) fulfilled expectations equilibrium is said to occur when the expected domestic output equals e times actual domestic output, and expected foreign output equals e* times actual foreign output. This corresponds to a fulfilled expectations equilibrium if both e and e* equal one. The profits of the domestic and foreign firms are given by (1) and

(1 - t) P(x,x*,NE)x - c(x)

291

(2)

High-Tech Trade Policy

F*(x,x\N*E)x* - c(jt*),

respectively. t is the tariff or subsidy imposed on the domestic firm, and c(.) and c*(.) are the cost functions of the two firms. The first order conditions when the networks of the two products are incompatible are given by (3)

(1 - t) [(/>, + 7^2 + eP3)x + P] - c' = 0,

(4)

(Pft* + P*2 + e*P*3)x* + P* - c*' = 0,

where subscripts denote partial derivatives. The subscript 1, for example, identifies the first variable as being the relevant one. If the products are compatible, they are given by (3')

(1 - t) {[Px + 7^2 + (e + e*7)P3]x + P} - c' = 0

(4')

[Pft* + P*2 + (e* + 7*e)]jc* + P* - c*' = 0.

In addition, the (e, e*) fulfilled-expectations equations, (5)

xE = e x

(6)

x*E = e*x*

must be met. Substituting (5) and (6) into (3) and (4) or (3') and (4') gives two equations to solve for the only two variables, x and x* in these equations. Equations (5) and (6) then define expectations. Therefore, the effect of a change in t on the endogenous variables x and x* in such an equilibrium can be found by performing comparative statics analyses on (3) and (4) or (3') and (4'), after substituting (5) and (6) into them. Let the actual change in x* as x changes be given by g. This is defined by equations (4) or (4'), after equations (5) and (6) have been substituted into them. It also equals the ratio of the comparative statics dxVdt terms ' ibdit • The problem is now fully specified. Firms maximize their profits taking any taxes and subsidies by the government as given. These firstorder conditions for thefirmsdefine their best-response functions which give two equations in JC, x*, xE, and x*E. The condition that the equilibrium be an (e,e*) fulfilled expectations equilibrium gives another two equations in these four variables. This allows solving for the endogenous variables x, x*, xE, and x*E. The question of optimal government policy can now be analyzed. Since the firms compete in a third market, welfare consists only of domestic profits,12 and is given by W = P(x,x*,NE)x - c(x). t does not affect welfare directly, since transfers between the government and firms cancel out, but it does affect welfare by its effect on the endogenous variables. Hence,

292

(7)

Kala Krishna

dW — = [(P, + P2g + P3e)x + P -

dx c']-

if the products are incompatible, and (7')

dW dx — = {[/>, + P2g + P*(e + g O l * + P ~ c'} —

if they are compatible. Using the first-order condition (3) in (7) and (3') in (7') allows the welfare changes to be expressed as (8)

dW [rn, ,-, x -^ = [[P2(g - y) + P3(e - e)]x

+

t ~\dx — C j -

if the products are incompatible, and by

(8')

f = ( {/>2fe - *> + P3[^ - £) + (ge* - 7e*)]}x + Y^rf Wt

if the products are compatible. The first-order conditions for an interior welfare maximum require (8) or (8') to be zero depending on whether the products are incompatible or compatible. This gives the optimal policy assuming that second-order conditions are met. Notice that the optimal value of t depends on the direction of both the wedge between conjectured and actual responses of the foreign firm to own output changes, (g - 7), a la Eaton and Grossman, as well as on the difference between the conjectured change in expected network size in response to own output and the actual change. This difference depends only on (e — e) if products are incompatible, but on the additional interaction term (ge* - ye*) as well when products are compatible. Consider first the case when products are incompatible. The difference between g and 7 defines a reason for the government to set t ¥^ 0. If g < 7 and e = e, the government should subsidize the domestic producer. This is because the domestic firm is too pessimistic in its conjecture about the foreign firm's behavior. If, for example, 7 = 0 and g < 0, the usual Cournot case, the domestic firm acts on the assumption that the foreign one will keep its output fixed. However, since its opponent actually reduces its output in response to any increase in domestic output, the domestic firm should produce more. The government can ensure this by subsidizing the domestic firm, as was pointed out by Eaton and Grossman. The interpretation of (e - e) is similar. If e > e, the domestic firm should be subsidized since the firm conjectures a smaller change in the

293

High-Tech Trade Policy

expectations of consumers about network size in response to output changes than actually occurs. This leads them to produce too little. If, for example, firms take the expected network size as given, and a fulfilled-expectations equilibrium is considered, there would be reason to subsidize the domestic firm. Expectations are likely to be fixed in the short run, and a fulfilled-expectation condition can be thought of as a static way of incorporating the longer run into the model.13 If firms tend to take a short-term view of the industry, as is implicit in e = 0, there is a role for a government that is aware of the long-run consequences to take appropriate action. The interpretation of the case with compatible products is similar. The only difference lies in the fact that any effect of a tariff via the network externality must include the effect on the foreign firm's expected and actual output as well as that of the domestic firm. Notice that e + ge* is the actual change in the network size, while e + ye* is the conjectured change in the network size. This makes e and 7 interact in the formula, but allows the same interpretations to be made as done previously. It is worth calling attention to a few points at this stage. First, it should be noted that g is the slope of the foreign firm's best-response function after imposing the expectations condition on it. In this it differs from the analogous concept in the absence of any expectations. Therefore, even if e = e, the introduction of expectations would tend to make the assumption g = 7 even stronger since g is not the slope of the best response function of the foreign firm for given expectations but when expectations are fulfilled in the (e,e*) sense. Second, if € = 1 = e and a firm fully takes into account the effects of its actions on the network size, then in the case with incompatible products the direction of optimal policy depends on the sign of (g - 7) only. Also notice that the direction of the optimal policy when e = e = 1 = e* = e* in the case of compatible products is dependent on more than just the sign of (g - 7). If P2 + P3 is positive, even if P2 is negative, the goods would become effective complements because of the presence of network externalities. In this case the incentives to subsidize would, of course, be reversed since a decrease in output by a competitor hurts rather than helps a firm when the goods are complements. Thus, if e = e* = e* = e = 1, P2 + P3 is positive, and g < 7 = 0, as in the Cournot case, the anticipation that the other firm would keep its output fixed would be too optimistic if it really lowered it. Obviously, the optimal policy would then be a tax which would correct the firm's over-optimism. Third, note that having compatible products tends to reduce the desirability of a subsidy or a tax when g = 7 < 0. Consider, for example, the case where e = e*, e = e* but e > e and g = 7 < 0. The direct effect of a subsidy is beneficial since the firms underestimate the

294

Kala Krishna

value of the network externality. This is captured by the P3(e - e) term. A subsidy, however, raises the foreign firm's network as well and tends to make it wish to produce more. If g < 0, this makes the domestic firm produce less. This is captured by the term P3g(e* - e*). This works against the desirability of a subsidy. Thus, network externalities and expectations regarding the size of the network affect optimal trade policy for oligopolistic industries in two ways. First, the presence of expectations effects creates a wedge if there are differences in the way the network externalities really work, as given by e and e*, and the way they are expected to work, as given by e and €*. This wedge creates a role for government taxation/subsidization to correct the "distortion." Second, when the goods are compatible the existence of network externalities in the absence of any expectations distortions tends to make goods complementary. If in fact this effect dominates, the direction of optimal policy is reversed. In the next section, the importance of network externalities is shown to create a reason why success in one market can help bring about success in another. The analysis also provides an example of the idea that import protection can act as export promotion, as in Krugman (1984). 11.3 Multimarket Interactions 11.3.1 Introduction Multimarket interactions are said to exist when decisions made in one market spill over into another by affecting optimal decisions there. The existence of such interactions is important for two reasons. First, it creates problems in identifying unfair trade practices since purely "domestic" policies could actually be trade policies in disguise. Second, such interactions are bound to exist in markets with network externalities, both because markets are linked by common networks and because expectations about network size are a determinant of demand and can be affected by domestic government policy. The topic of multimarket oligopolies has aroused a great deal of interest recently. The work of Bulow, Geanakoplos, and Klemperer (1985), Fudenberg and Tirole (1984), Krugman (1984), and Baldwin and Krugman (1986) is closely related to this part of the paper. Baldwin and Krugman (1986) focus on an important technological aspect of the production of 16K RAMS: the fact that experience lowers the effective cost of production. This creates multimarket interactions that are made important by the extremely large experience effects estimated to occur in this industry. They develop a model that is rich enough to capture the particular aspects of learning-by-doing in this industry, and then

295

High-Tech Trade Policy

use their theoretical structure in a simulation exercise to see if the effective closure of the Japanese market to imports played a critical role in developing Japanese superiority in this area. Their results indicate that this closure did indeed play a critical role. In this spirit, any domestic policy that raised Japanese domestic output of semiconductors could have significant trade effects. Bulow, Geanakoplos and Klemperer (1985) examine many examples of multimarket interactions. The core of their work is their definition of "strategic complements and substitutes" and they show that these play a crucial role in the results of many oligopoly models.14 The approach of Eaton and Grossman (1986) can also be extended, as is done in this paper, to better understand such interactions in markets with the characteristics previously mentioned. This approach provides a way of understanding multimarket interactions that is complementary to that of Bulow, et al. (1985). It focuses on the wedges created by differences between how firms expect the relevant variables to affect profits and how they actually do so. These wedges create an opportunity for possible government intervention. While many possibilities for multimarket interactions have been discussed, network externalities and expectations concerning network size have been neglected. In what follows, this paper will first discuss informally how such network externalities make "success" across markets positively correlated and then examine in a more formal way the differences in the roles played by expectations and network externalities in such interactions. 11.3.2 Network Externalities, Expectations, and Market Interactions Both the existence of network externalities and expectations about them create linkages between markets and give rise to the possibility that government policy in one market can affect a firm's competitive edge in another. Consider for example a firm operating in two markets, a home and a foreign market, and facing the same competitor in both of these markets. Assume for concreteness that the two firms are Cournot multimarket duopolists who produce incompatible products so that each firm's expected network size is the size of its expected output in all its markets. A fulfilled expectations equilibrium15 where firms take expectations about network size as given will be considered. Equilibrium is therefore characterized by each firm's maximizing profits by choosing its sales in each market, taking as given its competitors' output and expectations about network size. In order for expectations to be fulfilled, in equilibrium each firm's total output must equal the expected network size as well.

296

Kala Krishna

Any subsidy on domestic sales will tend to raise total output of the domesticfirmand lower the total output of the foreign firm. In a fulfilledexpectations equilibrium, greater output will lead to greater expectations about the network size of the domestic firm and smaller expectations of the foreign network size. This will help the domestic firm and hurt the foreign firm in both markets. Notice that domestic policy has international effects because of the role played by expectations. The same kind of argument works even iffirmsdo not take expectations as given but assume they are equal to their total output, since greater sales in the domestic market raise marginal revenue in the foreign market because of the effect of network externalities. Network externalities by themselves also create multimarket interactions. The preceding argument can be made clearer by using a series of diagrams to illustrate the process (see figure ll.l). 1 6 It can be verified that the same kinds of effects occur even when the effects of output changes on expectations are taken into account by firms. Some notation is required at this point to help follow the diagrams used to illustrate the process by which these multimarket interactions occur. Since the products are incompatible, each firm's network consists of its total expected output in the two markets. The firm's problem is, therefore, to Max Tr(x,y\x*,y*9NE) = r(x,x*,NE) + R(y,y*,NE) - c(x + y), x,y

where x,y are output levels of the domestic producer in the two markets; x* and y* are the competitor's outputs in the markets; and c is the constant marginal cost of the home firm.17 NE is the expected network size of the home firm. The home firm's revenue functions in the two markets are r and R and are assumed to have the usual properties, r2, f\u rn < 0 and r3 > 0. It is also assumed that R2, R22, Ru < 0 and R3 > 0. The home firm takes x*,y*, and NE as given. This is denoted by the profit function, IT, being conditional on given values of these variables. Similarly, the foreign firm's problem is Max TTm(x\y*\x,y,NE) = r*(x,x\N*E) + R*(y,y\N*E) x*y*

- c (x + y ). Therefore, equilibrium in the home market (depicted in figure 11.1(a)), where the revenue functions of the home and foreign firm are given by r and r*, is characterized by the intersection of the best response functions, b, b*, implicitly defined by TTX = 0 and ir** = 0. Similarly, equilibrium in the foreign market (shown in figure 11.1(b)) is defined by iry = 0 and TT** = 0. Both these equations hold at the intersection of B and B* which are defined by these equations. The usual stability

N*(N*E)

NE'

\

\N* E (N E )

\

^NE'

^E

\ . (e)

Fig. 11.1

,.*E

298

Kala Krishna

conditions are also assumed in figures 11.1(a) and (b). These are equilibria only for the given expectations, however. In addition, these expectations must be fulfilled. For a given N*E, it is easy to verify that increases in NE shift the home firm's best response function outwards in both markets and so raise its equilibrium output in both markets. Therefore total output, N, of the domestic firm rises with NE as shown in figure 11.1(c) by NiN*) = NE. What happens to A as N*E changes? Since an increase in N*E shifts B* and b* outward, it must reduce N for any NE. Thus, an increase in N*E shifts NiN^ inwards so that the NE that is self-fulfilling falls. This relationship between the self-fulfilling NE and N*E is depicted in figure 11.1(e) as N^N*^. Since the position of the two firms is symmetric, the same arguments produce another diagram, 11.1(d), which is analogous to 11.1(c), and another function, AT^A^), depicted in 11.1(e), which for any NE gives the expectation of N*E that is self-fulfilling. The intersection of these two loci, W^CN^) and N^N*5^, gives the set of expectations about NE and N*E which are jointly self-fulfilling. Once again, the relative slopes of these two functions are as shown for stability reasons. How would a subsidy program for production for the domestic firm affect a firm abroad? The direct effect of a subsidy with given expectations about network size would be to shift the best response function at home outwards to b'', as shown in figure 11.1(a). This raises total output of the domestic firm for the given expectations. In fact, it is easy to show that total output rises for any given expectations. Thus, for any N*E, NiN^ would shift to the right, to a line such as N' in figure 11.1(c). Hence, for any N*E, the self-fulfilling expectation of NE would rise, or in other words, N^N*^ would shift outwards to a line such as NEf, in figure 11.1(e). Moreover, since the subsidy reduces total foreign output with the original set of expectations, it must reduce total foreign output for any given set of expectations, or in other words it must shift AT^A^) inwards to AT' as in 11.1(d). This reduces the self-fulfilling expectation level of N*E for any NE, so that N*E(NE) shifts in as well to N*E', which is shown in 11.1(e). Both the shifts in figure 11.1(e) raise the jointly selffulfilling level of NE and lower that of N*E. This shifts demand for the domestic product out and the foreign product in in both domestic and foreign markets, shifting the equilibrium in these markets to the points / a n d Ffrom a and A in 11.1(a) and 11.1(b), respectively. A subsidy to the home firm in the home market raises expectations about its network size, which raises its output in all markets. A similar effect could easily be demonstrated even if firms completely take into account the effects of output changes on expectations, because the size of the network still connects the two markets.

299

High-Tech Trade Policy

Although this exercise shows how expectations and network externalities can link markets and how a subsidy to domestic sales can raise foreign and domestic sales, it does not say much about whether this subsidy is desirable. Fortunately, the model developed in the previous section can be extended to help answer these questions. The only modification that needs to be made is to allow the firms to compete in two markets. The case where firms make incompatible products is analyzed first. The results for compatible products are similar and are presented more briefly later on. The domestic (foreign) firm is assumed to behave as if it believed that a unit change in the domestic (foreign) network would lead to an e(e*) change in its expected network. 7 and T are the conjectural variations parameters for the domestic firm in the home and foreign market, and 7* and T* are the conjectural variations parameters for the foreign firm in the home and foreign markets, r and R are the revenue functions of the domestic firm in the domestic and foreign markets, respectively. r* and R* similarly denote the revenues of the foreign firm in these markets. An (e,e*) fulfilled expectations equilibrium is analyzed as before. Domestic consumption of the domestic firm's output is taxed or subsidized by the government at the rate t. Hence, the profits, TT, of the domestic firm are given by IT = (1 - t)r(x,x*,NE)

+ R(y,y\NE)

- c{x + y),

and TT*, the profits of the foreign firm, are given by TT* = r*(x,x*,N*E) + R*(y,y*,N*E) - c*(x* + / ) . The four first-order conditions are given by (9)

T 1 = 0 " 0(ri + 7>2 + *r3) + € * 3 - c = 0 dx

(10)

^ 7 = r\ + 7>I + e>; + e*R*3 - c* = 0

(11)

— = R{ + TR2 + eR3 + er3(l - i) - c = 0 dy

(12)

dX

^

dy

= R*2 + r/?; + e*R; + ev; - c* = 0.

The first two equations define equilibrium in the home market given y*,x*, and the second two define equilibrium in the foreign market given x,y. The condition that expectations be (e,e*) fulfilled requires that in addition NE = e(x + y) and N*E = e*(x* + y*). These conditions will give the equilibrium levels of x, x*t y and y* as a function only of t.

300

Kala Krishna

Also, comparative statics on the system will give the effects of changing t on these endogenous variables. These comparative statics results are dx dx* dy , dy* , ,, , . 1 denoted by —, ——, —, and —— and can be round by substituting & for dt dt dt dt the expectations conditions in equations (9) through (12) before performing the comparative statics exercise. For simplicity, the welfare function is broken into two components, consumer surplus, Wc, and profits, W7*. The effects of t on the two components are analyzed separately. This isolates the strategic multimarket profit-shifting effects, so that they can be analyzed clearly. Notice that in (13), welfare depends on / only indirectly via the effect of t on the endogenous variables. (13)

(14)

W* = r[x, x*, e(x + y)] + R[y, y\ e(x + y)] - c(x + y). dW" , —- = ( r , + dt

dx )_ + dt dy + (R{ + R3e + r3e - c)-£ + r3e

+ R3e _

c

dx* (ry— dt dy* (R2)-j^

Using equations (9) and (11) we can substitute for rx - c and Rx - c in equation (14). Also, let g be the actual change in x* relative to the actual change in x as t changes. In other words, g = „ (15)

, . . . Similarly, dxldt

dy*ldt dyldt ^ . J , dyldt , and n = dxldt . Now, equation (14) can be expressed as dW™ — = {[(c - R3e)t/(\ - t)] + her3t + (g - y)r2 dt dx + (G - Y)R2h + (r3 + R3)(e - e)(l + A)}—

The first-order conditions for a welfare maximum will give t to be zero optimally if actual changes due to a slight change in tariffs are equal to the conjectural changes, or if g = 7, G = T, and e = e. The previous Cournot example can be analyzed in this framework by setting 7 = 0 = T, G < 0, g < 0, and e = 0, e = 1, h > 0. Since (g - 7)^*2 > 0 and (G - F)R2 > 0, these effects call for a subsidy on domestic production. This is the standard effect that depends on the form of competition, as pointed out by Eaton and Grossman (1986). Moreover, notice that although the subsidy is imposed at home, the effects of a subsidy are desirable in both markets. This is evident in there being two terms, one for the domestic market and one for the

301

High-Tech Trade Policy

foreign market. A subsidy is desirable in the home market since the domestic firm is being too ''pessimistic" in its conjectures at home which leads it to produce too little and a subsidy alleviates this distortion. This effect is captured by the term (g - y)r2 being positive. A subsidy is desirable abroad because it causes the domestic firm's output to rise in both markets so that h is positive, and because the domestic firm is being too pessimistic abroad as well. This is captured by the term (G - T)R2h being positive. Expectations affect revenues in both markets. For this reason we see the term r3 + R3 in equation (15). Since the effects of a subsidy are direct at home and indirect abroad, we see the term (1 + h) multiplying the term (r3 + R3) in equation (15). Finally, since e - € > 0 in this case, firms are too conservative in their estimate of the network benefits of increased output. For this case, (r3 + R3) (e - e) (1 + h) is positive, since h is positive and the expectation effect just reinforces the previous effects. Therefore, the optimal policy would be a subsidy. This completes the analysis of profit shifting in the Cournot example. If e T^ 0, the sign of the terms multiplying t in (15) is ambiguous. dx However, as long as — < 0 at t = 0, similar arguments to those preat viously made indicate the direction of welfare-increasing policies from an initial state of no taxes or subsidies on domestic consumption of the domestic good. If, for example, firms overestimate the effect of output on network expectations so that e - e < 0, and g = y and G = T, then a small tax on domestic consumption of the domestic firm's product will raise welfare from the t = 0 welfare level. It is clear from the expression for the change in welfare that the optimal policy depends not only on the form of the strategic interaction, as parameterized by F and r relative to g and G, but also on the sign of h. In addition, it also depends on the distortions inherent in the expectations formulation.18 Now to turn to the effects on consumer surplus, Wc.19 (16)

w

= u&> x*> e(x + 30, e\x* + y*)] - r[x, x\ e(x + y)] -r*[x, x\ e*(x* + / ) ]

where U{.) is the utility function being maximized. Differentiating gives:

,**

(17)

dw

ldWc

dWc

dWc

,

dW',ldx

—r- = \ + -rrg + h + —-Gh \— at [_ dx ax dy dy \at = [(C/i - r, + eU3 - er3 - rj) + (U2 - r\ - r2 dx + U4em - r*3e*)g + (U3 - r3)eh + (U4 - r*3)e*Gh]— .

302

Kala Krishna

The four terms in brackets give the effect of changes in x, x*, y, and y* because of changes in t on consumer surplus. Since Ux = P, the inverse demand function facing the domestic producer at home, Ux - rx, is positive. Similarly, U2 - r\ is also positive. U3 - r3 is positive/ negative if an increase in the network size raises utility more/less than it raises revenues. The sign of this term is ambiguous and depends on the particular specification of demand used. A useful interpretation, along the lines of Spence (1976), can be made as follows. If increments in network externalities are valued less for marginal units than for all units on average, then U3 - r3 is positive. If the increments in network externalities are valued more for marginal units than for all units on average, then U3 — r3 is negative. This is because (18)

U3[x, x\ e(x + y), e*(x* + / ) ] - r3[x, x\ e(x + y)] = fcU3X[s, x\ e(x + y)]ds - P3[x, x*, e(x + y)]x = x\ - fx0P3[s, x*, e(x + y)]ds - P3[x, x\ e(x + y)]

The first term is the average willingness to pay for increments in the network size over all units purchased while the latter is the willingness to pay at the margin for increments in network size.20 The interpretation of U4 - r*3 is similar to this. Also, r\ and r2 are negative since the goods are substitutes. The effects of a tax or subsidy on consumer surplus can now be analyzed. Assume that a tax lowers x. Consumer surplus is affected through four channels, x, x*, y and y*, as shown. First consider the effect by means of x directly. A tax reduces x and since the fall in utility exceeds the fall in expenditure, this reduces consumer surplus. This is captured by the term Ux - rx being positive. The fall in x also reduces the network size. This is captured by U3 - r3. If U3 — r3 is positive, the fall in the network size will also reduce consumer surplus. If U3 - r3 is negative, it will raise consumer surplus. Since the goods are substitutes, r\ < 0 and the fall in x will raise the revenues of the foreign firm in the domestic market which, for a given level of x*, will reduce consumer surplus. Therefore, the effect of a tax via x reduces consumer surplus as long as U3 - r3> 0. Now turn to the effect via x*. If the fall in x raises JC\ i.e., g < 0, /dWc\ and U4 - r3 > 0, then gl — - I < 0. A tax then raises consumer surplus via its effect of raising x*. y effects consumer surplus only via its effect on network size. If h > o, so that a tax reduces sales in both markets, then the effect by means of y also reduces consumer surplus if U3 - r3 > 0.

303

High-Tech Trade Policy

y* also affects consumer surplus only through its effect on network size. If G < 0, h > 0, then the tax raises y* thereby raising the network for the foreign firm and raising consumer surplus if U4 - r\ > 0. The total effect of a tax on consumer surplus is therefore rather complicated and consumer surplus could rise or fall with a tax. For the Cournot example previously discussed, if the average increase in the willingness to pay associated with an increase in network size exceeds the marginal increase in the willingness to pay, and direct effects on surplus by x and y outweigh the relatively indirect ones by JC* and y*, a tax reduces consumer surplus. Thus, if a subsidy is called for because of strategic considerations in maximizing W17, it will also raise consumer surplus. Exactly the same procedure can be used to define the welfare effects of a tax or subsidy when the products are compatible. The analogous expression to equation (15) is somewhat formidable. With compatible products Wm is given by W17 = r[jc, JC*, e{x + JC* + y + y*)] + R[y, y\ e(x + x* + y + / ) ] - c(x + y). Then,

- = ^ — — — + M I + nj

(19)

+ (r3 + R3){e[(l + g) + (1 + G)h] - €[(1 + 7 ) + (1 + T)h]} + (g ~ 1>2 + KG - T)R2

dx dt

Notice that even with compatible products there is a role for subsidizing domestic consumption of the domestic firm. If, for example, dx e - e > 0, g = 7 < 0, G = 7 < 0, h > 0, — < 0 and [(1 + g) + at dW" (1 + G)h] > 0, then —— < 0 for t close to zero, so that a subsidy dt raises domestic profits. However, with compatible products, some of the benefits of a larger network accrue to the foreign firm, and even if e > e, it may not be worthwhile subsidizing domestic consumption in order to shift profits. This is captured by the fact that (1 + g) + (1 + G)h is required to be positive as well in this case. Also, Wc = U[x, JC*, e(jc + x* + y + / ) ]

- r[jc, JC*, 6>(JC + JC* + y + y*)]

- r*[jc, JC*,