International trade policy: A contemporary analysis

Grimwade, Nigel Book International trade policy: A contemporary analysis Provided in Cooperation with: Taylor & Francis Group Suggested Citation: Grimwade, Nigel (2006) : International trade policy: A contemporary analysis, ISBN 978-0-203-02788-2, Routledge, London, https://doi.org/10.4324/9780203027882 This Version is available at: https://hdl.handle.net/10419/251150 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by-nc-nd/4.0/ INTERNATIONAL TRADE POLICY The last decade has been witness to far-reaching and significant developments in international trade policy. At the multilateral level of trade these have included the conclusion of the Uruguay Round of trade negotiations and the creation of the World Trade Organisation. At the regional level this liberalism has been paralleled by the expansion and development of regional trading blocs. International Trade Policy: A contemporary analysis provides extensive, in-depth coverage of the theoretical and policy considerations, both old and new, which underlie these developments. The topics covered include: • key theoretical and policy issues, such as voluntary export restraints; antidumping and unfair trading practices; agricultural protectionism; regionalism; and services; • the issues which govern many current trade disputes, for example trade-related intellectual property rights; • the central issues involved in setting up regional trading areas; • trade-related investment measures and the developing world; • the future agenda of multilateral trade negotiations, which is likely to be dominated by the environment and labour standards. With trade negotiations becoming increasingly complex, International Trade Policy: A contemporary analysis presents a clear and up-to-date guide to contemporary policy and the theory upon which it is based. Written in an accessible style, the book assumes a good, basic knowledge of economics and will be invaluable to both students and policy makers in the area of international trade. Nigel Grimwade is Principal Lecturer in Economics and Head of the Economics Division at South Bank University. He is the author of International Trade (Routledge, 1989) and has written extensively on the subject of international trade policy. He is currently engaged in research on the trade policy of the European Union. INTERNATIONAL TRADE POLICY A contemporary analysis Nigel Grimwade London and New York First published 1996 by Routledge 11 New Fetter Lane, London EC4P 4EE This edition published in the Taylor & Francis e-Library, 2006. “To purchase your own copy of this or any of Taylor & Francis or Routledge’s collection of thousands of eBooks please go to http://www.ebookstore.tandf.co.uk/.” Simultaneously published in the USA and Canada by Routledge 29 West 35th Street, New York, NY 10001 Routledge is an International Thomson Publishing company © 1996 Nigel Grimwade All rights reserved. No part of this book may be reprinted or reproduced or utilized in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage and retrieval system, without permission in writing from the publishers. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging in Publication Data A catalogue record for this book has been requested ISBN 0-203-02788-4 Master e-book ISBN ISBN 0-203-15154-2 (Adobe e-Reader Format) ISBN 0-415-06878-9 (hbk) ISBN 0-415-06879-7 (pbk) To my wife, Charlotte, and two sons, Peter and Matthew CONTENTS List of figures vii List of tables viii 1 INTRODUCTION 1 2 INDUSTRIAL TARIFFS 18 3 QUANTITATIVE TRADE RESTRICTIONS AND SAFEGUARDS 43 4 UNFAIR TRADING PRACTICES: DUMPING AND SUBSIDIES 79 5 THE DEVELOPING COUNTRIES 126 6 AGRICULTURAL PROTECTIONISM 161 7 REGIONALISM 195 8 THE NEW ISSUES: SERVICES, TRIPs AND TRIMs 237 9 THE EMERGING AGENDA 276 Bibliography 302 Index 313 FIGURES 2.1 The small-country partial equilibrium model of the effects of a tariff 19 2.2 The large-country partial equilibrium model of the effects of a tariff 21 2.3 The welfare effects of a tariff on imports under imperfect competition 23 3.1 The effects of an import quota on the importing country 52 3.2 The effects of a voluntary export restraint on the importing country 60 3.3 The effects of a voluntary export restraint on the pattern of trade 62 4.1 Dumping with market power at home and abroad 82 4.2 Dumping with market power in the home market only 83 4.3 Illustration of EC dumping calculations 96 4.4 The effects of a domestic subsidy 110 4.5 The effects of an export subsidy on an exporting and an importing country 112 7.1 The effects of a customs union on the home and partner countries 202 7.2 The effects of a customs union when the rest of the world imposes a tariff 204 7.3 The effects of a customs union under decreasing costs 206 agreement will have on farm trade. Nevertheless, it seems unlikely that the past dichotomy between agriculture and manufacturing will come to an end. Chapter 7 addresses the issue of regionalism and the possible conflict between attempts to achieve regional trade liberalisation and the GATT objective of multilateral liberalisation. Recent years have witnessed a resurgence of regionalism to the extent that some observers have warned of the danger that the world trading system could fragment into a series of regional trading blocs. Although GATT rules permit the formation of customs unions and free trade areas subject to certain conditions, the intention was that these would be exceptions to the overriding objective of an open, multilateral trading system based on the principle of nondiscrimination. Certain kinds of regional trading arrangements were permitted on the grounds that these could be stepping stones towards global trade liberalisation. The concern is that the current fascination with regional trading blocs of even the big players such as the USA, which in the past was the major protagonist for multilateralism, could undermine rather than strengthen global trade liberalisation. Chapter 8 discusses the so-called new issues which were added to the agenda of multilateral trade negotiations at the commencement of the Uruguay Round. The three main new issues covered by the Round were trade in services, trade-related intellectual property rights (TRIPs) and trade-related investment issues (TRIMs). All three issues brought GATT into previously unchartered waters. No previous GATT round had sought to tackle these issues in any serious fashion. Yet they could no longer be omitted, given their crucial importance to the developed countries. The USA in particular was not prepared to embark on a new GATT round without these issues on the agenda. Finally, Chapter 9 makes a brave concluding attempt to identify the issues of the future. Already only six months into the life of the newly established WTO, the shape of a future agenda is emerging. Four key issues are likely to be important: trade policy and the environment; trade policy and labour standards; competition policy; and global investment issues. It is clear that the agenda for multilateral trade negotiations is changing rapidly. Negotiations are no longer concerned with the relatively simple matters of tariffs, over which countries could more easily bargain. As formal barriers have been lowered and international competition increased, trade negotiators are being forced to address a much wider range of issues. Market access is no longer concerned purely and simply with controls imposed at the border. Many other forms of government intervention, including seemingly innocuous forms of government regulation, can affect the ability of one country to sell goods in the market of another. Nor will the future agenda be concerned purely with trade matters as in the past. Rather, there will be growing demands that the newly established WTO broaden its concerns to cover issues affecting international factor movements, including direct investment abroad, labour movements and the transfer of technology. HISTORICAL BACKGROUND The basis for international trade policy over the last four and a half decades has been the General Agreement on Tariffs and Trade (GATT) signed in 1947. It is therefore desirable to begin with a brief survey of the historical background to the establishment of GATT International Trade Policy 4 and a summary of the role which it has played in world trade liberalisation over the past forty-eight years. In many respects, GATT was modelled on the prewar United States Trade Agreements Programme. This in turn came into being with the passage in 1934 of the US Reciprocal Trade Agreements Act (RTAA). The significance of the RTAA was that it gave to the US President a new and specific authority to enter into trade agreements with other countries whereby the US tariff would be reduced in return for equivalent concessions from trading partners. The overt intent was to use the US tariff as a weapon to gain easier access for US manufactures to the markets of other countries. Before the war, this took the form of bilateral agreements between the US and her major trading partners in which the US offered cuts in her own tariffs as a device for securing more open markets for US exports abroad. The impact of the programme was, however, rather limited because negotiations were largely bilateral and due to the outbreak of the Second World War. After the war, however, the US was keen to resume its tariff-cutting programme but this time on a multilateral rather than bilateral basis. GATT was largely the outcome of this process. This was strengthened by the political desire to assist the recovery of Europe following the devastation of the war and to contain the spread of communism. Before discussing the US Trade Agreements Programme and how it led to the creation of the GATT, it is necessary to examine the nature of US trade policy before 1934. Previously, US trade policy consisted of imposing mainly high tariffs to protect domestic industry and to generate revenue for the federal authorities. The tariff was regarded as largely a matter of domestic concern and, as such, non-negotiable. The significance of the tariff as a revenue-raising device was reduced with the passage of the 16th Amendment to the US Constitution in 1913 which made a federal income tax constitutional. This reduced the previous dependence on customs duties as a source of tax revenues. Nevertheless the policy of maintaining a high tariff as a device for protecting US producers continued throughout the 1920s and early 1930s. Indeed, in several respects, US trade policy became more protectionist. The 1921 Emergency Tariff Act resulted in higher duties on imported agricultural products, a response to the slump in agricultural prices which followed the end of the war. The 1922 Fordney-McCumber Tariff Act saw the protection granted to agriculture spread to manufacturing also. Then, in 1930, the Smoot-Hawley Act raised duties still higher. An important factor in this process was the phenomenon of logrolling whereby concessions granted to agriculture enabled Congressmen from urban areas to push for higher duties on industrial goods. However, historians seem in agreement that the policy of maintaining a high tariff during the interwar period was harmful to the US. If there had been some justification for high tariffs in the prewar period, this was not true of the interwar years. Firstly, by the time the war was over, the US had become a net exporter of merchandise. Moreover, in excess of one-half of her exports now consisted of finished and semi-finished manufactures compared with 30 per cent before 1900 (Kelly, 1963). Since many of these manufactures were produced under conditions of increasing returns or decreasing average cost, a large and growing market was important. As the domestic market for some of these goods became saturated, there was a need to seek out new markets overseas. A policy of maintaining high import tariffs was unhelpful in this respect. In the absence of a lowering of the US tariff, other countries were reluctant to grant US manufactures easier Introduction 5 access to their home markets. At the same time, import protection was of little benefit to the new, research-intensive growth industries pioneered by the US (Meyer, 1978). Secondly, by the time the war was over, the US had become a large net exporter of capital to the rest of the world. Before the war, she had been a net importer of capital. During the interwar years, the US played an important role in enabling other countries, especially the European economies, to finance current account balance of payments deficits. In the absence of US capital exports, these other countries would have faced major adjustment problems. Indeed, this is precisely what happened when US loans and investments were curtailed in the 1930s following the Wall Street stock market crash of 1929. This played a major role in the descent of the world economy into the Great Depression. However, the key point is that the changed status of the US from debtor to creditor necessitated a policy of low tariffs. For only by exporting more merchandise to the US could the rest of the world earn the foreign currency both to continue buying US goods and to meet the interest, amortisation and dividend payments on the loans and investments received from the US. The failure of the US to recognise this until it was too late was a major factor contributing towards the financial crisis which ensued in 1929. In economic life, policies are often slow to respond to changed circumstances. Governments often persist with policies even when changed conditions have rendered those policies obsolete. Thus, US trade policy was at odds with the changed position of the US in the world for much of the interwar period. However, there were a number of changes which did take place in US trade policy during this period which were important for what was to follow. Firstly, the principle of the ‘flexible’ tariff was established in both the 1922 (Fordney-McCumber) Act and the 1930 (Smoot-Hawley) Act. The President was empowered to vary the tariff within specified limits without resort to Congress. Of course, the intention was to give to the President the freedom to raise tariffs so as to equalise foreign and domestic costs of production. If, for example, foreign companies enjoyed a fall in costs of production, giving them a competitive edge when selling to the US, the President could raise tariffs to restore competitiveness to US producers. Such a policy was highly protectionist. If enforced by all countries, it could soon stop all trade. However, an important principle had been conceded, namely, that the President could change tariffs without seeking further Congressional approval. The intended upward flexibility could become a downward flexibility in the future. Secondly, in 1923, the US adopted an unconditional most-favoured-nation (MFN) policy. This meant that the US agreed not to discriminate in her tariff policy. All her trading partners would be treated equally. If a high tariff was imposed on imports of a certain product from one country, the same tariff would be applied to imports of that product from all other countries. Similarly, if the US entered into a trade agreement with another country whereby the US agreed to cut her tariff on a particular item, the tariff cut would be extended to all imports of that item regardless of where they came from. Before 1923, the US had adopted a conditional MFN policy. This meant that, if the US signed a trade agreement with another country, any tariff cut made by the US would only be extended to other countries if they offered concessions equivalent to those made by the country with whom the US had signed the agreement. Not surprisingly, such a policy did result in discrimination since some countries were unable to offer such concessions whenever the US entered into a trade agreement with a larger country. In fact, before 1923 actual US trade policy adhered more closely to an unconditional MFN policy. The International Trade Policy 6 adoption of an unconditional MFN policy in 1923 merely gave a legal basis for what was already practised. Furthermore, since at that time there were no trade agreements involving cuts in the US tariff, the change was not important. From 1922 onwards, the US tariff was to all intents and purposes non-negotiable. The situation could be described as one in which the US tariff was high and non-negotiable but nondiscriminatory. However, when in 1934 the US tariff did become negotiable, the unconditional MFN policy assumed a great importance. It meant that any tariff cuts offered by the US in a trade agreement with another country were now automatically extended to all other MFN trading partners. This made for much more rapid reductions in tariffs than would otherwise have been the case. Thirdly, after the First World War, the US largely renounced quantitative restrictions on trade. Protection took the form of high tariffs rather than import quotas. These had not been important before the war but had been extensively employed during the war. After the war, they were largely abolished except in Central and Eastern European countries where adverse conditions necessitated their retention. During the 1920s, US trade policy regarded quotas as an improper form of trade control and tariffs as the form which protection should primarily take. The 1934 Reciprocal Trade Agreements Act followed the 1933 Democrat election victory. The Roosevelt Administration viewed trade policy as one instrument for lifting the US economy out of the economic slump of the 1930s. By offering cuts in tariffs, the US could secure easier access for its exports to other countries. Not only would lower tariffs make other countries more willing to cut tariffs on US exports, but equally the enhanced ability of other countries to sell to the US would provide them with the foreign currency to buy US goods. This was to be achieved by the US President negotiating reciprocal tariff reductions with the US’s major trading partners. At the time, this was more acceptable to domestic economic interests than unilateral tariff reductions. For this purpose, Congress granted the President the authority to cut US tariffs by up to 50 per cent, initially for a period of three years, after which he could seek a renewal. This delegation of tariff-cutting authority by Congress to the President was of the utmost significance. Before 1934, Congress had jealously guarded its sole right to change the tariff. The President could propose cuts in the tariff but Congress had to give its approval. This meant that any trade agreement negotiated before 1934 could only be made effective if Congress approved. It made it difficult for any President to negotiate because neither he nor the country with whom he was negotiating could be sure that any concessions offered by the President would be approved by Congress. Not surprisingly, there were few trade agreements entered into before 1922 after which even the principle of a negotiable tariff was discarded. The 1934 Act changed the situation by giving the President the power to make meaningful offers in the course of negotiations. Moreover, the unconditional MFN policy was continued so that any concessions offered by the US in negotiations with any one country were automatically extended to all other MFN countries. The 1934 Act gave to the US for the first time in its history an effective foreign economic policy. The US tariff was to become a weapon which the US would employ to secure both easier and fairer access to overseas markets for its goods. This change in the position of the world’s largest trading nation was of enormous significance. The immediate impact was not so great. Some twenty-nine agreements were signed with various countries, the most important of which were with Canada and France in 1936 and Introduction 7 the UK in 1939. However, the outbreak of war meant that these agreements achieved little or nothing in expanding trade. Nevertheless, they were important as forerunners of the multilateral agreements reached after the war through the GATT. Following the end of the Second World War, there was a strong desire on the part of the Allied powers to ensure a rapid return to normal peacetime conditions. In particular, they were concerned lest a brief postwar boom should give way to a protracted slump and a return to the conditions of the 1930s. A free, open and stable world trade and monetary system was seen to be important in bringing this about. The Bretton Woods Conference of 1944 helped to create the necessary monetary rules and institutions in the form of the International Monetary Fund (IMF) and the World Bank or International Bank for Reconstruction and Development (IBRD). The US was anxious to shape a similar set of rules and institutions to cover the trading side, for which the model was to be the trade agreements programme of the prewar period. In 1945, Congress renewed the tariff-cutting authority granted to the President in 1934. Where tariff rates had been already reduced by 50 per cent under the 1934 authority, the President was permitted to make further tariff cuts of between 50 and 75 per cent of the 1934 level. Apparently, under the 1934 authority, tariff cuts of 50 per cent had been made on more than 40 per cent of US dutiable imports, so this fresh authority was very substantial (Kelly, 1963). As in 1934, the authority granted to the President was for three years. The aim of the US was to bring together a number of nations and simultaneously to negotiate tariff reductions. At the same time, the US took the initiative in a proposal for setting up a new ‘international trade organisation’ which was intended to be a counterpart on the trade front to the IMF and IBRD. The President’s 1945 tariffcutting authority did not include an agreement to create an ITO. Therefore, any charter for the creation of an ITO would have to be separately submitted to Congress for approval. For these reasons, the negotiations for a multilateral agreement to reduce tariffs became separated from the negotiations for setting up an ITO. The former led to the establishment of the GATT in October 1947. However, the GATT agreement was carefully drafted not to make any reference to the creation of any organisation. The GATT was a treaty not an organisation. This was to ensure that the agreement fell within the limits of the 1945 Congressional authority granted to the US President and so therefore did not require any further Congressional approval. The negotiations for the setting up of an ITO were not complete at the time when the GATT was signed. Rather than wait for these negotiations to be concluded, it was decided to proceed with the signing of the GATT. One reason was that the President’s authority was due to expire in mid-1948 so there was a need to reach agreement on cutting tariffs. The drafting of the ITO charter was completed at the Havana Conference of 1948. However, because the US Congress would not approve it, the ITO never came into being. These events leading to the signing of the GATT are important for understanding some of the features of GATT. Firstly, the GATT treaty contained no provisions for setting up any organisation or institutions. To begin with, there was not even a secretariat; eventually, one emerged. This was based in Geneva and headed by a Director-General. Secondly, the GATT signatories were known as ‘contracting parties’ not members, although the expression ‘GATT members’ was often used colloquially. Twenty-three nations signed the GATT initially. By the time of the Uruguay Round, there were 117 International Trade Policy 8 contracting parties. Thirdly, the agreement was only ever applied ‘provisionally’ by the contracting parties. Again, this goes back to the time when the GATT was first signed; it was expected that the drafting of the GATT would be followed by the setting up of an ITO, so it was decided to apply the treaty provisionally. Technically, what happened was that eight nations (Australia, Belgium, Canada, France, Luxembourg, the Netherlands, the UK and the USA) signed the Protocol of Provisional Application and applied it provisionally from 1 January 1948, while the other fifteen nations agreed to apply it soon after. Under the Protocol of Provisional Application, the contracting parties agreed to apply fully Parts I and III of the GATT and to apply Part II ‘to the fullest extent not inconsistent with existing legislation’. Part II contains most of the main substantial obligations. This meant that the contracting parties were free not to apply these provisions if they conflicted with legislation in existence at the time of becoming a GATT party. These so-called ‘grandfather rights’ still exist and are occasionally used to justify not applying Part II provisions. THE GATT FRAMEWORK GATT has served two purposes. Firstly, it has provided a set of rules to govern trade between the contracting parties. Rules are important for world trade because they create a degree of certainty for traders and hence stimulate investment and growth. Secondly, it has provided a multilateral forum for negotiating reciprocal reductions in trade barriers. Before GATT, trade negotiations were essentially bilateral affairs. This necessarily limited what they could achieve and the speed with which barriers could be lowered. Let us begin with the rules. They are set out in the major articles of the Treaty (see Table 1.1). They contain many of the principles on which US trade policy was based before the war. Most important was Article I, the Most-Favoured Nation Clause. This required contracting parties to treat goods coming from other contracting parties equally, that is, not to discriminate. It states that any advantage, favour, privilege or immunity granted by any contracting party to any product originating in or destined for any other country shall be accorded immediately and unconditionally to the like product originating in or destined for the territories of all other contracting parties. (Article I:1) In other words, GATT contracting parties were to accord nondiscriminatory (that is, most-favoured-nation) treatment to goods coming from (or destined for) the territories of other GATT contracting parties. Two important exceptions to the principle of nondiscrimination were the cases of customs unions and free trade areas. Customs unions involve the abolition of internal tariffs and the adoption of a common customs tariff. Free trade areas similarly involve internal free trade but the members are free to apply whatever rate of external tariff they choose. Thus, both result in preference or discrimination being granted to goods originating from inside the customs union/free trade area. Article XXIV states that Introduction 9 the provisions of this Agreement shall not prevent, as between the territories of contracting parties, the formation of a customs union or of a free trade area or of the adoption of an interim arrangement necessary for the formation of a customs union or of a free trade area provided that: (a) with respect to customs unions, duties and other regulations of commerce imposed at the institution of any such union…in respect of trade with contracting parties not parties to such union …shall not on the whole be higher or more restrictive than the general incidence of the duties and regulations of commerce applicable in the constituent territories prior to the formation of such union… Table 1.1 The GATT Articles of Agreement I Objectives II General most-favoured-nation treatment III Schedules of concessions IV National treatment and internal taxation and regulation V Freedom of transit VI Antidumping and countervailing duties VII Valuation for customs purposes VIII Fees and formalities connected with importation and exportation IX Marks of origin X Publication and administration of trade regulations XI General elimination of quantitative restrictions XII Restrictions to safeguard the balance of payments XIII Nondiscriminatory administration of quantitative restrictions XIV Exceptions to the rule of nondiscrimination XV Exchange arrangements XVI Subsidies XVII State trading enterprises XVIII Governmental assistance to economic development XIX Emergency action on imports of particular products XX General exceptions XXI Security exceptions XXII Consultation XXIII Nullification or impairment XXIV Customs unions and free trade areas XXV The organisation for trade co-operation XXVI Acceptance, entry into force and registration XXVII Withholding or withdrawal of concessions XXVIII Modification of schedules XXIX Tariff negotiations XXX Amendments XXXI Withdrawal XXXII Contracting parties International Trade Policy 10 XXXIII Accession XXXIV Annexes XXXV Nonapplication of the agreement between particular contracting parties XXXVI Trade and development: principles and objectives XXXVII Undertaking relating to commodities of special export interest to LDCs XXXVIII Outline of joint action on trade and development (b) with respect to a free trade area…the duties and other regulations of commerce maintained in each of the constituent territories and applicable at the formation of such free trade area…to the trade of contracting parties not included in such area…shall not be higher or more restrictive than the corresponding duties and other regulations of commerce existing in the same constituent territories prior to the formation of the free trade area… (Article XXIV:5) Customs unions and free trade areas were permitted, provided that they did not result in a higher level of restriction on imports from other contracting parties than existed before their formation. Article XI prohibits altogether one particular type of trade restriction, namely quantitative restrictions. Once again, this provision was a carry-over from US trade policy in the prewar period. The 1922 Tariff Act prohibited this method of restricting imports, regarding tariffs as the proper form of import protection. Article XI states that No prohibitions or restrictions other than duties, taxes or other charges, whether made effective through quotas, import or export licences or other measures, shall be instituted or maintained by any contracting party on the importation of any product of the territory of any other contracting party or on the exportation or sale for export of any product destined for the territory of any other contracting party. (Article XI:1) However, there were certain exceptions. These were ‘export prohibitions or restrictions to prevent or relieve critical shortages of foodstuffs’ or other ‘essential products’, ‘import or export prohibitions necessary to the application of standards or regulations for the classification, grading or marketing of commodities’ and ‘import restrictions on any agricultural or fisheries product…necessary to the enforcement of governmental measures which operate: (i) to restrict the quantities of the like domestic product to be marketed or produced…(ii) to remove a temporary surplus of the like domestic product…(iii) to restrict the quantities permitted to be produced of any animal product the production of which is directly dependent…on the imported commodity’. The exception granted to agricultural imports was especially important. It arose because of the existence in many countries of government policies unique to agriculture for regulating output. Introduction 11 Contracting parties were permitted to introduce trade restrictions additional to those already in existence in certain situations. These were specified in the Treaty. One important case was that of dumping. Article VI states that: The contracting parties recognise that dumping, by which products of one country are introduced into the commerce of another country at less than the normal value of the products, is to be condemned if it causes or threatens material injury to an established industry in the territory of a contracting party or materially retards the establishment of a domestic industry. (Article VI:1) Dumping is thus defined as a situation in which goods are sold on the foreign market at a price which is below their ‘normal value’. This is defined as ‘the comparable price, in the ordinary course of trade, for the like product when destined for consumption in the exporting country’. If no such ‘comparable price’ exists, ‘the highest comparable price for the like product for export to any third country in the ordinary course of trade, or…the cost of production of the product in any country plus a reasonable addition for selling cost and profit’ can be used. Contracting parties were permitted to impose antidumping levies on such imports provided that the duty did not exceed the margin of dumping, defined as the difference between the export price and the normal value. Similarly, there was a provision for imposing so-called ‘countervailing duties’ to offset any subsidy granted to the exporter. As with antidumping duties, this was not to exceed the amount of the subsidy granted. Another important exception was restrictions to safeguard a country’s balance of payments. Article XII states that any contracting party, in order to safeguard its external financial position and its balance of payments, may restrict the quantity or value of merchandise permitted to be imported. (Article XII:1) However, the restrictions imposed were not to exceed those necessary to remedy the situation and had to be progressively relaxed as conditions improved. Moreover, any country making use of this Article to introduce or intensify trade restrictions had to do so in consultation with other GATT contracting parties. Selective trade restrictions were also allowed under Article XIX, if imports of a particular product caused or threatened ‘serious injury’ to domestic producers. Article XIX states that If, as a result of unforeseen developments and of the effect of the obligations incurred by a contracting party under this Agreement, including tariff concessions, any product is being imported into the territory of that contracting party in such increased quantities and under such conditions as to cause or threaten serious injury to domestic producers in that territory of like or directly competitive products, the International Trade Policy 12 contracting party shall be free, in respect of such product, and to the extent and for such time as may be necessary to prevent or remedy such injury, to suspend the obligation in whole or in part or to withdraw or modify the concession. (Article XIX:1(a)) This is the so-called Safeguard or Escape Clause which allowed a contracting party to withdraw any tariff concession made in the past and impose a higher tariff for such time and to the extent necessary to remedy the injury to domestic producers caused or threatened by the imports. However, there was a requirement that notice of the intention to take any such action should be given in writing so as to provide opportunity for consultation. In exceptional circumstances, where delay would cause irreparable damage to the country concerned, provisional action could be taken without consultation. The intention was that, through consultation, it should be possible to reach agreement and avoid the necessity for such measures. However, if this was not possible, the importing country could proceed with its action. In this case, the affected parties could suspend concessions or obligations on trade with the party taking the action. A further ground on which a contracting party was permitted to impose trade restrictions was for the promotion of economic development in a developing country. Article XVIII states: The contracting parties recognise…that it may be necessary…in order to implement programmes and policies of economic development designed to raise the general standard of living of their people, to take protective or other measures affecting imports, and that such measures are justified in so far as they facilitate the attainment of the objectives of this Agreement. (Article XVIII:2) This applied ‘particularly to those contracting parties the economies of which can only support low standards of living and are in early stages of development’. These countries were permitted ‘to grant the tariff protection required for the establishment of a particular industry’ and to ‘apply quantitative restrictions for balance of payments purposes in a manner which takes full account of the continued high level of demand for imports likely to be generated by their programmes of economic development’ (Article XVIII:2). In addition to these rules governing the conduct of trade policy, GATT sought to bring about an expansion of world trade through a reciprocal lowering of tariff and other trade barriers. Not only did the contracting parties agree to refrain from certain kinds of trade restrictions and practices, they also undertook to meet periodically to negotiate a lowering of existing tariff barriers. This is provided for in Article XXVIII bis which states that The contracting parties recognise that customs duties often constitute serious obstacles to trade; thus negotiations on a reciprocal and mutually advantageous basis, directed to the substantial reduction of the general level of tariffs and other charges on imports and exports and in particular to the reduction of such high tariffs as discourage the importation even of Introduction 13 welfare loss. It is this last effect which makes tariffs harmful when viewed from a purely economic point of view. Following the work of Arnold Haberger (1963), the welfare costs of a tariff can be measured in the following way. (The basic idea behind welfare triangle analysis can be traced back to Jules Dupuit, 1844, and Alfred Marshall, 1920.) The loss for consumers is given by the reduction in so-called ‘consumers’ surplus’. For an individual consumer, this is the difference between the maximum price which a consumer is prepared to pay for the product (which measures the marginal utility of the product to the consumer) and the price actually paid. For consumers as a whole, it is equal to the area below the demand curve (which shows how much consumers were prepared to pay for different amounts of the product) and above the market price. At the pre-tariff price, OP0, in Figure 2.1(a), it is the area D0dP0. The effect of the tariff is, by raising the price, to reduce total consumer surplus by area (A+B+C+D). However, part of this loss to consumers represents extra income for both domestic producers and the government. The gain to domestic producers is equal to the increased ‘producers’ surplus’ generated by the rise in the price of the product sold. Producer surplus is equal to the difference between the price at which a supplier is prepared to supply a product (given by marginal costs) and the market price obtained. It is the sum of these amounts for each producer and is given by the area above the supply curve but below the market price. In Figure 2.1(a), at the price OP0, producers’ surplus is equal to area P0cS0. After the imposition of the tariff, this is increased by area A. As area C represents tariff revenue to the government, it follows that the net loss of welfare to the importing nation or the so-called deadweight loss of the tariff is areas B+D which is equal to area E in Figure 2.1(b). Conceptually, this deadweight loss can be divided into two parts: firstly, a consumption loss because consumers are unable to buy as much of the product as they would like; secondly, a production loss because the importing country must now devote more of its scarce resources to the production of the importable product than is optimal. The size of the welfare loss from any particular rate of tariff will depend on the slopes of the demand and supply curves, that is, the elasticities of demand and supply. The lower the elasticities, the less the welfare loss. The net welfare loss can be estimated using the formula: Net loss of welfare=0.5×tariff rate×reduction of imports or Net loss of welfare=0.5TP0{(Q0−Q1)−(Q2−Q3)} or Net loss of welfare=0.5TP0(Q5−Q4) Where the elasticities of demand and supply, ed and es, are known, the formula for estimating the welfare gain is: Net welfare gain=0.5T(edDT)+0.5T(esST) where T is the tariff reduction, D the original quantity demanded and S the original quantity supplied. The formula may be used to measure the cost of a particular tariff to an importing country provided that the value of elasticities is known. Equally, it may be International Trade Policy 20 used to estimate the potential gain to a country from lowering or eliminating a particular tariff. In a number of respects, this analysis of the effects of a tariff is over-simplified. Firstly, it is restricted to the case of a small importing nation which faces a world price over which it has no influence. The analysis needs to be modified for the case of a large importing nation which is able to influence the world price. In this case, the imposition of a tariff is likely to force down the world price of the good. If so, the loss in economic welfare from the higher tariff will be partially or even wholly offset by the gain from improved terms of trade. Figure 2.2 illustrates this case. D0D0 is the demand curve for the product in the importing country and S0S0 is the domestic supply curve. S0St is the total supply (domestic plus foreign) curve obtained by adding to domestic supply the amount which foreigners will supply at different prices. (S0St is flatter than S0S0 because world supply is more elastic than domestic supply but not perfectly elastic as in the case of a small importing nation.) Under free trade, demand is OQ1, domestic supply is OQ0 and imports are Q0Q1 which is equal to OQ4. Now, suppose a tariff is imposed at the rate of T. The effect is to cause foreigners to supply less at each and every price the total supply curve shifts vertically upwards to S0 S T The distance between the new and the old supply curve equals the amount of the tariff, that is, TP1. The new equilibrium price is O(1+T)P1. Consumption is reduced by Q3Q1 to OQ3. Domestic production is increased by Q0Q2 to OQ2. Imports fall from Q0Q1 to Q2Q3 which is equal to OQ5. The tariff reduces consumer surplus by the amount (A+B+C+D) as in the smallcountry model. However, the revenue effect is given by areas (C+E). Part of the revenue accruing to the government of the importing nation is a redistribution of income from foreign suppliers to that government. This is area E. It thus represents the increase in the economic welfare of the importing country resulting from the imposition of a tariff. It arises because foreign suppliers cut the price at which they supply the product from OP0 to OP1. Assuming no change in the importing country’s average export prices, the fall in Figure 2.2 The large-country partial equilibrium model of the effects of a tariff Industrial Tariffs 21 its average import prices leads to an improvement in its terms of trade. The loss of welfare from the tariff is then given by the difference between triangles (B+D) equal to area F and rectangle E, the gain to the importing country from the improvement in its terms of trade. It is possible for rectangle E to be greater than triangles B and D in which case a tariff could raise the welfare of the importing country. An ‘optimum tariff which would maximise the difference between area E and areas (B+D) can be constructed. It is given by the formula t=1/e where e is the elasticity of supply of imports (given by the slope of the import function in Figure 2.2(b)). (An import function shows the relationship between the quantity of imports supplied by the rest of the world and the price per unit in the importing country.) If this is known, a large importing nation could raise economic welfare by imposing a tariff. In this case, a tariff would not harm the importing country although it would reduce the welfare of the exporting country. It thus risks the danger that other countries would retaliate. If so, the importing country could lose. This case of retaliation was analysed by Harry Johnson (1953), who showed that, although one or other country might end up better off, both could not gain from a trade war. It also has limited practical usefulness as governments mostly lack the information required to be able to construct an optimum tariff. On the other hand, it may create an economic rationale for countries coming together in regional trading blocs with a common external tariff which could be used to force favourable movements in their combined terms of trade (see Chapter 7). A second and more important drawback with the orthodox model set out above is that it assumes perfect markets. There is an implicit assumption that both product and factor markets are perfectly competitive. One aspect of this is the assumption that imports are perfect substitutes for the domestically produced good with which they are competing. Put another way, the elasticity of substitution between imports and domestically produced goods is infinite. From this it follows that there will exist a single price for the product. The model can be made somewhat less restrictive by allowing for differences in quality such that goods of higher quality carry a price premium. However, the point remains that, when the price of imports rises due to the imposition of a tariff, the price of homeproduced substitutes rises by the same amount. What happens if imports are not perfect substitutes for importables; that is to say, if the elasticity of substitution between imports and importables is finite? Then, if the price of imports rises on account of a tariff, the price of home-produced goods may not rise by the same amount. Equally, if the price of imports falls, the price of home-produced substitutes need not fall by the same amount. Figure 2.3 sets out a model used by Batchelor and Minford (1977) for analysing the welfare effects of a tariff under imperfect competition. D0D0 and S0S0 are respectively the demand curves and domestic supply curves for import substitutes in the importing country. D2D2 is the demand for imports. Before the imposition of a tariff, the price of import substitutes is OP0 and of imports OP2. A tariff is imposed on imports at the rate T such that the price of imports rises to O(1+T)P2 and the demand for imports falls International Trade Policy 22 Figure 2.3 The welfare effects of a tariff on imports under imperfect competition to OQ2. This increases the demand for import substitutes causing D0D0 to shift to D1D1 and the price to rise to OP1. However, since import substitutes are only imperfect substitutes (the elasticity of substitution is finite), there is only a limited switch in consumer demand from imports to import substitutes. In the case of perfect competition, the switch is total, which is why the price of import substitutes rises by the same amount as the price of imports. The vertical distance between the two demand curves in Figure 2.3(a) represents the amount of the tariff, TP0, showing that the price of import substitutes has risen by less than the price of imports. It follows that consumers suffer a welfare loss from both a rise in the price of import substitutes (P1P0) and a rise in the price of imports (TP2). This will be less than in the orthodox model. On the other hand, domestic producers enjoy less of an increase in producers’ surplus. Diagrammatically, the loss to consumers is given by areas E+F of which area E represents increased government revenues. So the net loss of welfare is area F which is equal to areas C+D in 2.3(a). (In Figure 2.3(a) the total loss to consumers is areas A+B+C+D of which A and B go either to producers or to the government leaving a net loss of areas C plus D.) Mathematically, the formula for measuring the welfare loss can be written as either: Net welfare loss=0.5TP2(Q2−Q3) or Net welfare loss=0.5TP0(Q0−Q1). On the other hand, it has been demonstrated that, in such industries, the effect of a tariff reduction is to lead to more intra-industry trade (the simultaneous export and import of products belonging to the same industry). The welfare gains from intra-industry specialisation come in the form of an increased variety of goods for consumers to choose from rather than lower prices (see Greenaway, 1982; Greenaway and Milner, 1986.) Greenaway and Milner (1986) define the ‘pure’ gains from intra-industry trade as resulting from ‘the ability [of international trade] to permit some consumers to locate Industrial Tariffs 23 closer to their ideal variety than under autarky’ (p. 151). These gains need be no less than the gains which result from increased inter-industry trade. However, they are less amenable to measurement and will not be fully estimated by the conventional approach. On the other hand, intra-industry trade may lead to price reductions if the advantages of longer production runs and the stimulus of increased competition lead to lower costs. It is now a well-attested feature of many differentiated goods industries that production typically takes place under conditions of increasing returns (decreasing unit costs). Intraindustry trade in such goods makes possible longer production runs and thereby fuller exploitation of such scale economies. If trade also increases competition, these cost savings will be passed on in lower prices. Imperfections in factor markets may also modify some of the above analysis. Where factor markets are imperfect, short-run adjustment problems may result. If factor markets were perfect, resources released from import-competing industries as a result of tariff cuts would be immediately re-employed in newer expanding industries. Two general sources of market imperfection may be identified. Firstly, impediments to both the occupational and geographical mobility of labour segment the labour market such that workers displaced from the import-competing sector cannot immediately be re-employed in the expanding sector. This results in a bigger decline in the wages of workers employed in the import-competing sector than would have occurred otherwise. Secondly, the failure of wage-rates to fall sufficiently to ensure that those seeking work match the number of workers firms are prepared to take on at the going rate. This results in largescale structural unemployment. In these cases, trade expansion results in short-run private and social adjustment costs which may with difficulty be estimated. These will take the form of both declining capital values and wage-rates in the import-competing industry and the costs to individuals and society of higher transitional unemployment. These short-run adjustment costs may need to be deducted from the static welfare gains accruing from tariff reductions if the true gain from reducing tariffs is to be properly estimated. On the other hand, as Banks and Tumlir (1986) have convincingly argued, many of these so-called ‘costs’ are not costs in the strict economic meaning of the word. A cost is only a cost if it cannot be avoided. Since many of the so-called ‘adjustment costs’ arise from avoidable imperfections in the market, it remains questionable whether they should be so regarded. For example, governments can facilitate speedier adjustment by reforming the labour market in such a way as to ensure that wage-rates more fully reflect the demand-supply balance at any given time. However, where adjustment costs have been taken into account in empirical studies of the costs of protection, they have not generally been found sufficient to offset the potential gain from tariff liberalisation (for example, see Cable, 1981). Attempts at estimating the static welfare loss from tariffs have found it to be much smaller than is often thought to be the case. Equally, estimations of the welfare gain likely to result from eliminating or reducing tariffs have found this to be quite small. This is not altogether very surprising given the reduced importance of tariffs as a barrier to trade. Moreover, the static welfare gain from a lowering of tariffs captures only the immediate gain to countries resulting from an improved allocation of global resources. It takes no account of the longer-run dynamic gains which may be more important. As stated above, these are likely to be especially important where tariff reductions lead to increased intra-industry specialisation. They take the form of lower average costs International Trade Policy 24 resulting from both an expansion of the market facing exporters and cost savings brought about in response to increased competition. Whereas the static gains affect only the proportion of output which is traded, the dynamic gains are spread over the entire output of the firm or industry in question and are therefore potentially much greater. A particularly important aspect of the gain from lower tariffs is the guarantee which it gives to exporters that improved market access is permanent. This is especially the case where countries bind tariffs at a particular level. The assurance that tariffs will not be raised may encourage exporters to undertake costly investment in increased capacity which they might otherwise have considered too risky. Not only will such increased investment generate faster growth in the world economy as a whole, it should also bring further cost savings as efficient low-cost producers expand at the expense of less efficient high-cost competitors. Through this process of intensified competition, an important restructuring process may take place in which substantial high-cost excess capacity is eliminated, yielding significant cost savings (see Owen, 1983). REASONS FOR COUNTRIES TO IMPOSE TARIFFS If tariffs impose costs on countries, why do countries impose tariffs? There are several possible explanations. Firstly, governments often act irrationally because they are ill informed. In brief, they act in ignorance of the damage which tariffs are inflicting on the country. Mistakenly, they believe tariffs to be beneficial to the economy as a whole. Although not an implausible explanation, it lacks credibility. Governments employ advisers, who include professionally trained economists. It would seem improbable that governments could remain ignorant of the costs of tariff protection for very long. Some other motive must therefore exist. One possibility is the existence of some other noneconomic benefit which is considered sufficiently important to justify the economic cost. Governments pursue many objectives which are not part of the economists’ calculus. This is undoubtedly a reason for many of the tariffs which governments impose. An example is the tariff protection given to an industry deemed to be of vital strategic importance to the country (such as a tariff on imported steel products). Another example is tariff protection granted so as to raise the relative incomes of a particular sector or social group (such as farmers) thought to be at special disadvantage. Economists cannot comment on whether governments should or should not pursue such objectives. However, they can point to the costs of doing so and insist that these be set against any expected benefit. Moreover, they can ask whether a tariff is the best way of achieving the objectives being sought. There might be other policy instruments (for example, a subsidy) which could achieve the intended objective more efficiently or at less cost. Yet a third reason for tariffs is the possibility that they bring some economic gain which more than offsets the static welfare loss measured by the conventional model. Attention has already been drawn to the optimum tariff argument although it was argued that this has rather limited practical application. There are at least two other situations in which a tariff might conceivably be beneficial. In the first, the market for a product is imperfectly competitive such that firms are able to earn supernormal profits in the long run. A tariff might then be used to ‘shift’ profits from foreign firms to domestic firms and therefore from the exporting to the Industrial Tariffs 25 importing country (see Brander and Spencer, 1981, 1984; Krugman, 1986). Such profitshifting or ‘rent-snatching’ tariffs may enable the tariff-imposing country to raise its economic welfare at the expense of others. This is similar to the optimum tariff argument set out above. Like the optimum tariff, the argument holds only under certain fairly restrictive assumptions (see Grossman, 1986). Its application is confined to so-called ‘strategic’ industries dominated by a few sellers and in which entry barriers limit the potential for new firms to enter the industry and compete away excess long-run profits. There is the further problem of selecting or ‘targeting’ the right industries to protect. High long-run profits may be a return for greater risk rather than an indication that competition is absent. Moreover, in most cases a tariff is inferior to a subsidy as a method for supporting such an industry. This is because a subsidy does not raise the price of the imported good and therefore inflicts no consumption loss on the importing country. Finally, as with the optimum tariff, there is a danger that a profit-shifting tariff will provoke foreign retaliation, leaving both countries worse off. The second situation is where significant ‘externalities’ exist which do not enter into the private cost-benefit calculation of the orthodox model. This will be the case where the growth of the protected industry has important spillover effects on other sectors of the economy, such as high-technology, knowledge-intensive industries that generate knowledge which can be shared with other branches or sectors. For example, there are strong linkages of this kind between the various branches of the electronics industry, such as consumer electronics and electronic components, which might be used to justify protection. On the other hand, as with the preceding case, a tariff is nearly always inferior to a subsidy if such protection is considered desirable. Even then, there is a problem in determining the optimal level of subsidy since the excess of social over private return is not easily quantifiable. The case of a newly established or infant industry in a developing country constitutes a further extension of this argument. Once again, the preference must be for a subsidy rather than a tariff. A fourth reason why tariffs are imposed is that a sudden surge of imports can cause serious adjustment problems for the importing country. Adjustment difficulties arise because of imperfections in both product and factor markets. If markets were perfect, a sudden surge of imports need not cause any problem for an importing country. Resources would instantly shift out of the declining sector and into the expanding sectors of the economy with little or no cost. To the extent that the exchange rate is free to find its own level, it will fall as a consequence of the rise in imports and this depreciation will lead to an expansion of exports. If the declining sector were more labour-intensive than the expanding sector such that the number seeking work exceeded jobs available, a relative decline in the wage-rate would ensure that the labour market cleared. In reality, market imperfections mean that full adjustment only takes place in the long run. A temporary import tariff may buy time for the importing country to enable adjustment to take place. On the other hand, a tariff could equally well forestall adjustment, if the tariff is retained beyond the time needed. Moreover, an import tariff is no substitute for the importing country adopting adjustment measures to facilitate the necessary shift of resources. These may include removing particular types of market imperfection which prevent adjustment from taking place. Although tariffs may be imposed for economic reasons other than those listed above, most of these are much less soundly based. For example, most tariffs imposed to protect International Trade Policy 26 domestic producers from low-wage competitors in other countries have no rational economic justification. This is because low wage-rates are frequently offset by low labour productivity. Even if labour costs per unit of ouput are lower, tariffs merely serve to prevent specialisation taking place based on differences in comparative costs. Tariffs imposed for this reason are therefore based on ignorance. Alternatively, they are a response to pressures exerted by particular vested interest groups in the importing country which have succeeded in winning over the government of the day. The desire to placate the producers or workers employed in a particular industry faced with more intense foreign competition overrides the interests of the country as a whole. Many examples of this can be found. Much of the agricultural protection which the advanced industrialised countries grant to their farmers is the result of governments yielding to political pressures, disregarding the cost to the country as a whole. In some countries, farmers have a political influence which is disproportionate to their numerical weight in the population. Recently, economists have shown considerable interest in exploring this political dimension to making tariff policy. This has taken the form of attempts to construct politico-economic models of tariff determination. For example, Frey (1984, 1985) has explained how tariff policy is formulated in a political market place in which there exist opposing forces for and against protection. Because tariffs benefit some groups of society, albeit at the expense of the rest, there will always be some with interests who favour protection. Usually, these will be producers and workers in the import-competing industries. Opponents of tariff protection will comprise consumers and exporting firms who face higher costs from tariff protection. However, generally speaking, pro-tariff interests are better organised and therefore better able to influence the decision-making process. Furthermore, while the gain to society as a whole from free trade generally exceeds the loss to producers/workers in the import-competing industry, the societal benefit is diffuse while the loss to producers/workers in the import-competing industry is highly concentrated. When expressed per head, the benefit to consumers or buyers of the product from free trade may be quite small, while the loss to producers/workers in the import-competing industry will be quite large. It follows that pro-tariff groups may have a greater incentive to resist tariff reductions than anti-tariff groups have to strive for them. Thus, tariffs may be retained simply because political pressures make it difficult or impossible to remove them. Significantly, some empirical studies have found that conservatism is the major factor influencing the structure of tariff rates between industries (Lavergne, 1983). THE STRUCTURE OF TARIFFS Although the average level of industrial tariffs has fallen significantly in recent decades, the average level of a country’s tariffs can be deceptive in concealing a highly protectionist tariff structure. Indeed, a country’s tariff structure may become more protectionist at the same time as the average level of tariffs falls. This is because of the phenomenon of tariff escalation. Tariff escalation occurs whenever the nominal rate of tariff applied to a particular industry increases with the stage of production or degree of fabrication. The more nearly finished the product, the higher the level of tariff imposed. Industrial Tariffs 27 Table 2.1 gives some examples of tariff escalation. It shows the average tariff calculated from the trade-weighted tariffs of ten major developed countries and twentyone developing countries at each stage of the processing chain for a selection of commodities. Because a high proportion of the exports of developing countries are concentrated in unprocessed primary commodities, they face higher tariff barriers when exporting to the developed countries than the average rate of tariff might suggest. It follows too that reductions in nominal rates of tariff will be of little benefit to developing countries unless the degree of tariff escalation is also lowered. It should be noted that the tariff structures of developing countries also escalate. A further consequence of tariff escalation is that it creates a disincentive for developing countries to invest in processing capacity. For example, on the figures given in Table 2.1, a major sugar exporter such as Mauritius faces a 20.0 per cent tariff if she exports refined sugar but only a 1.0 per cent tariff if she exports raw sugar. It is often argued that the effect of tariff escalation is Table 2.1 Average tariffs applied by major developed and developing countries at different stages in the processing of various product groups Processing chain Developed country (%)Developing country (%) Meat Fresh and frozen meat Prepared meat 6.2 8.4 6.6 21.9 Fish Fresh and frozen fish Fish preparations 4.3 4.1 10.9 30.1 Vegetables Fresh vegetables Vegetable preparations 6.9 13.2 16.6 26.9 Fruit Fresh fruit Fruit preparations 7.4 17.1 17.0 11.1 Vegetable oils Oilseeds Vegetable oils 0.0 4.4 18.1 26.5 Tobacco Unmanufactured Manufactures 1.2 18.1 126.0 662.1 Sugar Sugar and honey Sugar preparations 1.0 20.0 23.5 24.3 Cocoa Beans, powder & paste Chocolate and products 1.0 3.0 11.6 29.7 Rubber Crude rubber Rubber manufacture 0.0 3.9 7.2 19.4 International Trade Policy 28 Leather Hides and skins Leather Leather articles 0.1 2.9 7.2 4.8 17.5 33.9 Wood Wood, rough Wood, shaped Veneer and plywood Wood manufactures 0.0 0.3 1.7 3.5 8.0 13.1 23.5 27.6 Cotton Raw cotton Cotton yarn Cotton fabrics 0.0 3.0 5.8 3.2 29.7 32.1 Iron Iron ore Pig iron Ingots and shapes Bars and plates 0.0 2.2 2.2 3.4 2.6 7.4 12.1 19.9 Processing chain Developed country (%)Developing country (%) Other metallic ores Ores, nonferrous Wrought and unwrought metals 0.0 2.4 4.1 18.2 Phosphates Natural phosphates Phosphatic fertiliser 0.0 3.2 12.8 9.4 Petroleum Crude petroleum Refined petroleum 0.5 1.0 5.1 12.8 Source: Finger and Olechowski (1987) to shut developing countries into an export structure that is heavily dependent on unprocessed primary commodities. Does this matter? It may do so if trade in unprocessed primary commodities grows at a slower rate than that of processed commodities. It is also frequently argued that that the prices of primary commodities have a long-run tendency to fall relative to those of manufactured goods. This argument was first put forward by Raul Prebish (1950) and H.Singer (1950). In fact, the empirical evidence for this proposition is mixed. Moreover, even if there is such a tendency at work, it is by no means clear that it should result in a welfare loss for developing countries because not all primary producers are developing countries and not all developing countries are primary producers. Another concern arises from the high volatility of primary commodity prices. Over-dependence on a few primary commodities for export earnings could mean that developing countries face a highly unstable balance of payments. This may further jeopardise long-run economic growth. It should be noted that fluctuations in primary commodity prices will only result in unstable export earnings if the prices of different export commodities are positively correlated. If, however, they are negatively correlated, a fall in the price of one commodity might be offset by a rise in the price of another with no adverse effect on the stability of export earnings. It should be pointed out that, even if Industrial Tariffs 29 Instead, it was for the new contracting parties to make concessions in order to gain admission to the GATT. The US made concessions on only 39 per cent of her imports; 80 per cent took the form of tariff bindings. The average tariff reduction was 37 per cent but affected only 6 per cent of dutiable imports (Finger and Olechowski, 1987). The third round at Torquay in 1951 was also concerned with the accession of new contracting parties, West Germany being the largest country to join. This time, the US made concessions on a mere 7 per cent of her imports. The average tariff reduction was 26 per cent, affecting 15 per cent of her dutiable imports (ibid.). In 1955, the US President was given a new tariff-cutting authority, although much less than in 1945. It was for tariff cuts of only 15 per cent of the rates applying on 1 January 1955, to be effected in three annual instalments of 5 per cent each. Moreover, the legislation contained a number of restrictive clauses which provided for increased protectionism. However, it was sufficient to cope with the main task of the fourth round held at Geneva in 1956. This was the accession of Japan. The United States was keen to gain the acceptance of other contracting parties to Japan becoming a new contracting party. A number were reluctant. To overcome their reluctance, the US made further concessions on 9 per cent of her total imports with an average tariff reduction of 15 per cent, affecting 20 per cent of her dutiable imports (Finger and Olechowski, 1987). Even then, fourteen out of thirty-five GATT members invoked Article XXXV which allowed them to withhold GATT treatment from Japan until they had negotiated with her themselves. It was some while before Japan was treated as a full GATT signatory. The fifth round, the so-called ‘Dillon Round’ followed the new US tariff legislation of 1958. This authorised the President to cut tariffs by up to 20 per cent of the rates prevailing on 1 July 1958 with no more than a 10 per cent reduction in any one year. It was designed to cope with the problems arising from the formation of the new European Community in 1958. The EC was a customs union involving internal free trade plus a common external tariff. Article XXIV of the GATT allowed the formation of customs unions provided that the arithmetical average of pre-union external tariffs was no lower than the average post-union common external tariff. However, if this involved some members increasing the tariff on any of their trade with other GATT contracting parties, the latter were entitled to compensation. The EC offered a 20 per cent cut in the Common Customs Tariff although this had not come into full operation. In return, the United States offered an ad valorem tariff reduction of 20 per cent on 19 per cent of her dutiable imports. The sixth round, the ‘Kennedy Round’, took place between 1964 and 1967 and followed the passage of the 1962 US Trade Expansion Act. The 1962 Act gave to the President greater authority than ever before and for a longer period to cut tariffs. The President was authorised to cut tariffs by up to 50 per cent of the rates applying on 1 July 1962 over five years. On products for which the EC and the US together accounted for 80 per cent or more of trade, tariffs could be reduced by more than 50 per cent or eliminated altogether. This envisaged the UK joining the EC, since it would have had very little application otherwise. The Trade Expansion Act was a response to the challenge posed by the formation of the EC and EFTA. Although she favoured European integration largely for political reasons, the US was afraid that the preferential nature of these two trading blocs would cause trade diversion from the US to Western Europe (see Chapter 7 for a definition and explanation of the concept of trade diversion). The Common International Trade Policy 36 Agricultural Policy was also seen as a threat to US agricultural trade while the proposed Common External Tariff was deemed likely to increase US investment in Europe, aggravating the US balance of payments. Unlike in earlier rounds, most of the concessions made by the US were in the form of tariff reductions rather than bindings since there were few tariffs left to bind. The average US tariff cut was 44 per cent on 64 per cent of dutiable imports (Finger and Olechowski, 1987). Unlike in the first round, the US insisted that other industrialised countries made equivalent concessions. Consequently, other countries also made substantial tariff cuts affecting an estimated 70 per cent of dutiable imports (excluding cereals, meat and dairy products). Two-thirds of the reductions were of 50 per cent or more, and around another one-fifth were between 25 and 50 per cent. In addition, some progress was made in tackling certain kinds of nontariff barriers. In 1967, for the first time since 1947, the President’s tariff-cutting authority was allowed to lapse, thus it was some while before a new round of trade negotiations could take place. There was a feeling in Congress that substantial concessions had already been made in the Kennedy Round and that no further concessions could therefore be afforded for the time being. At the same time, the worsening US balance of payments was seen as a constraint. Anxieties about the growing threat posed to US manufacturing by the emergence of the newly industrialising countries also dampened any enthusiasm for a further bout of tariff-cutting. Nevertheless, by 1974 opinion in Congress had changed. One reason for this was the challenge posed for the United States by the admission of the UK to the EC in 1973 and the enlargement of the EC from six to nine members. The US was anxious to draw the new enlarged EC back into fresh trade negotiations. The 1974 Trade Act empowered the President to make further tariff cuts of up to 40 per cent of the rate existing on 1 January 1975 over a five-year period. As under the 1962 Act, these were to be staged over five years (ten years in exceptional circumstances). However, in several respects the Act was much less liberal than that of 1962. Not only was the basic authority a smaller one, but there were many more qualifications permitting higher tariffs in certain circumstances. In particular, the Act reduced the control of the executive branch over trade policy and vested more power with Congress and the independent International Trade Commission. Since the latter two were more likely to be influenced by pro-tariff interests, the likelihood was that US trade policy would become less liberal. Nevertheless, the seventh round, the Tokyo Round, did result in further significant tariff reductions. Tariffs on industrial products were cut by a weighted average of 33 per cent. The United States reduced her tariff on industrial products by a weighted average of 30 per cent, the EC by a weighted average of 28 per cent and Japan by a weighted average of 46 per cent (GATT, 1979a). These were to be implemented over eight years commencing on 1 January 1980. This increase in the staging of tariff cuts in comparison with previous rounds clearly weakened the impact of the final agreement. Moreover, by the time the Tokyo Round took place, tariffs had become much less important than nontariff barriers. Although some progress was made in confronting this problem, the agreements reached fell a long way short of what had been hoped for. Table 2.4 sets out the average level of tariffs in leading trading countries following the completion of the Tokyo Round. The average applied tariff was generally lower than the average MFN (most-favoured-nation) tariff because of the various preferences countries granted to goods from other countries with whom they had special trading arrangements Industrial Tariffs 37 (for example, the preferences which developed countries granted to manufactures coming from developing countries). The divergence between MFN and applied rates measures the extent to which countries departed from the MFN (nondiscrimination) principle. Table 2.5 sets out the average level of tariffs in the developed countries by product groups following the completion of the Tokyo Round. This shows a much lower level of tariffs for food and raw materials than for manufactures. The average tariff for manufactures disguises a still quite high tariff rate applied to clothing and textiles and to footwear. THE URUGUAY ROUND Although the agenda of the Uruguay Round was noteworthy for its inclusion of a wide range of new issues, tariffs remained an important item. The 1986 Ministerial Declaration launching the Round stated that: Table 2.4 Post-Tokyo Round trade-weighted average MFN and applied tariffs in selected developed countries Country Average MFN tariff rate (%) Average applied tariff rate (%) United States 3.9 3.8 EECa 4.2 2.5 Japan 3.5 3.0 Canada 6.5 4.5 Sweden 3.5 0.8 Norway 4.8 1.0 Switzerland 3.0 1.0 New Zealand 13.6 10.9 Austria 9.9 2.0 Finland 4.8 1.0 Australia 12.4 8.2 Source: Finger and Olechowski (1987) Note: aThe trade-weighted rates are based on the external trade of the EEC. Negotiations shall aim, by appropriate methods, to reduce or, as appropriate, eliminate tariffs including the reduction or elimination of high tariffs and tariff escalation. Emphasis shall be given to the expansion of the scope of tariff concessions among all participants. (GATT, 1986) Thus, it was acknowledged that further progress could be made in the elimination of certain tariffs. There was to be a clear emphasis on dealing with the problem of high tariffs and tariff escalation. There was also agreement that, in contrast with previous rounds, tariff cutting should not be limited to the big developed market economies but should embrace a larger number of participants. At the 1988 Mid-term Review, four International Trade Policy 38 aspects of tariff liberalisation were highlighted as being necessary to address. These were tariff escalation, tariff peaks, low ‘nuisance’ tariffs and the need to increase the level of bindings (GATT, 1988). It was further agreed that the target should be an overall tariff reduction of ‘at the minimum, the [tariff] reduction achieved by formula participants in the Tokyo Round’, that the scope of tariff bindings should be widened and that special account should be taken of the needs of developing countries (GATT, 1988). Table 2.5 Post-Tokyo Round average MFN and applied tariff rates by product group in developed countriesa Product group Average MFN tariff rate (%) Average applied tariff rate (%) All food items 6.4 5.3 Food & live animals 6.5 5.3 Oilseeds & nuts 5.3 4.0 Animals & veg. oils 0.1 0.2 Agricultural raw materials 0.8 0.5 Ores & metals 2.3 1.5 Iron & steel 5.1 3.4 Nonferrous metals 2.3 1.3 Fuels 1.1 0.6 Chemicals 5.8 3.1 Manufactures (excluding chemicals) 7.0 7.9 Leather 5.1 11.9 Textile yarn & fabrics 11.7 9.0 Clothing 17.5 3.3 Footwear 13.4 3.0 Source: Finger and Olechowski (1987) Note: aDeveloped countries comprise Australia, Austria, Canada, EC, Finland, Japan, Norway, New Zealand, Sweden, Switzerland and the United States. A key issue was the method of tariff-cutting to be used. The majority of countries, including those in the EC, favoured a formula approach similar to that used in the Tokyo Round. This, however, was opposed by the US at an early stage. At the commencement of negotiations, the US showed reluctance to make large tariff reductions, arguing that it was now the turn of other countries to do so. At the same time, she made clear a preference for a request-and-offers approach alongside so-called reciprocal zero-for-zero deals in particular sectors. The latter entailed countries agreeing on sectors in which tariffs could be totally eliminated. However, the EC was not prepared to include many sectors in the zero-for-zero deals unless the US offered more cuts in its high tariffs, particularly on textiles and other sensitive products. The US zero-for-zero list initially included pharmaceuticals, construction machinery, medical equipment, steel (subject to reaching a multilateral agreement providing for the elimination of state subsidies), paper and wood products, nonferrous metals, electronics, fish and alcoholic drinks. Also, she Industrial Tariffs 39 proposed that tariffs on chemicals be harmonised at very low levels. The EC was strongly opposed to eliminating tariffs on electronic goods since EC chip manufacturers enjoyed a 14 per cent tariff on semiconductors (see The Financial Times, 18 December 1992). Disagreement between the US and the EC created a hurdle to completing the marketaccess negotiations as other participants were unwilling to make offers without the two major trading blocs establishing the essential framework. A breakthrough was achieved by the so-called Quad countries (US, EU, Japan and Canada) at the Tokyo economic summit in July 1993. A market-access package emerged which found common ground in the face of the seeming deadlock which existed between US and EU positions. It was agreed that tariffs should be completely eliminated on pharmaceuticals, construction equipment, medical equipment, steel, beer, furniture (subject to certain exceptions), farm equipment and spirits. Tariffs on chemicals would be harmonised at low levels. Tariffs on ‘high tariff products (carrying tariffs of 15 per cent or more) would be cut by up to 50 per cent, including textiles. Tariff cuts averaging at least one-third would be made on all other products. The latter included wood, paper and pulp, and scientific equipment which the US had originally earmarked for zero-for-zero tariff treatment (see The Financial Times, 9 July 1993). Throughout the autumn immediately preceding the conclusion of the Round, disagreements between the Quad countries continued to threaten the final agreement. The average tariff cut of only 26 per cent being offered by the EU was generally considered to be inadequate and certainly below that offered by other countries. On the other hand, the US was accused of offering 50 per cent tariff cuts on only one-half the tariff peaks identified as included in the July agreement. Instead, the US offered more zero-for-zero tariff deals, including electronics. Japan was also criticised for offering 50 per cent reductions on fewer than one-half of her tariff peaks (see The Financial Times, 13 October 1993). As the 15 December deadline for reaching agreement on tariff reductions approached, it became clear that a line-by-line tariff-cutting agreement could not be achieved. Instead, the plan was to finalise an agreement on tariff cuts for about fifteen to twenty countries which collectively accounted for the bulk of world trade. The main elements of the Final Agreement were reported in The Financial Times (16 December 1993) as: 1 Tariff bindings. The proportion of trade in industrial products subject to bound tariffs was to be increased from 78 per cent to 97 per cent in developed countries and from 21 per cent to 65 per cent in developing countries. 2 Extent of tariff reductions. Tariffs were to be reduced on an estimated US$464 billionworth of imports of industrial products of developed countries out of a total of US$612 billion worth not already tariff free. 3 Tariff elimination. Tariffs were to be eliminated on a wide range of goods, bringing the proportion of tariff-free developed country imports to 43 per cent. The major trading nations agreed to eliminate tariffs on all products listed for zero-for-zero treatment at the July summit plus wood and paper products, toys and some fish products. 4 Tariff cuts. A trade-weighted average reduction of 38 per cent was to be made in the tariffs of developed countries from 6.3 per cent to 3.9 per cent. Table 2.6 summarises the overall tariff-cutting results of the Round. The US and EU agreed that tariffs on chemical products were to be harmonised at around 3 per cent. Above-average tariff cuts were made on high-tariff products including industrial electronics. The US also offered to cut tariffs on certain textiles and some glass and ceramic products. In International Trade Policy 40 general, tariff cuts on textiles and clothing were proportionately smaller than on other industrial products. 5 Agricultural tariffs. Tariff equivalents on agricultural imports were also to be subject to a 36 per cent overall reduction. 6 Tariff escalation. Some progress was made in reducing tariff escalation. Tariff escalation was to be eliminated for paper products, products made from jute and from tobacco, and reduced for products made from wood and metals. There appears to be universal agreement that the tariff-cutting aspect of the Uruguay Round achieved more than looked probable at one stage. The overall reduction in tariffs was close to 40 per cent, which is more than was achieved in the Tokyo Round and more than the target of one-third set at the commencement of the Uruguay Round. The tariff cuts were of course to be staged, so the benefits will take a number of years to filter through. The staging period was six years for developed countries and ten years for developing countries, a little quicker than in the Tokyo Round. The increase in the proportion of tariffs which are now bound and the elimination of tariffs on certain products represent important gains. On the other hand, Table 2.6 Average tariff reductions achieved in the Uruguay Round for industrial goods Trade-weighted average tariff (%) Country group Imports from MFN origins (US$bn) Pre-Uruguay Round Post-Uruguay Round Average tariff cut (%) Developed countries 736.9 6.3 3.9 38 Canada 28.4 9.0 4.8 47 EU 196.8 5.7 3.6 37 Japan 132.9 3.9 1.7 56 USAa 420.5 4.6 3.0 34 Developing countriesb 305.1 15.3 12.3 20 Economies in transition 34.7 8.6 6.0 30 Source: Hoda (1994), quoted in Schott (1994) Notes: aBased on data provided by USTR bBased on bound rates, not applied rates there remain a number of high tariffs in particular sectors, most notably agriculture, which remain to be tackled in subsequent rounds. The Round notably failed to bring about substantial reductions in tariff peaks, in particular in the textiles and apparel sector. With regard to the problem of tariff escalation, some progress was made in reducing the difference between tariffs applied to processed as compared with unprocessed products, and for some products tariff escalation was eliminated altogether. Although more remains to be done in reducing tariff barriers, the Uruguay Round has gone a long way to reduce further the importance of tariffs as an impediment to world trade. Industrial Tariffs 41 CONCLUSION Tariffs represent the oldest form of protectionism. They inflict welfare losses on importing countries although the measurable loss is small relative to total trade. Nevertheless, countries still impose tariffs. Since there are few sound economic arguments for tariffs, it follows that governments must either be pursuing some noneconomic objective or have chosen to promote the particular interests of those benefiting from protection at the expense of the common good. It follows that countries can increase economic welfare by reducing or eliminating tariffs. Since this may be more difficult to bring about unilaterally, the preferred means is to negotiate reciprocal trading agreements with other countries by which all participants simultaneously cut their tariffs. The main forum in which this has taken place over the past forty-nine years has been the GATT. Successive rounds of multilateral tariff negotiations through the GATT have substantially reduced the importance of tariffs as a barrier to trade in industrial products. Nevertheless, it should not be concluded that tariffs no longer matter. Low average tariff levels may disguise high-tariff peaks on particular products. Moreover, tariffs can and often are raised. High rates of effective protection also mean that tariff structures may grant higher levels of protection to domestic producers than nominal rates of protection indicate. Moreover, up to the Uruguay Round, tariff reductions were largely confined to industrial products. Agricultural trade remained highly protected although mainly by nontariff measures. One significant result of the Uruguay Round is that these barriers must be converted into tariffs and then progressively lowered by amounts similar to other tariffs. This is discussed further in Chapter 6. In a similar fashion, so-called ‘grey area’ measures impeding trade in industrial products are to be subject to tariffication. This is discussed in the next chapter. Paradoxically, therefore, tariffs will become more important in future years as certain nontariff barriers are converted to tariffs. It follows that tariffs will remain an important issue in international trade policy in the immediate future. International Trade Policy 42 3 QUANTITATIVE TRADE RESTRICTIONS AND SAFEGUARDS INTRODUCTION In the previous chapter, we saw that much of the success of the GATT rounds in liberalising world trade after 1947 was in the considerable reduction in the average level of industrial tariffs. One result of this appears to have been a growth of other forms of protectionism. These have taken a variety of different forms, often grouped together under the general heading of ‘nontariff barriers’ (NTBs). The next two chapters examine some of the most important forms of nontariff restraint on trade. In this chapter, the focus is on quantitative restrictions. Two of the most important forms are import quotas and voluntary export restraints (VERs). The latter, in particular, have come to play an increasingly important role in what is variously referred to as ‘managed trade’ or ‘administered protectionism’. The following chapter will examine two other highly important forms of nontariff protectionism, namely antidumping policy and subsidies. These are both linked to the notion of so-called ‘unfair trading’. However, before examining the main forms of nontariff protectionism, it will be necessary to take a broader look at its nature and scope. It will be seen that there is a wide variety of different ways in which governments may grant protection to a domestic industry. It will also be apparent that many forms of government intervention in the economy have either secondary or incidental effects on trade flows. As government intervention in the economies of most countries increased in the 1960s and 1970s, the importance of nontariff distortions to trade has, not surprisingly, increased at the same time. It was not always the case that interference with trade was the primary or even secondary intention of such measures. However, as tariff barriers were being lowered at the same time, the effects of such measures on international competition could not be ignored. Moreover, to the extent that tariff rates were bound at lower levels than before, it was always tempting for a government wanting to grant protection to a domestic producer to use one or more of these measures for protectionist purposes. Indeed, it will be seen that attempts to measure both the extent and frequency of nontariff interventions in trade show that nontariff protectionism has become more important in recent decades. This has come to be referred to as the ‘New Protectionism’ to distinguish it from the oldstyle tariff protectionism of the past. Some of the forms of the New Protectionism such as import quotas and voluntary export restraints are potentially more damaging than tariffs. The original GATT agreement proved largely inadequate for coping with this new challenge. Furthermore, the forms of negotiation used to bring about a lowering of tariffs were generally inappropriate for dealing with nontariff barriers. One aspect of this is the difficulty of quantifying the impact of a nontariff barrier on trade. It therefore becomes impossible to negotiate balanced, reciprocal reductions in the level of NTBs in the same way as happens with tariffs. That is to say, where a country wishes to match concessions made with concessions received, there may be a problem of how to quantify the effects of any reduction in the level of a particular NTB. New approaches had to be explored. It was not until the Tokyo Round that any serious attempt was made to come to grips with the problem of nontariff barriers. The approach used was largely one of developing new codes dealing with particular types of NTBs which acted as extensions to the basic GATT agreement and to which countries had to agree to adhere. For example, in the next chapter the codes agreed to cover antidumping policy and subsidies will be discussed. A close link exists between some types of nontariff protectionism and the GATT rules for so-called ‘emergency protection’. When drafting the GATT Escape or Safeguards Clause, the intention of the architects of the GATT was to provide a route whereby countries could, in the event of an emergency, retreat from tariff concessions granted in previous negotiations. For example, if a domestic industry was threatened by a sudden surge of imports, a country may wish to raise a tariff which had been bound in the course of a previous round. Unless countries could be assured of an escape in an emergency from obligations entered into in the past, they would be unwilling to make meaningful concessions in tariff-cutting rounds. In practice, countries have often preferred to bypass the Safeguards Clause when faced with a demand from a domestic industry for protection. Instead, some form of quantitative restriction on trade has often been introduced. Many voluntary export restraint arrangements have often come into being for precisely this reason. Therefore, this chapter concludes with a discussion of the issue of safeguards and explores the debate which has surrounded the issue of its reform. TYPES OF NONTARIFF BARRIERS Olechowski (1987) has defined NTBs as ‘all public regulations and government practices that introduce unequal treatment for domestic and foreign goods of the same or similar production’. This covers a wide variety of different forms of trade restriction, including those where the intent is to reduce imports and those which serve some other purpose but where a reduction of imports is a secondary effect. Sometimes, a distinction is drawn between direct and indirect forms of nontariff intervention to distinguish between those where the primary intent is to restrict imports (direct) and those where there exists some other purpose but where imports are nevertheless affected (indirect) (Greenaway, 1983). Sometimes, a distinction is made between nontariff barriers and so-called nonborder measures (Finger and Olechowski, 1987). A nonborder measure is any measure other than border measures (for example, a tariff or quantitative import restriction) which also affects trade (Messerlin, 1987). For instance, subsidies to domestic producers are a nonborder measure which may distort trade. The expression managed trade is sometimes used with reference to trade that is subject to forms of nontariff intervention (see, for example, Page, 1981). An expression frequently used to refer to the use of nontariff measures for restricting imports is administered protection (Bhagwati, 1988). This is useful because it emphasises that the intensity of nontariff forms of protection can usually be altered without the need for the enactment of any new legislation. By way of contrast, International Trade Policy 44 any change in the level of a tariff does require the prior consent of the legislature. (Yet another often-used expression is that of ‘contingent protection’. This was first used by Grey, 1986, to refer to forms of protection which depend upon demonstrating that imports have caused injury to domestic producers. These include safeguard measures, which are discussed towards the end of this chapter, and antidumping policy, which is discussed in the next chapter.) One of the problems involved in any examination of the nature of NTBs is how to classify the wide variety of different types which exist. An approach often used in the classification of NTBs is to list them according to whether trade-distortion is the primary intention of the authorities responsible. Using this approach, Walter (1972) distinguished between three types of NTBs: those with a trade-distorting intent; those with only a secondary trade-restriction intent; and those with no trade-restriction intent but with spillover effects on trade. The different types of NTBs in each category are given in Table 3.1. For each type of NTB, a distinction is drawn between quantitatively operating measures and measures which operate through prices and costs. Thus, measures such as import quotas, voluntary export restraints or embargoes, which are clearly intended to restrict trade, operate by placing quantitative limits on imports/exports. Other measures such as variable import levies (common in agricultural trade), antidumping duties or subsidies to import competitors are also intended to restrict imports but work essentially by either raising the price of imports (variable import levies and antidumping duties) or lowering the price of domestically produced Table 3.1 Types of nontariff barriers classified according to the normal intention of the measure Type 1 measures (trade-distorting intent for imports) Type 2 measures (secondary trade restrictive intent) Type 3 measures (spillover effects on trade) A Qunatitatively operating 1 Global import quotas 2 Bilateral import quotas 3 Restrictive licensing 4 Liberal licensing 5 Voluntary export restraints 6 Embargoes 7 Government procurement 8 State trading practices 9 Domestic content regulations 1 Communications media restrictions 2 Quantitative advertising restrictions B Operating on prices/costs 1 Government manufacturing and distribution monopolies covering products such as armaments 2 Government structural and regional development policies 3 Ad hoc government balance of payments measures 4 Variations in national tax schemes 5 Variations in allowable capitaldepreciation methods 6 Variations in allowable capitaldepreciation methods 7 Spillovers from governmentfinanced defence, aerospace & nonmilitary projects 8 Scale effects induced by government procurement Quantitative Trade Restrictions and Safeguards 45 Figure 3.1 The effects of an import quota on the importing country importing nation. This leaves areas B+D=area E as the net welfare loss or deadweight loss from the quota. This is exactly the same as for a tariff. In view of the fact that the welfare loss from a quota is identical with that of a tariff, it may seem strange that quotas are widely regarded as being more harmful than tariffs. One reason is that quotas involve the state allocating licences to importers whereas tariffs rely on the price mechanism. If it is left to government officials to allocate licences, importers will seek to bribe them in order to get bigger quotas. Even if officials are not open to bribes, decisions must still be made about which firms are to be allocated licences and how much the licence should permit each to import. Government officials lack the information to make the right (that is, most efficient) decisions. One solution would be to auction licences to the highest bidder. This is preferable since it will ensure that licences go to the producers who can make the most efficient use of imports. Secondly, where the quota-restrained product is a raw material or intermediate good used as an input by other industries, quotas create rigidities in the structure of production within the importing country unless licences are marketable. Efficient producers who require more of the input cannot expand production while less efficient producers who are compelled to reduce production fail fully to utilise their quotas. Similarly, where the quota-restrained input is used by two or more industries, industries whose product faces increasing demand may be unable to expand output at the expense of other industries whose products face falling demand. Thirdly, where quotas are administratively allocated, the degree of protection and therefore the deadweight loss resulting from the quota will increase over time. Thus, in Figure 3.1, a rise in domestic demand causes the demand curve to shift from DD to D1D1. In the absence of any increase in the quota, all the increase in demand has to be satisfied by higher-cost domestic production rather than lower-cost imports. Consequently, price rises to OP2. In the case of a tariff, a rise in demand is met fully by imports and the price does not rise at all. International Trade Policy 52 Finally, to a greater extent than tariffs, quotas are discriminatory although they do not have to be. Too often, they are targeted at the world’s lowest-cost suppliers of the product. There would therefore seem to be merit in taking a tougher attitude towards import quotas than towards tariffs. This is indeed the position of the GATT. Article XI opposes all forms of quantitative restriction on trade except in special circumstances. However, the exceptions have been extensively used such that import quotas remain an important barrier to trade. One exception is where quotas are needed to enforce ‘standards or regulations for the classification, grading or marketing of commodities in international trade’ (Article XI, 2: b). Another is agricultural products where restrictions are required in order to restrict domestic supply or to remove a temporary surplus of the domestic product (Article XI, 2: c). Quotas affecting agricultural imports were also subject to the 1955 waiver granted to the United States that arose out of a statute passed by Congress in 1951 mandating quotas on certain agricultural imports. Subsequently, other countries used the US waiver to excuse similar practices. (See Chapter 6 for a fuller discussion of agricultural protectionism.) Not surprisingly, quotas are much more common in agricultural trade. Article XII also permits the use of quotas for the purpose of safeguarding a country’s balance of payments. However, any such restrictions must be progressively relaxed as the country’s balance of payments improves. In other words, they are to be temporary. Interestingly, the GATT rules authorise quotas rather than tariffs where trade restrictions are needed for balance of payments conditions. In practice, countries have generally preferred tariff ‘surcharges’ as a device when faced with such difficulties. Developing countries, which frequently encounter balance of payments problems, have made considerable use of this provision to impose import quotas. Particularly important in this respect is Article XVIII of the GATT under which developing countries have fairly broad freedom to impose quantitative restrictions on imports for general developmental reasons (see Chapter 5 for a fuller discussion). Finally, a major departure from the GATT rules on quotas was allowed in the case of textile and clothing products following the Long-term Cotton Textile Arrangement (LTA) of 1962. This was preceded by the Short-term Arrangement (STA) of 1961–2. The STA came into being following a GATT working party report to investigate the problem of textile protectionism. A variety of different measures, illegal under the GATT, already existed in a number of developed countries for controlling textile imports. The report found that there was a problem of ‘market disruption’ caused by import surges from lowwage countries. The STA permitted quotas on cotton fabrics and clothes as a temporary measure pending the completion of negotiations. The LTA, which lasted until 1973, required importing countries to drop their existing restrictions on imports of cotton textiles but allowed new ones only if and when they faced market disruption from actual or planned imports. These could take the form of import quotas, but the quotas must not be less than actual trade before the disruption and were to have a built-in growth factor of 5 per cent a year, that is, they were to allow for an expansion of trade of this amount. The Multi-Fibre Agrement (MFA) replaced the LTA of 1962. It extended the arrangements to cover noncotton textiles including man-made or synthetic fibres such as polyester and acrylic. The 1974 MFA was for three years only but was replaced by MFA2 in 1978. Each subsequent MFA similarly lasted for three years and was followed by a new Quantitative Trade Restrictions and Safeguards 53 agreement, each of which involved some modification of the previous one. Textiles trade is discussed more extensively in Chapter 5. VOLUNTARY EXPORT RESTRAINTS One of the commonest forms of quantitative restriction on trade in recent decades has been the voluntary export restraint (VER). Hamilon (1985b) has defined this as ‘the outcome of negotiation between two governments resulting in the exporting country limiting its export supply to the importing country’. In fact, a VER need not, and frequently does not, take the form of a government-to-government agreement. It may equally take the form of an agreement between industry groups in the exporting and importing country, for example, through an industry association. (Such agreements may, however, contravene antitrust laws.) The governments of the two countries, however, are likely to be tacitly in favour of the accord and may even have been instrumental in bringing it about. Yet another possibility is an agreement between the government of one country (usually, the importing country) and a nongovernment group in the other (for example, the exporting industry of another country). In a number of countries, government-togovernment agreements are referred to as orderly marketing arrangements (OMAs). In the United States, OMAs are legally distinct from arrangements involving industry participation. OMAs are legally binding, meaning that the terms of the agreement cannot be modified in any way without the consent of both of the parties. Typically, the agreement will include detailed rules about export supply, rights of consultation and the monitoring of trade flows. The term voluntary restraint arrangement (VRA) is frequently used to cover, in addition to OMAs, arrangements with industry participation. By way of contrast, there are many kinds of more ‘informal’ agreements which are not legally binding. These may take the form of some sort of statement by the exporting country designed to ensure that the exports of a particular product are kept below a certain limit. In this case, the exporting country has the right at any time to abandon or modify the restriction. In some cases, the arrangement merely entails exporters making some ‘forecast’ or prediction about export volume with some undertaking about the monitoring of exports. In effect, these amount to restrictive arrangements also. Usually, a VER will involve either a restraint on export volume or a minimum export price. Where there is a limitation on export volume, this may be expressed in terms of some maximum absolute level of exports or in terms of some maximum share of the market of the importing country. Kostecki (1987) distinguishes three different methods of dividing up the importing country’s market between domestic and foreign producers: the home-industry-first approach; the exporters-first approach; and the market-share approach. The differences are important for determining how expansion or decline of the market is shared out between domestic and foreign producers. In the home-industry-first approach, domestic producers are given a certain minimum level of sales. It follows that, if the market declines, all the risks are born by foreign producers. In the exporters-first approach, exporters are given a certain level of sales. If the market declines, the risks fall entirely on the domestic producers. In the market-share approach, exporters are given a certain percentage share of the market. If demand falls, domestic and foreign producers International Trade Policy 54 lose out equally so the risks are shared evenly. Usually, the government of the exporting country will undertake to allocate export quotas to its own producers. The export quota may be given in either volume or value terms but volume quotas are more common. Quotas may be allocated to exporters on the basis of some predetermined criteria or they may be auctioned. Usually, the export limitation is for some specified period of time, rarely more than five years, although it may be, and frequently is, subsequently renewed. It is apparent that there are many similarities between VERs and bilateral import quotas. Indeed, the economic effects of the two are similar. The obvious differences are that bilateral quotas are restrictions which are generally imposed by an importing country on the exports of another country and the task of enforcing these limitations resides entirely with the importing country. As we have seen, the GATT rules governing the use of import quotas are fairly clear even if the permitted exceptions to the rules mean that quotas are still widely used. However, the legal position with regard to VERs has always been more uncertain. In two respects, VERs would appear to involve a clear breach of GATT rules. Firstly, Article XI, which prohibits quantitative restrictions on trade covers both import and export restrictions. Secondly, since most VERs are discriminatory, they violate the most-favoured-nation rule set out in Article I. However, problems arise because, in many cases, the involvement of governments in bringing about the restriction is not always clear-cut. GATT rules do not cover the actions of private companies. In some cases, nongovernment bodies are the originators of the restraint. Further problems arise because VERs are negotiated agreements, not unilaterally imposed measures. It is therefore not obviously the case that the rights of another contracting party are being violated. Finally, the fact that trade is being restricted is not obvious in all cases, especially when the VER takes the form of a ‘prediction’ or ‘forecast’ of export trend. For these reasons, VERs fall into the category of what have come to be called ‘grey area’ measures: measures whose legality under existing GATT rules is uncertain. A major problem has been that few countries have sought to test the legality of a VER by making a complaint to the GATT. This is not surprising since the two parties directly affected have agreed to the arrangement, and presumably they would not have done so had there been a better alternative. Only some third party which feels it is being harmed by a VER agreed between two other countries is likely to make a complaint. Even this is improbable since other exporting countries stand to gain from the restriction by being able to export more to the importing country. On the other hand, one effect of a VER may be to cause the exporting country to divert exports to some third market. If so, producers in the latter may regard the increased competition as the direct result of the VER and, if injured, call upon the authorities to lodge a complaint against the VER. One case of a third-country complaint to the GATT was that lodged by the EC in 1987 against the USJapan Semiconductor Agreement. The agreement contained two aspects. Firstly, an undertaking by the Japanese government to increase the share of the domestic market taken by foreign producers. Secondly, an undertaking by both countries not to sell below agreed minimum prices in the US and third markets. The second aspect was the source of the EC complaint since the effect of the agreement was to raise the cost of memory chips to European computer manufacturers. The complaints panel found in favour of the EC and the two countries were forced to modify this aspect of the agreement. Although VERs have only been widely used in the last few decades, they date back to the 1930s and were applied to trade in textiles. The first example appears to have been an Quantitative Trade Restrictions and Safeguards 55 agreement reached in 1937 between the American and Japanese trade associations to limit Japan’s textile exports to the US. Recognising the existence of a ‘special’ problem facing the textile industry, the US administration allowed such an agreement to be reached. The only alternative would have been discriminatory quotas; this was not legally possible and would have run counter to the trade agreements policy of the Roosevelt Administration of that time (Wolf, 1989a). At the time, the VER was regarded as being a temporary measure to deal with an ‘exceptional’ situation and was not intended to set any precedent for other sectors in the future. In actuality, things turned out rather differently. After the Second World War, the problems of the US textile industry remained. Beginning in 1955, further voluntary export restraints were applied by Japan to her exports of cotton textiles to the US. In January 1955, Japan gave American officials details of a five-year programme of Voluntary’ export controls. In fact, the restrictions were extracted under pressure from the US administration, which in turn might have been forced by Congress to implement import quotas had restrictions not been offered. These controls did not end the matter as exports from other textile-producing countries expanded to take the place vacated by Japan, leading to demands to extend controls to other countries. The problems were not confined to the US. In 1959, Hong Kong, India and Pakistan reached a voluntary export agreement with the UK regulating trade in cotton fabrics. The US administration, however, was unable to persuade Hong Kong to apply similar restraints on cotton exports to the US. In the 1960s, political pressure built up on the Kennedy Administration to give support to the US textile industry. The problem was how to do this without the use of import quotas which would have violated the GATT rules and would have led to further demands for protection from other sectors similarly faced with severe import competition. The solution was, first, the Short-term Cotton Textile Arrangement (STA) of 1961 and then a year later the Long-term Cotton Textile Arrangement (LTA). In 1973, it was replaced by the Multi-fibre Arrangment (MFA). The MFA extended the LTA to trade in textiles and clothing made of synthetic fibres. Thus, trade in textiles and clothing has been subject to more or less permanent control for much of the period since just before the Second World War. Strictly speaking, the controls applied to trade in textiles and clothing under the LTA and MFA were bilateral quotas. Nevertheless, since they are agreed between exporting and importing countries after negotiation, they are in essence the same as a VER. Gradually, VERs were extended to sectors other than textiles. The three where VERs have been most widely used are automobiles, steel and consumer electronics. One of the earliest examples of a VER in automobiles was the agreement which the UK entered into with Japan in 1977, which froze Japan’s share of the UK automobile market at 11 per cent. This took the form of a market-sharing agreement between the industry associations of the two countries although with overt support from the two governments. At about the same time, France negotiated a similar agreement with Japan. In May 1981, the US negotiated a VER with the Japanese government which effectively reduced US imports from Japan by 140,000 units (about 7.5 per cent) from their 1980 level. As with textiles, this came into being under threat of statutory quotas. Following the US-Japan agreement, West Germany negotiated an agreement with Japan designed to limit the rate of increase of Japanese exports to the FRG to 10 per cent a year. The Netherlands and Belgium also negotiated agreements with Japan which froze the Japanese share of the market at its International Trade Policy 56 1980 level. Some of these agreements subsequently expired but several others were renewed. With the decision to establish a Single Market in which goods could no longer be checked as they crossed national borders, it became impossible for the EC to operate national VERs. In the absence of any border controls, quotas in ‘controlled’ markets would be undermined as cars were imported from ‘uncontrolled’ markets. Therefore, in June 1991, the various national VERs then in place were replaced with a new Community-wide VER freshly negotiated with Japan. Although the details of this agreement have been the source of much controversy, it appears that the agreement freezes the share of the EC market at its then existent level. This means that the volume of Japanese cars sold in the EC can only increase if demand for cars increases at the same time. The agreement also includes sub-ceilings for the various member state markets where national VERs previously operated. At the same time, the EC declared its intention to abolish all controls on imports of Japanese cars by the end of the decade. With regard to steel, many of the VERs which came into being were responses to alleged cases of dumping or subsidised trading and were presented as being alternatives to the imposition of antidumping or countervailing measures. The first VER was negotiated in 1968 between the US and both Japan and the EC. This lasted until 1974 when imports fell below the ceilings and the agreement was not renewed. In 1977 a socalled ‘trigger-price mechanism’ was introduced which provided the US industry with protection from imports sold in the US below a stated reference price based on estimates of costs of production in Japan. There also appears to have been a tacit agreement between the US and Japan since the late 1970s limiting Japanese exports to the US. In the case of the EC, a whole series of VERs were introduced as part of the EC’s crisis measures (the so-called Davignon Plan) for tackling the problems of overproduction, excess capacity, falling prices and mounting losses. Unable to sell all the steel they were producing in the depressed European market, many European producers backed by heavy state subsidies began to ship more steel to the US. The response of US producers was predictable. Rather then face antidumping measures, the EC preferred to sign a VER with the US. This came into being in 1982. It was followed by a whole spate of VERs between the US and virtually every other major foreign supplier (regardless of whether they were dumping and/or causing injury to domestic producers). The third sector which has been most affected by the spread of VERs has been the electronics industry, especially consumer electronics. VERs largely date back to the late 1970s and early 1980s and have involved Japan and other Far Eastern suppliers. In a manner similar to the steel industry, VERs have often been negotiated as an alternative to antidumping measures. Allegations of dumping by Far Eastern suppliers have been rife within Western Europe and the United States. In 1977, imports of television sets to the UK from Taiwan, South Korea and Singapore were also subject to a VER. In 1979, the US negotiated a VER with Japan restricting imports of television sets. Subsequently, this was widened to include Taiwan and South Korea. In 1983, the EC negotiated a VER with Japan covering imported video-cassette recorders (VCRs). The background to this was alleged dumping by Japanese producers in the European market. This particular VER was also significant in being the first-ever EC-negotiated VER, that is, a VER negotiated by the EC on behalf of all the member states with another country. In 1986, VERs spread to the industrial electronics sector with the negotiation of the US– Japan Semiconductor Agreement referred to above. As noted above, this had two aspects: (a) an agreement Quantitative Trade Restrictions and Safeguards 57 entered into by Japan not to sell various kinds of semiconductors (microchips) below a stipulated price both in the US and third markets; and (b) an undertaking by Japan to increase the United States’ share of the Japanese domestic market. Tables 3.5 to 3.7 show which sectors and which countries are most affected by VERs. Table 3.5 reveals that over one-half of all VERs were applied to exports of countries other than developed countries, a proportion far exceeding these countries’ share of world trade. Prominent among the restrained exporting countries were newly industrialising countries such as South Korea, China and Taiwan. Nearly one-half of all VERs applied to exports of developed countries affected Japan. Table 3.6 shows that, apart from agricultural products, which accounted for 18 per cent of all VERs, the majority of VERs were to be found in five sectors—iron and steel products, textiles and clothing, automobiles and transport equipment, electronic products and footwear. It should be noted that these are the sectors within the developed countries which, in recent decades, have experienced the greatest adjustment difficulties. Table 3.7 shows that, of the importing countries protected by VERs, the EC and the USA accounted for more than two-thirds of the total number of cases. How much of world trade is affected by VERs? One attempt to estimate the importance of VERs as a barrier/distortion to trade estimated that, by 1987, not less than 10 per cent of world trade and about 12 per cent of nonfuel trade was covered by VERs (Kostecki, 1987). However, this does not tell us how much trade is affected by VERs since there is clearly some unknown quantum of trade which would have taken place had these VERs not existed. It follows that the amount of trade affected is much greater. Moreover, for certain sectors, the proportion of trade covered by VERs is Table 3.5 The prevalence of voluntary export restraints, by restrained exporting country (excluding the Multi-fibre Agreement) 1986–7 Restrained exporting countryNo. of arrangements Percentage of total number of cases Developed countries of which: 56 40.9 Japan 27 19.7 EC 4 2.9 Australia 4 2.9 New Zealand 3 2.2 Sweden 3 2.2 Austria 3 2.2 Developing countries of which: 54 39.4 South Korea 17 12.4 China 6 4.4 Taiwan 5 3.6 Brazil 4 2.9 Pakistan 4 2.9 South Africa 4 2.9 Socialist countries of which: 27 19.7 Eastern Europe 18 13.1 China 6 4.4 International Trade Policy 58 Other 3 2.2 Total 137 100.0 Source: Kostecki (1987) Table 3.6 The prevalence of voluntary export restraints, by product group (excluding the Multifibre Agreement), 1986–7 Major known VERs No. of arrangements Percentage of total number of cases Iron & steel products 44 32 Textiles & clothing 25 18 Machine tools 6 4 Automobiles & transport equipment 15 11 Electronic products 10 7 Footwear 8 6 Agricultural products 24 18 Othera 54 Total 137 100 Source: Kostecki (1987) Note: aProducts involved were kraftliner, stainless-steel flatware, leather clothing and softwood lumber. Table 3.7 The prevalence of voluntary export restraints, by protecting importing country (excluding the Multi-fibre Agreement), 1986–7 Protecting importing country No. of arrangements Percentage of total number of trade cases Australia 1 0.7 Austria 1 0.7 Canada 10 7.3 EC 52 38.0 Finland 2 1.5 France 2 1.5 Italy 3 2.2 Japan 4 2.9 Norway 5 3.6 Portugal 1 0.7 Spain 2 1.5 UK 8 5.8 USA 45 32.8 W. Germany 1 0.7 Total 137 100.0 Source: Kostecki (1987) Quantitative Trade Restrictions and Safeguards 59 much higher. Kostecki estimated that 80 per cent of world trade in textiles and clothing is regulated by the MFA, with part of the remainder covered by bilateral export restraints involving non-MFA countries. An estimated 20 per cent of world trade in steel and steel products is subject to VERs (Kostecki, 1987). The 1986 US–Japan semiconductor agreement meant that 90 per cent of world trade in semiconductors was subject to a single VER. Finally, the proportion of trade covered by VERs is higher than average for certain countries. Kostecki puts the import-weighted coverage of VERs at 38 per cent for EC imports from Japan and not much less than 33 per cent for US imports from Japan. The economic analysis of VERs Hamilton (1984c, 1985b) has analysed the economic effects of VERs in partial equilibrium terms using two models: one for the case of a nondiscriminatory VER involving an importing country and all foreign suppliers of the product; and the other the case of a discriminating VER involving an importing country and one source of supply. The case of a nondiscriminatory VER is set out in Figure 3.2. DDDD is the demand curve for the product in the importing country and SDSD the domestic supply curve. SWSW is the combined domestic plus foreign supply curve which Figure 3.2 The effects of a voluntary export restraint on the importing country Source: Hamilton (1984a) International Trade Policy 60 is more elastic than the domestic supply curve. OP0 is the equilibrium price under free trade with domestic consumption equal to OQ4, domestic production equal to OQ1 and imports equal to Q1Q4. The importing country wishes to reduce the level of imports to Q2Q3. To do this, it enters into a VER with foreign suppliers. We can imagine either that it enters into a VER with all foreign suppliers simultaneously or that the only foreign producer is the one with whom it negotiates a VER. The effect of the VER is identical to that of an import quota: the equilibrium price rises to OP1, domestic consumption falls to OQ3, domestic production rises to OQ2 and imports fall to Q2Q3. However, although the market price in the importing country has risen to OP1, the foreign supply price has fallen to OP2. The logic behind this is that a nondiscriminatory VER applied to all suppliers will create excess capacity in the world industry, resulting in lower short-run marginal costs. This means that foreign suppliers can enjoy a windfall profit of P2P1 on every unit sold. In Figure 3.2, the shaded area C+E is the rent income which accrues to foreign suppliers. In the case of an import quota, the equivalent of this area constitutes economic rent to importers. In both cases, this represents a loss of income for consumers but, in the case of a VER, it is also a loss to the importing country. The loss of consumer surplus is areas A+B+C+D. The net welfare loss to the importing country is B+C+D+E. This is clearly much greater than for either a tariff or an import quota. The reason for this is that a VER worsens the terms of trade of the importing country by raising the cost of imports to the importing nation. This does not happen with either a tariff or a quota. The fact that foreign exporters enjoy a markup on every unit sold is one reason why exporting countries are willing to agree to restrain their exports. What is less obvious is why the importing country should prefer a VER to a tariff or quota as a device for restricting imports. Even the hoped-for improvement in the balance of trade is only assured if the demand for imports is elastic (such that the price increase is proportionately less than the volume decrease). The more realistic model is that of a discriminating VER. Most VERs are sourcespecific, covering some sources of supply (often the lowest-cost suppliers) but not all. They therefore have certain efffects on the pattern and not just the volume of trade and production. Figure 3.3 illustrates this case. There are three countries: the importing home country, an unrestrained partner country (or countries) and a restrained outside country (or countries). (The situation is analagous with that of a customs union made up of the importing country and the partner country. Exports from the partner country are not subject to the VER but exports from the rest of the world are.) DDDD is the demand curve for the product in the importing home country, SDSD is the supply curve in the importing home country and SdnSdn is the combined supply curve of the importing home country and the partner country. SWSW is the supply curve of the outside country. (The importing home country is assumed to be sufficiently small such that the world supply price is unaffected.) OP0 is the free-trade equilibrium price in the importing country. Domestic consumption is OQ4, domestic production OQ1 and imports Q1Q4, of which Q1Q2 comes from the partner country and Q2Q4 from the outside country. Now, the government of the importing home country decides that it wants to increase domestic production to OQ2. It therefore negotiates a VER with the outside country limiting its exports to DF. The partner country continues to enjoy free trade. The effect is to raise the equilibrium price to OP1 in the importing home country. Domestic consumption falls to OQ3, domestic Quantitative Trade Restrictions and Safeguards 61 exports of the country taking safeguard action. In order to avoid such an outcome, the importing country may prefer to enter into a VER with the country whose exports are causing the problem. VERs are generally source-specific and so avoid upsetting third countries in the way a nondiscriminatory tariff or quota would. In fact, they contain an element of ‘built-in’ compensation for the exporting country in the form of the rent income which exporters enjoy. Thus, the issues of compensation and of export restraint are simultaneously dealt with in the same negotiation rather than being the subject of two or more separate negotiations. There are also domestic political advantages for governments in using VERs in preference to other protectionist measures to deal with troublesome imports. Rising tariffs or import quotas are much more visible and thus inevitably attract public debate. In particular, they are more likely to generate opposition from consumer groups. Using a VER to appease domestic producers clamouring for protection may enable the government to pass the responsibility for the restriction on to the exporting country since the latter has the task of implementing the restraint. Tariffs or quotas may also take much longer to introduce because of the necessity for the measures to be first discussed and then approved by national legislatures. Where exports are being subsidised or dumped, a VER may again be preferred to a countervailing or antidumping duty. (A countervailing duty is a levy imposed on an imported product which has been subsidised such that the export price is lower than the domestic price, resulting in injury to producers in the importing country. An antidumping duty, which is a levy imposed on a product imported at a price below its ‘normal value’ (often taken to be the domestic price) and where dumping has been shown to cause material injury to domestic producers. (These practices are discussed further in Chapter 4.)) VERs avoid the lengthy and expensive procedures which an antisubsidy or antidumping action often require. Bhagwati (1988) has argued that VERs are a porous form of protection because they can be easily circumvented. For example, exporting countries may be able to get round a VER by exporting the goods through some uncontrolled third country or by setting up an assembly plant in a third country. Upgrading the product is another method of getting round a VER. This begs the question: if they are relatively ineffective, why do governments of importing countries favour them? One explanation may be that governments do not in fact desire effective restrictions on trade but must nevertheless be seen to respond to demands for protection from legislators and their constituencies. Bhagwati has suggested that the executive branch of the government is often biased in favour of freer trade but the legislative branch is more amenable to sectional pressures for protection from interest groups. According to this view, VERs are devices used by governments to resolve conflict between the different branches of the state. If this is so, they may have certain advantages rather than constituting a hindrance to freer trade. VERs also have certain attractions for exporting countries. Firstly, exporting countries are likely to prefer them to tariffs because they generate rent income for exporters whereas tariffs create revenues for the importing country. On the other hand, as Kostecki (1987) has argued, not all of the windfall gain for exporters will accrue to the exporting firm. Rather, these gains will be shared between exporters and distributors in the importing country, so they may not be as great as is often thought. Kostecki quotes one study of restraints on Japanese car exports to the US which found that 60 per cent of the windfall gains went to US dealers and only 40 per cent to Japanese car producers. International Trade Policy 68 Secondly, they may have an appeal to a high-cost exporting country because they guarantee their share of the market of the importing country. They are protected from low-cost exporters from another country. This makes VERs a costly form of protection for the importing country but can explain their appeal to the exporting country. The wellestablished exporting firms in the exporting country will find a VER especially appealing because it makes it more difficult for newcomers to compete with them. This will be the case where the exporting country enforces the VER by allocating quotas to exporters based on some criterion such as export volume/share in the period before the VER came into being. Thirdly, exporting countries may prefer a VER to an antidumping or antisubsidy investigation which are both costly and time-consuming. The outcome is uncertain for the duration of the investigation. Moreover, the result of such a case is often the imposition of an antidumping or a countervailing duty which may be quite punitive. As with a tariff, the revenue accrues to the importing country, whereas, with a VER, rent income is earned by the exporting country at the expense of the importing country. (However, as is explained in Chapter 4, dumping actions are sometimes settled by the exporter making price undertakings rather than by the imposition of duties. In this case, rents are transferred to the exporting country in much the same way as with a VER.) GATT SAFEGUARD PROVISIONS AND THE GROWTH OF VERs It is clear that there is a close connection between the growing use of VERs and the safeguard provisions of the GATT. In many respects, VERs have increasingly come into use because of deficiencies which importing countries consider to exist with the GATT Safeguard Clause (Article XIX). Consequently, the issue of how to contain the growth of grey-area measures such as VERs has been closely linked to discussions about reform of the GATT safeguard provisions. Hence, in the remainder of this chapter, the subject of safeguards is examined. This has been a key issue on the agenda of the last two rounds of the GATT. Attempts to reform the Safeguard Clause in the course of the Tokyo Round were unsuccessful largely because of disagreement between members of the European Community and other countries over the issue of selectivity. However, agreement was finally reached in the Uruguay Round in the form of a new Agreement on Safeguards. This is intended as a clarification and reinforcement of Article XIX. It is necessary to begin by asking why an agreement such as the GATT should need any Escape Clause. Is there any point in countries negotiating tariff reductions or tariff bindings if they are allowed, albeit in an emergency only, to withdraw such concessions? The reasons are both economic and political. At the purely economic level, the case rests on the so-called adjustment problem which can arise as a consequence of trade liberalisation. Lowering trade barriers leaves domestic industries more vulnerable to the sudden, unforeseen emergence of new sources of competition in another country. With perfectly competitive product and factor markets no problem need arise. In response to differences in prices and costs, resources would shift more or less instantaneously from declining to expanding sectors. In reality, markets are not perfect and so adjustment fails to take place sufficiently rapidly. As a result, additional costs may be created both for the Quantitative Trade Restrictions and Safeguards 69 owners of factors employed in the import-competing sector and for society as a whole. Temporary protection may be needed to allow time for the necessary adjustment to take place. The aim is to buy time in which factors can move out of the import-competing sector and into other expanding sectors of the economy. Furthermore, by increasing factor incomes in the protected sector, temporary protection helps to offset the private costs to factor owners that arise from adjustment. For example, higher profits may help producers to finance necessary rationalisation in the protected sector. Closer examination shows the economic argument for such protection to be a weak one. If trade expansion creates social costs, the economically most efficient solution is to seek to remove the source of market imperfection which gave rise to the adjustment problem in the first place. For example, if the problem is imperfections in the workings of the labour market, the best solution is to reform the way in which the labour market works so as to make wage-rates more responsive to demand and supply factors. Trade restrictions serve only to impose additional costs on the rest of society. These are rarely taken into account when import barriers are imposed for adjustment reasons. On purely economic grounds, temporary restrictions can only be justified if the marginal social costs (that is, the adjustment costs to society as a whole) of allowing increased imports are found to exceed the marginal social benefits (the welfare gains from trade expansion). In most cases where temporary protection is granted, it would seem improbable that this is the case. Certainly, little or no attempt is ever made to estimate costs and benefits in this way. If, in fact, costs are found to exceed benefits and temporary restrictions are deemed to be desirable, two further considerations need to be taken into account: firstly, how to ensure that adjustment does in fact take place during the period in which the restrictions are in place; secondly, how to ensure that import barriers are progressively lowered as adjustment takes place. The two points are related. Unless a definite timetable is established for the progressive lowering of barriers, there will be no incentive for producers in the protected sector to make the necessary adjustments. In this case, trade restrictions will delay rather than facilitate adjustment. If restrictions are retained beyond the period required, then protection will be positively harmful since the gains forfeited will exceed the costs saved. Clearly, given the tendency for temporary restrictions to remain in place for a longer period than was at first envisaged (for example, the Multifibre Arrangement), this is frequently the case. Indeed, the fact that temporary restrictions so often become permanent raises doubts as to whether the true motive for the restrictions in the first place is indeed the need to reduce adjustment costs. The political case for allowing countries to introduce emergency protection is stronger. Unless domestic producers in industries where trade liberalisation is taking place are assured of a possible escape route in the event of difficulties, there will be a reluctance on their part to agree to concessions being made. In other words, the inclusion of a Safeguard Clause may help governments to secure the agreement of producers, particularly in so-called sensitive sectors, to tariff cuts and/or tariff bindings. It may serve to allay any fears among producers in sectors which have in the past enjoyed high levels of protection that they will be defenceless in the event of a sudden, unforeseen surge of imports. At the same time, by allowing temporary protection if and when an expansion of trade causes adjustment problems, it can reduce private adjustment costs to factor owners and so weaken the case in favour of high levels of permanent protection. On the other International Trade Policy 70 hand, it creates a risk that emergency protection will be hijacked by producer interests in declining sectors as a device for increasing economic rents at the expense of producers in expanding sectors. Almost certainly, this risk is one that has to be taken if meaningful progress is to be made in lowering trade barriers and improving market access. The main provisions of Article XIX, as set out in the General Agreement and as it has been applied until now, can be summarised as follows: 1 It applies to ‘unforeseen developments’ and ‘the effect of obligations incurred by a contracting party’. 2 Increased imports must cause or threaten ‘serious injury’ to domestic producers, but nowhere is ‘serious injury’ defined. 3 Action is to take the form of the suspension of obligations (in whole or in part) or the withdrawal or modification of negotiated tariff concessions. 4 Such measures may be taken for ‘such time as may be necessary to prevent or remedy such an injury’, that is, they should be temporary, although the duration of any such measures is not specified precisely. 5 Although selective safeguards are nowhere specifically prohibited, it must be presumed that any measures taken should be nondiscriminatory in order to conform with Article I of the GATT, especially as there is no statement to the contrary. 6 Prior notice should normally be given of any safeguard measures which a country intends to take so as to allow for consultation with countries that have a substantial interest as exporters of the product. An exception is made where delay in introducing restrictions might cause damage to a country, in which case consultation should take place ‘immediately after taking such action’. 7 If, through consultation, agreement cannot be reached between the country taking the action and other interested parties, the latter are entitled to take retaliatory action against the country invoking Article XIX. This can take the form of affected parties, not later than ninety days after the action is taken, suspending ‘substantially equivalent concessions or other obligations’. A period of thirty days must elapse between notice of suspension and the implementation of the retaliatory measures. Article XIX has been widely been regarded as being inadequate as a Safeguards Clause for a variety of reasons. Firstly, the criteria for determining ‘serious injury’ (or the threat of it) are not specified. In many cases, any increase in imports relative to domestic production is considered to constitute serious injury. The absence of any definition of serious injury has meant that governments are more or less free to interpret it as they wish. It is left to the exporting country to prove that injury has not occurred by lodging a complaint with the GATT concerning action which has been taken. In practice, few countries have been prepared to do so. Secondly, the types of measure which are permissible for remedying an injury are not specified. The language used in the clause, which includes a reference to ‘suspension of obligations’ as well as ‘withdrawal or modification of concessions’, seems to suggest that restrictions could take the form of quantitative restrictions and not just a tariff increase. This indeed has become normal practice, although on theoretical grounds tariffs are to be preferred to quotas. If, however, quotas are used, there is a need to ensure that they are not too restrictive. Quantitative Trade Restrictions and Safeguards 71 Thirdly, there is no indication of how long such measures should be allowed, other than a vague reference to ‘for such time as may be necessary to prevent or remedy such injury’. This seems to imply that such measures should be temporary but there is no guidance on when the restrictions should be eliminated. As noted earlier, if no time limit is specified, such temporary measures will fail in their objective of encouraging adjustment. A related issue is concerned with whether the need for structural adjustment measures should be stipulated as an essential requirement if temporary protection is to be permitted. It has also been suggested that, in order to encourage adjustment, there should be a requirement that restrictions are progressively lowered as the need for protection decreases. Fourthly, the arrangements for compensating trading partners adversely affected by safeguard measures have been criticised. Because any safeguard measures must be nondiscriminatory, this may require a large number of separate negotiations with various countries whose exports have been affected. It could amount to a sizeable ‘compensation bill’ which may, in practice, be difficult to meet. Given that tariffs have fallen to very low levels, it may necessitate offering concessions on sensitive products still subject to high tariffs and invite opposition from domestic interests. This is one of the major reasons why countries often prefer to negotiate VERs as a way of providing safeguards to domestic producers since they avoid the need for extensive negotiations with or costly compensation being made to other trading partners. As Robertson (1992) points out, there is also an inconsistency between the fact that protection is to be temporary and agreeing compensation in the form of tariff concessions for other affected parties: once other affected parties have been compensated, there is no need to restore market access in the protected sector by eliminating the temporary restrictions. In other words, the incentive to ensure that safeguards are temporary is removed. Finally, Article XIX has been criticised for its failure to provide any multilateral machinery for supervising the use of safeguards. Although there is a requirement that the GATT should be notified of any safeguard measures to be taken, it plays no role in examining any such measures to ensure that they conform with the requirements of Article XIX, nor does it monitor their use to ensure that the measures are temporary and are removed as and when the situation permits. The GATT only gets involved if a particular action is the source of a complaint by another contracting party, which is rare. The clear implication of Article XIX is that any disputes arising from the use of safeguards are to be settled bilaterally. Even the notification of safeguards measures is not always undertaken. Given these weaknesses of the safeguard rules, it is not surprising that countries faced with a need to grant domestic industry temporary protection have preferred to bypass Article XIX altogether and negotiate separate safeguard measures in the form of VERs with the relevant supplier nation. Thus, by mid-1987, there had been only 134 actions under Article XIX which were notified to the GATT (Robertson, 1992), which can be contrasted with the 137 VERs identified by the GATT as being in force at about the same time (Kostecki, 1987). The majority (87) of the Article XIX actions were accounted for by the United States, Canada and Australia. The EC accounted for fourteen measures and the individual member states a further twelve. Most safeguard measures were targeted at the newly industrialising and other developing countries. International Trade Policy 72 Where countries did seek safeguard remedies by invoking Article XIX, they frequently did so in ways which were questionable on a strict interpretation of the provisions of the Article. Thus, many measures appear to have been discriminatory, being targeted at particular countries. Safeguard measures have often taken the form of quotas rather than tariffs. By their very nature, quotas are discriminatory because they deny market access to new suppliers. Also, notification of safeguard measures was frequently never undertaken and trade compensation was not offered. Finally, as noted above, there are grounds for doubting whether serious injury to domestic producers was taking place in many of the cases where measures were applied. In short, Article XIX was abused as well as being bypassed. Its provisions were both too lenient, such that misuse resulted, and too severe, often resulting in countries seeking remedies by other means. The issue of selectivity The inadequacies of the GATT Safeguards Clause has meant that the issue of its reform has occupied a central importance in international trade policy for several decades. The growth of VERs and other grey-area measures has made it clear that Article XIX needs to be changed. Consequently, the Safeguards Clause has been a major item on the agendas of both the Tokyo and the Uruguay Rounds. One of the key issues which has plagued the attempts of the GATT countries to reach agreement has been selectivity. In the Tokyo Round, the European Community plus some Scandinavian countries argued strongly that any new set of rules should allow countries to introduce discriminatory measures. This was strongly opposed by the developing countries, which in practice were the main targets of safeguard measures. Developing countries saw the EC proposals as an attempt to legitimise the use of VERs and other grey-area measures. The failure of countries to agree on the matter of selectivity ensured that by the end of the Tokyo Round no progress had been made in negotiating a new code. Despite attempts to achieve agreement in separate negotiations after the conclusion of the Tokyo Round, the matter remained unresolved by the time of the commencement of the Uruguay Round. In the Uruguay Round the issue of selectivity once again divided the participating countries. The case for a revision of the Safeguards Clause to allow selectivity is based on both economic and legal grounds. Firstly, there is a legal argument concerning the interpretation of Article XIX, which some see as allowing a departure from the nondiscrimination rule in the case of safeguards. Specifically, the reference to the ‘suspension of obligations’ could be taken to include Article I MFN obligations. But does this refer to a specific product or a particular country? Article XIX is clearly ambiguous in this respect. However, the stated position of the GATT appears to be that ‘suspension of obligations’ does not permit a departure from nondiscriminatory requirements. Secondly, advocates of selectivity argue that if GATT rules insist on nondiscrimination, countries will simply bypass Article XIX and seek remedies by other means. Indeed, as we have seen, the growth of VERs has in part resulted from what some countries see as the excessively strict requirements set out in Article XIX, including the rule that safeguard measures should be nondiscriminatory and trading partners offered compensation. Might it not be better to bring all safeguard measures under GATT discipline by allowing selectivity rather than encouraging the spread of VERs over which the GATT has little or no control? Finally, it is argued that, on economic grounds also, Quantitative Trade Restrictions and Safeguards 73 selective safeguards are preferable because they ensure that the exports of other countries whose trade is not contributing to the damage are not disturbed. The only country to be affected by the measures is the country whose exports have caused the market disruption. The counterargument in favour of upholding the nondiscrimination rule is as follows. Firstly, the economic case for selectivity is a weak one. Selective safeguards are invariably aimed at the most competitive suppliers of the product in question. This, indeed, is why imports from these countries have grown faster than those of other countries. Selective safeguards not only restrict imports of a product (which in any case involves an efficiency loss), they also divert trade away from the cheapest source (adding a further distortion). There will therefore be a greater loss of economic efficiency where safeguard measures are discriminatory. Implicitly, there is an admission by an importing country which takes selective measures that its domestic industry cannot compete. If the problem is one of adjustment, why is there the need to target low-cost suppliers? A second reason for upholding the MFN requirement is that it protects smaller countries, in particular developing countries, which might otherwise get ‘picked on’ by bigger, developed nations. This is because they lack the power to retaliate. Nondiscrimination is important to ensure fairness in trade policy. It is also the case that, where measures are applied equally to all suppliers, importing countries will be less inclined to resort to safeguard measures as a way of appeasing uncompetitive domestic producers. In short, it should help to ensure that the safeguard option is not abused. Finally, selectivity makes it easier for importing countries to delay adjustment. Clearly, the incentive to adjust is less when the most competitive suppliers are excluded from the market. Moreover, suppliers who continue to enjoy access cannot be relied upon as a source of pressure on the importing country gradually to dismantle barriers after a suitable period of time. By the time of the 1988 Mid-term Review, the Trade Negotiations Committee reported that progress had been made towards reaching agreement on many aspects of a new safeguards code. However, ‘significant movement on the central issue of nondiscrimination [had] not yet materialised’ (GATT, 1988), but by the end of the negotiations there had been some movement by the EC which eventually made agreement possible. Although the EC abandoned insistence on selectivity, certain aspects of the new agreement clearly do allow for discrimination in everything but name. The opposition of the developing countries appears to have been bought off by provisions which exempt the products of developing countries from safeguard measures in certain circumstances. The new safeguards agreement The main aspects of the new safeguards agreement are: 1 a requirement that any safeguard measure may only be applied after a proper investigation by the importing country in which all interested parties should be able to give evidence. This is to include views as to whether the measure is in the public interest. In critical circumstances, a measure may be taken provisionally, before a full investigation has been conducted, subject to a preliminary determination. However, in this case, the provisional measure should not last for more than 200 days and should take the form of a tariff increase refundable if the subsequent investigation determines against the measure; International Trade Policy 74 2 a new definition of serious injury and the threat thereof. Serious injury is defined as ‘a significant overall impairment of the position of a domestic industry’. With regard to the threat of serious injury, this must mean ‘serious injury that is clearly imminent’ and must be ‘based on facts and not merely on allegation, conjecture or remote possibility’. Criteria to be used in the assessment of serious injury include ‘the rate and amount of the increase in imports of the product concerned in absolute and relative terms, the share of the domestic market taken by increased imports, changes in the level of sales, production, productivity, capacity utilisation, profits and losses, and employment’; 3 restrictions on the intensity of any safeguard measure. This should be only so much as is needed to ‘prevent or remedy serious injury and to facilitate adjustment’. If quantitative restrictions are used, these should not reduce the level of imports below the average level of the previous three representative years unless it can be shown that a different level is needed; 4 a stipulation that any safeguard measure should be applied irrespective of the source. This would seem to rule out selective measures. In practice, this is less certain because of the arrangements made for quota allocation, where quantitative restrictions are used. The agreement states that, where quotas are allocated among exporting nations, these should be based on the proportions supplied by the countries in question during a previous representative period. However, a country may depart from these provisions if imports from certain countries ‘have increased in disproportionate percentage in relation to the total increase of imports of the product concerned in the representative period’. This sounds like discrimination under a different name; 5 a limit on the duration of safeguard measures: normally, safeguard measures should be applied for no longer than four years. This may be extended for another four years if necessary to prevent or remedy serious injury and provided that there is evidence that the industry is adjusting. Measures should also be progressively liberalised at regular intervals. If measures are for more than three years, they should be subject to a midterm review to consider their withdrawal or an increased pace of liberalisation. Finally, there are provisions to ensure that no product which has been subject to safeguard measures shall again be subject to such a measure for a period of at least two years and in many cases longer; 6 provisions for consultation on trade compensation and retaliation where consultations have been unsuccessful. The Agreement states that the aim should be ‘to maintain a substantially equivalent level of concessions and other obligations’ towards other countries. To achieve this, countries taking safeguard measures should consult with their relevant trading partners and agree adequate means of trade compensation. Where agreement cannot be reached, there is provision for retaliation in the form of the suspension of ‘substantially equivalent concessions or other obligations’ but not in the first three years that a safeguard measure is in effect; 7 special provisions for developing countries. Safeguard measures are not to be applied against products coming from developing countries so long as the share of imports coming from a particular country does not exceed 3 per cent and that developing countries (each having a share of less than 3 per cent) collectively account for no more than 9 per cent of total imports. Developing countries may also extend the duration of safeguard measures for a period of up to two years beyond the maximum; Quantitative Trade Restrictions and Safeguards 75 8 the prohibition of all existing grey-area measures. The Agreement provides that all grey-area measures (including VERs, OMAs, export moderation schemes, price monitoring systems, export/import surveillance, and so on) shall be phased out or brought into conformity with the Agreement within a period not exceeding four years following the establishment of the WTO. An exception can be made for one specific measure per importing country where the phase-out date is to be 31 December 1999. In the case of the EU, the exception is to be the VER governing imports of cars from Japan; 9 provisions for notification of safeguard measures and their surveillance. The Agreement requires countries to notify the Committee on Safeguards of any investigation to be initiated, finding of serious injury or decision to impose a safeguard measure. The Committee on Safeguards is assigned a monitoring and surveillance role to ensure that the provisions of the Agreement are applied. How may the new Agreement be assessed? There can be no doubt that in several respects it represents a major improvement on the existing situation. Perhaps the two most significant changes are the provisions for eliminating grey-area measures over a four-year period and the introduction of a sunset clause for safeguard measures. If implemented, the Agreement means that all existing VERs will be scrapped or replaced with other measures which conform with the Agreement before the end of the decade. Presumably, this could mean their replacement with tariffs or quotas but subject to the time limits and other requirements set out in the Agreement. Nevertheless, this would represent a significant improvement on the existing situation, and prohibit the future use of such measures. At the same time, the sunset clause should ensure that temporary restrictions do not become permanent and that adjustment takes place during the period in which the measures are in force. In addition, the requirement that barriers should be progressively lowered at regular intervals should further encourage the necessary adjustment. The attempt to establish stricter conditions on the use of safeguard measures, including a more precise definition of serious injury and the threat thereof and the requirement that measures be preceded by an investigation, bring the safeguard rules into line with similar rules which apply to the use of antidumping policy. It is also to be welcomed that a limit has been put on retaliation, if countries are to be persuaded to act legally when dealing with problems of market disruption rather than going outside the Agreement. Finally, the provisions to exempt developing countries from safeguard measures in certain circumstances will give these countries some protection against discriminatory measures. Two areas where the Agreement might be considered less than satisfactory both concern the nature of the safeguard measures that are permissible. The provisions which are designed to limit the intensity of any measure applied are to be welcomed. On the other hand, there is no attempt to restrict such measures to tariffs even though tariffs are to be preferred on purely economic grounds. Although provisional measures must take the form of a tariff, this is not the case for measures imposed following a full investigation. The undesirable characteristics of quotas were discussed above. In particular, they freeze market shares and make it extremely difficult for a new entrant to gain access to the market. Moreover, they are inherently discriminatory. This means that the rejection of the principle of selectivity is effectively undermined. Although quotas are to be allocated on the basis of each exporter’s share of the market over the previous three years, an importing country may depart from this approach in special circumstances. If it International Trade Policy 76 can demonstrate that imports from a particular country have increased disproportionately during this period, it could impose lower quotas. This seems to allow discrimination in special cases although the country would need the approval of the Committee on Safeguards. The concession was probably necessary to secure the agreement of those countries keen to retain some provision for selectivity. It is also true that the exemptions granted to developing countries go some way towards protecting these countries from an over-use of discriminatory measures which might be targeted at them. The crucial test will be whether or not the incentives which the Agreement provides for developed countries to make greater use of safeguards provisions to deal with adjustment problems will indeed have that effect. If they fail to do so, countries will continue to seek other less desirable methods to obtain import relief, such as antidumping. The elimination of greyarea measures will then serve merely to increase the proliferation of other forms of nontariff protection. CONCLUSION In this chapter, we have seen how, with the gradual lowering of tariffs, nontariff barriers in the form of quantitative restrictions have emerged to take their place. Empirical studies of NTBs clearly show that the proportion of trade covered by such measures has been increasing. An economic analysis shows that, beyond doubt, such quantitative restrictions are more harmful in terms of economic efficiency than are tariffs. It is also clear that, for a variety of reasons, these measures are more convenient and attractive as a device for controlling troublesome imports than the use of tariffs and that this accounts for their popularity. It is also apparent that the increased use of grey-area measures such as VERs is intimately bound up with the inadequacies of the GATT safeguard provisions. When faced with market disruption, countries have preferred to bypass Article XIX and seek remedies by extra-legal means. This has resulted in a proliferation of measures outside the arrangements created for this purpose and therefore not subject to any system of multilateral monitoring or surveillance. The issue of how to contain this ‘New Protectionism’ has therefore been linked to the reform of the safeguard rules. After a slow start because countries were divided largely over the acceptability of selective safeguards, the negotiations to reform the safeguard rules were finally and successfully concluded. What has emerged is a new Safeguards Agreement which, although inadequate in certain respects, does represent some improvement on the previous situation. In particular, the provisions for eliminating all grey-area measures by the end of the century is important. It can only be hoped that the terms of the Agreement will be fully implemented. For newly industrialising countries, which have so often been the target of such measures, the Agreement is most welcome. It nevertheless contains inadequacies. In particular, it does not ensure the elimination of discriminatory quantitative restrictions on trade. These will continue to be used to cope with adjustment problems even though import quotas are generally the least efficient way of doing so. The still fairly vague criteria for determining serious injury or the threat thereof are likely to mean that safeguard measures will be used to gain protection for essentially Quantitative Trade Restrictions and Safeguards 77 possible. There can be little doubt that predatory pricing is harmful since, if successful, it will lead to the disappearance of any element of competition in the foreign market. In the long run, consumers will suffer from having to pay prices well above marginal costs. In many countries, such practices are illegal and can result in prosecution. Cases of predatory dumping could therefore be adequately dealt with under national competition laws. What is less clear, however, is the extent to which predatory dumping takes place in international trade. It seems improbable that, in recent years, anything more than a small fraction of antidumping cases have been concerned with such behaviour. This is because the conditions for such a strategy to be worth pursuing are highly restrictive. Firstly, the exporter must possess larger financial resources than its competitors in the foreign market or it will fail in its attempt to drive them out of the market; so it would end up with massive losses which it cannot recover. Secondly, there must exist barriers to entry in the foreign market, otherwise when prices are raised in the long run, new firms will enter the industry and excess profits will be competed away. Even where these exist, they must be sufficient to offset the prospects for higher profits which might otherwise tempt potential entrants to gain entry. Otherwise, the predator will need to hold down long-run prices to deter entry and may therefore fail to recover his short-run losses for some while to come (if at all). These two conditions mean that cases of predatory pricing will be confined to concentrated industries characterised by significant entry barriers and in which one firm is dominant. However, not even these conditions are sufficient. Even if entry from domestic firms can be prevented, it will also be necessary to deter entry by foreign firms. This necessitates persuading the host government of the importing country to restrict access to new competitors from other countries. It is highly improbable that any government would agree to do so. Only if the exporter already enjoys global market dominance does it seem probable that he could ensure that no new competitors ‘spoil the show’ by entering the foreign market. These conditions suggest that cases of predatory dumping are likely to be confined to industries in which one firm accounts for a substantial share of the world market and in which the number of other firms is very small. Even then, there are cheaper ways of obtaining monopoly profits. The firms could collude to enforce a minimum price, or the more powerful firm could buy control of the others. These are likely to cost less money than a policy of holding price below cost for any length of time. A third cause of dumping is the existence of excess capacity arising from a combination of demand uncertainty and short-run adjustment costs (see Ethier, 1982; Davies and McGuiness, 1982; and Bernhardt, 1984). This situation can arise in competitive industries where demand fluctuates a great deal but where in the short run firms face large adjustment costs in changing output to match demand. This is the case in many intermediate goods industries such as steel and chemicals where production occurs in continuous-run plants involving considerable changeover costs and necessitating constant use of capacity. In these industries, demand is also subject to considerable cyclical fluctuations. Because demand is uncertain, producers face the problem of choosing the right amount of capacity. If they underinvest in capacity, they will rarely overproduce but will often incur adjustment costs to increase output when demand is higher than expected. If they overinvest in capacity, they will frequently overproduce and International Trade Policy 84 face the necessity of making costly reductions in output when demand is lower than expected. If, however, firms can dispose of any excess output abroad, they will err on the side of overinvestment in capacity. Then, if demand is lower than expected, the surplus production can be exported abroad at a price below cost and lower than the price charged domestically. If the producer unloaded his excess output on the domestic market, the price fetched on each unit of nonexcess output would be depressed. However, by disposing of the excess output abroad, the price can be maintained on domestic sales. Some of the profits on domestic sales can then be used to offset the losses on export sales. This may still be cheaper than incurring the adjustment costs of trying to reduce output in the face of lower than expected demand. Ethier (1982) has developed an interesting model of dumping which includes all of the ingredients referred to above. Demand is uncertain. Factors of production employed in the dumping industry are highly specific and so cannot easily be switched to some other activity if demand falls. Wages are ‘sticky’, which means that firms adjust to fluctuations in demand by laying off workers. One part of the workforce comprises ‘essential’ workers (called managers) who have to be employed whatever the state of demand. Another part comprises nonessential workers (called labourers) who can be laid off if demand falls. However, if nonessential workers are laid off, it increases the job insecurity of those who remain. To secure the employment of these workers, producers have to pay them a higher wage. Hence, faced with a fall in demand, the preferred solution of producers is to export their excess output abroad at a price below marginal costs. The model throws up a number of possible determinants of dumping in such industries. Firstly, in countries where for legalistic/institutional reasons, the ratio of essential to nonessential workers is high, the propensity to dump may be greater. Thus, historically (although this is changing), Japanese workers have enjoyed lifetime employment guarantees such that, in Japan, adjustment can less easily be achieved by laying off workers. Similarly, most European workers have traditionally enjoyed greater employment protection rights than American workers. One effect is to make it more costly for employers in Europe to sack workers if demand falls. Secondly, the higher the rate of unemployment benefit (or other income transfer) paid to workers laid off, the easier it is for employers to lay off nonessential workers if demand falls. In such countries, the propensity to dump is accordingly lower. Thirdly, the extent to which demand fluctuations is symmetrical between countries will determine the extent of dumping. If demand fluctuates simultaneously in opposite directions in any two countries and if the fluctuations are equally pronounced, it can be shown that no dumping will take place. Where, however, fluctuations are in the same direction (for example, all countries experience a downswing at the same time) and/or fluctuations are more pronounced in one country than another, dumping will result. Are antidumping measures justifiable in such cases? A variety of different arguments is put forward in favour of antidumping measures. Firstly, there is the argument that this is ‘unfair’ competition: perhaps the least convincing argument of all. Firms frequently sell a proportion of their output below cost whenever the market for the good is depressed. This is sound business practice whenever the revenues earned on such sales can be used to contribute towards the recovery of some fixed costs. It is not clear why this should be deemed ‘unfair’ when it takes place in the context of international trade. Unfair Trading Practices 85 A second argument is based on the existence of different cost structures in different countries, which results in some countries having a greater propensity to dump. As in Ethier’s model, it may be the case that, in some countries, fixed costs account for a higher proportion of total costs than in others. In some cases, these differences may be by products of government policy, in which case offsetting measures may be considered desirable. On the other hand, it is necessary to demonstrate that imposing antidumping duties is the best way of dealing with the problem. Suppose that a particular country is more prone to being dumped on because laying-off workers is less costly. Deardorff (1990a) argues that the best policy would be either a tax on lay-offs or a subsidy to continued employment rather than antidumping measures. Thirdly, it is argued that antidumping measures are needed to reduce employment variation. Frequent upward and downward changes in domestic ouput and employment impose considerable adjustment costs on the importing country, so the argument goes. However, the employment variations largely arise from the invariability of wage-rates. In a perfect market, wage-rates would fluctuate with little or no employment variation. A best policy would be measures to achieve greater flexibility of wage-rates. The disadvantage of less stable wage-rates should be offset against the cost to the importing country of imposing antidumping duties on imports leading to a higher price of the dumped product. Deardorff (1990) argues that, even when the problem of adjustment is considered justifiable grounds for protection, antidumping policy is not the most appropriate instrument. If an industry needs temporary protection, it would be more appropriate to do so under the GATT Safeguards or Escape Clause since the issue is an adjustment problem rather than a dumping one per se. Antidumping authorities are likely to be less well equipped to deal with such issues. Fourthly, transitional dumping may occur when an exporter needs to price below marginal cost in order to maximise sales and expand market share. In this case, belowcost pricing is a kind of investment in the marketing of the product, deemed to be worthwhile if profits can be earned in the long run. Because this may require fixing price below marginal cost, it may be treated as predatory pricing, yet clearly it is not. One form of this occurs where a new entrant to an industry must initially set export price below those of established firms in order to attract consumers away from traditional brands. Such below-cost pricing is temporary. The intention is not to eliminate rivals but rather to gain entry to the industry. Having done so, the firm will hope to raise price and recuperate the costs incurred in the entry period. Clearly, there is nothing harmful about this kind of behaviour. On the contrary, in so far as it enables new entrants to an industry to generate more competition for established firms, the consumer should gain. A second form of this kind of dumping may occur in high-technology industries where new products and processes are continually being developed. Very often in such industries, there will be considerable savings to be reaped in the early stages of production as a result of so-called ‘learning-by-doing’. A good example of this type of product is that of DRAM—dynamic random access memory—chips where significant reductions in costs are associated with increased volume of output. Apparently, every doubling in the volume of output is associated with a 30–40 per cent reduction in costs (Tyson, 1992). Not surprisingly, the industry is one in which there have been constant allegations mainly by US and European producers of dumping by Japan and South Korea (see Yoffie, 1991; and Tyson, 1992). This may make it worthwhile for a producer International Trade Policy 86 initially to price the good below marginal cost in an effort to increase sales and achieve a volume of output high enough to generate such learning effects. It may result in dumping if the price is fixed lower in the foreign market than at home. For example, if sales cannot be increased any further at home, a firm might embark on an export drive based on below-cost prices designed to increase overseas sales sufficiently to achieve the required volume of output. Is this kind of dumping harmful? The main argument usually advanced for taking measures against it is to ensure that national producers get further up the experience curve than foreign firms. It is frequently argued that such industries bring special advantages to a country either because they enable domestic factors of production to earn higher returns than in other sectors of the economy or because they generate externalities or spillover benefits for the rest of the economy. Even if these arguments are accepted, it must still be demonstrated that antidumping policy is the best instrument for achieving these objectives. Clearly, it is not. Theory shows that a superior instrument would be a production subsidy granted to domestic producers sufficient to correct for the market distortion. Finally, it is possible for dumping to appear to be taking place when in fact it is not, due to exchange-rate variations. Suppose that US$1=Y100. Suppose that a Japanese product has an identical ex-factory price of Y100,000 (US$1,000) when it is sold in Japan and exported to the US. Now suppose the dollar depreciates against the yen so that $1=Y90. Suppose also that initially Japanese exporters make no change to the price of their exports to the US. The price of the product in the US is still $1,000 which, when converted into yen at the new exchange rate, is Y90,000. Clearly, this is not dumping. Eventually, Japanese exporters will have to raise their prices to correct for the change in the exchange rate. However, because there may be a time lag in exporters adjusting export prices, it may show up as dumping. This position could arise where goods are sold under contract and where the currency appreciation takes place shortly before the period of investigation and before exporters have had time to raise prices. A particular problem will occur when a sudden rise in the exporter’s currency takes place during the period of the investigation followed by a sudden fall. In this case, exporters leave their prices unchanged and dumping seems to be taking place. In practice, this may not be as important a cause of dumping as is sometimes thought. In the United States, allowance is often made for time lags in the response of exporters to rises in the value of the export currency as well as for so-called ‘spikes’ in the exchange rate. With regard to the EC, Messerlin (1989) found no evidence for a strong and positive relationship between exchange-rate variations and antidumping cases. The economic rationale for antidumping policy is regarded by many economists as being rather weak. Most would accept that antidumping measures are justifiable in the special case where predatory pricing is found to be taking place. However, as we have seen, this is comparatively rare in practice. Therefore, one approach would be to confine antidumping to cases where export prices are below costs of production. But which costs? Clearly, not average costs, since profit-maximising firms may well sell below average costs if average costs are falling. Clearly, the relevant costs are marginal costs. Even then, as we have seen, below-cost pricing need not imply predatory intent. There would be a need to examine other criteria such as the exporter’s share of the market, the existence of any entry barriers, the effects on competition, and so on. Nevertheless, if antidumping were confined to cases of below-cost pricing, this would be preferable to current practice. Unfair Trading Practices 87 An alternative approach would be to tackle the problem of dumping through competition rather than trade policy. Since the major concerns are the implications of dumping for competition in the domestic market, it might be more appropriate for dumping cases to be investigated by the competition authorities. The requirement would then be simply to determine whether or not dumping is likely to result in reduced competition and higher prices to consumers in the long run. This would ensure that antidumping is not hijacked by domestic producers for protectionist purposes. If an industry needs temporary protection on grounds of adjustment, this is best dealt with under the Safeguards Clause. Antidumping policy should be restricted to the one single, theoretically valid case for intervention, namely, predatory pricing. However, what is desirable on economic grounds is unlikely to be acceptable to policy makers or regulators in most western industrialised economies: for political reasons, no government will agree to such a restriction on the use of antidumping measures. On the contrary, the demands of these countries is for tougher measures to combat dumping. Given the fact that antidumping measures benefit producers in the protected industry and nearly always harm consumers, this is a reflection of how successfully antidumping policy has been captured by producer interests. A more promising approach may be for economists to press for measures which will curb the power of antidumping authorities and limit the misuse of antidumping. For example, any reforms which strengthen the role of consumer organisations and therefore act as a counterweight to the influence of domestic producers are to be welcomed. In the next sub-section, the actual implementation of antidumping policy is discussed and an attempt will be made to explore some of the options for reform. Antidumping policy The GATT’s Antidumping Policy is set out in Article VI of the General Agreement and further elaborated in the Antidumping Code (GATT, 1979b). The Code was negotiated as part of the Kennedy Round and further revised in the course of the Tokyo Round. A new Code is contained in the Final Act Embodying the Results of the Uruguay Round (GATT, 1994a, 1994b) and the details of this are discussed below (pp. 120–7). The Code contains a set of rules governing antidumping policy, including the finding of dumping, the measurement of the dumping margin, the determination of injury, the imposition of duties and procedures covering the investigation of dumping by the authorities in the importing country following a complaint. It should be pointed out that it is up to any individual country to decide whether it wishes to operate an antidumping policy. The only requirement is that this should conform to the rules set out in the Article VI of the GATT and the Antidumping Code. This means that antidumping policy may and very often does differ between one country and another, which may create problems for exporters who face different rules and procedures in different countries. These may in themselves create a barrier to trade. Until recently, it was mainly the developed market economies who made use of antidumping policy. Table 4.1 shows the number of antidumping cases initiated by countries over the years 1980–93. Since these figures say nothing about the amount of trade affected by the cases in question, it is not possible to decide whether or not the incidence of antidumping has increased. Nevertheless, for much of the period the frequency of antidumping International Trade Policy 88 investigations did increase. In particular, the years since 1990 show a marked increase in antidumping activity. Research by Baldwin and Steagall (1994) has shown that, over the period 1980–90, there occurred a significant increase in the number of US antidumping and countervailing duty cases. In 1980–90, the US International Trade Commission investigated 494 antidumping and 306 countervailing duty cases compared with only 172 antidumping and 10 countervailing duty cases in 1970–9. It can be seen that, for the period as a whole, most antidumping cases were initiated in four areas: Australia, Canada, the European Community and the United States. For most of the period, the United States and Australia appear to have been the leading users of antidumping policy. However, in recent years, there has been a growing number of cases originating in other developed countries and, most notably, in some developing countries. Before 1985, there had been no antidumping cases initiated by a developing country; by 1990, there were no less than forty-one cases or 23 per cent of the total number for that year. From being mainly recipients of antidumping measures imposed by developed countries, developing countries have increasingly become users of antidumping policy. Not all cases result in antidumping measures being imposed, therefore, it may be more useful to examine which countries were the most active. Antidumping measures may take the form of either the imposition of an antidumping duty or the extraction of price undertakings from exporters. GATT figures show that, over the period from 1 July 1980 to 1 July 1990, the EU compelled 150 exporters to give price undertakings and imposed 84 definitive antidumping duties, the United States imposed 156 antidumping duties and exacted 6 price undertakings, Australia imposed duties on 174 occasions and obtained 41 price undertakings, and finally Canada imposed 156 duties and secured 11 price undertakings (GATT, 1990). On this measure, and taking duties and undertakings together, the EU emerges as the greatest user of antidumping policy. However, because the EU places much greater reliance on undertakings, it is less important in terms of duties imposed. Table 4.1 Antidumping cases initiated, 1980–93 Australia Canada EC US Other developed countries Developing countries Total 1980–1 61 48 37 24 3 0 173 (35%) (28%) (21%) (14%) (2%) 1981–2 54 64 39 51 2 0 210 (26%) (37%) (23%) (29%) (1%) 1982–3 71 34 26 19 0 0 150 (41%) (20%) (15%) (11%) 1983–4 70 26 33 46 1 0 176 (40%) (15%) (19%) (26%) (1%) 1984–5 63 35 34 61 0 0 193 (33%) (18%) (18%) (32%) 1985–6 54 27 23 63 2 3 172 (31%) (16%) (13%) (37%) (1%) (2%) 1986–7 40 24 17 41 5 4 131 (31%) (18%) (13%) (31%) (4%) (3%) 1987–8 20 20 30 31 9 13 123 Unfair Trading Practices 89 (16%) (16%) (24%) (25%) (7%) (11%) 1988–9 19 14 29 25 12 14 113 (17%) (12%) (26%) (22%) (11%) (12%) 1989– 90 23 15 15 24 5 14 96 (24%) (16%) (16%) (25%) (5%) (15%) 1990–1 46 12 15 52 9 41 175 (26%) (7%) (9%) (30%) (5%) (23%) 1991–2 76 16 23 62 21 39 237 (32%) (7%) (10%) (26%) (9%) (16%) 1992–3 61 36 33 78 8 38 254 (24%) (14%) (13%) (31%) (3%) (15%) Source: GATT Secretariat; quoted in The Financial Times, 15 December 1992 and 25 November 1993 Table 4.2 shows which exporting countries were worst hit by antidumping duties imposed over the period 1980–9. Japan heads the list with 74, followed by the United States with 51, Korea with 40 and China and West Germany each with 32. Even these figures tell us nothing about the extent to which trade was affected by the measures imposed. However, what is clear is that antidumping policy frequently results in duties being imposed well in excess of the average level of tariff applied to the products in question. According to a study carried out by the World Bank, average tariffs in the entire US manufacturing sector would be 23 per cent today, compared with a nominal level of 6 per cent, if they were adjusted to take account of the cost of antidumping duties, in particular on steel, textiles and cars (World Bank, 1992). In other words, antidumping measures have, in effect, wiped out much of the gain achieved by tariff liberalisation. Morkre and Kelly (1994) estimated the average margin of dumping (on which the rate of duty is based) at 33.2 per cent for all US antidumping cases over the period 1980–9. Bourgeois and Messerlin (1993) similarly estimated the average dumping margin for all EC antidumping cases at 37.4 per cent for the period 1980–8. However, in the EC rates of duty were often set below margins of dumping. Messerlin and Reed (1995) estimate the average rate of antidumping duty in the EC over the period 1980–9 was 17.8 per cent (although only for cases which were terminated by ad valorem duties). This compares with the average MFN tariff rate for industrial products of 7.8 per cent. Any country wishing to make use of Article VI is required to follow a carefully prescribed procedure. This requires the importing country to demonstrate by an investigation that (a) dumping, as defined by the GATT, has taken place; (b) material injury or the threat of material injury to the domestic industry exists; and (c) dumping is the cause (although not necessarily the sole cause) of the alleged injury. In the United States, the investigation into whether dumping has taken place and the investigation into whether it has resulted in injury (or the threat of it) are separated. The International Trade Commission (ITC) is charged with the responsibility for the latter and the International Trade Administration (ITA) makes an entirely separate decision about the former. By way of contrast, in the case of the EU, both aspects are considered simultaneously by the Antidumping Unit of the European Commission. International Trade Policy 90 Table 4.2 Countries most frequently subject to antidumping duties, 1 July 1980–1 July 1989 Exporter Countries imposing duties Australia Canada EC US Total EC 35 41 2a29 107 Japan 21 13 12 28 74 US 16 25 10 – 51 S. Korea 14 14 2 10 40 China 12 6 3 11 32 W. Germany 11 14 – 7 32 Taiwan 11 4 – 13 28 Brazil 4 7 6 10 27 Italy 8 6 – 8 22 France 8 8 – 6 22 Canada 6 – 5 10 21 UK 6 9 – 2 17 Spain 2 4 2 6 14 Source: GATT Secretariat; quoted in The Financial Times, 1 July 1990 Note: a Spain. During the investigation, all interested parties are allowed the opportunity to submit relevant evidence. Provisional duties may be imposed after a preliminary finding that dumping has occurred and injury resulted. These must not be greater than the provisionally estimated margin of dumping. They should not be imposed for more than four months, or six months in exceptional cases. Provisional duties are repayable if the full investigation subsequently finds no evidence of dumping or injury or a causal link between the two. There is a provision for suspending or terminating proceedings without any duty being imposed if exporters are willing to make price undertakings, that is, undertakings to raise prices so as to eliminate the injurious effects of dumping (Article VII, GATT Code). In the case of the EU, price undertakings have been more common than antidumping duties. Where the investigation does result in a positive finding, definitive duties may be imposed provided that these do not exceed the dumping margin. The 1979 GATT Code set no limit on the duration of duties except that they should not remain in force for any longer than is necessary to counteract the dumping causing injury. An important feature of US legislation is that it imposes strict time limits on the various stages of investigation. The ITC must reach a preliminary decision within 45 days of the filing of the petition. If its decision is affirmative, the ITA has 160 days from the filing of the petition to reach its preliminary decision. Once the ITA has reached its preliminary decision, it then has 75 days to make its final determination. If its decision is affirmative, the ITC then has 45 days to make its final determination. Following an affirmative decision by the ITC, the US Customs has 7 days in which to issue an order (see Devault, 1990, for an analysis of the procedure). This strict timetable may tend to deliver the provision of evidence into the hands of the plaintiff. Unfair Trading Practices 91 The determination of dumping The first requirement is to provide evidence that dumping has taken place. This entails a comparison of the export price with the normal value. Article VI states that the normal value may be determined in one of three ways: 1‘the comparable price, in the ordinary course of trade for the like product when destined for consumption in the exporting country’. Much controversy has surrounded the phrase ‘in the ordinary course of trade’. In some cases, this has been taken to exclude any domestic sales made at below costs of production. Sales between associated parties, that is, when an exporter sells on the domestic market to a related company, are also not generally considered as being ‘in the ordinary course of trade’. Also, when goods are not sold in sufficient quantities on the domestic market, this method is deemed to be inappropriate. 2 ‘the highest comparable price for the like product for export to any third country in the ordinary course of trade’ or the export value method. The prices charged for the product when exported to some third market can serve as a proxy for the domestic price. The European Commission makes little use of this method on the grounds that there is a strong probability that if an exporter dumps on one market he will do so on other markets. Bellis (1990), however, believes that it has more to do with administrative convenience. By way of contrast, the export value method is widely used by the US Department of Commerce. (One case where this method was used by the Commission was in 1983 and involved low-density polyethylene (LdeP) imported from the Soviet Union, Poland, East Germany and Hungary. It was settled by the exporters making price undertakings. In this case, because the dumping countries were non-EC economies, the Commission used prices in the Swedish market as a proxy for domestic prices. However, as Messerlin (1991) has shown, prices on the Swedish market were themselves distorted by the cartelisation of the EC market which itself became the subject of an antitrust decision by the EC authorities in 1988.) 3 ‘the cost of production of the product in the country of origin plus a reasonable addition for selling cost and profit’ or the constructed value method. Costs of production are determined by adding up all costs, fixed and variable, incurred in the course of producing the good, both the costs of materials and of manufacture. The procedure for determining the ‘reasonable addition’ for selling, administrative and other general expenses is controversial. The US provides for a minimum level of 10 per cent for general expenses but the EC bases this on actual costs incurred. Where a product is sold through a related sales company, the general selling expenses of the sales company are also included. These are allocated on the basis of some criterion such as turnover. The Antidumping Code states that ‘as a general rule, the addition for profit shall not exceed the profit normally realised on sales of products for the same general category in the domestic market of the country of origin’. The US practice is to add a minimum 8 per cent whereas the EC seeks to determine a reasonable margin for profit based on the average profitability of the exporter in his own home market. In some cases, where no profit is realised, the EC may use the profit rate realised by other producers of the product in their home market. The constructed value method necessarily throws up a highly arbitrary estimate of the normal value because of the International Trade Policy 92 difficulties of estimating costs and knowing what addition to make for selling costs and profit. Where dumping is taking place in a nonmarket economy, the practice of using a ‘surrogate domestic price’ has evolved. In nonmarket economies, domestic prices are fixed by the state and therefore cannot be meaningfully compared with prices in market economies; there is thus a need to find some other way of determining normal value. The procedure is to calculate a surrogate or reference price from costs of production for the like product in another country with the usual addition for selling expenses plus profit. The procedure has been much criticised. The estimate of normal value is necessarily arbitrary and makes no allowance for any cost differences between the exporting and surrogate country. Moreover, there is nothing to stop the importing country from choosing the surrogate country with the highest costs. Of course, there is a problem involved in determining normal value of a product in a nonmarket economy for the reasons given. But this raises the question as to whether it makes any sense even to try to apply antidumping laws to such countries. If there is a need on grounds of injury to domestic producers to grant protection from imports coming from such economies, Bellis (1990) has suggested that it would be better to do so through means other than antidumping policy, which necessarily results in arbitrary calculations of an intangible concept. It is interesting to compare the frequency with which these different methods for determining normal value have been used in different countries. According to Messerlin (1989), nearly 68 per cent of all antidumping cases initiated by the EC over the period 1980–5 used either the constructed value method or the third market (surrogate country) method applicable to nonmarket economies. The export value method was not used at all. The heavy reliance on methods which seek to construct normal value from information regarding costs with an addition for selling expenses and profits is a cause for concern for the reasons given above. Even if nonmarket economies are excluded, 40 per cent of all cases initiated were based on constructed estimates. The concern is justified, given some evidence that, where constructed estimates were used, there was a greater likelihood of restrictions being imposed (Messerlin, 1989). Devault (1990) carried out a similar exercise for the United States covering the decade 1980–9. Only 45.6 per cent of all cases (measured by the combined home and US market value represented) used the domestic price method whereas 9.4 per cent used the export-value method, 15.9 per cent used the constructed-value method, 10.4 per cent used the surrogate-country method and a further 18.4 per cent used the ‘best information available’. The number of cases based on best information available is disturbing. This is often used where the foreign firm is unwilling to co-operate in providing the necessary information or provides inadequate responses. Best information available can come from a number of sources including the petitioner. Not only was this was the second most frequently used method, but Devault found the average dumping margin to be much higher in these cases than in others. Having estimated normal value, the antidumping authorities must similarly estimate the export price in order to determine the margin of dumping (if any). GATT rules state that the two prices must be compared ‘at the same level of trade, normally at the exfactory level, and in respect of sales made as nearly as possible at the same time’ (Article II:6, GATT Antidumping Code, 1979). This requires that the export price be reduced by the amount of any costs of transportation or any import duties payable. Where goods are Unfair Trading Practices 93 country not subject to antidumping measures for re-export to the country imposing duties. Yet a third posssibility is to make physical alterations to the product so that it no longer attracts antidumping duties. It is often argued that, as production has become increasingly globalised, it is easier for exporters to circumvent antidumping duties. To deal with this problem, in July 1987 the EC introduced an important but controversial amendment to its antidumping regulations to allow the imposition of antidumping duties on imported components and parts where screwdriver plants were used to circumvent EC antidumping measures. The amendment comes into effect whenever (a) companies start up or substantially increase local assembly or production operations after the opening of antidumping investigations; and (b) imported parts or materials exceed the value of all other parts or materials by at least 50 per cent. There is a further requirement to take account, on a case-by-case basis, of the amount of research and development carried out by the assembler within the EU as well as the degree of technology applied. These provisions were strongly criticised when they were introduced. In particular, the requirement that no more than 50 per cent of the value of component and parts should come from the dumping country was seen as an attempt to compel foreign companies investing in the EC to buy parts and materials locally. It was also argued that 50 per cent was unreasonably low in the case of a company at the early stages of investment. As Bellis (1990) argues, ‘what is ostensibly an anticircumvention provision is being deviated from its initial purpose to become a “buy European” instrument’. Two of the first uses of the screwdriver plant regulation were to impose duties on Japanese manufacturers of electronic typewriters and photocopiers (see NCC, 1990, for further details). The two cases resulted in Japan lodging a complaint with the GATT. In March 1990 a GATT panel ruled that the EC’s anticircumvention provisions were illegal. Firstly, the duties imposed were internal charges and not customs duties, and as such infringed Article III of the GATT. The latter requires that no internal tax be applied to imported products in such a way as to give more favourable treatment to products of national origin than like products imported from abroad. Secondly, the panel rejected the EC’s argument that duties were permissible under Article XX. The latter allows deviations from GATT obligations to prevent enterprises evading obligations imposed on them which are consistent with the GATT such as the evasion of an import duty. The panel drew a distinction between action taken by a company to evade and action taken to avoid a legal obligation. The decision by a company to transfer the production of a good on which an import duty is levied to the importing country did not constitute evasion of a legal obligation and therefore was not covered by Article XX. The adoption of the panel report by the GATT Council left the EC’s then-existing anticircumvention provisions in doubt. Consequently, it remained an important negotiating aim of the EC in the Uruguay Round to ammend the GATT Code so as to permit anticircumvention measures. Finally, with regard to the duration of antidumping duties, the original Code merely stated that they should ‘remain in force only so long as, and to the extent necessary to counteract dumping which is causing injury’ (Article IX:1). In fact, the new revised Code has introduced new limits on the duration of measures, as will be explained below (p. 126). The EC regulations have always had a ‘sunset clause’ which limits the duration of duties to five years. After five years, any duty still in existence automatically expires unless an interested party can show that expiry would again lead to injury or the threat of injury. Messerlin (1989) sees merit in such a clause in weakening ‘the collusive impact of International Trade Policy 100 antidumping actions’ and allowing competition ‘to surface’ a few years before the duties are due to expire. On the other hand, he sees a risk that it might generate ‘a race to undertakings’ because the earlier an exporter can gain acceptance of undertakings, the earlier it can benefit from the sunset clause. In addition, any exporter may request a review provided that at least one year has passed since the conclusion of an investigation. In the US, an exporter who can demonstrate that no sales have taken place at less than fair value for two years and that there is no likelihood that such sales will be resumed may get an order revoked. The GATT negotiations and the Final Act The 1986 Ministerial Declaration launching the Uruguay Round stated that one of the objectives was ‘to improve, clarify or expand, as appropriate, agreements and arrangements negotiated in the Tokyo Round of multilateral negotiations’. The Antidumping Code referred to above was one such agreement. In the negotiations which ensued, there was a clear difference between those countries which made great use of antidumping policy and others which more often than not were the victims. Countries such as the USA, those of the EU, Canada and Australia favoured changes to the Code which would make it easier for countries to catch dumpers. By way of contrast, countries such as Japan and the newly industrialising countries, which generally were on the receiving end of antidumping measures, were concerned that the rules should be made stricter. A key issue for the former group was the problem of circumvention. They were concerned with the way in which companies subject to antidumping duties could circumvent these measures either by setting up assembly plants in the importing country or by switching production to some third country and exporting to the country imposing the duties. As noted above, Japan lodged a complaint with the GATT concerning the EC’s anticircumvention provisions and secured a ruling declaring these provisions to be illegal. The EC was keen to agree the inclusion in the new Code of a provision allowing countries to take measures where circumvention was found to be taking place. In July 1990 Charles Carlisle, the Deputy Director-General of the GATT and chairman of the negotiating group dealing with antidumping, tabled a paper which proposed a compromise between the two opposing camps. On the one hand, it contained several changes which would significantly tighten the existing antidumping rules. On the other hand, in deference to the USA and the EC, it included a proposal to allow countries to act against circumvention under strict conditions. The paper met with strong resistance from Japan, which regarded the new rules as insufficiently strict and opposed the proposals for tackling circumvention. In August 1990, a second version of the Carlisle paper was prepared which was much more vague than the first. It succeeded in attracting wider support than the first version but failed in its attempt to conclude the antidumping negotiations. The antidumping code proposed in Arthur Dunkel’s draft Final Act, published in December 1991, adopted a similar approach but included new concessions in an attempt to maximise agreement. In deference to Japan, it recognised that selling below cost in the launch phase of a new product was a legitimate business practice and allowable under strict conditions. Action against circumvention would be allowed but only when the cost of parts imported for assembly was more than 70 per cent of total costs. In November 1993, a matter of months before a last attempt was made to conclude Unfair Trading Practices 101 the Round, the United States raised a number of further demands which temporarily threatened the conclusion of the negotiations. Some last-minute concessions were made to the US and this proved sufficient to secure agreement. Perhaps the most significant feature of the new Code is that the provisions regarding circumvention contained in both the original Carlisle proposals and the Dunkel Draft Final Act have been omitted. Instead, the Final Act incorporates a two-sentence statement: The problem of circumvention of antidumping duty measures formed part of the negotiations which preceded this Agreement. Negotiators were, however, unable to agree on specific text, and, given the desirability of the applicability of uniform rules in this area as soon as possible, the matter is referred to the Committee on Antidumping Practices for resolution. In other words, it has for the moment proved impossible to reach agreement on this issue. However, in a number of other areas, significant changes have been made to the Antidumping Code. These are incorporated in the Agreement on the Implementation of Article VI of GATT 1994 (GATT, 1994b). The main changes are set out in the following sub-sections. The determination of dumping Clearer and more detailed rules are stipulated for determining if dumping has taken place. Article 2.2.1 of the new Code clarifies the conditions under which sales of a product in the domestic market of the exporting country at below cost may be treated as not being ‘in the ordinary course of trade’ and therefore disregarded in the determination of normal value. Three conditions must be satisfied. Firstly, such sales must be made ‘within an extended period of time’ which should normally be one year but in no case less than six months. Secondly, such sales must be made ‘in substantial quantities’ (not less than 20 per cent of the volume sold in transactions). Thirdly, such sales must be ‘at prices which do not provide for the recovery of all costs within a reasonable period of time’. It is made clear that the prices must be below weighted average costs for the investigation period as well as below costs at the time of sale for this condition to be met. These provisions will help to eliminate the practice of excluding all below-cost transactions from the estimation of normal value which can contribute towards countries obtaining an overinflated dumping margin. Article 2.2.1.1 contains a reference to the need to make adjustments for ‘start-up operations’ in the determination of costs. A problem with many new products is that costs are very high in the early stages of production with the result that price is temporarily fixed below cost. The new Code states that costs must be adjusted ‘for circumstances in which costs during the period of investigation are affected by start-up operations’. A footnote states that ‘the adjustment made for start-up operations shall reflect the costs at the end of the start-up period or, if it extends beyond the period of investigation, the most recent costs which can reasonably be taken into account by the authorities during the investigation’. International Trade Policy 102 Article 2.2.2 contains a statement that, in the calculation of normal value, where the constructed price method is used, ‘the amounts for administrative selling and any other costs and for profits shall be based on actual data pertaining to production and sales in the ordinary course of trade for the like product by the exporter or producer under investigation’. As noted above, the US currently uses a fictitious 10 per cent addition for selling costs and 8 per cent for profit, which inevitably yields a highly arbitrary estimate of normal value. However, Article 2.2.2 does allow for ‘any other reasonable method’ where ‘such amounts cannot be determined’, subject to the condition that ‘the amount for profit so established shall not exceed the profit normally realised by other exporters or producers on sales of products of the same general category in the domestic market of the country of origin’. This merely repeats the wording of the 1979 Code. Article 2.4.1 contains a useful provision for ensuring that ‘exchange-rate dumping’ is not subject to antidumping measures. It states that ‘fluctuations in exchange rates shall be ignored and in an investigation the authorities shall allow exporters at least 60 days to have adjusted their export prices to reflect sustained movements during the period of investigation’. Article 2.4.2 addresses the issue of how the export price should be compared with normal value in the determination of the dumping margin. It was shown above (p. 112) that where the export price is compared with a weighted average normal value, dumping can be found when in fact no dumping is taking place. Article 2.4.2 states that ‘the existence of margins of dumping during the investigation phase shall normally [my emphasis] be established on the basis of a comparison of a weighted average normal value with a weighted average of prices for all comparable export transactions or by a comparison of normal value and export prices on a transaction to transaction basis’. It continues, ‘a normal value established on a weighted average basis may be compared to prices of individual [my emphasis] export transactions if the authorities find a pattern of export prices which differ significantly among different purchasers, regions or time periods and if an explanation is provided why such differences cannot be taken into account appropriately by the use of a weighted average-to-weighted average or transaction-to-transaction comparison’. In other words, in exceptional circumstances, it is possible to compare the normal value ‘established on a weighted average basis’ with individual export prices. Hindley (1994) has argued that this comes very close to authorising the procedures referred to above which impart an upward bias to the calculation of the dumping margin. He argues that the requirement for the authorities of an importing country to provide an explanation for using this procedure is too weak and will almost certainly allow it to be used so long as a country can provide an explanation that has ‘rudimentary plausibility’. The determination of injury One of the demands of Japan and the Asian NICs was for clearer definition of material injury or the threat of injury. Article 3 of the new Code goes some way in this direction. Article 3.3 states that cumulation is only permissible when (a) the margin of dumping in relation to imports from each country is more than de minimis (defined as less than 2 per cent of the export price) and the volume of imports from each country is not negligible; and (b) ‘is appropriate in the light of the conditions of competition between imported Unfair Trading Practices 103 products and the conditions of competition between the imported products and the like domestic product’. Some sort of de minimis cut-off had been advocated by critics of cumulation as a way of protecting smaller exporters from unfair exposure to antidumping policy. Article 3.5 reiterates the requirement stipulated in the 1979 Code that there must exist a ‘causal relationship between the dumped imports and the injury to the domestic industry’. It also goes a little further in listing factors that possibly cause injury, other than dumped imports, which might be taken into account; namely, ‘the volume and prices of imports not sold at dumping prices, contraction in demand or changes in the pattern of consumption, trade restrictive practices of and competition between the foreign and domestic producers, developments in technology and the export performance and productivity of the domestic industry’. With regard to the threat of material injury, Article 3.7 goes a little further than the original Code in listing factors which the authorities should consider in this respect. These are: (i) a significant rate of increase of dumped imports into the domestic market indicating the likelihood of substantially increased importations; (ii) sufficient freely disposable or an imminent, substantial increase in capacity of an exporter indicating the likelihood of substantially increased dumped exports to the importing country’s market, taking into account the availability of other export markets to absorb any additional exports; (iii) whether imports are entering at prices that will have a significant depressing or suppressing effect on domestic prices, and would likely increase demand for further imports; and (iv) inventories of the product being investigated. This is similar to the definition of ‘threat of material injury’ already given in the antidumping regulations of some users of antidumping policy. However, it is unlikely to allay the fears of the newly industrialising countries that any increase in investment in an exporting industry could trigger an antidumping investigation. An important issue in the negotiations concerned the definition of ‘domestic industry’ to be used in the determination of material injury. The GATT antidumping code has always required that injury to a major proportion of the domestic industry must be established. The EC had wanted its Single Market to be divisible into regions for the purpose of determining injury. This was strongly opposed by other countries. Article 4.3 of the new Code states that where countries have achieved a degree of integration such that their combined domestic market has the characteristics of a single market, the industry of the entire area is to be taken as the domestic industry. However, Article 4.1 says that: in exceptional circumstances the territory of a Member may, for the production in question, be divided into two or more competitive markets and the producers within each market may be regarded as a separate industry if (a) the producers within such market sell all or almost all of their production of the product in question in that market and (b) the International Trade Policy 104 demand in that market is not to any substantial degree supplied by producers of the product in question located elsewhere in the territory. In such circumstances, injury may be found to exist even when a major portion of the total domestic industry is not injured, provided there is a concentration of dumped imports into such an isolated market and provided further that the dumped imports are causing injury to the producers of all or almost all of the production within such market. In this case, duties should normally only be levied on imports to the market in question. If that is not possible and only after the exporter has been given the opportunity to resolve the matter through undertakings, the importing country may impose duties without limitation. Thus, under strict conditions, the EU can presumably divide its market for the purpose of determining injury. Moreover, if undertakings cannot be secured from exporters, duties could be applied at an EU-wide level although injury may only apply to producers in one part of the EU. Antidumping investigations One change proposed by some countries in the negotiations was for stricter rules governing the evidence that must be provided before an antidumping investigation can be started. It was also argued that it should be made more difficult for countries to impose provisional antidumping measures without first giving the dumpers the opportunity to defend themselves and without a preliminary finding of dumping and injury. Article 5 of the new Code includes some tougher conditions which must be met before an investigation can be initiated. Article 5.4 states that an application for an investigation may only be considered if the application has the support of domestic producers collectively accounting for 50 per cent of total production ‘of that part of the industry expressing either support for or opposition to the application’. … No investigation shall be initiated where domestic producers expressly supporting the application account for less than 25 per cent of total production of the like product produced by the domestic industry.’ An important footnote to Article 5.4 states that ‘members are aware that in the territory of certain Members, employees of domestic producers of the like product or representatives of those employees, may make or support an application for an investigation’. This codifies an understanding which has existed since the 1967 Antidumping Code. Article 5.8 provides for immediate termination of an investigation where the margin of dumping is de minimis (defined as less than 2 per cent of the export price) or the volume of dumped imports (actual or potential) or the injury is negligible (generally defined as less than 3 per cent of total imports). In the negotiations, the US unsuccessfully fought for the de minimis margin to be 0.5 per cent. Article 6 of the new Code contains new procedures designed to make it easier for interested parties—defined as the exporters or foreign producers subject to investigation, the government of the exporting country and producers in the importing country—to present evidence. There is also a requirement that industrial users of a product and representative consumer organisations be given the opportunity to provide relevant information to the investigation. Article 7 states that provisional measures may only be Unfair Trading Practices 105 applied if there has been a proper preliminary affirmative determination of dumping and injury. The imposition of duties Article 9 of the new Code contains some new and stricter procedures for reimbursing exporters who pay antidumping duties which turn out to be greater than the antidumping margin. A further important provision in the new Code concerns the case of companies in an exporting country subject to antidumping measures which were not exporting the products during the period of the investigation but which are likely to face even higher rates of antidumping duty when they commence exports to the importing country. This is because the practice of countries applying antidumping measures is often to set a higher rate of duty for imports coming from companies in the dumping country which did not provide evidence for the investigation. Article 9.5 requires the authorities in the importing nation ‘to promptly carry out a review for the purpose of determining individual margins of dumping’ for such exporters. The review must be carried out on an accelerated basis compared with normal proceedings, and no duties may be imposed during the review period. If, however, the review shows that such exporters have been dumping, duties may be imposed retroactively. An important new addition to the antidumping rules is a sunset clause setting a fiveyear limit to the imposition of duties. Article 11.3 states that ‘any definitive antidumping duty shall be terminated on a date not later than 5 years from its imposition…unless the authorities determine, in a review …that the expiry of the duty would be likely to lead to continuation or recurrence of dumping and injury’. The same rules apply where price undertakings are preferred to duties. This is more or less identical to the sunset clause in EC antidumping rules. Placing some limit on the duration of antidumping duties must be welcomed but, as the experience of the EC demonstrates, it is not entirely a panacea. The US was strongly opposed to the sunset clause. Apparently, over 10 per cent of US antidumping duties had been in place for more than twenty years. A new Article 12 seeks to bring about greater transparency and openness in antidumping investigations by requiring countries to give proper public notice of investigations and of preliminary and final determinations. The same provisions apply for reviews of existing antidumping measures. Consultation and dispute settlement Article 17 sets in place procedures for resolving any disputes arising between members regarding antidumping. Any disputes which cannot be resolved by bilateral consultations may be referred to the Dispute Settlements Board (DSB) of the WTO for examination. However, Article 17.6 makes clear that, in its assessment of the facts, the panel should be confined to determining ‘whether the authorities’ establishment of the facts was proper and whether their evaluation of these facts was unbiased and objective’. In other words, the panel is not permitted to make a judgement as to whether in its evaluation the antidumping authorities came to the right conclusion. Furthermore, ‘where the panel finds that a relevant provision of the Agreement admits of more than permissible interpretation, the panel shall find the authorities’ measures to be in conformity with the International Trade Policy 106 Agreement if it rests upon one of those permissible interpretations’. These guidelines regarding the settlement of disputes arising under the new Code were included at the insistence of the US and are widely seen as leaving importing countries with considerable discretion. It seems unlikely that exporting countries will have much success in getting decisions overturned by the DSB except where measures are clearly GATT-inconsistent. * * * How are we to assess the new agreement on antidumping? Before the completion of the Uruguay Round, a tightening up of the Antidumping Code was generally regarded as one of the most important tasks facing negotiators. A widely held view was that any liberalisation package would be of little value unless new rules were introduced to restrict the ease with which countries can use antidumping measures to interfere with trade. However, it became clear as negotiations proceeded that the two main users of antidumping, the United States and the EC, were not prepared to permit any reduction in the strength of their antidumping armoury. On the contrary, they pushed hard for a strengthening of the rules to permit quicker and more effective action against dumpers— for example, through the inclusion of new anticircumvention provisions within the Code. This made it almost inevitable that no agreement was likely to be very satisfactory from a free-trade point of view. The fact that no new rules were introduced to permit anticircumvention measures is surely a relief, although the present uncertainty regarding the permissibility of such an extension of antidumping is far from wholly satisfactory. In other respects, however, the new Code is a disappointment. The main improvements are procedural. The new rules make it somewhat harder for domestic producers to bring an antidumping action. The rules also reduce slightly the degree of discretion which the antidumping authorities in the importing country currently enjoy both in the finding of dumping and the determination of injury. Although the Code fails to tackle many of the highly dubious methods which countries use to prove dumping, there are some restrictions on the methods which are acceptable for calculating the margin of dumping. There is some improved protection for smaller exporting nations and the new sunset clause ensures that, after five years, antidumping measures will automatically expire unless the importing country can demonstrate that their removal would lead to renewed dumping and injury to domestic producers. However, these changes are unlikely to be sufficient to deter producers from seeking import relief through antidumping. The opportunities which antidumping creates for bringing relatively swift action to bear against troublesome imports will continue to make it an attractive option. Indeed, the outlawing of VERs in the new Safeguards Agreement (discussed in the previous chapter) may lead to increased resort to antidumping. The methods used to determine dumping will continue to be of questionable meaning or objectivity and are likely to continue the familiar pattern of grossly inflated dumping margins. Indeed, in several ways, the effects of the new Code are malign by actually codifying certain practices which countries have in the past used to inflate dumping margins. Although the new Code makes greater provision for other groups of producers and consumers or users of the product to be consulted and for exporters to put their case before the antidumping authorities, an opportunity has been largely missed to ensure that antidumping decisions take into account the interest of the whole of the importing country and not just the domestic producers petitioning for protection. Worse still, the Code fails to provide a satisfactory mechanism for the monitoring of antidumping Unfair Trading Practices 107 decisions. Although exporting countries can appeal to the Disputes Settlement Board, the terms of reference of any panel set up to investigate such a dispute render it improbable that many cases will be reversed. In this respect, also, an opportunity has been missed to establish proper multilateral machinery for adjudicating antidumping cases which could have been used to protect exporters from the discretion of antidumping authorities taken captive by beleaguered import-competing domestic producers. SUBSIDIES AND COUNTERVAILING DUTIES Subsidies and their effects A second type of unfair trade practice is the subsidy. At the outset, it is necessary to draw a distinction between an export subsidy and a domestic subsidy. In the case of an export subsidy, a producer receives a subsidy only on the amount which is exported. A domestic subsidy is paid to a producer on all that is produced regardless of whether the output is for export or the home market. In a manner similar to dumping, export subsidies allow an exporter to sell the good in a foreign market at a lower price than at home and possibly below costs of production. As with dumping, GATT rules deem such a practice to be ‘unfair’ if it causes or threatens material injury to producers in the importing country. Article VI allows countries to impose countervailing duties on such imports, provided that the rate of duty does not exceed the element of subsidy. As is explained later (p. 139), agricultural exports constitute an important exception to the rule. Export subsidies may be disguised in various ways. One form of this is the use of export credit subsidies whereby governments provide subsidised credit to foreign importers who purchase goods from the exporting country using loans taken out with a bank in the exporting country. In the past, such export credit subsidies have been the subject of a special ‘gentlemen’s agreement’ between OECD countries. The approach has been to agree limits on the amount of interest rate subsidy permissible for exports to different markets of the world. In other words, export credit subsidies have in the past not been regarded as a GATT issue. Domestic subsidies are in some respects a more complex issue since their purpose is often not a distortion of trade. In addition to financial aid granted to a particular producer or industry, they include total or partial tax exemption, remission of tax, provision of credit on special terms, and preferential treatment in the provision of public infrastructure. Although the primary intention may not be to restrict trade, either exports or imports may be indirectly affected. If the producer or industry which is subsidised exports part of its output, the subsidy will enable it to export at a lower price than would otherwise be possible. Foreign producers may therefore regard such a subsidy as a form of ‘unfair’ competition. Alternatively, if the producer or industry being subsidised sells all of its output domestically but is competing with imports from abroad, the subsidy may enable it to undercut foreign exporters. In this case, the effect is similar to a tariff in discriminating against foreign-produced goods. The GATT Treaty contained no provisions to control domestic subsidies. However, in recent decades, countries have become increasingly concerned about the trade-distorting effects of these measures. The concern was greatest in countries which adopted more laissez-faire policies. They argued International Trade Policy 108 that their producers were at an increasing disadvantage when competing with imports coming from countries where governments adopted more interventionist measures. There was a sense in which, as tariffs were gradually lowered, the impact of such measures on trade was more strongly felt. It may also have been the case that growing government intervention in industry in the 1960s and 1970s meant that subsidies played a more tradedistorting role than in earlier years. More recently, however, mounting deficits have caused governments to reduce the overall level of subsidies to industry, although more careful targeting of subsidies has been an accompanying factor (Ford and Suyker, 1989). Both export and domestic subsidies were the subject of a new Subsidies Code agreed in 1979 as part of the Tokyo Round. Its significance was that, unlike the GATT Treaty itself, it included domestic as well as export subsidies. Moreover, it went much further than the GATT Treaty in elaborating and interpreting the GATT provisions. A major deficiency of the Code was the failure to include agricultural subsidies, which have subsequently become a major source of trade-distortion (see Chapter 6 for an explanation for why agricultural subsidies were treated differently). The Leutwiler Report of 1983 listed revision, clarification and more effective rules on subsidies as one if its fifteen recommendations (GATT, 1985). Subsidies and so-called countervailing measures were included on the agenda of the Uruguay Round and a special negotiating group was set up to deal with this issue. A new Agreement on Subsidies and Countervailing Measures (SCM) was contained within the Uruguay Round Final Act. The new Agreement is discussed below. However, before doing so, it is necessary to examine at a theoretical level the effects of the two types of subsidy on trade. The economic effects of subsidies Domestic subsidies Figure 4.4 illustrates the effects of a subsidy on domestic production of a particular good of which the country is an importer. OPW is the world price. Domestic production equals OQ0 and domestic consumption OQ2 and the volume of exports Q0Q2. The effect of the subsidy of PWPS (=RS) is to push the domestic supply curve SDSD to the right to Domestic production increases by Q0Q1 and imports fall to Q1Q2. Domestic producers gain increased producer surplus equal to area PWPSRT. However, the subsidy costs the government area PWPSRS. So there is a net welfare loss to the importing country equal to area RTS or A. The difference between a subsidy and a tariff is that a subsidy results in no loss to consumers since the domestic price is unaffected. This means that a domestic subsidy is always to Unfair Trading Practices 109 an international agreement not to use export subsidies with a view to achieving strategic advantage. The problem with any such agreement is that each country has an incentive to cheat and grant covert illicit subsidies to its own exporters. For this reason, a country may be unwilling to co-operate in an agreement to control subsidies. This leads to the argument that countries which are seeking co-operation must resort to subsidies in a titfor-tat manner in order to goad other countries to co-operate. In this case, the export subsidy is being used to punish other countries for engaging in practices harmful to producers in the retaliating country. Strategic trade policy suffers from a further problem. A policy of targeting strategic export industries assumes that governments are capable of behaving in an objective manner. Even assuming they have all the necessary information at their disposal for making a rational decision, will they act dispassionately in maximising national economic welfare? The political economy of trade policy formation suggests this is unlikely. Instead, it is more likely that they will respond to whatever industrial pressure groups are most effective in lobbying for subsidies. Past experience shows that politicians will favour those groups with the greatest lobbying clout rather than act on available objective criteria. In general in the lobbying process, experience shows that producer groups tend to gain at the expense of consumer groups and the more organised industrial lobbies at the expense of the least organised. Industries located in regions containing a large number of marginal constituencies may well do better than industries in other regions. As Grossman (1986) has put it: ‘the market failures in the political realm might easily outweigh those in the economic realm, leaving us with a set of strategic trade policies that would serve only the interests of those fortunate enough to gain favor’. GATT rules on subsidies The original GATT Treaty contained very little in the way of discipline to control the use of subsidies which might distort trade. The issue of export subsidies was dealt with as part of Article VI of the GATT. In addition to the right of a country to impose antidumping duties to counteract dumping, Article VI authorises the imposition of countervailing duties on imports which have been subsidised by the exporting country. As with antidumping duties, countervailing duties must not exceed the amount of the subsidy granted. As with dumping, the subsidy must cause or threaten material injury to an established domestic industry or retard the establishment of a domestic industry. In addition, Article XVI requires countries to notify the GATT of any trade-distorting subsidy. It further added that where a subsidy seriously prejudiced the interests of another contracting party, the subsidy-granting country should be prepared to discuss ways of limiting the subsidy. But this did not amount to very much. At the first review session in 1955, more substantive obligations were added. These are now set out in the second part of Article XVI but they relate to export subsidies only. A distinction is drawn between primary and nonprimary products. In the case of primary products, contracting parties are required to ‘avoid the use of subsidies’ on exports. However, where they are granted, they should not result in the contracting party ‘having more than an equitable share of world export trade’—whatever that might mean. In the case of nonprimary products, it is stated that, with effect from 1 January 1958 ‘or the earliest practicable date thereafter’, countries were to cease granting export subsidies which resulted in a price lower than the International Trade Policy 116 domestic price of the good. Because of the different treatment accorded to nonprimary products, developing countries saw this as a form of discrimination against their exports and therefore refused to adopt this aspect of the 1955 amendments. The only country in the world to have made extensive use of the Article VI provisions has been the United States. Before the US Trade Act of 1974, although US importers sought to make use of countervailing law against allegedly subsidised imports, they were not very successful. Out of 191 investigations between 1934 and 1968, only 30 resulted in the imposition of countervailing duties (Destler, 1992). However, in 1974 US countervailing law was changed in a way which made relief easier to obtain. It required final action to be taken within a year of any petition for relief being received and provided for any decision that denied relief being subject to judicial review. The result was a significant increase in the number of countervailing investigations. A greater number of these resulted in affirmative decisions (35 between 1976 and 1978). However, due to a special Congressional waiver permitting the President not to impose countervailing duties for four years if the foreign government took steps to reduce its subsidy support, a significant number of these cases did not result in duties being applied. The reasons were political. The US was anxious to get the agreement of other countries on a new subsidies code as part of the Tokyo Round. In the US, subsidies were widely regarded as a means whereby other countries were able to gain an unfair advantage in trade. Existing GATT rules were considered as inadequate for coping with this situation. On the other hand, other countries had legitimate grounds for complaint against the way in which countervailing duties could be imposed under US law. Because the countervailing laws of the US had been established before the GATT, the US was entitled to so-called ‘grandfather rights’ under the Protocol of Provisional Application. This meant that it was not bound to apply all the provisions contained in Article VI when these differed from US law. A specific aspect of this concerned the material injury test which makes a demonstration that imports have caused or threatened material injury to domestic producers a precondition for the imposition of countervailing duties. US countervailing law contained no such requirement, so for several decades the US was able to impose countervailing duties on imports without the need to demonstrate that imports were causing or threatening material injury to domestic producers. Other countries were insistent that, as a quid pro quo for any new discipline governing subsidies in trade, US countervailing laws should be brought into line with GATT law in this respect. The 1979 Subsidies Code The issue of subsidies was a key issue in the Tokyo Round. In particular, the US was anxious to introduce new disciplines on other countries and was prepared to bring its own countervailing laws into conformity with the GATT in order to secure an agreement. The approach adopted was to negotiate a separate stand-alone agreement or ‘code’ rather than seek to amend the GATT. The same device was used for tackling ten other types of nontariff barriers including antidumping (see above, p. 105). As codes are additions to the GATT rather than amendments, they only bind those countries which agree to sign the Code. The 1979 Subsidies Code had two parts or ‘tracks’. The first covered countervailing duties and specified clearer rules, including the requirement that the subsidy be causing or threatening material injury to domestic producers. The 1979 US Unfair Trading Practices 117 Trade Agreement Act revised US countervailing statutes accordingly. On the other hand, it also made clear that the full benefits of the Code would only be extended to those countries which had signed the Agreement or generally accepted its obligations. Moreover, because the Code provided for a general exception to its obligations for developing countries, the US deemed such developing countries to be ineligible to receive its full benefits. This meant that countervailing duties could be imposed on these imports without a material injury test. Alternatively, these countries could render themselves eligible to receive the full benefits of the Code by reaching separate bilateral agreement with the US. Subsequently, a number of developing countries did so. Nevertheless, a large number of countervailing duty cases which followed the 1979 Act were resolved without any injury test. A further weakness of this part of the Code was the absence of any definition of a ‘subsidy’ for countervailing duty purposes. There was a widely held view that this left governments with too much latitude in applying countervailing duties to imports. The second part of the Code dealt with the obligations of countries with regard to subsidies affecting trade. The Code prohibited export subsidies on nonprimary products and went somewhat further than Article XVI. However, export subsidies affecting primary products were allowed so long as they did not result in a ‘larger than equitable’ share of trade (as stated in Article XVI) or depress prices unduly. The US had wanted limits imposed on both industrial and agricultural export subsidies but the EC was not prepared to accept this. The limits on primary product export subsidies were clearly much softer than those imposed on industrial products and fell a long way short of what the US had been seeking. However, for the first time, an agreement was reached on subsidies other than export subsidies. It made clear that where domestic subsidies were desirable for social or economic reasons, they should be allowed. On the other hand, it was recognised that domestic subsidies can cause or threaten injury to domestic producers and ‘nullify or impair’ benefits extended by one contracting party to another. Therefore, the Code required countries to avoid subsidies that have these effects. Clearly, for this to work, countries need to supply information about the various kinds of subsidy granted, but countries proved reluctant to do so. There is also the question of what kinds of assistance to industry should be classified as a subsidy. Almost every kind of government activity can be regarded as giving some assistance to domestic industry and therefore could conceivably be included. The Code failed to provide guidance. Nevertheless, the attempt to establish some discipline in relation to domestic and not just export subsidies was widely regarded as a significant step forwards. As stated above, a particular difficulty with the Code was the exception granted to developing countries. Under the Code, they were permitted to use export subsidies on industrial products. But, where these were inconsistent with their competitive and development needs, they were expected to reduce or eliminate such subsidies. Although the Subsidies Code proved a useful first step towards bringing subsidies in trade under some form of international discipline, it left a number of issues unresolved. These were left over to the next round of GATT. Meanwhile, the incidence of US countervailing measures increased following the passage of the 1979 US Trade Agreements Act and despite the fact that a material injury test was now incorporated into US countervailing rules. In part, the reason was that the 1979 Act actually helped US firms that were petitioning for countervailing measures against allegedly subsidised International Trade Policy 118 foreign imports. It did so by setting strict time limits on countervailing duty cases under investigation, and this tended to favour petitioners rather than foreign suppliers, who had less time to work on and present their defence. It also provided for temporary measures to be taken against subsidised imports if there was a preliminary finding of injury, thus making possible earlier action against imports. Furthermore, responsibility for remedying unfair trade practices was passed from the Treasury to the Department of Commerce which was certain to be more sympathetic to the concerns of industry. The early 1980s witnessed a surge of countervailing cases in the US. Between 1980 and 1984 there were 249 investigations; of which 135 or 54 per cent resulted in the imposition of duties or suspension of the subsidy (Destler, 1992). The majority of these involved steel products, and were in fact resolved by the US negotiating a series of voluntary export restraint arrangements with offending countries. After 1985, the number of countervailing duty cases fell. Thus, between 1985 and 1989, there were only 96 cases. One reason for the decline was that the cases involving steel had been resolved by other countries agreeing VERs with the US. The other reason was the declining use of subsidies as interventionist policies fell out of favour and governments sought ways of reducing budget deficits. Civil aircraft subsidies A classic example of the problems involved in regulating the use of subsidies in trade is provided by the long-standing dispute between the US and the EC over civil aviation subsidies. Civil aircraft were the subject of a separate code negotiated as part of the Tokyo Round. This aimed to reduce tariffs and nontariff barriers affecting this trade. It also identified domestic subsidies and other special assistance given by governments to producers as a factor distorting competition between countries. The civil aircraft industry was also subject to the provisions contained in the more general Subsidies Code. However, the US expressed disappointment with the Aircraft Code in particular because it failed to establish adequate discipline over the use of subsidies by governments to gain unfair advantage in the world market. Specifically, the US was aggrieved by the extent to which Europe was subsidising the production of new aircraft as part of its Airbus programme. A particular issue, which became the subject of a US complaint to the GATT and led to the setting up of a disputes panel, concerned subsidies from the German government to Deutsche Aerospace to cover potential exchange-rate losses on the sale of Airbus aircraft as a result of a fall in the value of the dollar. The US regarded this as a blatant export subsidy, while the EC argued that it was a currency insurance scheme. The EC argued further that US companies enjoyed substantial indirect support from government subsidisation of military R&D funding and direct help from civil government budgets, such as NASA. In part, the problem arose because EC subsidies to Airbus were production and not export subsidies and therefore did not violate existing rules. Moreover, the option of using countervailing duties against the EC was hardly appropriate in this case. Although the US threatened to do so, this would only have affected Airbus sales in there and not sales in third markets. Nevertheless, the possibility of this sanction being employed did serve to persuade Europe to seek a new agreement. On 1 April 1992, the US and the EC succeeded in reaching a bilateral agreement on civil aircraft subsidies. The US agreed to withdraw its threat to take the issue to the GATT or to take unilateral action against Airbus sales in the US. However, the parallel Unfair Trading Practices 119 complaint against the German scheme for protecting Deutsche Aerospace from exchangerate fluctuations was treated as a separate case and would not be dropped. The agreement covered all new civil aircraft with over 100 seats but did not affect subsidies granted before the agreement. Direct subsidies for aircraft production were to be banned and direct subsidies for new aircraft programmes were to be limited to 30 per cent of total development costs. This was well below the 45 per cent ceiling sought by the EC but above the 25 per cent cap demanded by the US. The US implicitly admitted that its manufacturers benefited from indirect support by agreeing, at the request of the EC, to limits on indirect subsidies also. Indirect grants were not to exceed 5 per cent of the manufacturer’s civil aircraft turnover. Other aspects of the agreement included meetings twice a year for the mutual exchange of information about current and future projects and the level of government support attached to these programmes; a ban on government inducements to third countries to buy their aircraft; and provisions for the agreement to be temporarily suspended if a manufacturer were adversely affected by external factors (see The Financial Times, 2 April 1992). The intention was to make the bilateral agreement the centrepiece for a multilateral agreement on aircraft subsidies the following year. However, in February 1993 the newly elected Clinton Administration indicated its intention to reopen the issue of aircraft subsidies. Faced with job losses among domestic aircraft producers, the US Trade Representative, Mickey Kantor, approached the EC for fresh consultations over Airbus subsidies. A specific issue was lack of transparency. The US claimed that loans made to members of the Airbus consortium were at significantly lower rates of interest than in the past and well below market rates. At the same time, it threatened to bring countervailing duty complaints against Airbus, an action which would entitle Airbus to terminate the agreement. Later, the US added the demand that the ceiling on direct subsidies should be lowered to 20 per cent and the agreement be multilateralised to include other countries, most notably Japan. The 1992 agreement stated that no review of the agreement or withdrawal from the agreement could happen before July 1994. On the other hand, there was pressure on both sides to reach a new multilateral agreement within the framework of the Uruguay Round before these negotiations were concluded. Not least, the EU was keen to reach a separate agreement on civil aircraft which would give it greater flexibility for aircraft subsidies than any general subsidies agreement contained in the Uruguay Round. However, in December 1993, shortly before the Uruguay Round was concluded, talks broke down. At the time of writing, agreement has still not been achieved. Interestingly, the two sides have changed their posture. The US is reported as favouring reliance on the new tighter Subsidies Code negotiated as part of the Uruguay Round (see below). It has argued that a separate agreement for aircraft is no longer necessary. By way of contrast, the EU has demanded a toughening of the Aircraft Subsidies Code, particularly with a view to placing tighter controls on indirect government support not covered by the broader general Subsidies Code. It also wants subsidy disciplines extended to aero-engines and parts, which it maintains are especially subject to such indirect support (The Financial Times, 19 May 1994). International Trade Policy 120 The new Subsidies Code The issue of subsidies was covered in the Uruguay Round by the same negotiating group that dealt with antidumping. Progress was initially slow because of a basic difference between those countries, in particular the United States, which were concerned about the growing use of subsidies to distort competition in trade, and other countries which were more concerned about the abuse and excessive use of countervailing duties against allegedly subsidised imports. The former wanted either the complete abolition of tradedistorting subsidies or, at least, much tougher rules to control subsidies. The latter wanted GATT rules governing the use of countervailing duties to be made stricter and in particular to avoid harassment of exporters who were trading fairly. At the Mid-term Review, it was agreed that subsidies should be divided into three different categories: subsidies which should be prohibited; nonprohibited subsidies but against which countervailing action could be taken if they are shown to be injurious; and subsidies against which no action would be allowed. The argument was over the category into which different kinds of subsidy should be put. The US favoured putting most subsidies into the first category, whereas the EC favoured including many types of subsidy in the second and even third categories. There was agreement that export subsidies should be prohibited. However, the US argued that many other kinds of subsidy could achieve the same effects as a straight export subsidy. She wanted this category extended to include those contingent on firms meeting domestic content or local sourcing requirements, those going to firms that are predominantly exporters, and domestic subsidies that exceed a given percentage of a company’s total sales. The EC approach was to emphasise two principles: firstly, that a subsidy should only include government actions which impose a cost on the granting government; and, secondly, that a subsidy should be specific to a firm or industry. Only such subsidies would be regarded as ‘actionable’ and therefore subject to GATT discipline. The US argued that some kinds of subsidy might impose no cost on the government yet still confer benefits on an exporter. For example, if a government can borrow money more cheaply than a private borrower, it may be able, at no cost to itself, to lend funds to an exporter at a rate of interest below the market rate. The specificity concept, however, was more generally acceptable to the US, being already embodied in US countervailing duty law. This would exclude all general measures used by governments which benefit all producers and therefore do not distort trade. It can reasonably be argued that, where a country makes greater use of general subsidies than other countries, any distorting effect on overall competitiveness will be eliminated through an upward appreciation of the exchange rate. Nevertheless, there is still a problem that many nonspecific subsidies might in effect bring disproportionate benefits to particular exporters or industries and thus still distort trade. The new Subsidies Code in the Uruguay Round Final Act is entitled Agreement on Subsidies and Countervailing Measures. In Part I of the agreement, Article 1.1 defines a subsidy as ‘a financial contribution by a government or any public body…or…any form of income or price support’ where ‘a benefit is thereby conferred’. This can include direct transfer of funds (such as grants, loans and equity infusion), tax allowances or concessions, government provision of goods and services other than general Unfair Trading Practices 121 infrastructure or government assistance to a funding mechanism. This adopts the cost-tothe-granting-government approach to the definition of a subsidy favoured by the EC. Article 2 incorporates a specificity requirement. Significantly, paragraph 2.1(c) goes most of the way to meeting possible objections to a narrow interpretation of specificity by allowing other factors to be taken into account, namely, ‘use of a subsidy programme by a limited number of certain enterprises, predominant use by certain enterprises, the granting of disproportionately large amounts of subsidy to certain enterprises, and the manner in which discretion has been exercised by the granting authority in the decision to grant a subsidy’. Part II identifies subsidies which are prohibited under the agreement and establishes a clear procedure for remedying a situation where one party considers that such a subsidy is being granted. Prohibited subsidies are defined as ‘subsidies contingent, in law or in fact, whether solely or as one of several other conditions upon’ either export performance or the use of domestic over imported goods (Article 3.1). Annex I of the agreement contains an illustrative list of export subsidies. They include governmental insurance against exchange-rate risks such as was provided by the German government to Deutsche Aerospace and which was the source of a US complaint to the GATT. They also include export credits at below-market rates of interest except where a member is party to an international agreement on official export credits (or in practice applies the provisions of such an agreement). Part III is concerned with ‘actionable subsidies’ which are defined as subsidies as defined in Article 1 which cause ‘adverse effects to the interest of other Members’. This may happen in one of three ways: injury to domestic producers; nullification or impairment of concessions made by countries in the course of the Uruguay Round; or ‘serious prejudice’ to the interests of another member country. Except for civil aircraft, the threshold for a subsidy to be actionable is set at 5 per cent of the value of the product. (Annex IV of the agreement contains rules for calculating the ad valorem subsidisation.) Article 6 makes clear that ‘serious prejudice’ exists whenever subsidies cover losses sustained by an industry or firm, or write-off government debt and have a cross-border effect. The latter is defined as (a) displacing or impeding either the imports or exports of another member; (b) resulting in significant price undercutting in the market of another member; (c) resulting in an increase in the world market share of a particular primary product or commodity compared with its average share of the previous three years. Agricultural subsidies are, however, not included but are dealt with separately. As with prohibited subsidies, there is a set procedure for remedying situations arising whenever one country believes that actionable subsidies are having the above effects. Part IV identifies a third category of nonactionable subsidies. These cover all nonspecific subsidies, assistance to research activities (providing that it does not cover more than 75 per cent of the costs of industrial research or 50 per cent of the costs of precompetitive development activity), regional aid (subject to certain strict conditions) and assistance to help existing facilities adapt to new environmental requirements. However, there is a requirement that members must notify the WTO of any nonactionable subsidy programmes which it intends to implement plus any subsequent changes to make sure that they conform to the conditions and criteria stipulated. Part V deals with countervailing measures. Article 11 sets out new stricter requirements for countervailing duty investigations. There is a requirement that sufficient International Trade Policy 122 evidence must be provided by petitioners for both the existence of a subsidy and a causal link between subsidised imports and injury before an investigation is initiated. Simple assertion which is not substantiated is not sufficient. A de minimis requirement that the amount of the subsidy must exceed 1 per cent of the value of the product must also be satisfied. Article 12 sets out the rights of interested members and all interested parties during the investigation itself. Article 14 contains rules regarding the calculation of the amount of the subsidy. There is a general requirement that the method used must be set out in national legislation and be applied in each particular case in a way that is transparent and properly explained. In addition, further guidelines are included which make clear that government provision of equity capital, a government loan or loan guarantee, or government provision of goods or services cannot normally be treated as conferring a benefit. Article 15 sets out rules for the determination of material injury as required under Article VI. These are similar to those stipulated for antidumping. Many of the other provisions relating to countervailing duties are much the same as those provided for dumping. As with antidumping duties, there is a new sunset clause providing for duties to be terminated after five years unless it can be demonstrated through a review that subsidisation and injury would reoccur with the removal of the duty. Finally, Parts VIII and IX contain some important provisions for developing countries and countries in transition. It is recognised that subsidies play an important role in the economic development of developing countries. The poorest of them are exempt from the general prohibition affecting all export subsidies. Other developing countries are allowed eight years in which to phase out all export subsidies, with provision for longer in exceptional cases. Where export subsidies are inconsistent with development needs, however, they must be eliminated within a shorter period of time and not increased. Any developing country which reaches ‘export competitiveness’ in a particular product— defined as a share of at least 3.25 per cent in world trade for two consecutive years— must phase out export subsidies within a two-year period. Developing countries are also exempt from some of the provisions relating to actionable subsidies, although not if nullification or impairment of tariff concessions or other GATT obligations occurs. Countervailing duty investigations against products from developing countries must also be terminated if the overall level of subsidy is less than 2 per cent of the value of the product (3 per cent for countries which have eliminated export subsidies within the stipulated eight-year period) or the volume of subsidised imports is less than 4 per cent of total imports. Part IX makes clear that countries in transition from a centrally planned to a market-based economy are also subject to special arrangements because of the need to use subsidies in the interim period. They have seven years in which to eliminate export subsidies. As with developing countries, actionable subsidies are not subject to the provisions for developed market economies except where there is a nullification or impairment of tariff concessions or other GATT obligations. The new subsidies agreement clearly represents a considerable improvement on the former 1979 Code. In particular, it contains a much clearer definition of what constitutes an actionable subsidy. The prohibition of all kinds of export subsidy, with an illustrative list of what might be included, accords with US demands for the abolition of the most obvious kinds of trade-distorting subsidy. The identification of a second category of actionable subsidies with a specificity requirement and a need to demonstrate a crossborder effect provides a clearer set of rules for tackling subsidies other than export Unfair Trading Practices 123 subsidy. It also meets the concerns of those countries opposed to any kind of outright ban on subsidies which are considered desirable for other reasons. The identification of a third category of nonprohibited subsidies further ensures that subsidies with no obvious trade-distorting effect are not subject to the new measures. The agreement also contains a much clearer statement of the method to be used in calculating the level of subsidy. The provisions relating to the use of countervailing duty investigations should also significantly reduce the risks to exporters who are trading fairly of being subject to harassment. Finally, there would appear to be a consensus that developing countries should be treated differently while recognising that these countries cannot forever be allowed to enjoy special treatment. An important merit of the new Subsidies Code is that it will apply to all WTO members unlike the Tokyo Round Code which applied only to those countries that chose to accept its obligations. There remain a number of loose ends, which must be left to future rounds to knit up in the light of the experience of the new agreement. It is also the case that agricultural subsidies continue to be subject to different treatment even though the Uruguay Round has succeeded in providing for a substantial reduction in their levels. As with the antidumping provisions, the adequacy of these new arrangements will only be fully known in the light of subsequent experience. CONCLUSION The gradual lowering of tariff barriers by the developed market economies has brought in its wake increasing demands for a so-called ‘level playing field’ in international trade. Free trade, the argument goes, is not possible between countries which play by different rules. Dumping and subsidisation are seen as two ways by which some countries compete ‘unfairly’. Closer scrutiny of the arguments about dumping suggest that much of this concern is ill-founded. The one case where a clear-cut argument exists for combating dumping is a situation where a dominant supplier engages in below-cost pricing with predatory intent. Such cases are certainly fairly rare. Where, however, predatory pricing is found to be taking place with harmful consequences, the matter could readily, and probably more effectively, be handled by the antitrust authorities in the importing country. Regardless, however, of whether or not an economic rationale exists for antidumping, it seems unlikely that countries will dispense with the armoury which they have created for dealing with it. If this is so, the need then becomes one of devising rules which will prevent antidumping from becoming an easy means for producers who are unable to cope with increased competition from abroad from gaining back-door protection. The frequent use of antidumping by developed market economies in recent decades suggests that it has become too easy an option for rent-seekers unable to get protection by other means. The new Antidumping Code goes some way to tackling some of the anomalies which have existed up to now. Nevertheless, in other respects, it is woefully inadequate. It is likely that countries will continue to make considerable use of the latitude which the existing rules allow. It is already the case that a growing number of investigations are originating in developing countries clearly ready to play the developed countries at their own game. Subsidies raise somewhat more complex problems. Although the existence of an element of subsidy in the price of an imported product may be used by an importing International Trade Policy 124 country to justify countervailing measures equivalent to those of antidumping policy, governments may also deliberately use subsidies to distort trade. Therefore, rules are required to control the use of trade-distorting subsidies; they need to distinguish those subsidies which are clearly damaging to producers in other countries from those which are not. It is desirable that these should cover domestic as well as export subsidies. Given the absence of such rules in the past, except the provisions for countervailing measures against export subsidies, the new Subsidies Code is to be welcomed. One immediate result is certain to be an increase in the number of disputes which concern the use of subsidies. This need not matter if, as a result, WTO member states make less use of unilateral threats. Multilateral rules and procedures are to be preferred as they permit a more objective appraisal of whether a particular practice constitutes a genuine infringement of another country’s trading rights. It remains to be seen how well the new Code will work. A developing belief that, in high-technology industries at least, subsidies are needed to get first-start advantages can also be expected to generate a growing number of disputes of this kind. Unfair Trading Practices 125 Zambia 1.2 India 1.1 Ghana 0.4 Cameroun −0.1 Senegal −0.6 Bangladesh −1.4 Sudan −1.9 1973–85 Singapore 6.5 Malaysia 4.1 Cameroun 5.6 Bangladesh 2.0 Hong Kong 6.3 Thailand 3.8 Indonesia 4.0 India 2.0 S. Korea 5.4 Tunisia 2.9 Sri Lanka 3.3 Burundi 1.2 Brazil 1.5 Pakistan 3.1 Turkey 1.4 Yugoslavia 2.7 Dominican Republic 0.5 Israel 0.4 Colombia 1.8 Uruguay 0.4 Mexico 1.3 Chile 0.1 Philippines 1.1 Kenya 0.3 Honduras −0.1 Ethiopia −0.4 Senegal −0.8 Sudan −0.4 Costa Rica −1.0 Peru −1.1 Guatemala −1.0 Tanzania −1.6 Ivory Coast −1.2 Argentina −2.0 El Salvador −3.5 Zambia −2.3 Nicaragua −3.9 Nigeria −2.5 Bolivia −3.1 Ghana −3.2 Madagascar −3.4 Source: World Bank (1987) significant element in its new programme of structural adjustment lending. Although structural adjustment loans contained many different elements, almost 80 per cent have had trade policy reform as a condition (Greenaway and Milner, 1993). Key elements were the removal of quantitative restrictions on imports, the lowering of tariffs, reductions in exchange rates which are clearly overvalued and export promotion. In 1990, a World Bank study examined some thirty-six examples of trade policy reform in nineteen countries over the entire period from 1945 to 1984 (World Bank, 1990). It found that fifteen had been successful, nine were partially reversed and twelve had collapsed. All successful programmes involved a mix of the following: reduction or elimination of import quotas, currency devaluation and tight fiscal policy. Of these measures, elimination of import quotas was found to be even more important than cutting tariffs and certain to yield positive results. An early and substantial devaluation was also found to be an important ingredient of a successful development programme but only if accompanied by tight fiscal and monetary policy. Expansionary fiscal and monetary policies were found to be the most important cause of the abandonment of trade reforms. Two potentially adverse effects of a trade policy reform programme concern the effects of cutting tariff rates on government revenues and of lowering import barriers on unemployment. Developing countries often depend on import tariffs for revenues so that a policy of cutting tariffs can create budgetary problems. (Export taxes are also an International Trade Policy 132 important source of fiscal revenues in many developing countries.) However, because lower tariff rates lead to a higher volume of imports, tariff revenues need not fall, and may even rise. Moreover, if accompanied by a devaluation which causes the local currency price of imports to rise, lower tariffs may still yield more in local currency terms. It is also the case that trade reform programmes typically involve a switch from quantitative import barriers which yield no revenue to tariffs which do. Trade policy reform may also lead to a small rise in unemployment in the short run due to a decline of employment in the import-substitution sector. On the other hand, this will be offset by increased employment in other sectors (such as agriculture) which were previously subject to negative discrimination. A key factor here is whether or not trade policy reform is accompanied by a lowering of the exchange rate which boosts output and employment in the tradable goods sector. The growing awareness among developing countries that economic growth is generally best promoted by a policy which emphasises export promotion and is mostly harmed by inward-looking policies which seek import-substitution has caused these countries to adopt a different approach towards trade negotiations with the developed countries. As noted above, for much of the postwar period, developing countries played little or no role in GATT rounds. They were largely content to reap the benefits of any tariff concessions made by the developed countries and which were automatically extended to developing countries on a most-favoured-nation basis but were unwilling to make concessions themselves as a bargaining counter to gain benefits of more interest to them. The GATT was widely viewed as a ‘rich man’s club’ bringing few if any benefits to the developing world. Indeed, in the early years, a number of developing countries (such as Mexico) chose not to sign the General Agreement because they considered the rules biased towards industrial countries. Those which did sign took little or no part in any of the earlier GATT rounds. In the recently concluded Uruguay Round, however, a significant change took place. For the first time, many developing countries took an active part and were prepared to offer concessions of value to developed countries and to accept more GATT obligations than in the past in order to achieve their particular objectives. Matters of special concern to developing countries were the need to secure improved access for products of special interest to them, such as textiles and agricultural goods, and the need to obtain a tougher and more effective disputes-settlement mechanism which would protect them against the imposition of new restrictions on their exports. This change reflected an awareness on the part of developing countries that their own interests were not served by maintaining high barriers against goods coming from the developed countries. Such barriers are more likely to impoverish the country than benefit it. Instead, the need was to secure guarantees of improved access for their exports from the developed countries in order to attract foreign investment to their economies and expand exports. To achieve this, there was a need to offer developed countries something in return. The outcome of this changed approach is discussed towards the end of the chapter, when the results of the Uruguay Round are examined. First, it is necessary to discuss how, if at all, the GATT has sought to incorporate the particular interests of developing countries. The Developing Countries 133 DEVELOPING COUNTRIES AND THE GATT The General Agreement contains only two provisions for special treatment of developing countries. Firstly, Article XVIII, entitled Governmental Assistance to Economic Development, begins by recognising that the attainment of the objectives of this Agreement will be facilitated by the progressive development of their economies, particularly of those contracting parties the economies of which can only support low standards of living and are in the early stages of development. (Article XVIII:1) Further, it states that: it may be necessary for those contracting parties, in order to implement programmes and policies of economic development designed to raise the general standard of living of their people, to take protective or other measures affecting imports… (Article XVIII:2) Section A permits a developing country to ‘modify or withdraw a concession’ if it ‘considers it desirable, in order to promote the establishment of a particular industry with a view to raising the general standard of living of its people’ (para. 7). This is the familiar case of infant-industry protection. There are the usual provisions requiring prior consultation of other contracting parties and compensatory adjustment. Section B recognises that when a country is in the process of rapid economic development it may ‘experience balance of payments difficulties arising mainly from efforts to expand their internal markets as well as from the instability in their terms of trade’ (para. 8). Therefore, a developing country is permitted, ‘in order to safeguard its external financial position and to ensure a level of reserves adequate for the implementation of its programme of economic development…[to]… control the general level of its imports by restricting the quantity or value of merchandise permitted to be imported’ (para. 9). Once again, there are strict conditions, requiring consultation with other contracting parties, review of restrictions imposed by the contracting parties and, if necessary, modification of restrictions if they are found to be inconsistent with the provisions of Section B. Sections C and D set out provisions that allow developing countries to grant infantindustry protection through quantitative restraints on imports. Again, there are the usual provisions requiring notification and consultation before measures are imposed (except where the industry requiring protection has already started production). Despite the fairly strict conditions set out in Article XVIII, developing countries have imposed quantitative restraints on imports almost at will using the existence of the Article to give legitimacy to such measures. Hindley (1987) argues that, in the past, one effect of this has been ‘to International Trade Policy 134 deter developed countries from entering into normal GATT reciprocal bargaining with developing countries’. Since in practice a developing country can easily impose quantitative restrictions on imports under Article XVIII, the effect is to weaken any concessions made by a developing country when negotiating with a developed one. The second set of rules within the GATT providing for special treatment of developing countries is contained in Part IV, entitled Trade and Development, which was added to the General Agreement in 1964. Perhaps the most important provision which this contains is the nonreciprocity commitment. Article XXXVI:8 states that: the developed country parties do not expect reciprocity for commitments made by them in trade negotiations to reduce or remove tariffs and other barriers to the trade of less-developed contracting parties. This is generally taken to mean that there is no obligation on developing countries to offer any tariff concessions in GATT negotiations although they automatically receive any concessions offered by one developed country to another and which must be extended in the usual way to all other contracting parties. In other words, they are allowed to ‘free ride’. As stated above, this provision has proved to be of questionable value given that most of the tariff cuts negotiated between developed countries in subsequent GATT rounds and extended to all other countries were on products of little or no importance to developing countries. In order to gain concessions from the developed countries of value to developing countries, developing countries needed to offer something else in return. As Hindley (1987) has put it: It might be legislated that shopkeepers can sell their goods to members of some group in the community only at half price. If that is all that is legislated, however, it is likely to mean simply that members of the favoured group will have great difficulty in buying anything. Subsequently, three additional agreements have been signed which contain provisions for special treatment for developing countries. Firstly, the various Tokyo Round codes dealing with nontariff barriers include special provisions for developing countries. These cover antidumping, subsidies, technical standards, government procurement, customs valuation, import licensing and civil aviation. These extend GATT discipline to the particular areas in question and/or provide greater detail about how GATT discipline is to be applied in these areas. Secondly, in June 1971 it was decided through the GATT to grant developed countries a waiver from Article I (the nondiscrimination rule) to enable them to introduce a so-called Generalised System of Preferences (GSP). The GSP scheme had originated in the 1960s with the United Nations Conference on Trade and Development (UNCTAD). At its first conference in 1964, a group of less developed countries known as the Group of 77, moved a resolution calling for changes in the international economic order, including a lowering of tariffs on goods coming from developing countries. In 1968 at UNCTAD 2, agreement was reached on a scheme for granting tariff preferences on imports from developing countries. The 1971 waiver authorised the GSP programme initially for a period of ten years, provided that any preference granted to any one developing country was extended to all. The Developing Countries 135 Thirdly, in 1979, as part of the Tokyo Round, a so-called Enabling Clause was included in the Framework Agreements (entitled Differential and More Favourable Treatment, Reciprocity and Fuller Participation of Developing Countries). This provided a legal basis for extending the GSP beyond ten years without the need to secure a further waiver from Article I. In other words, it gave to the GSP scheme a permanent legal basis. Two aspects of the Enabling Clause are particularly important. Firstly, there was some reference to what has subsequently come to be known as a ‘graduation’ requirement. Developing countries are expected, as they grow and become able to do so, to participate more fully in the rights and obligations of the GATT, including making negotiated concessions in GATT rounds. Secondly, the Enabling Clause did not impose any legal obligation on GATT countries to extend such preferences. It merely made it legally possible for them to do so if they wished. Because such preferences were offered unilaterally and not as negotiated concessions, any developed country can at any time abandon or modify its GSP scheme without the need to provide compensation to the countries affected. The 1979 Enabling Clause also exempts developing countries from the requirements of Article XXIV, which deals with customs unions and free-trade areas. This means that developing countries may form preferential trading areas with each other which involve less than 100 per cent preferences. TARIFF PREFERENCES Following the GATT waiver of 1971, most of the developed countries introduced GSP schemes. Both the European Community and Japan did so in 1971 and the United States followed in 1976. The USA was the last country to do so, having been the main opponent of preferences in the 1960s. In most cases, they grant duty-free entry for all industrial products. However, this is nearly always qualified by provisions denying certain countries entitlement to preferential treatment, restricting the range of products covered and placing limits on the degree of preferential treatment permitted. For these reasons, the schemes fall a long way short of being a system of ‘generalised’ preferences as was originally envisaged. Firstly, the lists of developing countries entitled to preferences are far from being allembracing. The US excludes communist countries, countries participating in international commodity cartels such as OPEC (the Organisation of Petroleum-exporting Countries), countries expropriating US property without granting compensation, and countries refusing to co-operate in preventing narcotics entering the US. The EC scheme is formally more comprehensive in its coverage. It includes all countries which belong to the Group of 77 (some 125 countries) plus Romania and China (except textiles, where preferences were confined to countries belonging to the Multi-fibre Arrangement which had agreed bilateral voluntary export restraints with the EC). Some of these countries also enjoy special preference as members of the Lomé Convention which the EU has signed with some seventy African, Caribbean and Pacific (ACP) countries. Secondly, the preferences do not extend to all products exported by developing countries. Most countries exclude so-called sensitive products which are often the products of most interest to developing countries. (Sensitive products are usually taken to be those which, if imported, are likely to cause a relatively high employmentInternational Trade Policy 136 displacement effect often concentrated in a particular region and therefore giving rise to a serious adjustment problem.) The US scheme excludes certain import-sensitive products; namely, textiles and apparel articles subject to textile agreements, watches, importsensitive electronic articles, footwear articles and import-sensitive glass products. Any interested party may petition for articles to be removed or added to the list. The EC scheme includes most manufactured and semi-manufactured products but excludes many processed agricultural products. Thirdly, all schemes are qualified by various kinds of quantitative limitation. In order to prevent exports from countries not entitled to preferences from being diverted through qualifying countries, strict rules of origin are prescribed. The US scheme requires that either the product must be imported directly from the beneficiary country or that the value of materials produced in the beneficiary country plus direct costs of processing must exceed 35 per cent. The EC requires that either the product be wholly produced within the beneficiary country or that imported materials used have been subject to ‘substantial transformation’. This is defined as a transformation which brings them into a new four-digit heading of the Brussels Tariff Nomenclature (BTN). (Systems of trade classification aggregate goods at different levels. The one-digit level is the most aggregative. The degree of disaggregation increases with the number of digits such that the four-digit level represents quite a high degree of disaggregation.) The US scheme contains a ‘competitive need limitation’ under which a country may lose its duty-free treatment if its exports to the US exceed either 50 per cent of the total value of US imports of the product or a certain stated dollar value adjusted annually in accordance with the growth of US GNP. The EC scheme operates a system of individual tariff quotas for sensitive goods whereby any imports in excess of the quota become subject to the full MFN tariff. Processed agricultural goods are subject to a special safeguard clause which entitles the EC to reimpose tariffs if imports enter the EC in quantities or at prices which place EC producers at a serious disadvantage. Table 5.2 shows the value of imports of OECD countries which are both eligible for and receive preferential treatment. These indicate that in 1984 US$63,899 million, or roughly 50 per cent, of the MFN dutiable imports of the OECD countries were eligible for preferential treatment. However, because of the various kinds of product exclusion and limits on preferential treatment, only US$32,341 million, or 26 per cent, received preferential treatment. Roughly two-thirds of these imports were accounted for by the United States, the EC and Japan. Considerable variations existed between Table 5.2 Imports of preference-giving countries from beneficiaries of their schemes, 1984 Imports (US$ million) Shares (%) (1) (2) (3) (4) (5) (6) (7) (8) Market Total MFN-dutiableGSP-eligiblePreferential(4)/(3)(5)/(4)(5)/(3) Australia 4,881 2,797 1,689 1,689 60.4 100.0 60.4 Austria 2,178 1,854 1,732 320 93.4 18.5 17.3 Canada 6,980 2,914 1,728 1,295 59.3 74.9 44.4 EC 80,505 30,462 23,719 8,667 77.9 36.5 28.4 Finland 1,726 680 330 285 50.8 86.3 43.8 The Developing Countries 137 Japan 32,553 15,268 10,042 6,037 65.8 60.1 39.5 New Zealand 1,002 297 260 260 87.7 100.0 87.7 Norway 934 293 132 68 45.0 51.3 23.2 Sweden 1,954 772 390 266 50.5 68.4 34.5 Switzerland 2,947 2,855 1,277 453 44.7 35.5 15.9 United States 89,600 65,925 22,600 13,000 34.4 57.0 19.7 Total OECD 225, 259 124,087 63,899 32,341 51.5 50.6 26.1 Source: UNCTAD, quoted in Page and Davenport (1994) Note: Australian and New Zealand data are for fiscal year 1983–4. countries. At one extreme, 87 per cent of New Zealand’s MFN-dutiable imports received preferential treatment whereas, at the other, only 15.9 per cent of Switzerland’s imports were subject to preferences. In the case of some countries (notably, Australia and New Zealand), all eligible imports received preferential treatment. In other words, these last two schemes were unique in being truly generalised. In the case of other countries (notably, Austria, Switzerland and those of the EC), the ratio of GSP-eligible to GSPpreferential trade was much lower. In these cases, the true benefit was significantly reduced by exceptions, quantitative limitations, strict origin rules and safeguards. It is interesting to contrast the EC and US schemes. In the case of the EC, a large proportion of trade is eligible because few countries and few industrial products are excluded. It therefore appears more generous. However, quantitative limitations on preferences combined with stricter administrative rules and the exclusion of agricultural products mean that, in practice, it is much less generous than it appears. By way of contrast, the US scheme covers a smaller proportion of imports because more countries and products are excluded but there are no quantitative limitations on preferences. A system of tariff preferences has economic effects equivalent to those of a regional trading bloc. At the static level, they will give rise to both trade creation and trade diversion. (See pp. 237–8 for a fuller discussion of these concepts.) The reduction of tariffs on imports from beneficiary countries will cause imports from developing countries to displace some higher-cost domestic production in preference-granting countries. At the same time, tariff discrimination against imports from nonbeneficiaries will cause some higher-cost imports from beneficiaries to displace lower-cost imports from nonbeneficiaries. From the global economic point of view, tariff preferences will only increase economic welfare if the trade-creating effect exceeds the trade-diverting effect. On the other hand, from the point of view of developing countries receiving preferential treatment, it is the total trade effect which matters regardless of whether it is due to trade creation or trade diversion. At the time when the GSP was introduced, it was argued that the net effect would be mainly trade-creating. This was because the effective rate of tariff protection in developed countries was found to be significantly and positively correlated with the comparative advantage of developing countries. As a result, a comprehensive GSP would allow developing countries to expand those industries in which they enjoyed a comparative advantage (Iqubal, 1974). But this took no account of the various exceptions and limitations which were subsequently built into the preference schemes created by the developed countries. Various studies have been conducted to estimate these effects. Baldwin and Murray (1977) sought to estimate the impact of the GSP in the US, the EC and Japan using 1971 International Trade Policy 138 trade flows and theoretical preference margins. This was an ex ante study in which the effects of preferential tariff reductions were being estimated in advance. Their results showed that, for the United States, the expansion of trade amounted to nearly 30 per cent of 1971 trade flows, of which 80 per cent could be accounted for by trade creation. For the EC, the trade-creation effects were 20–25 per cent of 1971 trade flows. However, because the study assumed that all eligible imports would receive preferential treatment, the trade effects were overestimated. For the same reason, the study overestimated the trade-creation effect. At the same time, trade diversion was probably underestimated because it was assumed that elasticities of substitution between imports from GSP beneficiaries and nonbeneficiaries were the same as between imports from beneficiaries and domestic production (Langhammer and Sapir, 1987). Sapir and Lundberg (1984) used an econometric model to estimate the effect of tariff preferences on US trade flows for the period 1975–9 using both theoretical and actual preference margins. They found that trade-creation effects were about two and a half times larger than trade-diversion effects. Using a similar method to that of Baldwin and Murray, they estimated the tradecreation effect for all GSP-eligible products at US$2.2 billion using theoretical preference margins and US$1.3 billion using actual preference margins. The latter figure amounted to 21 per cent of GSP duty-free imports by the United States in 1979. Langhammer (1983) sought to estimate the trade effects of the EC scheme using changes in import-consumption ratios between 1972 and 1975. Paradoxically, he found that imports from nonbeneficiary countries increased in the EC in comparison with the US and Canada, suggesting negative trade diversion. Other studies have suggested that the impact of the EC GSP was less than that of the US. This may not be surprising given that the EC, unlike the US, had already entered into a series of trading agreements with a number of other trading partners. As a consequence, the actual preferential margins enjoyed by GSP beneficiaries were significantly lower than if all EC imports from nonbeneficiaries had been subject to an MFN tariff. Furthermore, the extent of EC preferences was reduced by the quantitative limits imposed on the amount which could be imported at the preferential rate. Karsenty and Laird (1986) estimated the trade effects of the GSP schemes of all OECD countries for 1983. Because imports of textiles and clothing were subject to restrictions under the MFA, the GSP was largely inoperative for this product group and hence it may be more appropriate to exclude these products from the estimates. With textiles and clothing excluded, Laird and Sapir (1987) estimated the total trade effect of the GSP in 1983 for all OECD countries at US$4.6 billion or 3.2 per cent of MFN dutiable imports. Most of this was attributable to trade creation. However, this compares with a potential gain of US$20.6 billion if all eligible imports were included without any quantitative limitation. All studies confirm that the benefits from the GSP scheme are heavily concentrated among a few developing countries. Langhammer and Sapir (1987) estimate that three countries—Taiwan, South Korea and Hong Kong—account for about two-thirds of the trade expansion effects of the GSP. Some ten developing countries share 90 per cent of the gain. In a similar manner, most of the trade expansion resulting from the EC scheme went to a comparatively small number of developing countries. In both cases, the major beneficiaries have been newly industrialising countries. The poorest, least developed countries are found to have benefited proportionately less. The Developing Countries 139 The results of these studies of the effects of the GSP may be summarised as follows. Firstly, the GSP has resulted in an expansion of developed-country imports from developing countries over and above what would otherwise have taken place. Secondly, this has largely taken the form of trade creation rather than trade diversion. However, the gain has not been very great largely because not all GSP-eligible imports are subject to preferential treatment. This is due to limited product coverage, limited country coverage and, in some cases, limitations on the extent of preferential treatment afforded. For these reasons, the trade effect is substantially below what would have resulted had the scheme been truly general. Fourthly, the gains are very unevenly distributed with the resultant trade expansion accruing to a small number of newly industrialising countries. The poorest LDCs derive very little benefit from the scheme largely because their exports consist of primary commodities rather than manufactures. These deficiencies have led to calls for a reform of the GSP. Proposals for reform have focused on (a) achieving uniformity between countries and across products to ensure simplicity, and (b) eliminating or reducing the amount of administrative discretion available to the importing country for denying GSP-eligible imports preferential treatment. With regard to (a), it is argued that inter-country differences in both the countries and the products granted preferences makes for complexity. With regard to (b), it is argued that the benefits of the scheme are reduced because exporters can never be sure that a consignment of goods eligible for preferential treatment will in fact qualify when they arrive in the importing country. The main effect of complexity and uncertainty is to discourage investment in manufacturing export industries in developing countries and thereby weaken the potentially positive effects of preferences on economic growth and development. An alternative view is that developing countries should not expend time and energy in seeking to extract improvements to tariff preferences from developed countries for two reasons. Firstly, improved tariff preferences may be less valuable to developing countries than securing reductions in MFN tariffs. Although nondiscriminatory (MFN) tariff cuts erode margins of preference, it is argued that the loss from smaller preferences would be offset by the gain from lower tariffs. The reason is that the benefit of the GSP resides largely in its tariff-cutting rather than in its preferential element. This is born out by the evidence that preferences have resulted in significantly more trade creation than trade diversion. In some cases (for example, the EU), there is some evidence for negative trade diversion. Although a truly generalised system of preferences could yield significant trade gains for developing countries, developed countries are unlikely to make sufficient concessions in this respect to make that possible. By way of contrast, MFN tariff cuts, depending on the formula adopted, would affect a larger volume of trade because they would not be subject to the various kinds of product exceptions, country exceptions and quantitative limitations which characterise the GSP. Furthermore, although as a result of the 1979 Enabling Clause the GSP now has a permanent legal basis under the GATT, preferences granted do not have the same legal status as negotiated MFN tariff reductions. Whereas the latter are normally bound against future increases, this is not the case with preferences, which are regarded as gifts or favours extended by developed to developing countries. This weakens their value since they can more easily be withdrawn (and indeed frequently have been) and because there is no obligation on the importing country to offer equivalent concessions in compensation. One of the weaknesses of preferences has been that they have encouraged developing countries to hold back from International Trade Policy 140 seeking MFN tariff reductions because these were seen as eroding the margin of preference granted under the GSP. The World Bank (1987) has criticised developing countries for seeking ‘a Faustian exchange’ in which they have ‘given up a voice in reciprocal trade negotiations’ in exchange for the granting of special and differential treatment. It is argued that the bargain was not worthwhile for the developing countries. Secondly, the potential benefit from preferences is in many cases negated by the proliferation of nontariff barriers. In Table 3.3 (p. 58), an estimated 19 per cent of the value of imports of developed countries from developing countries and an estimated 21 per cent of import categories were subject to nontariff barriers, in both cases higher than the equivalent for imports from developed countries. The proportion is significantly higher in specific sectors of special interest to developing countries, such as textiles, steel products and agricultural products. It is argued that securing reductions in nontariff barriers through negotiation would bring greater benefits to developing countries than seeking improvements in preferences. Of special importance to developing countries in this context are the various bilateral quotas which have regulated much trade between developed and developing countries in textiles and clothing. As was discussed in Chapter 3, trade in textiles and clothing products was granted a special dispensation from normal GATT rules and disciplines under the Long-term Cotton Textile Arrangement (1962) and later the Multi-fibre Arrangement (MFA). The effect has been to reduce significantly the value of any tariff cuts (preferential or nonpreferential) granted by developed countries on these products. This is discussed further below. However, before doing so, it is necessary to discuss an issue which has occupied a central place in the debate about preferences; namely, the graduation issue. GRADUATION When, in 1979, the developed countries conceded the principle that developing countries should be afforded on a permanent basis ‘differential and more favourable treatment’, a quid pro quo was that such treatment would be gradually withdrawn as a country developed. Put differently, a developing country would be expected to participate progressively in GATT rules and obligations as it graduated towards developed-country status. It was, however, left unclear how this principle was to be applied in practice. Subsequently, both the United States and the EC incorporated the graduation principle in their GSP provisions. The United States 1984 Tariff and Trade Act, which renewed the authority first granted in 1974 enabling the President to grant preferences to developing countries, introduced fresh provisions for graduation. Under the provisions, a country is withdrawn from GSP benefits if its per capita income exceeds a specified dollar value (fixed annually). The 1984 legislation also expanded the list of conditions which may disqualify a particular country from the GSP. For example, a country which fails to provide adequate protection of US intellectual property rights may be treated as ineligible. In 1987, Chile had GSP status withdrawn because it had failed to afford its workers ‘internationally recognised workers’ rights’. Similarly, products can be withdrawn from the GSP programme under various systems. The competitive-needs limitation contained in the 1974 Trade Act was referred to above, whereby a product can be withdrawn if exports exceed more than 50 per cent of The Developing Countries 141 lowering of existing barriers to ensure improved market access; on the other hand, a rules-based approach is also necessary in which countries agree to adhere by certain international principles designed to ensure fairness and efficiency in the global allocation of resources. GATT rules applying to trade in goods are based on the principle of unconditional most-favoured-nation (MFN) treatment which each contracting party is required to extend to every other. A key issue in both academic debate and, above all, in the actual services negotiations within the GATT has been whether this principle should be applied to trade in services. This would mean that each country undertook to treat both the traded services and the service-providers of other countries equally and not to discriminate against those coming from one particular country. Grey (1990) has urged caution in seeking to apply this principle to trade in services in the same way as to trade in goods. To begin with, as we have seen in earlier chapters, there have been so many exceptions to the unconditional MFN rule as applied to goods that we must doubt whether it can practically be applied to services. He concludes: With regard to the concept of unconditional MFN treatment, there would appear to be little scope for the application of such a clause, given the emerging concern for a measure of reciprocity, particularly as regards services delivered by establishments. On the other hand, he adds: ‘It would be a serious mistake, though, to set this key policy concept aside.’ Reciprocity makes nondiscriminatory treatment conditional upon the other country treating the traded services or service-providers of the country imposing the restrictions in an equivalent way: ‘I will treat your service firms in the same way as you treat mine.’ It is worth pointing out that the principle has a precedent within the GATT. It was adopted as the basis for the GATT Codes on Subsidies and Countervailing Duties and Government Procurement in the Tokyo Round. A country is required to extend the benefits contained in the two codes only to those countries adhering to the Agreement and can withdraw benefits from any country which ceases to do so. It has come to be adopted as a key principle by a number of countries and regions with respect to their service sectors. For example, the European Community had an important clause in its 1989 Second Banking Directive which threatened to withdraw certain benefits from foreign banks operating in the Single Market (for example, suspension or delays in authorisation) and coming from a country which denied national treatment to EC banks. Reciprocity is also widely practised in the civil aviation industry. For example, one country may reduce or withdraw landing rights from the carrier of another country which is deemed to be discriminating against the airline of the former in its route allocations. This was illustrated by the recent conflict between the US and Japan which began when Japan refused to allow United Airlines to extend its New York–Tokyo flights to Sydney (so-called ‘beyond rights’) which, the US alleged, violated the 1952 bilateral aviation agreement between the two countries. The US responded with threats to withdraw equivalent rights from Japanese airlines. Since countries with a fairly open policy towards their service industries (such as the US) are unlikely to be willing to grant International Trade Policy 244 [Document text truncated for crawler view.]