STUDIES OF ECONOMIES IN TRANSFORMATION - 15G o 4 AqCi I jq Foreign Trade in the Transition The International Environment and Domestic Policy Bartlomiej Kaminski, Zhen Kun Wang, and L. Alan Winters 20 THE WORLD BANK RECENT STUDIES OF ECONOMIES IN TRANSFORMATION No. 1 Country Department III, Europe and Central Asia Region, Food and Agricultural Policy Reforms in the Former USSR: An Agenda for the Transition No. 2 Michalopoulos and Tarr, Trade and Payments Arrangements for States of the Former USSR No. 3 Country Department III, Europe and Central Asia Region, Statistical Handbook: States of the Former USSR No. 4 Barr, Income Transfers and the Social Safety Net in Russia No. 5 Country Department III, Europe and Central Asia Region, Foreign Direct Investment in the States of the Former USSR No. 6 Wallich, Fiscal Decentralization: Intergovernmental Relations in Russia No. 7 Michalopoulos, Trade Issues in the New Independent States No. 8 The World Bank, Statistical Handbook 1993: States of the Former USSR No. 9 Holt, Transport Strategies for the Russian Federation No. 10 Fong, The Role of Women in Rebuilding the Russian Economy No. 11 de Melo and Ofer, Private Service Firms in a Transitional Economy: Findings of a Survey in St. Petersburg No. 12 Chu and Grais, Macroeconomic Consequences of Energy Supply Shocks in Ukraine No. 13 Michalopoulos and Tarr, Trade in the New Independent States No. 14 The World Bank, Statistical Handbook 1994: States of the Forner USSR No. 15 The World Bank, Russia : Creating Private Enterprises and Efficient Markets No. 16 Lieberman, Ewing, Mejstrick, Mukherjee, and Rahuja, Mass Privatization in Central and Eastern Europe and the Former Soviet Union: A Comparative Analysis No. 17 Le Houerou, Investment Policy in Russia No. 18 Belkindas and Ivanova, Foreign Trade Statistics in the USSR and Successor States No. 19 The World Bank, Statistical Handbook 1995: States of the Former USSR STUDIES OF ECONOMIES IN TRANSFORMATION Foreign Trade in the Transition The International Environment and Domestic Policy Bartlomiej Kaminski, Zhen Kun Wang, and L. Alan Winters 20 The World Bank Washington, D.C. Copyright © 1996 The International Bank for Reconstruction and Development/THE WORLD BANK 1818 H Street, N.W, Washington, D.C. 20433, U.S.A. All rights reserved Manufactured in the United States of America First printing April 1996 Papers in the Studies of Economies in Transformation series present the results of policy analysis and research on the states of the former USSR. The papers have been prepared by World Bank staff and consultants and issued by the World Bank's Europe and Central Asia Department III under the supervision of Constantine Michalopoulos. In light of the worldwide interest in the problems and prospects of these countries, dissemination of these findings is encouraged for discussion and comment. The findings, interpretations, and conclusions expressed in this paper are entirely those of the author(s) and should not be attributed in any manner to the World Bank, to its affiliated organizations, or to members of its Board of Executive Directors or the countries they represent. The World Bank does not guarantee the accuracy of the data included in this publication and accepts no responsibility whatsoever for any consequence of their use. The boundaries, colors, denominations, and other information shown on any map in this volume do not imply on the part of the World Bank Group any judgment on the legal status of any territory or the endorsement or accep- tance of such boundaries. The material in this publication is copyrighted. Requests for permission to reproduce portions of it should be sent to the Office of the Publisher at the address shown in the copyright notice above. The World Bank encour- ages dissemination of its work and will normally give permission promptly and, when the reproduction is for non- commercial purposes, without asking a fee. Permission to copy portions for classroom use is granted through the Copyright Clearance Center, Inc., Suite 910, 222 Rosewood Drive, Danvers, Massachusetts 01923, U.S.A. The complete backlist of publications from the World Bank is shown in the annual Index of Publications, which contains an alphabetical title list (with full ordering information) and indexes of subjects, authors, and countries and regions. The latest edition is available free of charge from the Distribution Unit, Office of the Publisher, The World Bank, 1818 H Street, N.W., Washington, D.C. 20433, U.S.A., or from Publications, The World Bank, 66, avenue d'1ena, 75116 Paris, France. Bartlomiej Kaminski is Director of the Center for the Study of Post-Communist Societies at the University of Maryland and a consultant to the International Trade Division in the International Economics Department of the World Bank. Zhen Kun Wang is a consultant working on the World Bank's World Development Report and was for- merly a consultant to the Transition Economics Division in the Policy Research Department of the World Bank. L. Alan Winters is Division Chief of the Intemational Trade Division in the Intemational Economics Department of the World Bank. ISSN: 1014-997X Library of Congress Cataloging-in-Publication Data Kaminski, Bartlomiej, 1944- Foreign trade in the transition: the international environment and domestic policy / Bartlomiej Kaminski, Zhen Kun Wang, and L. Alan Winters. p. cm - (Studies of economies in transformation, ISSN 1014-997X; 20) Includes bibliographical references. (p. ). ISBN 0-8213-3611-8 1. Europe, Eastem-Commercial policy. 2. Europe, Eastern- Economic policy-1989- 3. Former Soviet republics-Commercial policy. 4. Former Soviet republics-Economic policy. 5. Intemational trade. 1. Wang, Zhen Kun, 1963- . 11. Winters, L. Alan. III. Title IV. Series: Studies of economies in transformation ; paper no. 20. HF1532.7.K36 1996 382' .3' 0947-dc2O 96-14513 CIP Contents Foreword v Preface vii 1. Introduction 1 2. The Collapse of Central Planning and the CMEA 5 The Trade Legacy 6 Foreign Trade Under Central Planning 6 The Unraveling of the CMEA 8 The TradeAdjustment Problem Facing the NIS 9 The Trade Potential of Transition Economics: Inherited Distortions 10 3. Shifting Patterns of Foreign Trade 15 Export Performance in OECD Markets in the 1980s 16 Shifting Patterns, 1988-94 17 An Index of Export Performance 22 Changes in Imports 25 iv FOREIGN TRADE IN THE TRANsrnON 4. Changes in Market Access 27 Incomplete Normalization of Relations 28 Preferential Market Access 31 The Significance of Improved Market Access 32 5. Policy and Trade Performance 37 Exogenous Factors Accounting for Differences in Trade Reorientation 37 Trade Policy Reform 38 Fast Versus Slow Reformers 39 Use of Export Controls 41 Changes in Domestic Institutional and Policy Settings 42 Settings of Good and Poor Trade Performners 42 6. The Impact of Transition on Developing Countries' Trde 47 Trade Displacement in EU markets 48 Primary Commodities Markets 50 7. Condusion 55 Notes 57 References 59 Foreword Despite a common legacy of central planning, transi- tion economies have differed widely in their foreign trade performance. The dissolution of the Council for Mutual Economic Assistance and the demise of the Soviet Union required them to reorient their trade suddenly and dramat- ically, and the extent to which they succeeded in doing so has varied considerably. This study provides strong evi- dence that the vigorous implementation of economic reforms aimed at macroeconomic stability and openness to international trade is the single most important variable explaining differences in the pace of trade reorientation. The countries which made the strongest progress in mov- ing to a market-based economic system have been precise- ly those which have most successfully reoriented their for- eign trade in line with comparative advantage and eco- nomic opportunities. The link between domestic policy and export perfor- mance has been long suspected, but has not previously been carefully documented. Given that the World Bank has supported and advised countries in their efforts at transformation and that it has consistently argued the vi FOREIGN TRADE IN THE TRANsmoN case for an open trade regime, it is appropriate on a close collaboration with staff in the that this evidence be widely disseminated. The Europe and Central Asia Regional Vice- research reported here was undertaken by Presidency. In view of its relevance to the latter's economists in the Bank's Development operational work, I am pleased to indclude it in Economics Vice-Presidency, but it was based this series. Johannes Linn Vice President Europe and Central Asia Preface This paper started as one of a series of overview articles initiated by the Transition Economics Division of the World Bank's Policy Research Department to review and assess the growing body of evidence on the determinants of a successful transition. It was intended to survey the inter- action between international trade and trade policy on the one hand and progress in transition on the other. It had long been recognized that sound economic policy played a crucial role in both development and transition, but the extent to which this factor appeared to dominate the deter- mination of trade performance was a surprise. As this pic- ture began to emerge from the survey, the objective of this paper evolved beyond a review of the literature into a more substantial research exercise designed to test the proposi- tion and document more systematically the evidence behind it. The paper considers three broad sets of determinants of the speed with which transitional economies in Europe and Central Asia have been able to normalize their trading per- formance and boost their exports: their initial, pre-transi- tion conditions, the changes in their access to western mar- kets, and their policy stances. The first two factors clearly have some influence over trade performance but do not ViiI FOREIGN TRADE IN THE TRANsrnON seem to distinguish strong from weak export International Trade Division of the International performers: there are weak performers who had Economics Department. It has benefited from relatively good initial conditions and market the broad-based discussion of several drafts and access, and strong performers who were less well the authors are particularly grateful to Stijn endowed in these respects. Policy, on the other Claessens, Alan Gelb, Martha de Melo, Costas hand, appears to offer almost perfect discrimina- Michalopoulos, Martin Schrenk, David Tarr and tory power between classes of export perfor- Ulrich Zachau for their insightful comments. mance, with sound stabilization and thorough- In addition they are grateful to Costas going domestic price liberalization characteriz- Michalopoulos for guidance in preparing the ing almost exactly the same set of transitional paper for 'Studies in Economies in countries as strong export performance. Transformation," to Minerva Patefia for secretar- The research on this paper has been sup- ial support, to Jeff Hayden and Alyse Zable for ported by the Transition Economics Division of editing the paper and to Jeff Hayden for oversee- the Policy Research Department, and the ing its publication. 1 Introduction Contrary to expectations, several formerly centrally planned economies have exhibited strong export perfor- mance in OECD markets. According to the conventional wisdom held prior to the collapse of central planning, their capacity to compete in "real" markets was limited by their legacy of misallocated resources and low-quality products. This generalization was refuted by the experience of several economies, although, at the same time, the performance of others strongly supported it. Our challenge is to explain why some countries have been successful at integrating into the world economy and reviving their economies while oth- ers have not. This study focuses almost exdusively on the economics that entered transition because of the sudden collapse of communism: the European and Central Asian (ECA) members of the former Council for Mutual Economic Assistance (CMEA).1 China is rather different, for its cen- tral planning did not suddenly collapse and it was not a member of the CMEA; thus it will be included in our analyses only occasionally and as a comparator to other transition economies. Despite sharing a legacy of central planning, there have been significant differences among our sample countries in 2 FOREIGN TRADE IN THE TRANSITION foreign trade performance, both before and sub- trade regimes, macroeconomic policies and price sequent to transition. All of the economies now policies. Almost all of the good export perform- in transition suffered the collapse of their tradi- ers had made important strides in macroeco- tional trade links within the CMEA or within nomic stabilization, in establishing market sup- the Soviet Union. But the extent to which they porting institutions, in dismantling administra- have succeeded in reorienting their trade has var- tive price controls, and in unifying exchange ied widely. For example, former Czechoslovakia rates. increased its trade with OECD countries at an The countries which showed poor export average annual rate of about 20 percent over performance, however, had not yet corrected 1988-94; that of Bulgaria and Romania grew at a macroeconomic fundamentals or fully liberal- modest 2 percent per year in the same period. ized domestic prices. Although they had liberal In searching for the factors underlying such import regimes, domestic producers and con- differences in export performance, the obvious sumers were insulated from international mar- place to start is countries' initial (pre-transition) kets by extensive export controls-designed to conditions. These conditions are described in maintain subsidized products at home-and the Section 2, which examines policy-related lega- high level of protection offered by undervalued cies as well as inherited distortions in foreign domestic currencies. These slow and vacillating trade patterns. Section 3 compares countries' reformers could not take advantage of improved export performance in OECD markets in 1980- access to OECD markets. 87 and in 1988-94. Using an index of export One caveat needs to be made clear: our performance designed to account for differences main analysis was based on information avail- in initial conditions, these countries are grouped able in mid-1995. Thus our categorization of in terms of their export performance. Section 4 "good" and "bad" performance and policy may describes how access to OECD markets changed no longer be correct. But the analytical conclu- following the collapse of central planning. sion about the link between macroeconomic Section 5 looks at the links between trade per- policies and foreign trade performance is time- formance on the one hand and foreign trade independent and seems, if anything, to be more regimes and domestic institutional and policy strongly supported by recent developments. settings on the other, and then compares the We find little support for the often-voiced influence of domestic policies and institutions concerns about disruptions in developing coun- with that of market access. Section 6 briefly dis- tries trade caused by the emergence of new cusses the impact of transition on trade in devel- players of transitional economies in internation- oping countries. al markets. The redirection of Russian primary The study offers strong empirical evidence commodities from the former CMEA markets that the pace of implementing a liberal econom- to OECD markets has contributed little to ic reform, including government commitment depressing world prices in 1990-92. The impact to macroeconomic stability and openness to of transition economies in manufactured trade international trade, is the single most important has been much smaller than that associated variable explaining differences in the pace of with the resurgence of China. This small impact trade reorientation. Good and poor trade per- is a result of two factors. First, aggregate exports formers are set apart not only by export perfor- from the ECA transition economies have mance but also by fundamental differences in remained flat as good and poor performers have their policy environments, including foreign offset each other. Second, the falls in the market INTIRoDUC1ON 3 shares of developing countries that have 'grad- Kong) left plenty of room for transition uated" from exporting simple manufactures economies without other suppliers suffering (such as the Republic of Korea and Hong absolutely. 2 The Collapse of Central Planning and the CMEA Economies in transition are heterogeneous both in terms of their economic characteristics and the challenges they have faced. In this paper we distinguish three broad groups of countries: the new independent states (NIS),2 which sprang from the dissolution of the Soviet Union, the Central and East European Countries (CEECs)3, and China. The NIS were part of the Soviet "common eco- nomic" space, and their trade with the "outside" world was minimal. The CEECs were less dependent on CMEA I The Trade Legacy trade, and started their transition earlier. China's retreat * The Trade Potential of from central planning has been gradual; it was poorer and Transition Economies: was not a heavy trader with the CMEA. Thus China was Inherited Distortions spared most of the shocks that the demise of the CMEA inflicted on its former members. Even within the broad groups the conditions prevailing immediately after the collapse of communism varied widely across transition economies, a variance that had implica- tions for their ability to take advantage of opportunities offered by international markets. Several of the CEECs had partially dismantled the state monopoly over foreign trade before the collapse of communism. Although the foreign trade regime remained a source of distortions and ineffi- ciencies, this early decentralization may have made the 6 FOREIGN TRADE IN THE TRANSmON eventual adjustment to market forces easier 1980s-even as the European transition because firms had made direct commercial con- economies' performance in international mar- tacts with western traders and had experience of kets deteriorated. This tension drove several gov- international marketing. The NIS had none of ernments to experiment with reforming their these benefits, and indeed, their foreign trade foreign trade regimes in the 1980s. The general regimes still display many of the features seen in approach to reform induded linking domestic the CEECs in the late 1980s. Thus the latter prices to international prices; establishing direct offer a useful benchmark for assessing the for- links between enterprises and international mar- mer's progress. kets; bypassing foreign trade organizations; establishing a larger number of intermediaries The Trade Legacy with less-restricted trading profiles; introducing The two historical factors that have most shaped currency auctions; and reducing the number of European transition economies' trade-central exchange rates and devaluing them to more real- planning and the economies' involvement in the istic levels. CMEA-are intertwined. Their effects were felt Before the collapse of communism, not only directly in terms of the pattern and vol- Hungary and Poland had made the most ume of trade with western markets, but also progress among CMEA members in decentraliz- indirecdy in the way in which they curtailed the ing their foreign trade regimes. Both countries development of the physical and institutional were highly indebted to the West and were the infrastructure necessary for such trade. The first to orient their economies away from the peculiarities of the CMEA as a regional econom- CMEA (Hillman and Schnytzer 1992:253). ic grouping can only be explained in terms of its With the decline of Soviet capability to sustain members' adherence to bureaucratic mecha- intra-CMEA trade in the late 1980s, other nisms of coordinating and stimulating economic countries also began to undertake foreign trade activity. reforms, but they were less comprehensive than in Hungary and Poland. Foreign Trade Under Central Planning The reforms in Hungary and Poland shared In a "dassical" planned economy the state, oper- two distinct features. First, they granted limited ating through foreign trade organizations, financial autonomy to state-owned enterprises, retained monopoly power over foreign trade allowing them to conduct their own foreign activity. Shortages, administrative rationing, and trade. Second, they created incentives for state- foreign trade price equalization schemes insulat- owned enterprises to expand non-CMEA ed producers from international markets. Firms exports through hard currency retention faced competition neither at home nor abroad schemes and exchange rate policies. In 1986 the and remained largely indifferent to prices in Hungarian government adopted "parallel" trade both markets. But because full autarchy would licenses, which were made available to all firms, have been economically damaging, even for a not just dominant exporters and importers, and bloc as large as the CMEA, links to internation- covered most products. As a result, the number al markets were not completely cut. In fact, with of firms operating in international markets had the growing technological gap relative to OECD increased dramatically by the end of the countries and growing consumer pressures, 1980s-see Mizsei (1991:15-20). In Poland sig- international markets became more important nificant steps to dismantle the state monopoly to their economic wefare during rhe 1970s and over foreign trade were undertaken in the early THE COLLAPSE OF CENTRAL PLANNING AND THE CMEA 7 1980s, when the authorities liberalized condi- Moreover, despite the gradual disintegration tions to obtain foreign trade licenses. Between of the central distribution system and erosion of 1982 and 1985 the number of state-owned the state monopoly over foreign trade during enterprises empowered to conduct their own 1988-89, the involvement of local (republic) foreign trade operations increased from 109 to governments in foreign trade remained minimal. 361 in Poland. The real exchange rate depreciat- As the Soviet Union was falling apart, these gov- ed by 30 percent over the same period (Roe and ernments tried to control interrepublic trade Roy 1989:6), and this started to affect firms' through a web of bilateral agreements, and sev- export performances once they were allowed to eral official proposals on republic self-finance retain a portion of their foreign exchange earn- kept foreign economic relations firmly in the ings (Tarr 1990; Winiecki 1991). purview of the center (Hewett and Gaddy Other CMEA countries, except for 1992:110). But their levels of foreign trade were Romania, also began to change foreign trade tiny and they had few official instruments to policy, but few of the changes were effective. control it. Thus when Moscow surrendered the Bulgaria, Czechoslovakia, and the Soviet Union right to trade on behalf of the whole Soviet introduced foreign currency auctions, and cur- Union, there was almost no decentralized expe- rencies were devalued by 19 percent in rience to replace it. Czechoslovakia and by a factor of 12 in Bulgaria In short, no matter how radical the reform (Nuti 1991:50). Bulgarian exporters were during the 1980s, foreign trade regimes under allowed to retain 60 percent of their export earn- central planning remained a source of distor- ings. But these measures had less impact on for- tions and inefficiencies throughout the whole eign trade than the measures taken by Hungary CMEA region, isolating domestic producers and Poland because the state-owned enterprises from changes in relative prices in world markets. in these countries still functioned as administra- While the reforms allowed the development of tive units of the state. some marketing expertise in some enterprises Institutional arrangements shielded Soviet and provided incentives to boost exports, they exporters from the world economy to a much failed to introduce the critical discipline of the larger degree than did those in the CEECs. market-market clearing at single prices, free of Although after 1987 some larger state-owned ex post and ad hoc subsidies and levies, and enterprises were allowed to retain a small portion tightly linked to profits or losses, which were of their foreign currency earnings, trading of solely the responsibility of the enterprise (Nuti retained foreign exchange was proscribed until 1991:50). That is, they did not make foreign late 1989. The monopoly over foreign trade was trade an effective conveyor of international effi- effectively abolished only in late 1991, with a ciency standards. In order to achieve this end, decree drastically simplifying licensing of foreign countries would have had to remove the anti- trade activities. When the Soviet Union dissolved export biases (import restrictions, administra- two months later, firms had had little time to tive allocation of raw materials and foreign develop independent contacts with foreign sup- exchange, and price controls) and the soft bud- pliers and buyers. Enterprises had little or no get constraints that cushioned inefficient pro- experience in international markets and the lack ducers. But these changes could not be made of expertise, especially in the smaller states locat- without abandoning central planning: effective ed far from the main world markets, was a seri- foreign trade reforms first required a reformed ous impediment to export performance. economy. 8 FOREIGN TRADE IN THE TRANSITION In some respects the approach taken by emerged from bilateral negotiations and were Chinese reformers has followed a logic similar to often different across countries. Finally, it was that in the CMEA economies. In China, the not a payment union because national curren- state monopoly over foreign trade was loosened cies were excluded from CMEA transactions. by bestowing greater autonomy on the branch The standard box of tools for analyzing the trade offices of foreign trade organizations. Line min- impact of regional integration is thus of litde rel- istries and lower levels of general governments evance for analyzing the CMEA. were allowed to establish their own foreign trade The nature of the CMEA-at least until the organizations; foreign exchange retention rights early 1980s-can best be captured in terms of were granted, import-licensing authority was the political and economic dominance of the decentralized, and domestic currency was signif- Soviet Union, based on its ample endowment of icantly devalued (Panagariya 1991). In other natural resources and system of central planning. respects, however, China's approach has This dominance gave rise to bilateral ties diverged. Aided by its much lower level of between the Soviet Union and the other CMEA industrialization than in other transition members, creating a radial pattern of economic economies (Sachs 1995), China has laid greater relations within the bloc. Over time the links emphasis on decentralization, competition, con- between the CMEA states and the Soviet Union servative macroeconomic policies, and a credible intensified, while those with the rest of the government commitment to opening the econo- world atrophied. my. These factors, combined with removing This form of organization resembled a some coastal regions from the yoke of central North-South trading relationship; the South- planning and establishing an investment regime the Soviet Union-specialized in exports of raw friendly to foreign investors, explain China's materials and, having superior technologies in superb performance in foreign trade.4 Exports military production, had a dominant political from the coastal provinces, home to 40 percent position, while the remaining members (mainly of China's population, accounted for 80 percent the CEECs) adjusted their productive structures of total exports in 1990 (Dadush and Hue to the needs of the Soviet Union by specializing 1995), and Special Economic Zones and Open in manufactures, usually of low quality. The Cities contributed 40 percent to total exports CEECs' CMEA-oriented export baskets were that year (Panagariya 1995). Thus it is arguable distinctly different from their non-CMEA ones: that the key to China's export success has been a in the former, machinery and other manufac- selective abandonment of central planning tures dominated while in the latter energy, raw rather than its refining. materials, agricultural goods, and material inten- sive products (steel, petrochemicals, fertilizers) The Unraveling of the CMEA constituted the bulk of exports. The CMEA The CMEA-serving as the center for coor- relieved central planners from international dinating plans, launching joint investment pro- competitive pressures and provided mutual assis- jects, and negotiating bilateral trade agree- tance in sustaining inefficiencies. Eastern ments-is not easily definable in economic Europe had never been at the forefront of tech- terms. Although it was a discriminatory trading nological development and central planning fur- arrangement, it was not a customs union ther inhibited the development of new tech- because tariffs were irrelevant to trade decisions. nologies,5 while "socialist industrialization" Nor was it a common market, as relative prices completely disregarded energy and raw material THE COLLAPSE OF CENTRAL PLANNING AND THE CMEA 9 intensity. The Soviet connection made it possi- CMEA actually represented a weaker shock to ble for the centrally planned economies to sur- the CEECs than might at first be imagined. vive, however, because consumer demand was suppressed and the Soviet Union had energy and The Trade Adjustment Problem Facing the NIS raw materials available in abundance. The effect of the disintegration of the "common In the 1980s these arrangements began to economic space" on the NIS was much larger unravel because of the growing indebtedness of than that of the breakdown of the CMEA on most of the CEECs and the stagnation of Soviet CEECs. Although the CEECs were "faithful" raw material and energy supplies. During the members of the Soviet bloc, they had consider- 1982-84 credit squeeze that followed the impo- able discretion in their foreign economic policy sition of martial law in Poland, the CMEA and maintained some level of autonomy in offered no "mutual assistance." Faced with domestic economic policies. They were not part declines in domestic oil output and in oil prices of the same command system, as Moscow had in world markets, the Soviet Union chose to not succeeded in imposing supranational plan- expand exports to the West in order to maintain ning on the CMEA. Nor were they part of a the hard currency revenues it badly needed to unified budget system that redistributed income revive its ailing economy; it began to pressure among republics of the Soviet Union. As a result other CMEA members to dismantle the pay- their economic structures were less distorted ments mechanism based on artificial prices and than those in the NIS, and there was at least a the transferable ruble and to expand their own weak link between CMEA and world prices. exports to the west. The Soviet move to world In the Soviet Union, by contrast, prices had prices and convertible currencies (January 1, no influence over the distribution of resources 1991) removed any rationale for the existence of among republics because distribution was deter- the CMEA, and it was officially dissolved at its mined by non-price-related criteria. Soviet cen- forty-sixth general meeting on June 28, 1991. tral planners fostered specialization rather than The cement that had bound the economies of diversification within each republic and encour- the CMEA together was the exchange of Soviet aged interdependence by locating different raw materials for Eastern European manufac- stages of production in different regions. This tures. The organization collapsed because the arrangement produced very high levels of inter- Soviet Union could no longer afford to sustain republic exchanges. Moreover, central planners this exchange, and the CEECs sought to liberate also established giant industrial agglomerations themselves from economic links inherited from that enjoyed full or almost full monopoly privi- the past. lege in the production of particular products For political economy reasons the CMEA (Schroeder 1993). It is estimated that between was not successful in integrating its member 30 and 40 percent of total industrial output in economies (Kaminski 1989). As a result, trade the USSR was in products for which there was flows were less diverted and productive struc- only a single manufacturer and another third in tures less distorted than they would have been products with only two manufacturers (Kroll had supranational planning been in force. The 1991). Under these circumstances, even though major sources of distortion in members' trade the notoriously low reliability of supplies of pro- patterns were their own central planning and duction inputs had prompted some state-owned easy access to the Soviet raw materials, which firms to substitute their own products for was beginning to erode. Thus the demise of the bought-in inputs (Winiecki and Winiecki 10 FOREIGN TRADE IN THE TRANSITION 1992:15-18), disruptions to former interrepub- Association (EFTA), other OECD countries, lic trade links had very serious consequences. developing countries, CMEA, and other Soviet An important barrier to the foreign trade of republics-and compared them with actual the NIS is the lack of adequate banking, trans- exports from the mid-1980s. The ratios of actu- portation, and telecommunication infrastructure, al trade to potential trade illustrate the extent to as well as the relative remoteness of important which trade flows differed from levels deter- markets. Unlike Visegrad countries (the Czech mined by economic factors, including trans- Republic, Slovakia, Poland, and Hungary), none portation costs, and thus the potential for of the NIS has a common border with a highly expansion once economic factors displace industrialized economy.6 The transportation and noneconomic considerations as determinants of communication infrastructure was developed to trade. The inherited trade distortions were large serve the division of labor within the common and once they are removed, trade, especially (autarkic) economic space of the Soviet Union. with OECD countries, can expand rapidly. Thus transportation, organized mainly around The gravity model explains the size of bilat- railways, was focused on Moscow, while road eral trade flows by a combination of supply and transport, used mainly to feed the railways, demand factors-proxied by population and accounted for only a small portion of total GDP-and other stimulating or restraining fac- freight movement. Transportation problems are tors affecting specific trade flows. We used it to particularly significant for land-locked Asian estimate the determinants of bilateral trade NIS, and the limited capacity of pipelines and among seventy-six market economies in 1984- ports has constrained exports from Kazakhstan, 86, and then applied the resulting model to the Russia, and Turkmenistan. transition economies to estimate what their Compounding these difficulties as they trade flows would have been had they been mar- entered the transition, the NIS' supplies of key ket economies in that period.7 We then assume traditional exportables, such as oil, had been that the ratio between actual and potential trade falling for some time before the collapse of cen- measures the extent of preexisting distortions tral planning and these countries did not have and, by extension, the scope for increasing trade the institutional capacity to expand nontradi- over the transition. tional exports. They were also hit by gradually The gravity model's predictions, which pro- liberalizing energy prices, which created a major vide a long-term equilibrium view of the volume terms-of-trade shock for energy importers (Tarr and direction of trade, depend heavily on the 1994). For all these reasons the NIS were ill- estimates of GDP used in projecting the poten- equipped to expand trade with the rest of the tial trade flows. The estimates of GDP in for- world when they were set free in 1992. merly centrally planned economies in the mid- 1980s differ by factors of up to five among vari- The Trade Potential of Transition Eonomies: ous sources. Since an increase of 1 percent in a Inherited Distortions country's estimated GDP augments its predicted What was the overall effect of these legacies? In exports by 1.2 percent and its imports by 1.0 order to assess the extent of inherited distortions percent, it is evident that large margins of uncer- in geographical patterns of foreign trade, we tainty attend these predictions. have used estimates derived from the gravity The interpretation of the gap between actu- model of potential exports to selected markets- al and predicted trade as the likely change in the European Union (EU), European Free Trade trade volume over the transition also hinges crit- THE COLLAPSE OF CENTRAL PLANNING AND THE CMEA 11 ically on the accuracy of existing trade statistics. The results of the estimation of the trade Data on intra-CMEA trade were distorted by potential of transition economies can be sum- the use of arbitrary exchange rates for transfer- marized: CEECs' trade with CMEA partners able rubles (TR). Hungary used the conversion was slightly above "normal" levels, whereas that rate of TR 1.94 per US dollar in 1985; Poland, with market economies, especially OECD coun- TR 2.16; Bulgaria, TR 0.64; and Czechoslovakia, tries, was far below its potential; the republics of TR 1.38. As a result the same volume of CMEA the USSR overtraded with each other and trade yields different dollar equivalents. For undertraded with outside partners, both on a example, Bulgaria appears to have nearly twice massive scale; and China "overtraded" with the total trade of Hungary, although their developing countries, especially in Asia, but economies are of roughly equal size in terms of "undertraded" with industrial countries in the 1985 GDP More formidable are the data prob- mid- I 980s. The results are broadly compatiblc lems for the former Soviet republics. In addition with those obtained in other studies using gravi- to the conversion problem, which contributed ty analysis to assess the long-term potential for to the overstatement of both interrepublic trade shifts in trade patterns (see box 1). and trade with some developing countries based The CEECs' trade patterns with non- on special payments arrangements, statistics on CMEA (or non-Soviet) partners reveal the lega- extrarepublic trade in the NIS are sketchy, cover cy of past commercial and political linkages a short time span, and remain subject to consid- (table 1). Trade is least restricted relative to its erable error. The earliest comprehensive data on potential with developing countries, followed by the direction of the republics' trade are for 1990. EFTA and the EU, and then other industrial With all these problems, the results described countries. The CEECs' mutual trade was about below can at best give only a general indication one-third higher than that justified by economic of the possible changes in trade patterns. considerations alone. The shortfall for the Table 1: Distortions in Geographical Patterns of Exports, mid-1980s (realization ratios in percent) To Other Total Inter- European OECD OECD Developing repubik From Union EFTA markets markets Countres CEECs China trade Baltic Rep. 9 12 3 8 59 34 36 551 European Rep. 52 31 11 36 249 178 95 666 Caucasian Rep. 14 18 2 9 41 59 25 768 Central Asian Rep. 32 21 8 20 61 147 35 894 CEECs 20 31 10 16 95 133 n/a n/a China 66 66 48 52 1187 n/a n/a n/a Note: n/a. Not available. Values less than 100 suggest 'undertradec and those more than 100 suggest 'overtrade." Predicted and actual data refer to 1985 for the CEECs and China, while for the republics predicted data are for 1985 and actual data for 1990. Realization ratios = actual exports/predicted exports. Source: Authors' own calculations. Data from World Bank (1993); IMF (1989); and World Bank staff estimates. 12 FOREIGN TRADE IN THE TRANSMON Box 1: Review of Gravity Studies of the Trade Potential of European Transition Economies The disappearance of systemic barrers hobbling East-West trade before the demise of central planning in 1989-91 triggered interest in assessing their impact on future trade pattems. Collins and Rodrk (1991) conduct an exercise similar to ours for Eastem Europe. They fit an openness relationship across ninety- one countries, regressing the export-to-GNP ratio on GNP, log(GNP), log(population) and a series of dummies, and then apply them to the estimates of GNP for the European transition economies as given by the PlanEcon (a consultancy group). Their predicted trade levels are very similar to ours, although- because their results refer to 1988 following a substantial dollar devaluation-the European transition economies share in world trade is lower than in our estimate: 10 percent compared with 18 percent. The more marked contrast between our results and those of Collins and Rodrik lies in the geo- graphical distribution of trade. Their model implied considerable similarities between the predicted pat- terns of the various Eastern European countries and, because of the choice of their benchmark-six Westem European countries-predicted a greater concentration of Eastern trade with the EU. (For more detail of these comparisons, see Winters and Wang 1994.) Havrylyshyn and Pritchett (1991) also estimate the gravity model of trade and use it to simulate post-transition pattems of trade in Eastem Europe. Their results show that intra-CMEA trade was exces- sive and that it would decrease in the future. They too emphasize the change of geographic direction of Eastern European trade and predict large shifts toward the West. However, like Collins and Rodrik (1991) they find the increase much more concentrated on Westem Europe than we do; they suggest, in fact, a fall in the share of trade with North America. This result probably stems from their inclusion of land area in their model-it affects trade negatively-and also from their greater elasticity on distance (around -1.5 compared with our -0.75). Rosati (1992) uses the gravity model to estimate bilateral trade flows of seventeen West European countries (all EU countries, all EFTA countries except Iceland and Yugoslavia) and then to compute hypothetical trade flows among former CMEA members. He compared his estimated results with CMEA trade data for 1989-90 and assumed that the switch to a market regime initiated in 1989-90 would force these countries into a similar pattern as the one existing in Westem Europe. His results show that intra- CMEA trade flows in 1989-90 exceeded corresponding theoretical values, with intra-CMEA trade link especially strong in case of the Soviet Union, Czechoslovakia, and Bulgaria. Rosati's study includes only industrial Westem European countries. Their bilateral trade flows are strongly influenced by intraindustry trade and policy preferences. CEECs' trade with other industrial countries, of republics was internal trade. They undertrade which the United States and Japan are the prin- with the outside world; trade of Russia with cipal members, was by a factor of 10. OECD countries is much less suppressed than The ratios of actual to potential trade flows that of other republics. Differences notwith- for the former Soviet republics, aggregated into standing, the potential for increasing trade with four geographic regions, show both similarities market economies is immense: in OECD mar- and differences. All overtrade with each other, kets, the smallest ratio of actual to potential reflecting not only the anti-external trade bias of exports was for the Baltic states (8 percent) and the Soviet Union under central planning but the Transcaucasian NIS (9 percent), followed by also the fact that trade between the former Central Asian NIS (20 percent) and the THE COLLAPSE OF CENTRAL PLANNING AND THE CMEA 13 European NIS (36 percent).8 These ratios sug- Turning to table 2, which deals with indi- gest that, over the medium term, exports from vidual republics, the biggest market predicted by Baltic and Caucasian states might increase more the gravity model for the NIS' exports is than tenfold, from Central Asia by a factor of Western Europe, especially for those NIS locat- five, and from the European NIS by a factor of ed in Europe, such as Estonia, Latvia, Moldova, about three. All groups undertraded more with and Russia. Their predicted export shares to "other industrial countries" than with Western Western Europe account for more than 40 per- Europe. Central Asian and European NIS trad- cent of total exports. The importance of trade ed relatively more with the EU than with EFTA, with the CEECs varies with geographical loca- while the other two regions had trade patterns tion. The share of the CEECs in total exports of similar to that of the CEECs. Kazakhstan, Kyrgyz Republic, and Tajikistan The NIS' trade potential with market hovers at around 6 percent, while for Ukraine it economies seems likely to be realized by shifting is 15 percent. Further, contrary to expectations, from interstate to third country trade, rather than predicted exports to China are not large. This by increasing the current level of total trade. The prediction may be explained by China's low openness ratios-(exports + imports)/GDP-of income level used in the prediction exercise; its all the republics were very high, well above those large population, which suppresses its trade; and for other countries of a similar size and GDP. the long distances between the buoyant com- They ranged from 52 percent for Lithuania to 19 mercial centers of China (such as Guangdong) percent for Russia in 1989, with most republics and those of most republics.9 Finally, the share around 35 to 40 percent. of interstate trade will inevitably fall once trade Table 2: The Geographical Composition of NIS' Market-Driven Exports (percent) Actual Predicted share of share of Other interrepublic interstate EFTA Other Total developing exporls exports and EU OECD OECD CEECs China countries Armenia 97 27 35 20 55 8 1 9 Azerbeijan 94 24 35 22 57 8 1 9 Belarus 90 32 37 12 49 13 1 5 Georgia 94 20 31 32 63 8 1 8 Kazakhstan 90 27 30 23 53 6 4 10 Kyrgyz Rep. 98 37 26 20 46 5 3 9 Moldova 95 26 39 14 53 13 1 7 Russia 68 16 44 20 64 10 2 8 Tajikistan 86 26 30 22 52 6 3 11 Turkmenistan 91 22 35 23 57 7 2 10 Ukraine 84 33 33 13 46 15 1 5 Uzbekistan 84 26 32 23 55 7 2 11 Estonia 93 27 49 12 61 7 1 4 Latvia 93 23 47 14 61 10 1 5 Lithuania 92 36 37 11 48 11 1 4 Source: Authors' calculations. For sources, see table 1. 14 FOREIGN TRADE IN THE TRANSmON is driven by market forces; under the model, the gravity model cannot explain such bilateral share of the NIS in Russia's total exports trade, however, because it is not materially dropped from 68 percent to 16 percent, while affected by Hong Kong's income level, which is for other NIS this share tumbled from more the main determining variable in the gravity than 90 percent to less than 30 percent. model. Second, China's trade with developing Returning to table 1, the fact that the real- countries, particularly Asian developing coun- ization ratios for China's trade in the mid-i 980s tries, may reflect differences in natural resource are much larger than for other transition endowments. These differences can be as large as economies might suggest that the Chinese eco- those with industrial countries. Hence, "Ricardo nomic system uniquely combined central plan- goods" can be an important part of trade ning in some areas with stronger trading incen- between China and other developing countries. tives in others. China began curtailing central Third, the low estimate of China's GDP used for planning earlier and on a larger scale than most 1985 lowers trade potential. As noted above, a CMEA countries. Further, it has never been part 1.0 percent increase in GDP will boost exports of the CMEA soft payments arrangements. by 1.2 percent and imports by 1.0 percent. We Thus it is tempting to draw the conclusion that used US$330 per capita for China's GDP in central planning, as it stood in the mid-1980s, 1985. The current World Bank estimate is had less trade suppressing effects in China than US$470 per capita for 1992, and arguments in other centrally planned economies. have been made that it should be much higher. This conclusion is not fully warranted: These findings also confirm that the transi- China may have seemed to be overtrading for tion economies could become very important several reasons. First, consider China's unique, exporters and importers (Winters and Wang trade-creating relationship with Hong Kong. 1994). Trade between the transition economies Hong Kong has absorbed more of China's and OECD countries is likely to expand dra- exports than have the industrial countries as a matically, while, given the high ratios of actual whole throughout the 1980s and the early to potential trade, trade between developing 1990s. But Hong Kong acts as a clearing house, countries and former CMEA countries will not as an ultimate market (Yeats 1995). The remain relatively stable. Shifting Patterns of Foreign Trade Since the collapse of communism foreign trade patterns have changed drastically. Some of these changes were sim- ply the result of the contraction in import demand trig- gered by the collapse of central planning and the Soviet Union, but others followed successful efforts to reorient trade in line with economic incentives and comparative advantage. *Export Performance in Following the collapse of central planning, both output OECD Markets in the and import demand in all ECA transition economies con- 1980s tracted, triggering the collapse in intra-CMEA and inter- * Shifting Patterns, 1988-94 republican trade. Because of these circumstances, the share of trade with other markets has increased dramatically. This * An Index of Export section compares export performance before and after the Performance collapse of central planning, assessing the extent to which inherited distortions in trade patterns have been corrected. * Changes in Imports We make one qualification: because intra-CMEA trade peaked around 1987-88, 1988 is selected as the cutting point rather than 1991, when the CMEA was officially clis- solved. There was no significant variation in the CEECs' and the Soviet export performance in OECD markets during 1980-88. Their competitive position declined and they shifted toward exporting low-value-added, resource-inten- 16 FOREIGN TRADE IN THE TRANSITION sive products. Soviet efforts to expand their oil OECD imports, fell each year except 1984.10 exports to OECD markets resulted in reduced Their average annual export growth rate was 2.9 supplies to CMEA partners, which in turn percent, compared with 5.6 percent for all depressed CMEA exports to the Soviet Union. OECD imports. The shares of Bulgaria, former As a result, the share of CMEA trade in total Czechoslovakia, and Poland in total OECD trade contracted, although until the introduc- imports reached their peak in 1980, and that of tion of stabilization cum transformation pro- Romania peaked in 1984 (at the height of grams there was little increase in their trade with Ceaucescu's policy of paying off the external OECD countries. The dissolution of the Soviet debt). Thereafter, the two Balkan countries Union caused interstate trade to collapse, and experienced the largest deterioration in export the significance of other markets increased by performance-in 1989 their shares were less default. than three-fifths of their peak performance We begin our analysis by discussing changes years-while the former Czechoslovakia and in the geographical direction of trade and in the Poland managed around three-quarters of their composition of exports to OECD countries. 1980 peak.11 Hungary, which revived its reform Contrary to earlier predictions about the non- effort after joining the IMF and the World Bank marketability of CMEA-specific products else- in 1982, failed to noticeably improve its com- where, almost all transition economies have petitive position between 1980 and 1989. The managed to greatly increase their trade with Soviet Union's share increased slightly in the OECD countries. Manufactures contributed 1980s, but mainly because of increased gas and heavily to this increase. Yet, CMEA members oil supplies sold to Western European countries. differed in their export performance, reflecting Thus by the end of the 1980s, despite domestic not only economic policy differences (discussed political pressure and active policy interventions in Section 5) but also initial structural condi- to expand exports (in order to obtain much- tions. In order to capture differences in trade needed hard currency), exporters from transition performance arising from these conditions, we countries had failed to maintain even the limited construct an index of export performance and market shares that they claimed at the end of the use it to identify countries according to their 1970s. success in normalizing foreign trade. Because The declines in the CEECs' market shares foreign trade is ultimately about getting imports, in OECD were greatest in manufactures. The we also briefly discuss developments in transi- former Czechoslovakia, once a renowned tional economies' imports from OECD, reveal- exporter of machine tools and other high-quality ing the growing significance of former-CMEA industrial products, became increasingly special- markets, particularly to exporters from the EU. ized in agricultural products, ores, and nonfer- rous metals. Similarly, Hungary, a country much Export Performance in OECD Markets in praised for its reform efforts in the 1980s, tend- the 1980s ed to shift away from manufactures and food During the 1980s the export performance products toward minerals and raw materials. of the transition economies in OECD markets Poland, offered credits by Western governments was unimpressive and displayed some disturbing and banks, should have had a relatively modern trends. Although their exports generally industrial base by the beginning of the 1980s, increased in absolute terms, their competitive but it still experienced the second largest loss position, as measured by their shares in total among the CEECs in OECD market share for SHIFTING PArrERNS OF FOREIGN TRADE 17 manufactured products. Correspondingly, the ucts were low quality and they lacked long-term shares of agricultural products, raw materials, commercial contacts. ores, and nonferrous metals in total exports The picture that emerges from this analysis expanded widely, indicating an overall shift contrasts with the export performance of China toward low value-added products in trade with after reforms began in the late 1970s. First, the West. The region's share of OECD imports China's exports to OECD markets expanded of these products also increased, suggesting that rapidly, rising almost 50 percent in value in both the ECA transition economies' comparative 1979 and 1980. Through the 1980s the value of advantage was moving away from even mildly OECD imports from China grew at an average sophisticated manufactures. annual rate of around 25 percent-well in excess Another disturbing trend for the CEECs of total OECD import growth. Second, the was their growing marginalization in OECD Chinese export bundle expanded. For example, markets in the 1980s. This trend was seen in during 1976-78 only two out of eight broad their falling shares of OECD imports and in product groups had revealed comparative advan- large annual fluctuations in their exports, relat- tage indices exceeding one; this number doubled ed to OECD business cycles. Especially outside over 1984-86. Third, unlike developments in the EU, the cyclical contraction in OECD the exports of ECA transition economies, the import demand for products from CEECs gen- composition of Chinese exports shifted from erally exceeded that for imports from other natural resource products to labor-intensive countries. The notable exceptions were ores, manufactures. nonferrous metals, and mineral fuels. Vulnerability to fluctuations suggests that Shifting Pattems, 1988-94 CMEA economies were essentially marginal The trends in overall trade performance in (swing) suppliers, probably because their prod- 1988-94 varied among transition economies, Figure 1: Exports of ECA Transition Economies, 1981-94 100 - 120 90- 80 100 70 80 Zito CtVA (percent) B650 L; ; t lzm to OED (percent) 250. 60 C _D-0--to CEk (US$ billion) ,S40A 30XA 40 Total exports (XUS$ billion) 20 20 10- 0 0 n~~~~~~6 cne ne n nc nc c n cn_ a O~0) 6)6 ) 0) 0) 0) 0) )a ) Source: Derived from data in the United Nations COMTRADE foreign trade records; various PlanEcon, Inc. publications; UNECE, Survey ofEurope, various issues. 18 FOREIGN TRADE IN THE TR'sInoN reflecting their different international and 1994 (figure 1). Exports to OECD took up the domestic circumstances. China, which did not slack. The same patterns are evident in figures 2 have to endure the collapse of central planning, and 3, which separate the CEECs and the (for- continued its steady, rapid trade expansion. In mer) Soviet Union (FSU) The share of exports 1978-88, China's average annual exports growth sold outside the CMEA and OECD countries rate was 23 percent. During this period the has contracted. This change was most obvious value of Chinese exports to OECD countries for exports from the FSU to developing coun- increased more than sevenfold-from US$4 bil- tries after 1990, mainly because subsidies were lion to US$29.4 billion (as reported by OECD abolished: developing countries' share of countries). Export growth to OECD markets exports from the FSU fell from around 33 per- increased slightly in 1988-1994 to an annual cent in 1988-89 to 21 percent in 1991 and to 9 average rate of 24 percent, and the value of percent in 1992. exports increased to US$71 billion. According The extent to which the fall in intra-CMEA to partner trade statistics, China reoriented its trade was absorbed by international markets var- exports toward OECD markets, whose share in ied arnong countries. The Balkan countries, total exports increased from 45 percent in 1978, even more than the FSU, registered a huge the eve of economic reform, to 57 percent in decline in total exports between the mid-80s 1994. and early 1990s (figure 4). Exports to OECD The trade patterns of ECA transition countries fell from US$5.5 billion in 1988 to economies changed dramatically but, interest- US$3.6 billion in 1992, but recovered thereafter ingly, the shift began before the recent political to achieve a share in the total exports of 57 per- changes. Between 1980 and 1986 the share of cent in 1994. The Visegrad countries, on the total ECA transition economies' exports sold to other hand, showed strong increases in total the CMEA (excluding non-European members) exports in the mid-1980s and again in 1992 and increased steadily from 38 percent to 44 percent, 1994; exports to OECD countries more than but it declined steeply thereafter to 14 percent in replaced those to the CMEA, increasing their Figure 2: Exports of CEECs, 1981-94 100 . 60 80 -s 60. 40 C-3 ~~~~~~~to CMEA (proent) ~ito OECD (p.rcen) 50 - 30~~~~~~~ _ -0-to CMEA (US$ blivio) 40D 20 -1 To exort (US$ btilon) 10-~~~~~~~~~~~1 Source: See figutre I SHIFTING PArrERNS oF FOREIGN TRADE 19 Figure 3: Exports of the FSU, 1981-94 100*. 70 60 80 70 50 r__to CMEA (percent) 'E ~~~ ~~~~~~~40 111Into OEGD (percent) 0~ ~ ~~~~~~~~~~~3 o 50 C; -0- to CMEA (US$ billion) A 40 .o --Ttlexports (US$ billion) 30 D 20 20 10 10 0)0 0 )0)0 ) )0 Source: See figurec1 share in the growing total from 33 percent in reorientation to induce a commodity shift 1986 to 75 percent in 1994 (figure 5). toward the OECD basket (primaries). To what extent did the switch in the direc- The ECA transition economies show tion of exports coincide with changes in their increased specialization in manufactured prod- commodity composition and to what extent ucts, according to calculations of comparative were these shifts in the geographical composi- advantage indicies in these markets. Of the com- tion of trade due to re-orientation of existing modities in which European transition exports or the differential growth of two distinct economies had comparative advantage (a bundles of exports? It is not possible to answer revealed comparative advantage index exceedting these questions directly, but some indlications unity), manufactures accounted for about 60 exist. First, exports of primary goods to non- percent of the value of total exports in 1988, but OECD markets fell faster than exports of manu- about 70 percent in 1994. Furthermore, export factures (65 percent compared with 52 percent concentration indices for OECD-oriented over 1990-92 for Russia, Kazakhstan, and exports remained unchanged for Bulgaria, the Turkmenistan), which presumably reflects the FSU, and Hungary, and fell significantly for the greater futngibiity of primaries, i.e. that they can formeir Czechoslovakia and Poland. These be recdirected. changes were also associated with shifts toward Second, consider the CEECs, which sup- labor-intensive exports to OECD; see table 3, plied manufactures eastward andi primary goods which shows such products accounting for westward. Since quality cannot be significantly between 25 percent (Bulgaria and V-isegxiLd improved in the short-term and since exporters countries) and more than 50 percent (Romania from transition economies had to compete on and China) of total exports in 1994.12 the same footing as other suppliers to western In fact, table 3 shows that the share of man- markets, one might expect the geographical ufactures in exports to OECD countries 20 FOREIGN TRADE IN THE TRANSITION Figure 4: The Balkans; Reorientation of Trade, 1980-94 90 18 80 16 70 14 =Ito OECD (percent) 60 12 2E 50 12 02 to Central Europe and the FSU Z(percent) *40 -86 6 co -i-to Central Europe and the FSU 20 6(US$ billion) 10 - 42 --*Total (in US$ billion) 10 0 2 _ 2 0_ Source: See figure 1. Figure 5: The Visegrad Countries; Reorientation of Trade, 1980-94 100 45 90 40 80 - a3 -IOECD (percent) 70 -3 z'60 30 g Central Europe and the FS U 25 a(percent) 50- CL 40 - S | | | | lt W l 20 Z -x-Central Europe and the FSU ~9 40 - o(US$ billion) 30 - 15 a --*-Total (USS billion) 20 10 10 5 0 0 Source: See figure 1. increased between 1988 and 1994 for all the Within manufactures there is also a marked transition economies. Although this is partly increase of exports of machinery and transport due to the lower level of primary commodity equipment and of the miscellaneous category prices, falling energy exports, and the fall in agri- "other manufactures." For the Visegrad coun- cultural exports from Bulgaria, Hungary, and tries it seems reasonable to suppose that at least Poland (because of adverse weather conditions some of this came from the redirection of sales and cuts in agricultural subsidies), the changes from former CMEA markets and the replace- appear too large to be dismissed as arising from ment of some East German exports to the EU.13 changes in the terms of trade or the collapse of There are no reliable data on Czechoslovak trade primary exports. with the CMEA in the 1980s, so it is not possi- able 3: Changes in the Composition of Exports to OECD Markets, 1988 and 1994 'ercentage of total exports to OECD countries) Primary Products Crops Manufactures Raw Forest Tropical Animal Labor Capital Petroleum materials Total products agriculture products Cereals Total intensive intensive Machinery Chernicals Total 3ulgara 1988 11 6 17 3 11 10 9 32 15 12 10 15 51 1994 6 13 18 4 8 6 8 26 25 11 8 12 56 FCSK 1988 5 7 12 13 3 5 1 22 19 21 14 12 66 1994 2 8 10 6 2 3 2 13 27 21 23 8 78 Hungary 1988 5 7 12 4 7 16 5 31 18 14 12 12 57 1994 4 5 8 3 5 11 4 22 25 11 24 9 70 Poland 1988 2 24 25 5 6 14 2 28 17 10 13 8 47 1994 1 17 18 5 5 7 1 17 29 12 19 6 65 Romania 1988 28 8 36 3 2 3 1 9 30 12 6 7 56 1994 3 3 5 1 2 3 2 9 54 17 9 7 86 China 1988 8 5 13 1 5 6 7 19 41 14 8 5 68 1994 3 2 4 1 3 3 2 9 58 9 17 3 87 c) USSR/FSU > 1988 41 24 65 10 1 3 2 15 7 3 4 6 20 1994 36 30 67 1 1 5 3 9 8 5 4 7 24 0 Note: The agegates are based on Leamer's (1984) dassification. 0 Source: Derived from UN COMTRADE trade files. z 22 FORIGN TRADE IN THE TRANsrnON ble to assess directly the extent of redirection, OECD markets. The extent to which countries but redirection appears to play an important role embrace the inevitable redirection of trade will in Hungary and Poland (for which data are depend partly on foreign aggregate demand, but available). mainly on their efforts to liberalize foreign trade Although the share of machinery and trans- practices and prices and improve marketing. We port equipment in Poland's exports to OECD must examine the change in the CMEA share countries remained unchanged over 1988-94, because the extent of countries' former depen- these exports grew substantially in absolute dence on these markets differs significantly. terms, at the expense of exports to former- Using the change in the value of manufac- CMEA markets. The total value of machinery tured rather than total exports to OECD as a and capital equipment exports from Hungary measure of trade reorientation excludes exports and Poland fell by around 15 percent each (fig- that are largely insensitive-at least in the short ures 6 and 7).14 But while the value of exports run-to the type of economic system. Raw to the FSU fell by 68 percent for Hungary and materials were successfully marketed by state 78 percent for Poland over 1988-94, exports to foreign trade organizations in the past and, by the EU from Hungary increased by 327 percent the same token, the changes in their exports may and from Poland by 355 percent for 1992. If indicate only changes in domestic consumption exports to the EU had remained at 1988 levels, and supply, rather than in the incentives associ- Hungarian exports would have been 38 percent ated with the introduction of market-oriented lower than their actual value in 1994, and reforms. Poland's about 49 percent lower. These figures The inevitable fall in domestic demand asso- suggest that the reorientation to OECD markets ciated with macroeconomic stabilization makes involved both increasing capacity from new and the survival of many firms dependent on interna- existing exporters as well as diverting trade from tional sales. We reflect this and its role in mod- former CMEA markets. ernizing the economy by examining the share of manufactures sold to Western markets in GDP. An Index of Export Performance This measure also helps to redress the bias inher- In this section we compare transition economies' ent in using the growth in exports of manufac- progress in foreign trade readjustment using a tures in OECD markets, which "overvalues" synthetic index of export performance. We huge percentage increases from very low initial ranked transition economies by four criteria: the levels. Although the share of manufactures sold change in the dollar value of total exports, the to Western markets in GDP tends to inflate the change in the share of CMEA countries (for the performance of small economies located near CEECs) or interstate trade (for the NIS) in total OECD markets, proximity alone does not exports, the percentage increase in manufactured explain the variation. For instance, by this stan- exports to OECD, and the ratio of OECD-ori- dard the share for Belarus (2 percent) should be ented manufactured exports to GDP in 1994. as high as that for Lithuania (32 percent), and The index is the sum of these four ranks. Ukraine's share (3 percent) should be higher than Export growth is a natural index of trade that of Russia (8 percent). One may thus argue performance, but the other measures may that some portion of this variation reveals an require some justification in terms of what they effort (or the absence of an effort) to reorient tell us about transition and trade. The gravity trade. Subsequent sections explain the variation analysis forecasts a large shift of trade toward in the pace and scope of trade reorientation. SHIFTING PATTERNS OF FOREIGN TRADE 23 Figure 6: Hungary's Exports of Machines and Transportation Equipment (SITC 7) to the EU and FSU, 1985-94 3500- 3000 2500 2000 - European Community E Soviet Union Do 1500 -X- Total ~ 500 100 LO cO N O0 0) 0 N CO t C CD CD co c o 0) 0) 0) 0) Source: Derived from United Nations COMTRADE files. Figure 7: Poland's Exports of Machines and Transportation Equipment (SITC 7) to the EU and FSU, 1985-94 4000 33500° - t 3000 .o 2500 - _ _ - .European Community E 2000 - - -9 Soviet Union 1000 t1 M 1i X pS-| * | -X-Total 1500- 100 0 0 -0 0 00 C0 0.) 0) 0) 0) 0 0) 0) 0) 0) 0 Source: Derived from United Nations COMTRADE files. Table 4 presents data on the four compo- to OECD, on the other hand, finds the NIS nents of the performance index. In terms of total heading the list, although, as the fourth critcri- export growth Czechoslovakia performed the on shows, their relative levels of sales in OECD best, and the Transcaucasian NIS the worst. are much below those of the Baltics and the Similarly, the reorientation away from CMEA CEECs. These last data are somewhat uncertain, separates the CEECs and the Baltics from the however, for they are sensitive to the estimates of other NIS. The growth of manufactured exports GDP. Estonia, with an estimated GDP almost rable 4: Export Reorientation of ECA Transition Economies Value of total exports Share of inter- Exports of mfrs. 7 (US$ million) republic exports to OECD (US$ m) 0 Change in Share of Index share Index OECD exports Index of export Z 1991 1994 1991=100 1990 1994 (percent) 1991 1994 1991=100 in GDP, 1994 reorientation H Baltic States Estonia 2,624 1,103 42 98 34 -66 150 479 320 10.5 24 Latvia 4,171 1,101 26 97 52 -46 107 334 311 5.7 37 Lhhuania 5,556 1,686 30 94 55 -42 138 445 322 8.5 33 H Central Asian NIS Kazakhstan 13,191 8,339 63 89 84 -6 65 177 272 1.0 49 Kyrgyz Republic 3,493 817 23 97 86 -11 1 11 1,309 0.4 49 Z Tajikistan 2,310 660 29 82 52 -37 1 15 2,413 0.7 40 Turkmenistan 5,029 3,049 61 96 88 -9 6 6 97 0.1 60 Uzbekistan 10,485 4,030 38 89 57 -36 3 26 960 0.1 44 European NIS Belarus 21,638 8,018 37 92 87 -6 100 387 386 1.9 47 Moldova 2,732 1,342 49 93 91 -2 11 36 323 1.0 48 Russia 168,455 84,351 50 68 44 -35 4,937 4,195 85 1.5 49 Ukraine 51,647 20,535 40 83 77 -7 682 1,279 187 1.7 52 Transcaucasian NIS Armenia 1,952 453 23 97 91 -6 8 39 501 1.5 51 Azerbaijan 6,654 1,094 16 92 67 -27 6 11 179 0.3 64 Georgia 2,493 404 16 91 79 -13 8 6 78 0.2 71 CEECs Bulgaria a 5,230 3,500 67 53 33 -38 545 1,117 205 11.3 29 FormerCSKa 11,330 15,700 139 37 15 -59 3,457 9,237 267 19.1 17 Hungary a, b 9,672 10,000 103 41 23 -44 2,558 5,159 202 12.6 25 Poland a, b 14,670 16,970 116 41 13 -68 2,955 8,730 295 9.2 20 Romania a 4,571 6,120 134 25 15 -40 1,895 2,812 148 9.6 31 a. Share of CMEA exports in total exports. b. The reference year for the share of the CMEA and value of total exports is 1989. Source: World Bank staff estimates and UN COMTRADE records. SHIFTING PATrERNS OF FOREIGN TRADE 25 twice as high as that of the other Baltics, has a and shortage economies, generating sufficient surprisingly low ratio as do the remaining NIS. exports to buy necessary imports represents the Countries were ranked according to each of major challenge and constraint. We should not these criteria and the ranks summed. The maxi- lose sight of the fact, however, that imports are mum (worst) scores that a country can obtain is the key link between foreign trade and econom- 80 (see the last column in table 4). Countries ic welfare. All else being equal, rising or higher- scoring below 40 are classified as good export quality imports indicate increased consumption performers and those with scores above 40 as or investment, and importing under competitive poor performers. The group of good performers conditions stimulates local producers to improve may be further broken down into top perform- their performance. In both respects the transi- ers-the Visegrad countries and Estonia with tion economies' imports from OECD countries scores below 25-and satisfactory performers- are probably the critical factor. Bulgaria, Lithuania, Latvia, and Romania. The Poland and the former Czechoslovakia have poor performers are broken into two groups shown the strongest import growth. Both coun- depending on whether their scores fall between tries' initial liberalizations included substantial 40 and 50 or are above 50. The first group import expansion, as pent-up demand and includes Belarus, Kazakhstan, Kyrgyz Republic, spending power were released, followed by a Moldova, Russia, Tajikistan, and Uzbekistan, period of relative calm, as the shock and stabi- and the second group, Armenia, Azerbaijan, lization took hold. As the transition became well Georgia, Turkmenistan, and Ukraine. established (1991 for Poland and 1992 for These rankings confirm earlier observations Czechoslovakia), imports increased by no less about both the better starting points and the than two-thirds in one year. A similar pattern substantial trade reorientations of the Central could be observed in Estonia: having successful- European countries and the Baltic states. Given ly stabilized its economy by 1993, the value of Tajikistan's limited progress in dismantling cen- Estonia's imports from OECD countries tral planning and liberalizing trade, its border- increased by 81 percent in 1994 alone. In line index is surprising, but easily explainable. It Hungary the expansion was steadier and more reflects a surge in chemicals exports to hard cur- moderate, in keeping with the slower pace of rency markets, which shows up as both a redi- reform. Romania is a special case. Although it rection of its exports to OECD markets and a did not move beyond the initial release phase of relatively high percentage of OECD exports in transition until 1993, its imports-drastically GDP by 1994.15 We would predict-see suppressed during Ceaucescu's "successful" drive below-that the prospects for further expansion to pay off foreign debt-expanded strongly as ofTajikistan's exports are limited unless compre- new debt was incurred. Only by 1994 was hensive economic reforms are made. Indeed, import growth underpinned by significant whether Tajikistan can keep up this level of per- export growth. formance provides one with a good test of the Bulgaria shows the most disappointing per- importance of reform.16 formance in import terms, probably a result of its extremely tight connections with former Changes In Imports CMEA-markets and the high level of sovereign Until now we have discussed trade adjustment as debt it inherited from the communist regime. if it were purely a matter of export performance. Excluding the Baltic states, the NIS also appear We have focused on exports because for poor to be locked into the same bind although some 26 FOREIGN TRADE IN THE TRANSInON Table 5: Transition Economies Imports from OECD Countries, 1988-1994 (million US$) Share of the EU in OECD exports Average (percent) growth rate, 1988-94 1988 1989 1990 1991 1992 1993 1994 1988 1994 (percent) Bulgaria 2,385 2,399 1,567 1,669 1,906 1,674 1,889 82 97 -2 FCSK 3,553 3,612 4,815 6,205 10,692 10,535 12,826 70 81 27 Hungary 3,952 4,608 5,372 6,592 7,789 8,199 8,678 69 71 15 Poland 4,905 6,128 7,587 12,481 13,513 13,780 15,564 65 84 25 Romania 1,258 1,222 2,395 2,273 3,146 3,329 3,742 57 83 25 China 25,775 24,253 21,130 25,857 32,837 42,285 48,667 26 32 13 FSU 24,684 28,638 26,894 28,189 26,667 27,512 31,038 48 68 4 Total 66,512 70,860 69,760 83,266 96,550 107,314 122,405 44 58 11 Source: Derived from UN COMTRADE trade files. countries have shown signs of advances recently. grouping is beginning to emerge as the Asian Included in this set are the land-locked Asian NIS become important trading partners for NIS, which significantly expanded imports in Turkey. Turkey's exports to Asian NIS more than 1993 and 1994 albeit from a very modest base. doubled between 1992 and 1994, with Asian The major suppliers of the expanding mar- NIS now absorbing around 3 percent of Turkey's kets in the transition economies were exporters total exports. from Western Europe, especially the EU, which Thus, despite the contraction in the ECA moved quickly to deepen relations with the transition economies' GDP in 1989-93, they CEECs and the Baltics. The ECA transition provided expanding markets for OECD prod- economies have provided exceptional export ucts. Once the "transitional recession" is over, a market growth for the EU countries recently (17 large increase in import demand in other ECA percent per year over 1988-94 compared with 6 transition economies can be expected. If nothing percent for total extra-EU exports). Their share else, this growth potential and the growth rates of EU extra-bloc exports has risen from 7 per- already obtained suggest that OECD countries cent in 1988 to 12 percent in 1994 and is likely have a serious stake in the success of the transi- to continue to grow. A second geographical tion. 4 Changes in Market Access T o what extent did changes in access to OECD markets following the collapse of central planning account for the shift in trade patterns of the ECA transition economies? The value of improved market access depends on several factors, including existing discriminatory arrangements, geographical proximity, the transportation network, and the domestic policy setting. In this section we argue that *Incomplete Normalization improvements in market access were relevant only to coun- of Relations tries that chose a radical approach to dismantling central planning and that had relatively good transportation access * Preferential Market to OECD markets. Access The access of ECA transition economies to OECD markets has evolved in three stages: the removal of discrim- * The Significance of Improved Market Access inatory measures aimed specifically against state trading economies; the granting of preferential access under the General System of Preferences (GSP), which puts transition economies on a par with developing countries; and the negotiation of free trade and association agreements. Access varies widely among transition economies and OECD markets, depending on the transition economies' commitment to reform, their pre-collapse access, and their political relations with the various Western governments. China and several former Soviet republics have yet to move 28 FOREIGN TRADE IN THE TRANSITION beyond the first stage, whereas other transition Where MFN status really counted-in the economies have reached second and third stages United States-progress was slower. For (table 6). Romania MFN treatment was delayed for four years after the fall of Ceaucescu, while it has Incomplete Normalization of Relations generally taken more than a year after indepen- Before the collapse of communism exports from dence for the NIS to receive MFN treatment. the transition economies were subject to some The U.S. government had granted MFN status restrictions imposed only on centrally planned to the CEECs (excluding Romania) and the economies. Because of the state monopoly over Baltic states by December 1991, and to foreign trade, centrally planned countries- Armenia, Kyrgyz Republic, Moldova, Russia, including those that were GATT members- Tajikistan, and Ukraine by August 1992 (table Czechoslovakia, Poland, and Romania- were 6). The remaining NIS followed in 1993, except defined as "state trading countries."'7 As such for Azerbaijan, whose exports still face U.S. gen- they were exempt from the GATT's Article 13 eral tariffs averaging about 30 percent compared forbidding the use of quantitative restrictions with the average MFN rate of 5 percent. (Tovias and Laird 1991:15), but subject to non- Because of these delays, the move to MFN treat- tariff barriers in OECD countries with greater ment appears to have had little effect on the frequency and with a more restrictive impact observed trade performance of NIS in the U.S. than were market economies (Olechowski and markets. Yeats 1982). In the United States, only China, In 1990 (1991 for Romania) the EU abol- Hungary, and Romania (until 1987) received ished specific quantitative restrictions-that is, most-favored nation (MFN) treatment through- quotas applied only on exports originating in out the 1980s; other exporters were subject to CMEA economies. It also suspended nonspecif- prohibitively high tariffs for at least part of the ic restrictions (except in the Spanish and decade. Portuguese markets) on nonsensitive industrial The trade normalizing measures introduced imports from Hungary and Poland on January by OECD governments after the collapse of 1, 1990, from Bulgaria and Czechoslovakia on central planning involved removing these dis- October 1, 1990, from Romania on May 1, criminatory practices at a pace determined by 1991, from the Baltics on January 1, 1992, and the Western government's perception of the from all other former Soviet republics on transition economy's progress toward democracy January 1, 1993.18 Other measures for the and a market system. Normalization meant CEECs included the elimination of some quan- granting MFN status (to those countries that titative restrictions on iron and steel (since part- did not have it) and eliminating quantitative ly replaced by antidumping actions) and increas- restrictions applied specifically against state trad- es in textile and clothing quotas, especially on ing countries. The CEECs obtained MFN status outward processing trade. in EU markets by January 1, 1990 (except for Despite the steps just outlined, trade rela- Romania, which received this status in May tions between transition and OECD countries 1991) and the NIS by January 1, 1992. Apart have not yet been completely normalized. In par- from worries about the security of such access, ticular, some OECD countries continue to das- however, this measure was of no additional con- sifyr ECA transition economies as state trading sequence because the EU already offered de economies, making them more vulnerable to facto MFN status. quantitative trade restrictions and subjecting ible 6: Changes in Position of the European Transition Economies in OECD Import Regimes tatus as of December 1994) European Union EFTA Austria Switerzland Finland Norway MFN GSP FTA MFN GSP FTA MFN GSP FTA MFN GSP FTA MFN GSP FTA MFN GSP FTA altic States Estonia 1/92 1/92 1/95 n.a. n.a. no 1/92 7/92 no 5/26 no 04/93 12/92 no 12/92 1/92 no 7/92 Latvia 1/92 1/92 1/95 n.a. n.a. no 1/92 7/92 no 5/25 no 04/93 5/93 no 5/93 1/92 no 7/92 Lithuania 1/92 1/92 1/95 n.a. n.a. no 1/92 7/92 no no no 04/93 1/93 no 1/93 1/92 no 7/92 3uropean NISs n.a. n.a. Belarus 1/92 1/93 a no n.a. n.a. no 1/92 3/93 no Yes b no no 2/93 no no 1/92 no no Moldova 1/92 1/93 a no n.a. n.a. no 1/92 3/93 no Yes b no no 2/93 no no 1/92 no no Russia 1/92 1/93 a no n.a. n.a. no 1/92 3/93 no Yes b no no 2/93 no no 1/92 no no Ukraine 1/92 1/93 a no n.a. n.a. no 1/92 3/93 no Yes b no no 2/93 no no 1/92 no no rranscaucasian NiSs n.a. n.a. Yes b Armenia 1/92 1/93 a no n.a. n.a. no 1/92 3/93 no Yes b no no 2/93 no no 1/92 no no Azerbeijan 1/92 1/93 a no n.a. n.a. no 1/92 3/93 no Yes b no no 2/93 no no 1/92 no no Georgia 1/92 1/93 a no n.a. n.a. no 1/92 3/93 no Yes b no no 2/93 no no 1/92 no no ,entral Asian NISs n.a. n.a. Yes b Kazakchstan 1/92 1/93 a no n.a. n.a. no 1/92 3/93 no Yes b no no 2/93 no no 1/92 no no Kyrgyz Republic 1/92 1/93 a no n.a. n.a. no 1/92 3/93 no Yes b no no 2/93 no no 1/92 no no Tajikistan 1/92 1/93 a no n.a. n.a. no 1/92 3/93 no Yes b no no 2/93 no no 1/92 no no Turkmenistan 1/92 1/93 a no n.a. n.a. no 1/92 3/93 no Yes b no no 2/93 no no 1/92 no no Uzbekistan 1/92 1/93 a no n.a. n.a. no 1/92 3/93 no Yes b no no 2/93 no no 1/92 no no CEES countries Bulgaria 11/90 1/91 2/94 n.a. n.a. 1/93 8174 4/72 1/93 4/73 no 1/93 1/75 c 7/73 1/93 Yes no 1/93 former CSK 11/90 1/91 3/92 n.a. n.a. 1/93 Yes 7/91 1/93 7/71 no 1/93 1/75 c no 1/93 Yes no 1/93 Hungary 12/88 1/89 3/92 n.a. n.a. 1/93 Yes 7/88 1/93 1/74 no 1/93 1/750c no 1/93 Yes no 1/93 B Poland 12/89 1/89 3/92 n.a. n.a. 1/93 Yes 1/90 1/93 12/73 no 1/93 4/78 c no 1/93 Yes no 1/93 Romania 5/91 1/74 6/93 n.a. n.a. 1/93 5/76 4/72 1/93 4/73 no 1/93 1/75 c 12/73 1/93 Yes no 1/93 -1 n.2. not applicable. Continued on next page a. Granted on an exceptional and temporary basis b. Continuation of the 1948 agreement with the Soviet Union. c. Under eligibility review by the U.S. government. 0 able 6:(continued) Sweden Canada United States Japan Memorandum: v 0 MFN GSP FTA MFN GSP FTA MFN GSP FTA MFN GSP FTA GATTg status C) z 3altic States E Estonia 10/91 no 7/92 1/92 4/92 no 12/91 2/92 no no no no A Latvia 10/91 no 7/92 1/92 4/92 no 12/91 2/92 no no no no A z Lthuania 10/91 no 7/92 1/92 4/92 no 12/91 2/92 no no no no A European NIS Belarus no no no 1/92 11/92 no 2/93 no d no Yes e no no A Moldova no no no 1/92 11/92 no 7/92 no no Yes e no no A 0 Russia 1/92 no no 1/92 4/92 no 6/92 10/93 no Yes no no A Z Ukraine no no no 1/92 4/92 no 6/92 3/94 no Yes e no no A Transcaucasian NIS Amnenia no no no 1/92 4/92 no 4/92 no no Yes e no no A Azerbeijan no no no 1/92 11/92 no No no no Yes e no no 0 Georgia no no no 1/92 11/92 no 8/93 no no No no no none Central Asian NIS Kazakchstan no no no 1/92 11/92 no 2/93 4/94 no Yes e no no 0 Kyrgyz Republic no no no 1/92 11/92 no 8/92 12/93 no Yes e no no 0 Tajikistan no no no 1/92 11/92 no 11/93 No no Yes e no no none Turkmenistan no no no 1/92 11/92 no 10/93 No no Yes e no no 0 Uzbekistan no no no 1/92 11/92 no 1/94 no d no Yes e no no A CEES countries Bulgaria Yes 1/75 1/93 1/74 1/74 no 11/91 12/91 no 8/70 4/72 no A formerCSK Yes no 1/93 4/48 12/91 no 11/90 5/91 no yes 5/92f no M Hungary Yes no 1/93 9/73 12/89 no 7/78 11/89 no 9/76 4/86 no M Poland Yes no 1/93 10/67 12/89 no 2/87 1/90 no 10/80 1/90 no M Romania Yes 1/75 1/93 11/71 1/74 no 11/93 11/93 no 7/70 4/72 no M d. Applied de facto since January 1992. e. Agreement providing for no tariffs on industrial products and MFN rates on agricultural products. f GSP applied between May 1, 1992 and January 1, 1993. g. M=member; A-accession; and O=observer. Source: United States Trade Representative Office, EFTA Secretariat, United Nations Economic Commission for Europe. CHANGES IN MARKET ACCESS 31 them to different treatment in antidumping Only the EFTA and the EU have gone actions. The EU removed the Visegrid countries beyond GSP to sign free-trade agreements with from this category in 1993 and the Baltics in various European transition economies. The 1995, but other transition economies-regard- Baltic states signed bilateral free trade agree- less of how advanced they are in their transi- ments with Finland, Norway, Sweden, and tion-still face this extra threat to market access. Switzerland in 1992-93. These agreements Moreover, most European transition economies called for immediate duty free treatment of are not protected by GATT rules and procedures industrial products subject to rules of origin, (see the last column of table 6). Not a single and were among the EFTA countries' most lib- transition economy has been admitted to the eral agreements in terms of market access (Sorsa GATT since 1989, while members admitted ear- 1994c). In addition, separate agreements with lier on special terms (former Czechoslovakia, Sweden and Norway covered agricultural prod- Poland, and Romania) have yet to complete ucts and fish. The EFTA agreements with the negotiations on their new terms of membership CEECs-signed in 1992-were EFTA-wide, that reflect the disappearance of central planning not bilateral as was the case with the Baltic and of the state monopoly over foreign trade. states; they called for the gradual elimination of all tariff and nontariff barriers on industrial Preferential Market Access products, in line with the industrial component In line with its predilection toward trade prefer- of EU's association agreements. The association ences and its role as the ECA transition agreements, frequently referred to as the Europe economies largest trading partner, the EU took Agreements, were signed between the EU and the lead in offering the emerging economies individual CEECs in 1991, 1993, and 1994, preferential market access. First, the EU offered and came into force in the form of interim trade them generalized system of preferences (GSP) agreements after March 1992. They overshad- treatment, which put them on a par with most owed GSP arrangements by making preferential developing countries with quota-limited tariff- tariffs permanent rather than subject to an free access for most products. Bypassing the state annual review of their GSP status, and dominat- trading qualification, the EU granted GSP status ed the EFTA-EU free trade agreement by pro- to Hungary and Poland in 1990, to Bulgaria and viding for somewhat improved access to agricul- Czechoslovakia in 1991, to Estonia, Larvia, and tural markets, similar to that granted in the Lithuania in 1992, and to other former Soviet Lome Convention and the Mediterranean republics on an exceptional and temporary basis Agreements. They also called for extensive coop- in 1993. Romania's and China's exports have eration and for the CEECs' economic legislation been subject to GSP treatment in EU markets to move toward EU standards. Following the since 1974. Except for Canada and Austria, 1993 Copenhagen EU summit they also explic- which granted GSP status either before the col- itly foreshadowed accession to the EU. All of lapse of central planning (to Bulgaria and these measures take them well beyond 'ordi- Romania) or immediately after it, other OECD nary' free-trade agreements. governments delayed this action and in some The Baltics signed cooperation agreements cases have not yet taken it. For example, in 1994 with the EU in 1992 (Estonia) and 1993 (Latvia Japan had yet to grant GSP status to a NIS, and Lithuania), then free trade agreements cov- while the United States had done so only for ering industrial products, and eventually Europe eight of the fifteen states. Agreements in 1995. 32 FOREIGN TRADE IN THE TRANSITON Trade relations between the EU and the goods was not included in the agreement, except non-Baltic NIS have been governed by the non- for a provision promising negotiations on sepa- preferential 1989 Trade and Cooperation rate trade regimes for coal, steel, and nuclear Agreement with the Soviet Union. In October materials. Trade in textiles and clothing, covered 1992 the EU Council of Ministers adopted a by the Multifiber Arrangement (MFA), is sub- framework for negotiating more extensive agree- ject to separate agreements already in force. A ments with those countries, with Ukraine being similar trade liberalizing agreement has been the first to sign a new Partnership and subsequently concluded between the EU and Cooperation Agreement in March 1994. The Russia. Agreement calls for the mutual granting of MFN status, the removal of quantitative restric- The Significance of Improved Market Access tions, and the establishment of a free trade area. The value of a preferential trade agreement is However, a free trade area will be established closely linked to the cut in protection it makes only if Ukraine progresses in establishing a mar- and the degree of preference it offers. Current ket economy. Its progress will be jointly OECD trade regimes are characterized by: low reviewed in 1998. Market access for sensitive tariffs; common use of nontariff barriers, such as Table 7: Trade Measures Affecting European Transition Economies' Exports to the EU Average Average Nontariff tariff rate tariff rate Nontatiff barlers barriers (frequency) Country (simple) (trade weighted) (coverage ratio) (by items) Bulgaria 7.5 7.0 33.6 26.5 FCSK 7.3 6.3 32.7 26.7 Hungary 1.2 1.6 38.9 29.4 Poland 1.3 1.8 25.1 27.2 Romania 0.9 0.4 25.1 32.3 First Layer Austria (EFTA) 1.2 0.3 10.4 11.8 Finland (EFTA) 0.8 0.3 7.4 9.9 Second Layer Cote d'lvoire (ACP) 0.3 0.1 8.1 13.2 Swaziland (ACP) 0.5 0.6 67.1 18.1 Third Layer Israel (Mediterranean) 1.7 2.0 15.0 11.0 Turkey (Mediterranean) 0.8 0.8 53.2 24.5 Fourth Layer Japan 7.4 7.4 36.2 10.9 United States 7.4 4.6 7.0 9.7 Developing Countries 2.9 2.4 11.3 15.9 World 3.8 2.9 8.0 15.9 a. The coverage ratios were computed using 1990 non-tariff barriers data against 1988 trade flows. Source: UNCTAD-World Bank SMART database. CHANGES IN MARKET AccEss 33 quantitative restrictions (MFA, agricultural tion ultimately to lead to full membership. They products) and voluntary export restraints; and are probably not directly responsible for the frequent recourse to remedies against "unfair" recent growth of exports to the EU, however. trade such as antidumping investigations. While the Europe Agreements will remove the GSP status offers preferential treatment over quota limits on preferential access that charac- MFN status. In the EU MFN tariffs on indus- terize the GSP, they have a number of remaining trial products average around 6 percent, whereas restrictions, including delays in liberalizing average GSP tariffs are around 2 percent, and imports of sensitive products, tight rules of ori- most GSP items (94 percent) are subject to zero gin, continuing threats of antidumping, and the rates.19 GSP preferential rates cover 63 percent virtual exclusion of agriculture (Winters and of all Combined Nomenclature tariff lines. For Wang 1994 and box 2). Despite their generous- industrial products the shares are even higher: sounding provisions and the fact that Europe 74 percent of tariff lines, all with zero rates. Agreements have been more friendly toward the The EU system of preferences may be CEECs than trade regimes in other OECD thought of as a pyramid, with the EFTA coun- countries, the concessions now in force do not tries at the apex (most preferred) passing currently go much further than the GSP through the Lome, and Mediterranean agree- The free-trade agreements between the ment countries and other developing countries EFTA and the CEECs have now partly been (GSP), to non-European OECD countries superseded by the accession of Austria, Finland, (MFN) at the bottom (Schumacher and M6bius and Sweden to the EU, although they continue 1994). In 1990 Bulgaria and Czechoslovakia still for the other EFTA members. They are similar faced MFN rates, as did Japan and the United to and subject to some of the same reservations States (table 7). Soviet exports were in the same as the Europe Agreements. The same arguments class as those from Bulgaria and Czechoslovakia, also apply to the free-trade agreements with the but, with a favorable composition of exports, Baltics, but with less force because GSP cover- faced a much lower weighted average rate than age of imports varies widely among the coun- those countries. Average tariffs on imports from tries of the EFTA, ranging from 1 percent for China, Hungary, Poland, and Romania were Norway to 45 percent for Switzerland. Free- close to those on imports from EFTA, African, trade agreements moved the Baltics to an inter- Caribbean, and the Pacific, and Mediterranean mediate level in the EFTA preference pyra- countries, reflecting their GSP status. They were mid-above GSP-eligible developing countries much lower than those levied on developing but below the EU. The value of these free-trade countries in general as this group includes a agreements is eroded somewhat by their bilater- number of countries that do not have GSP sta- al nature, which makes them more restrained by tus. GSP appears to represent a big step in tariff the rules of origin than would be a single liberalization, but there is little evidence that, EFTA-wide agreement. Only Estonia seemed with its limitations and exclusions (quantity lim- able to reap the new opportunities offered by its, special treatment of sensitive products, EFTA's preferential arrangements. The Baltics' uncertainty of access), it alone can explain exports to the EU, however, doubled between changes in transition economies' shares in 1991 and 1994, even initially without a free- OECD imports. trade agreement. This doubling is probably best For the CEECs the Europe Agreements explained in gravity terms-the major market deepen integration with the EU, with the inten- in the EU, Germany, is not only larger but also 34 FOREIGN TRADE IN THE TRANSITION more conveniently located than Norway or ments has been limited to European countries, Switzerland. however, and has not been directly responsible Preferential access to OECD markets has for much of the growth of exports from CEECs been advocated in both the East and the West or the Baltics. This is so because, in addition to as a means of fostering industrial restructuring their inherent limitations (antidumping, tight among the European transition economies. But rules of origin, and so on), the agreements were even putting aside the issues of creating a new slow to improve upon GSP concessions. constituency of countries with an apparent Moreover, the extent to which countries with interest in higher MFN tariffs and of the trade similar access have performed differently-for diverting effects of the CEECs offering the EU example, Romania vs. the other CEECs, reciprocal preferences, this view must be quali- Estonia vs. the other Baltics-clearly indicates fied. The removal of discriminatory measures that even if some level of access is necessary for has clearly been important in terms of both export reorientation, it is certainly not suffi- trade volume and welfare. Putting the European cient. The critical element in performance is transition economies on a par with developing domestic reform. It is in this respect, not that of countries by granting the GSP improved their relative access rules, that the agreements have market access and has, despite the shortcomings been important in signaling the CEECs' com- of the GSP, probably stimulated trade. The mitment to the transition and in enhancing the extension of preferences to free-trade agree- credibility of their market-oriented reforms. Box 2: EU Concessions for Industrial Products, Interim Trade Agreements (ITAs), Market Access, and CEECs Exports, 1991 Industrial products, as defined in the ITAs, cover products falling within Chapters 25 to 97 of the CN (com- bined nomenclature). They include not only manufactures (SITC. 5 to 8-68) but also some agricultural materi- als, mineral fuels, and ores and metals. The groups are presented in descending order of their sensitivity to EU trade measures, as specified in the ITAs, with the exception of coal which falls under the authority of the European Coal and Steel Communities. The trade liberalizing measures include concession granted at the EU Summit, Copenhagen, 21-22 June 1993. 1. The "immediate" free-trade group. It includes industrial products for which all restrictions, both tariff and nontariff, are eliminated upon passage of the ITAs Former Czecho- Bulgaria slovakia Hungary Poland Romania Value of EU imports, 1991(US$ million) 335 2,097 1,627 2,761 602 Share in total exports to EU (percent) 36 42 36 36 33 Share in industrial exports (percent) 50 45 49 44 35 NTB coverage ratio, 1990 3.6 3.8 3.7 3.8 3.4 Average tariff rate (simple) 5.6 5.7 0.0 0.0 0.0 Range of tariff rates (percent) 37.9-0 39.7-0 14-0 14-0 6.2-0 Continued... CHANGES IN MARKET AccEss 35 Box 2: Continued 2. The "one year-delayed" free trade group. It includes mainly industrial raw materials. Duties reduced by 50 percent upon passage of the ITAs and eliminated next year. Former Czecho- Bulgaria slovakia Hungary Poland Romania Value of EU imports, 1991 (US$ million) 10 52 23 27 1 Share in industrial exports (percent) 1.4 1.1 0.7 0.4 0.0 NTB coverage ratio, 1990 33.3 40.0 52.6 15.4 40.0 Average tariff rate (simple) 5.1 4.4 3.0 6.2 2.0 Range of tariff rates (percent) 6-0 18.3-0.9 7-0.9 18.3-0 3.2-0 3. The "two-year-delayed" free trade group. It includes low-processed primary intermediate products. There is significant differences in product coverage in the ITAs. Duties reduced by 20 percent of the base rate the first year that the ITAs are in force and fully abolished by end of the second year. Forrner Czecho- Bulgaria slovakia Hungary Poland Romania Value of EU imports, 199,(US$ million) 1 2 17 49 27 Share in industrial exports, (percent) 0.1 0.1 0.5 0.8 1.6 NTB coverage ratio, 1990 20.0 100.0 40.0 13.3 0.0 Average tariff rate (simple) 3.3 3.1 3.6 3.3 3.2 Range of tariff rates (percent) 6-0 6.2-0 6.2-0 6-0 6-0 4. The "quota/five-year-delayed" free trade group. It includes some chemicals, some leather goods, cork and wood products, optical goods, glass, steel products not covered by the European Coal and Steel Community (ECSC), copper products, electric machinery, footwear, clothing accessories, furniture, motor vehicles, toys, and so on. There are substantial differences in coverage in individual ITAs-4ists for Hungary and Poland are more extensive than for other CEECs. Trade liberalizing measures include an increase in tariff quotas or ceilings (20 percent per year for Hungary and 25 percent for other CEECs and a reduction in tariffs (10 percent) per year for Hungary and 15 percent for other CEECs). Former Czecho- Bulgaria slovakia Hungary Poland Romania Value of EU imports, 1991 (US$ million) 111 1230 810 1477 535 Share in industrial exports, (percent) 16.3 26.5 24.3 23.6 31.4 NTB coverage ratio, 1990 18.8 20.6 21.0 21.7 23.7 Average Tariff Rate (simple) 8.6 8.7 0.0 0.0 0.0 Range of tariff rates (percent) 25.8--0 25.8--0 0--0 0--0 0--0 5. Includes the group covered by the ECSC. Trade liberalizing measures vary across products and markets. Three subgroups are distinguished: Continued ... 36 FOREIGN TRADE IN THE TRANSITION Box 2: Continued 5.1. Steel products. QERs eliminated immediately and tariffs reduced to 80 percent of the base rate at the beginning of the first year; by 60 percent in the second year; by 40 percent in the third year; by 20 percent in the fourth year; and eliminated at the beginning of the fifth year. Former Czecho- Bulgaria slovakia Hungary Poland Romania Value of EU imports, 1991 (US$ million) 77 489 139 280 61 Share in industrial exports (percent) 11.4 10.5 4.2 4.5 3.6 NTB coverage ratio, 1990 74.6 64.4 58.2 57.4 68.2 Average tariff rate (simple) 5.4 5.6 0.0 0.1 0.0 Range of tariff rates (percent) 10--0 10--0 3.2--0 4--0 0--0 5.2. Coal products. QERs and custom duties levied on imports from Bulgaria, the FCSK and Poland elimi- nated at the beginning of the second year of the ITAs, and on imports from Hungary and Romania reduced by 50 percent on January 1, 1994 and eliminated by December 31, 1995. Imports into Germany and Spain are excluded. QER duties must be abolished no later than for years after the ITAs in Germany and Spain. 5.2.A. QERs and duties abolished during the first year or on December 31, 1995. Former Czecho- Bulgaria slovakia Hungary Poland Romania Value of EU imports, 1991 (US$ million) 0.9 1.1 2.7 268.5 0.1 Share in industrial exports (percent) 0.1 0.0 0.1 4.3 0.0 NTB coverage ratio, (1990) 0 0 0 0 0 Average tariff rate (simple) 4.2 5.9 2.0 4.3 0.0 Range of tariff rates (percent)) 8.3%-0% 8.3%-0% 2%--0% 8.3%--0% 0%--0% 5.2.B. QERs and duties abolished after four years of the ITAs. Value of imports into Germany and Spain, 1991 (US$ million) 0 140 1 285 0 Share in industrial exports (percent) 0.0 3.0 0.0 4.6 0.0 6. The group covered by the Multifiber Agreements. It includes textiles and clothing products. Custom duties reduced in the first year by two-sevenths and subsequently by one-seventh annually except for the last year (two-sevenths) so that they are eliminated by end of the fifth year. Quotas are negotiated annually and new arrangements are tied to the outcome of the Uruguay Round. Former Czecho- Bulgaria slovakia Hungary Poland Romania Value of EU imports, 1991(US$ million) 142 630 712 1099 480 Share in industrial exports (percent) 20.9 13.6 21.4 17.6 28.2 NTB coverage ratio, 1990 90.6 87.6 85.1 88.8 86.2 Average tariff rate (simple) 10.8 10.7 0.1 0.0 0.1 Range of tariff rates (percent) 17--0 17--0 9.3--0 8.6--0 9.3--0 Sources: Derived from the European Association Agreements, the UN COMTRADE database and the UNCTAD- World Bank SMART database. 5 Policy and Trade Performance T he reorientation of transition economies' trade away from other CMEA states seems quite inevitable, but the speed with which it is achieved will depend on the freedom and incentives that firms have to exploit foreign markets- that is, on the transition economies' progress in establishing a market economy, balanced by constraints in foreign demand, domestic capacity, and transport capacity. 1 Exogenous Factors In this section we explain the differences in trade reori- Accounting for Differences entation among ECA transition economies by looking at in Trade Reorientation structural impediments, both domestic and international, * Trade Policy Reform and at the reform measures that have been implemented to alleviate them. We start with a series of exogenous factors, * Changes in Domestic which explain little, and pass through trade policy to Institutional and Policy macroeconomic and institutional factors, which appear to Settings offer the best explanation of differing degrees of reorienta- tion. Exogenous Factors Accounting for Differences in Trade Reorientation The initial conditions in the CEECs were more favorable than those in the NIS. First, the demise of the Soviet com- mon economic space was a more formidable adjustment challenge to the NIS, except for Russia,20 than was the 38 FOREIGN TRADE IN THE TRANSITION breakdown of the CMEA to the CEECs. The access accorded to the Baltic states might have CEECs had domestic currencies, whereas the had some influence, but, at least in EFTA, only NIS, excluding Russia, which inherited the Estonia seems to have benefited. Furthermore, ruble, had to introduce them. The NIS did not although most trade restrictions directed toward have the institutional infrastructure to sell their the Soviet Union remained in place during the products abroad because foreign trade organiza- first year of NIS' independence, the differences tions tended to be located in Moscow. The between the CEECs and the NIS diminished CMEA had not succeeded in imposing suprana- significantly in 1993, when the NIS obtained tional plans on its members, but the Soviet GSP status. Thus only a proportion of the varia- republics operated under a central planning and tion in export reorientation by 1994 can be budget system. In short, the level of inherited explained by differences in access to OECD mutual interdependence was much higher for markets. This is not to deny the importance of the NIS than for the CMEA members. This dis- market access. Rather, it is to point out that parity has had two consequences. First, the access is far from sufficient, especially in the breakdown in supply linkages both within the early phases of transition. FSU and among former-CMEA members has Neither the trends in competitiveness in affected supply in the NIS, including Russia, OECD markets nor the length of the adjust- more than in CEECs. Second, the share of prod- ment to market institutions can fully explain ucts, mostly noncompetitive in terms of quality, changes in export performance. The export that had been developed solely for the protected expansion of the Visegrad countries in 1990-92 CMEA or FSU markets was probably larger in was a reversal of trends prevailing in the previous the NIS than in the CEECs. two decades. Similarly, former Czechoslovakia, The Visegrad group began to reorient their which had been slow to reform under central trade toward OECD markets in the mid-1980s, planning and to expand commercial relations which gave them more time to adjust than the with the OECD, recorded greater export growth NIS. Given these circumstances, the trade than did Hungary or Poland, albeit from a records of the Baltic states and of Bulgaria look smaller base. Furthermore, although the initial especially impressive. Finally, because commu- conditions in terms of foreign trade reforms nism collapsed earlier in the CEECs than in the were fairly uniform across the NIS, their trade FSU, the CEECs benefited earlier from the response varied substantially. Thus while geo- trade concessions offered by OECD govern- graphic location and earlier links with OECD ments. The OECD trade policy response to the importers and marketing expertise may explain dissolution of the Soviet Union was slow- some of the difference in response between except for the Baltic states, which quickly countries, the bulk of it seems due to something obtained preferential access to OECD markets. else. Despite the differences in access to OECD markets in 1991-92, Western trade policy does Trade Policy Reform not seem to be a major discriminating factor Trade and exchange rate reforms are crucial for between countries. The CEECs obtained much any transitional economy dismantling central greater access to EU markets during 1990-92, planning, but different countries have moved at thanks to their GSP status. Yet Romania, which different paces and in different directions. Some had GSP status since 1973, never took advan- have merely toyed with the inherited system tage of it. As for the NIS, the better market while others have achieved a complete overhaul. POLICY AND TRADE PERFORMANCE 39 In some instances trade liberalizing measures countries that inherited very rigid foreign trade have been neutralized by others imposing greater regimes (other CEECs and the Baltic states). direct control, and even in the boldest of The alternative path was gradualism. The mea- reformers, liberalizing measures have been sures taken by the countries following this track phased in over time and have experienced some resembled (and in some cases still resemble) the reversals. A detailed discussion of these develop- foreign trade "reforms" made under central ments is beyond the scope of this study,21 but planning by the early "reformers," although we can give a general account of developments their scope was more extensive, the erosion of in foreign trade policy. The transition in foreign the state control more complete, and the liberal- trade policy can be assessed in terms of several ization of imports deeper. measures. The extent of state trading is the most The slow-reforming NIS also share another important, although several other indicators structural feature with the CEECs of the late mark the progress toward achieving a genuinely 1980s: institutional arrangements underlying market-based system. On the export side, these interstate trade have been strongly reminiscent indicators include the coverage of export con- of those governing commercial interaction trols and licenses and, to the extent that the offi- among CMEA countries. While for the CEECs cial exchange rate diverges from the black mar- the demise of the CMEA put an end to a dual ket rate, the requirements for surrendering for- system of international trade, the dissolution of eign exchange. On the import side the main the Soviet Union led to the emergence of trade indicators are the ability of firms to obtain for- arrangements among the NIS. Because of defi- eign exchange and the levels of tariff and nontar- cient payments arrangements, trade among NIS iff barriers. was largely conducted on a bilateral basis for Despite reforms in the CEECs in the 1980s, some time. However, countries that succeed in the continuing centralized control of foreign replacing state trading with enterprise-to-enter- exchange transactions effectively shielded prise trading will go a long way in moving domestic producers from international competi- toward a full-fledged, market-based foreign tion. On the import side the centralized alloca- trade regime. tion of foreign exchange served as a nontariff barrier, while on the export side, multiple Fast Versus Slow Reformers exchange rates distorted prices and foreign Taking into account the differences in the exchange surrender requirements undermined scope and pace of the foreign trade reform mea- the incentives for firms to engage in foreign sures implemented so far, we can distinguish trade. Thus the foreign trade regime was explic- between two groups of countries. The first group itly biased against exports and implicitly biased includes those that have followed a fast track and against imports regardless of tariffs or other spe- have fully dismantled the vestiges of central plan- cific nontariff barriers. Its dominant force was ning. It indudes the CEECs and the three Baltic administrative rather than market-based. states. The second group includes the remaining Changes in trade policies since the collapse ECA transition economies. Michalopoulos and of central planning have taken two different Tarr (1994) further disaggregate the latter group tracks. In the first all spheres of foreign trade by distinguishing between countries that have were rapidly liberalized. Not only did not completed the reform process but have "advanced" reformers (that is, Hungary and already liberalized some areas pertinent to foreign Poland) take this path, but so too did some trade-Kyrgyz Republic, Moldova, and Russia- 40 FOREIGN TRADE IN THE TRANSITION from those that have yet significantly to liberalize have varied among the countries that have made foreign trade at all-Georgia, Turkmenistan, substantial progress but there have also been Ukraine, and Uzbekistan. The countries in the some important similarities: all quickly adopted first subgroup have liberalized trade and a unified exchange rate; all made their currencies exchange rate policies except for a few key com- convertible in the current account; all gave firms modity exportables, while those in the second full autonomy to operate in international mar- have retained extensive administrative controls kets; and all abolished remaining controls on and pervasive export restrictions, with foreign exports. They all introduced tariffi, rules of cus- trade organizations controlling most foreign toms valuation, and contingent protection pro- trade. cedures, which are more or less compatible with What sets the CEECs and the Baltics apart GATT standards, and tariffs and exchange rates from other ECA transition economies is their became the most important policy variables virtual absence of export controls and state affecting foreign trade. In addition, following a trading. While in the former tariffs and dramatic liberalization of their import regimes, exchange rates have become effective policy most of these countries subsequently increased tools directly influencing microeconomic deci- their tariff and, sometimes, nontariff protection. sions, in the latter their impact is significantly Turning to specifics, Poland removed all its weakened by direct controls and distorted export taxes and subsidies, and nearly all quanti- exchange rate regimes. These differences have tative import restrictions and suspended import created a deceptive trade liberalization asymme- tariffs in the second quarter of 1990 (Wellisz try. Countries in the "fast track" group, with the 1994). Czechoslovakia adopted a more evolu- possible exception of Estonia, tend now to have tionary course, with liberalization in 1991 mod- higher tariffs than those in the second group, erated by restrictions on access to convertible which have had low tariffs and no import currency markets and licenses on exports of licenses or quotas. As the CEECs and the Baltic essential inputs until 1992(Oblath 1993). By countries have removed the mechanisms that the end of 1991 Bulgaria and Romania had also shielded firms from international competition liberalized their foreign trade regime along simi- and as their initial real currency undervaluation lar lines to the former Czechoslovakia (Rodrik have been eroded, strong pressures for protec- 1992), although Romania temporarily main- tion have grown up from some domestic indus- tained some export controls. Hungary only tries. These pressures indicate that these coun- gradually (and reluctantly) liberalized its foreign tries are well advanced in their move toward a trade regime by dismantling foreign trade orga- market economy in that they reveal that domes- nizations and slowly decreasing the share of tic producers have felt the chill winds of compe- imports covered by licenses: licenses covered tition and have become sensitive to develop- about 50 percent of domestic output in ments in international markets. None of this, Hungary. however, amounts to saying that the NIS offer Taking into account their worse initial con- easier access to their markets than the CEECs ditions, the Baltic states' transition to a market- and Baltics. On the contrary, poor export per- based trade regime was particularly impressive. formance generates insufficient revenues for In a decisive liberalization in mid-1992, Estonia imports, and there are other barriers to entry dismantled virtually all export controls, than tariffs. refrained from introducing quantitative import The timing and pace of trade liberalization restrictions, and virtually abolished import POLICY AND TRADE PERFORMANCE 41 taxes.22 Latvia and Lithuania adopted a more mented faster and in the presence of less-con- gradual approach. Latvia replaced export quotas trolled prices, more autonomous (often private) and licenses with export taxes in mid-1992. By economic actors, and significantly larger macro- the end of 1993, the export tax rates had been economic disequilibria. In addition, all main- lowered significantly. In Lithuania a consider- tained controls designed to hinder exports able proportion of exportables remained subject throughout 1991: these included pervasive to the general excise tax at rates significantly export registration and licensing, foreign higher than those levied on products domestical- exchange surrender requirements, and taxes on ly consumed until mid-1993. In early-1994, hard currency earnings. What separated one for- however, export tax rates were slashed, as was the mer republic from another were the extent and number of products subject to tax (Sorsa pace at which these were either reduced, 1994b: 162). replaced with more transparent measures, or With the decrease in protection provided by fully abolished. undervalued domestic currencies (see next sec- Export controls and licenses are a remnant tion), liberalization in terms of low tariffs and of centralized trade and their use is a means of the absence of nontariff measures has turned out cross-subsidization and, sometimes, ad hoc state to be a short lived. First in Poland and later in interventionism to raise revenue. The trend has almost all other countries of this group (Oblath been to reduce the share of non-NIS trade han- 1993), an increase in protection followed the dled centrally. For instance, this share fell from liberalization. Its scope depended on the extent 40 percent of total non-NIS exports from Russia of the initial liberalization. Except for Estonia, in 1993 to around 20 percent in 1994 the countries that liberalized their foreign trade (Konovalov 1994:39); in Ukraine it decreased regimes most during the first stages of transition from 100 percent to 60 percent in early 1994 (for example, Poland) were the ones to make the (Le Gall 1994:74). In April 1993 the Moldovian largest reversals.23 Those countries that did not government scrapped the system of generalized open as abruptly experienced less significant export licensing (Walters 1994:174), as did the reversals but still increased protection of domes- Kyrgyz government in April 1994 (Krumm tic markets-see Csaba (1995) and Messerlin 1994:192). Although in Uzbekistan the number (1995) on Hungary. All the Baltic states had of products covered by export quotas was abolished essentially all import restrictions by reduced to twenty-six in January 1994, they still the end of 1991, but thereafter-as price-related accounted for the bulk of non-NIS exports border measures began to affect trade flows- (Connolly and Vatnick 1994:207). Latvia and Lithuania began to increase domestic In a scheme reminiscent of retention quotas protection. The Lithuanian import regime has in some CEECs, many NIS eased the foreign proven to be particularly vulnerable to protec- currency surrender requirement and opened for- tionist pressures: its provisions were changed ten eign exchange auctions. For example, Ukraine times between July 1993 and May 1994 (Sorsa reduced its surrender requirements from 100 1994b: 163). percent to 50 percent in September 1992; Moldova reduced its rate to 35 percent in Use of Export Controls December 1992. Uzbekistan introduced a 30 Although in some important respects the percent surrender requirement in April 1994. changes in the other NIS resembled the reforms The surrender requirement may be justified dur- made under central planning, they were imple- ing the early stages of a stabilization program in 42 FOREIGN TRADE IN THE TRANSMON terms of encouraging economic actors to shift to inate their ostensible liberalism toward imports, using domestic currencies in domestic transac- with the result that domestic firms feel little tions, but it has to be accompanied by current competition. What matters during the first account convertibility of domestic currency. As stages of the transition is not the average tariff long as there exist "black" foreign exchange mar- rate, but whether international competition real- ket premia, a surrender requirement is a tax on ly affects domestic producers. This requires mar- exports. For example, the three-fold difference ket-based foreign trade policy tools, which in between the official and market exchange rates turn depend on price signals operating domesti- in Ukraine in the second half of 1993 was an cally. In other words, in order to reap the bene- equivalent of "the levy of an additional exchange fits of a liberal foreign trade regime the internal tax on exports equal to two thirds of export economy must be reformed. earnings."24 The trend in the NIS is to curb disincentives to exports, but powerful restraints Changes in Domestic Institutional and Policy remain in many countries. Thus, although the This section considers domestic reform and market is starting to provide incentives to export because state-owned enterprises are achieving finds strong empirical evidence of the link because ~~~~~~~~between radical domestic reforms, which bring greater autonomy, the private sector is develop- X ing, soft budget constraints are hardening, and about macroeconomic stability, competitive domestic demand is contracting, the NIS' con- markets, and liberal foreign trade regimes, and tinuing export controls result in undervalued successful foreign trade reorientation. Reforms currencies and eliminate the pressure for other in one area reinforce reforms in other areas: protectionist measures. hardening enterprises' budget constraints, liber- This helps to explain why countries with alizing foreign trade, and making the domestic highly protected economies can appear to have currency convertible all play a vital role in sup- liberal import regimes. For example, Uzbekistan, porting export reorientation. Countries that where export restraints are common and state scored the highest in terms of trade reorienta- trading dominates (Michalopoulos and Tarr tion also recorded the greatest progress in 1994:9), abolished duties on hard currency reform. It is also shown that liberalizing the import regime is virtually irrelevant unless imports In 1993 (Connolly and Vatnick mareo mi fud enls re lo 1994:204), as did the Central Asian NIS (IMF acrec i d al 1994). In Ukraine-another slow reformer- addressed. most tariff rates lie in the range of 0 to 10 per- Settings of Good and Poor Trade Performers cent (Le Gall 1994). All ECA transition economies have moved Table 8, which reports data on ECA transition away from the "old" system. The CEECs and countries' progress in reforming their price the Baltics have foreign trade regimes firmly regimes, macroeconomic stances, exchange rates rooted in the market economy. Even though sev- and trade regimes, orders countries by export eral have recently increased their overall protec- performance. It is easy to see that good and bad tion, their export performance has been main- export performers differ in their reform progress. tained and their domestic producers exposure to The most readily visible feature relates to effective international competition has been price liberalization, suggesting that the elimina- much higher.25 The NIS, however, remain in tion of administrative control over prices is nec- transition. Their pervasive export controls dom- essary (albeit not sufficient) to break away from Table 8: Macroeconomic Developments and Change in Exchange Rate and Foreign Trade Regimes of European Transition Economies Macroeconomic Situation Exchange Rate System Trade Regime Inflation Ratio of PPP exchange (Consumer Pnce Index) Cumulative rate to nominal exchange rate change Current Surrender State Direct Price in GDP Domestic account requirement monopoly of exports Import liberalization 1990 1991 1992 1993 1994(es) 1989-94(f) currency convertibilty 1991 1992 1993 1994 (export tax) foreign trade controls controls FCSK F-1991 10 57 11 n/a n/a -31 n/a 1991 3.4 2.8 n/a n/a Jan-91 N.Sig.(90) 0 1 Czech Rep. 21 10 n/a 3.4 2.8 2.8 2.5 Slovakia 23 14 n/a 3.2 3.0 Poland F-1990 586 70 43 37 32 -8 n/a 1990 2.1 1.9 2.2 2.2 Jan-90 N.Sig.(89) 0 1 Estonia F-09/92 283 968 88 50 -17 06/92 1992 2.3 2.2 Jan-92 N.Sig.(92) 0 1 Hungary F-1991 29 35 23 23 19 -19 n/a 1991 1.9 2.0 1.8 1.6 N.Sig.(89) 0 1 Bulgaria F-1991 334 83 75 96 -28 n/a 1992 3.4 2.4 3.6 3.7 Jan-92 N.Sig.(90) 0 1 Rornania F-1992 175 211 256 180 -32 n/a 1993 2.7 3.6 2.5 2.4 N.Sig.(92) 1 1 Lithuania F-9/1992 216 1021 410 70 -57 10/92 1993 2.8 2.4 2.4 Jun-93 Lim. (91) 0 1 Latvia F-06/92 172 950 109 40 -48 05/92 1993 2.4 2.5 2.3 Feb-92 N.Sig.(91) 0 1 Tajikistan N/A 85 913 2351 349 -50 no date set not appl. 4.1 2.9 3.3 Extensive 1 0 Uzbekistan P-1993 106 599 885 1627 -20 11/93 L-1993 3.1 2.6 2.5 6/93(35%) Extensive 1 0 Belarus controlled 94 1016 1682 2324 -37 no date set L-1 993 2.3 2.3 yes-20% Extensive 1 0 Moldova P-1993 98 941 1576 1086 -49 07/93 no 3.0 2.7 yes-35% Moderate 1 0 Russia P-1992 1100 1100 323 -40 n/a 1993 1.3 9.7 2.2 2.0 yes-50% Moderate 1 0 Kazkhstan P-1993 91 885 1162 2202 -45 11/93 L-1994 2.8 2.4 2.6 yes-30% Moderate 1 0 Kyrgyz Rep. 107 906 1146 305 -50 05/93 1993 3.4 2.7 2.8 N.Sig. (94) Armenia 174 729 2260 8680 -60 3.1 3.2 0 Ukraine P-1993 84 1240 4730 891 -40 04/92 L-1993 1.6 11.5 2.0 2.1 2/92(20-70%) Extensive 1 1 Turkmenistan controlled 89 629 4271 3400 -41 01/93 no 3.2 Extensive 1 1 > z Azerbaijan 102 1063 981 1650 -51 01/93 3.6 3.0 3.4 Extensive 0 Georgia 768 11647 -86 04/93 2.9 3.0 Extensive 1 0 n.a. Not applicable; F - liberalization of prices on (almost) all tradables; P - partial liberalization of prices, i.e., not all tradables freed; L - limited. Note: The date indicates when a meaningful move to a "demand-constrained" economy with a market-related price structure occurred. For countries that launched their transformation programs before 1992, data on inflation and fiscal (im)balances cover the first two years of the program. For the NIS, data refer to 1992 (their first year of independent existence) and 0 1993. With a significant proportion of prices administratively controlled in 1992 in many NIS, their CPls tend to underreport the real scope of inflation. Sources: IMF: Annual Report, various issues, IMF: Trade Policy Reforms in the Countries of the former Soviet Union, IMF Economic Review, 2,1994; EBRD: Economic Outlook, September, 1993, PlanEcon, various issue, Michalopoulos and Tarr 1994, European Commission: Economic Reform Monitor, July/August 1994, Balcerowicz and Gelb (1994), national n statistics, Havlik er al. (1995), UNECE, Economic Survey of Europe, various annual publications. Exchange rate data for 1993-94 derived from World Development Indicators, World Bank Atlas (1995). 44 FOREIGN TRADE IN THE TRANsInON supply-constrained central planning. The coun- inflation in the second year of their programs tries with decent export performance all liberal- and subsequently kept it below 50 percent per ized prices. In Poland (1990), Bulgaria (1991), year. We interpret this as indicating a credible the former Czechoslovakia (1991) and the Baltic commitment to fight inflation. A corrective states (1991-92), prices were liberalized almost price adjustment during the first year of a at a stroke and price subsidies were cut drastical- reform program seems of litde relevance for the ly to achieve fiscal balance. Hungary stuck to its pace of export reorientation: indeed, a step traditional evolutionary approach, but still liber- increase in prices is frequently a necessary part of alized prices quite rapidly. Romania was slowest, stabilization and marketization, and can be a but at least on paper achieved liberalization by positive rather than a negative signal. Thus 1993. despite an almost five-fold increase in domestic All the poor export performers, by contrast, prices in 1990, Polish exports to OECD markets either retained administrative controls and/or increased by almost 50 percent while a 200 per- adopted a gradual approach to price reform. cent increase in prices in Romania in 1991 was Although the Russian government freed most accompanied by a 15 percent fall in its exports wholesale and retail prices in January 1992, cen- to OECD countries. But in 1991 inflation in tral controls were retained on energy, basic con- Poland fell to 60 percent while Romanian infla- sumer goods and services, many producer goods, tion increased to 210 percent in 1992 and 256 and the state distribution sectors. These controls percent in 1993. The improvement in stabiliza- were gradually removed throughout 1992 and tion with inflation falling to 61 percent in 1994 1993, but major exportables, such as energy and coincided with the surge in exports to OECD precious metals were still under state control. markets that moved Romania into the group of The experience of poor export performers good performers. Countries from the poor shows that using administrative controls, even export performance group experienced increases gradually eroding ones, to set the prices of trad- in inflation in the second year with the excep- able goods below world market prices under- tion of Russia, where hyperinflation fell but did mines adjustment and restructuring. Implicit not cease. subsidies deriving from distorted foreign Lax monetary and interest rate policies exchange allocation, corruption, and rent-seek- encourage the continuation of soft budget con- ing associated with export licenses proliferate in straints, which in turn undermine incentives to such an environment. Moreover, if the domestic shift production toward import substitutes and prices of tradable goods are held below world profitable exportables (Bruno 1994; Balcerowicz prices, there is a strong incentive to divert them and Gelb 1994; Michalopoulos and Tarr 1994). from domestic to foreign markets, and export Furthermore, inflationary expectations provide controls appear to become essential. The hybrid an extra incentive to hold hard currencies of quasi-market and quasi-administrative disci- instead of rapidly depreciating domestic curren- pline has a negative impact on export perfor- cies. Such "dollarization" of the economy exacer- mance and contributes to persistendy underval- bates shortages of convertible currency and con- ued local currencies, which increase the cost of tributes to the depreciation of domestic curren- imports and reduce competition in the economy. cy. These changes raise import prices and offer Another feature shared by countries falling extra protection to inefficient domestic produc- into the "good" export performers group is that ers. Under these circumstances, regardless of their stabilization policies significantly reduced how liberal an import regime is, the benefits of POLICY AND TRADE PERFORMANCE 45 trade liberalization in the form of competition ed with administrative distortions and the near- and better price signals fail to materialize. collapse of the domestic money and credit sys- Undervaluation generates high returns to tem. All these factors tend to prevent firms from exporting if firms can produce for export and producing goods for export, even if there are keep their receipts, but the weakness of import huge potential profits from doing so. competition and the unprofitability of nontrad- Even in less extreme cases a rising, if not ables curtails the development of effective sup- high, real exchange rate can be constructive. ply chains.26 Following Poland's sharp devaluation the state- The extent of undervaluation of the domes- owned enterprises, which held large reserves of tic currency can be represented crudely by the all important inputs, were the dominant ratio of the nominal market exchange rate to the exporters. They had better access to foreign purchasing power parity (PPP) rate.27 Following equipment-purchased by their distribution Balcerowicz and Gelb (1994), a ratio exceeding companies in the West; and social and econom- two is taken to indicate an undervaluation of the ic dogma encouraged the hoarding of labor. domestic currency for middle-income develop- Under these circumstances the state-owned ing countries. The lack of reliable data on PPP enterprises, despite their inefficiencies, were exchange rates makes it impossible to draw firm able to meet the increased demand for exports conclusions for the transitional economies, but a brought on by the strong depreciation. Winters pattern is evident in which good export per- and Wang (1994, Chapter 6) show, however, formers move quickly away from large underval- that in the Polish clothing sector the state- uation. For example, at first the nominal-PPP owned enterprise export boom subsided after ratio was higher for the former Czechoslovakia about a year, as the real exchange rate rose and than for many NIS, including those in the poor the (unreformed) state-owned enterprises performer group, but not for very long. became uncompetitive. This left room for small A commonly expressed fear was that real private exporters to expand and exploit the new appreciation in the transitional economies markets in inputs unencumbered by the ineffi- would curtail export growth and boost imports ciencies of the old system. Here the rising (but to unsustainable levels. There is some truth in not yet overvalued) real exchange rate was a this in the later stages of the transition-witness, source of pressure for reform. All told, there- for example, Hungary's recent difficulties-but fore, it appears that undervaluation is as much in the early stages it is not supported. Poland of a hazard as overvaluation in early transition. and Czechoslovakia, for example, both imple- Particularly in the second year of transition, the mented strong devaluations followed by real successful exporters in table 8 have lower (i.e. appreciations as inflation took off. But in nei- less undervalued) exchange rates than the less ther case did export growth suffer unduly. The successful ones. essential point is that even if mild undervalua- If the artificial price (measured in units of tion boosts exports by increasing profitability local money) of foreign currency is much higher and even if this boost has very long-lived conse- than market clearing levels, the authorities are quences via hysterisis effects, extreme underval- sometimes tempted to appropriate some of the uation is different. Extreme undervaluation rents from exporting by imposing export taxes weakens the domestic production sector by hin- or a foreign currency surrender requirement. dering non-tradables industries and protecting Except for Tajikistan and Lithuania, where poli- import competing industries. It is often associat- cy vacillated from the abolition of the surrender 46 FOREIGN TRADE IN THE TRANSITION requirement (April 1992) to partial reintroduc- over both exports and imports do not produce tion (in February 1993, but only on state enter- the expected outcome in countries where stabi- prise earnings) to final abolition in June 1993, lization has either not been undertaken (such as successfil exporters did not resort to such poli- Ukraine and Turkmenistan) or has been spo- cies. Similarly, the successfiul exporters, as well as radic and incomplete (such as Romania until some unsuccessful ones, generally had unified 1993, Russia, and Moldova), and where the exchange rates and fairly free currency convert- state has retained direct, albeit informal, controls ibility for current account transactions. over distribution channels. In these countries A further difference between "good" and either the incomplete liberalization of prices or "bad" export performers is the extent of state export controls, which held domestic prices trading. The state monopoly of foreign trade below world prices, compounded the shortages was formally terminated in all countries, and of hard currency earnings stemming from weak yet, in the poor-performing countries foreign export performance, which, in turn, restrained commercial transactions for a vast number of import purchases. products were either direcdy or indirectly con- The common denominator among the top trolled by the state. These state controls were not export performers is price decontrol combined necessarily limited to trade with other NIS but with strong and credible government commit- extended to other markets as well. Thus, while ment to macroeconomic stabilization and export licenses and quotas were hardly used by broadly conceived liberalization. While resource the CEECs and the Baltic states, they were endowments, past patterns of production, and extensively used by other NIS creating large the ease of adjustment in different industries implicit taxes on some key exports. determine the trade patterns of a country, Trade reforms followed different paths in progress in macroeconomic stabilization and in different transition economies, but overall the establishing market-supporting institutions was experience shows that trade liberalization is not perhaps the single most important factor deter- sufficient to expand exports. We have already mining foreign trade performance over the tran- seen that poor performers appear to have more sitional period. Countries in the poor performer liberal import regimes than good performers, group, despite progress in other areas of trans- and the experience in several countries shows formation (such as privatization in Russia), all that relaxing trade controls alone does not gen- display macroeconomic chaos and a vacillating erate export growth. Again, macroeconomic and microeconomic reform program that sets them institutional environments are linked. Controls apart from the successful group.28 6 The Impact of Transition on Developing Countries' Trade Under central planning foreign trade was regarded as an evil necessary to obtain unavailable inputs and technolo- gies. Although chronic domestic shortfalls, stagnation, and decline in the 1980s pushed the CMEA countries toward openness, they neither made the effort nor succeeded in integrating into the world economy. Thus although the Soviet Union significantly expanded its exports of oil and natural gas following the 1973 oil shock, the CMEA coun- tries remained marginal players in world trade. With the Markets demise of the Soviet economic bloc, this situation changed, raising fears that the diversion of intra-CMEA traded prod- * Primary Commodities ucts to world markets would disturb existing trade flows, Markets especially if, simultaneously, those countries' access to OECD markets was improved. This section focuses on two aspects of the foreign trade impact of the emergence of transition economies. First, it considers the extent to which, given their newly granted GSP status and the Interim Trade Agreements, exports from transition economies to EU markets have displaced those from other developing countries, especially the African, Caribbean, and Pacific (ACP) countries. This analysis deals only with manufactures because of the limits on preferential access in agricultural products for both 48 FORPIGN TRADE IN THE TRANSITION Lome Convention and Europe Agreements labor-intensive, lightly processed manufactured exporters. Second, it considers the impact of the goods in which the CEECs and some NIS share collapse of the Soviet Union, a major exporter comparative advantage with developing coun- and importer of many nonrenewable natural tries. Within this category, the largest increases resources, on primary commodity markets.29 in CEEC exports to EU markets have been in The approach we adopt is to examine textiles, clothing, and footwear, which jointly briefly changes in trade and trade shares since accounted for around 20 percent of their EU- the mid-1980s to see if there is prima facie evi- directed exports in the early 1990s. Since the dence of major impact from the transition on degree of protection for these products of the other countries. A full analysis of the welfare EU is relatively high, changes in preferential effects and of what developing countries' trade access may have a great impact. Having exam- would have been in the absence of transition is a ined imports of these products, however, we find much larger task and must be held over to little evidence of trade displacement triggered by another occasion. Our preliminary conclusion is the normalization of East-West trade relations. that the demise of the CMEA and the Soviet The most likely case of displacement is in Union had a more limited impact than one textiles. Exports of textiles and clothing from might have expected on the basis of past perfor- developing countries are determined within the mance. framework of the MFA, based on quotas deter- mined at the product level. The European tran- Trade Displacement in EU Markets sition economies' shares of total EU quotas were The trade liberalizing measures embodied in the raised unilaterally from between 3 percent and 5 Europe Agreements raised concerns that exports percent in 1989 to between 4 percent and 7 per- from the CEECs would displace those from cent in 1992 (Cadot and de Melo 1994:14). It is developing countries. The granting of GSP sta- not clear that this precluded increases for other tus eliminated the preferential tariff margins that suppliers; rather it probably just increased aggre- most developing countries had with respect to gate EU imports. the European transition economies in EU mar- Turning to total EU textile imports (rather kets. Although the Europe Agreements and their than just those under MFA quantitative restric- precursor Interim Trade Agreements currently tions), the share of the European transition offer relatively little preference over the GSP, economies in imports from low-cost producers other measures-such as increases in textile and fell from 17 percent in 1981-82 to 11 percent in clothing quotas (regular quota as well as out- 1990, and then increased to 15 percent in 1992, ward-processing-traffic quota for clothing)- and to 19 percent in 1994 (table 9). Both appeared to pose a significant threat to develop- China's and Hong Kong's share contracted sig- ing country market shares. nificantly: China's from around 20 percent in Our analysis suggests that fears that the 1985-89 to 12 percent in 1993-94, and Hong transition economies would displace developing Kong's from 3 percent to 1 percent. The fall in countries have been exaggerated. The number of the share of ACP countries from around 10 per- manufactured products defined in terms of cent in the late 1980s to 5 percent in 1993-94 eight-digit Harmonized System (HS) items in was quite substantial, although it would be dif- which ACP countries compete directly with the ficult to attribute this fall fully to the European transition economies is relatively small. The transition economies because it follows an notential for trade displacement is largest for apparent trend decline since 1984. It is not able 9: EU Textils Imports from Developing Countries and Transition Economies 1980-94 n million ECU; percentage share in parenthesis) Xowby 190 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 ,EECs 190(7) 175(6) 195(7) 236(7) 318 (8) 345 (8) 336 (9) 304(7) 297 (7) 311 (7) 330 (6) 374 (7) 462(10) 456 (7) 590 (8) -SU 208 (8) 310 (11) 297 (10) 231 (7) 151 (4) 136 (3) 190 (5) 266 (6) 273 (6) 330 (7) 281 (5) 343 (7) 223 (5) 575 (9) 862 (11) Zhria 363(13) 406(15) 512(18) 582(18) 708(17) 873(21) 725(19) 778(19) 858 (19) 1,064(22) 830(16) 783(15) 753(16) 770(12) 909(12) HongKong 99(4) 112(4) 110(4) 132(4) 135(3) 133 (3) 127(3) 123(3) 128(3) 126(3) 128(2) 123(2) 92(2) 61(1) 47(1) ACP Countries 239 (9) 253 (9) 231 (8) 337 (10) 535 (13) 493 (12) 464 (12) 434 (10) 458 (10) 504 (11) 505 (10) 454 (9) 362 (8) 331 (5) 364 (5) India 305(11) 309(11) 297(10) 365(11) 424(10) 410(10) 372 (10) 530(13) 573 (13) 601 (13) 691 (13) 742 (14) 748 (16) 855(13) 913 (12) OtherLDCs 1,303(48) 1,134(42) 1,191 (42) 1,405(43) 1,838(45) 1,811 (43) 1,656 (43) 1,727(41) 1,813(41) 1,821 (38) 2,472(47) 2,353 (45) 2,047(44) 3,378(53) 3,961 (52) Total 2,706 2,699 2,832 3,288 4,108 4,201 3,870 4,163 4,401 4,756 5,238 5,173 4,688 6,427 7,646 Source: Dcrived from EUROSTAT data. Table 10: EU Clothing Imports from Developing Countries and Transition Economies, 1980-94 H (in million ECU; percentage share in parenthesis) Country 1920 1981 1982 1963 1984 1985 1986 1967 1988 1989 1990 1991 1992 1993 1994 5 0 z CEECs 194(19) 168(12) 174 (13) 188 (13) 182(11) 185(11) 202 (10) 161 (8) 209 (6) 208(6) 254(7) 374 (8) 500(10) 614 (8) 627 (9) 0 FSU 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 1 (0) 2 (0) 1 (0) 31 (0) 45 (1) China 31 (3) 52 (4) 55 (4) 66(5) 86 (5) 108 (6) 153 (6) 172 (9) 303 (9) 319 (9) 458 (12) 825 (19) 980 (19) 1,195 (16) 1,176 (17) m Hong Kong 339(34) 449(32) 420(30) 466(32) 548 (33) 536(32) 598 (30) 540(28) 806(25) 805(23) 810(22) 990(20) 1,005 (19) 1,005 (13) 847(12) c ACP 29(3) 58(4) 51 (4) 58(4) 80(5) 105(6) 136 (7) 100 (5) 220(7) 239 (7) 279(7) 328 (7) 395(8) 420(6) 346(5) z India 11 (1) 16(1) 20(1) 21 (1) 27(2) 31 (2) 36(2) 28(1) 112 (3) 140(4) 196 (5) 222 (5) 294 (6) 366 (5) 372 (5) ^ Other LDCs 407 (40) 64 (47) 668 (48) 659 (45) 759 (45) 727 (43) 866 (44) 924 (48) 1,589 (49) 1,735 (50) 1,749 (47) 2,168 (44) 2,079 (40) 3,887 (52) 3,592 (51) ° Total 1,010 1,391 1,389 1,458 1,682 1,693 1,990 1,926 3,240 3,445 3,746 4,910 5,256 7,518 7,007 Source: Derived from EUROSTAT data. Hn 50 FOREIGN TRADE IN THE TRANSITION inconceivable, however, that the surge in FSU ing as other developing countries, the latter's exports in 1993-94 might have triggered some shares of EU markets should fall. That does not displacement of ACP goods. On the other hand indicate an adverse impact, however, if total other LDCs probably contributed more to imports grow sufficiently to allow all exporting crowding out exporters from ACP countries countries to increase their sales absolutely. In than European transition economies; they regis- textiles and clothing this appears to have been tered the largest increase in exports to the EU in the case, while in footwear at least part of the the 1990s. observed pattern arises from China and the Turning to clothing, we found no evidence CEECs moving into the markets forcibly vacat- of negative impact on ACP exports (table 10) ed by Korea and Taiwan under their Voluntary except perhaps to the (small) extent that Export Restraints with the EU. It would seem, between 1989 and 1992 the CEECs took up therefore, that there is little evidence that the the shares vacated by Hong Kong and the reemergence of the former CMEA economies NICs, and thus prevented the ACP countries has adversely affected developing countries' sales from doing so. However, CEECs' share in the EU. By extension, because they have remained flat in 1993-94, while that of the offered no greater preferential access than the ACP countries fell from 6 to 5 percent. Given EU, we might reasonably conjecture that such the strong growth in exports from other low- displacement effects are also absent in other cost suppliers (other LDCs), however, it is diffi- OECD countries. cult to see the ACP countries benefiting much from the NICs' decline in this sector even in the Primary Commodities Markets absence of the CEECs. The surge in EU import The Soviet Union was an important player in demand for footwear (by 44 percent in 1993 many of the markets for primary commodities, alone) was mainly taken by exporters from both as a buyer and a seller. Many of the NIS other LDCs, whose share increased from 40 have increased their world market shares in percent in 1992 to more than 50 percent in many commodities. Even though production 1993-94 (table 10). levels have sometimes fallen in the chaos sur- The increase in footwear exports from the rounding independence, the collapse of domes- CEECs, from around US$420 million in the tic demand has reduced consumption of indus- late 1980s to US$691 million in 1994, did not trial inputs, leading to cuts in imports and/or appear to have any impact on ACP countries. increased surpluses directed to international The value of the latter's exports increased five markets. In some products, the FSU has turned times, albeit from a very low level (table 11). from being a net importer to being a net The CEECs' share of total imports increased by exporter. In addition, the dissolution of the 10 percentage points, from 11 percent in 1988- Soviet Union has also had an indirect influence 89 to 21 percent in 1992, but fell subsequently on the prices of certain mineral fuels and non- to 17 percent in 1994. The share of other LDCs ferrous metals, as the end of Cold War has led to dropped from 68 percent to 40 percent. With its cutbacks in military outlays and hence in the share increasing from 10 to 29 percent, China demand for the minerals used in military pro- appeared to be the major driving force behind duction. the change in EU imports of footwear. The most important primary market is the It is to be expected that, as the transition market for oil. According to some analysts,30 economies are put on the same competitive foot- without the contraction in oil exports from the rable 11: EU Footwear Imports from Developing Countries and Transition Economies, 1980-94 ,in million ECU; percentage share in parenthesis) 'ount8y 19t0 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 _EECs 143(25) 130(20) 132(21) 144(19) 191 (23) 222(24) 221 (22) 215(16) 195(11) 212(11) 274(13) 388(14) 535(21) 604(15) 691 (17) FSU 0 (0) 1 (0) 1 (0) 2 (0) 2 (0) 2 (0) 3 (0) 2 (0) 3 (0) 5 (0) 7 (0) 17 (1) 7 (0) 42 (1) 44 91) China 32 (6) 49 (8) 50 (8) 61 (8) 77 (9) 80 (9) 79 (8) 130 (9) 172 (10) 224 (12) 267 (13) 526 (19) 728 (29) 887 (22) 875 (22) HorigKonfg 43(8) 58(9) 56(9) 5497) 64(8) 59(6) 51 (5) 54(4) 48(3) 39(2) 32(2) 33(1) 30(1) 54(1) 53(1) ACP 1 (0) 2 (0) 2 (0) 2 (0) 1 (0) 1 (0(0 ) (0) 1 (0) 2 (0) 2 (0) 3 (0) 8 (0) 8 (0) 8 (0) 10 (0) Inrdia 38(7) 45(7) 46(7) 44(6) 62(7) 74(8) 69(7) 110(8) 125(7) 135(7) 188(9) 215(8) 203(8) 263(7) 297(7) Other LDCs 307(54) 364(56) 352(55) 435(59) 432(52) 494(53) 561 (57) 856(63) 1,205 (69) 1,303(68) 1,332 (63) 1,581 (57) 1,002(40) 2,140(54) 2,045(51) i Total 563 649 640 741 828 932 983 1,367 1,750 1,922 2,103 2,768 2,513 3,999 4.015 H Source: Derived from EUROSTAT database. C z 0 z tn 52 FoREIGN TRADE IN THE TRANSInON Box 3: Trade Distortions; The Case of Aluminum Exports The developments in aluminum exports, mainly originating in Russia, show in a nutshell peculiarities of the transition process, the extent to which international markets are managed (despite a free-trade rhetoric), and the threats associated with too fast an increase in exports. Distorted incentives. The incomplete liberalization of prices and of foreign trade regime, compound- ed by the collapse of former interrepublic links and falling domestic demand, have contributed substan- tially to increased exports of aluminum from the former Soviet Union. This increase took place despite the contraction in domestic output. Although the FSU may have a comparative advantage in aluminum production, the export expansion was driven largely by unique circumstances that will disappear as the centrally planned regions are further dismantled. Subsidized transportation costs, bank credits, and energy made producers from FSU competitive in international markets. Furthermore, inflation has pro- vided an extra incentive to maximize earnings in hard currencies (rather than in rapidly depreciating domestic currencies) and, therefore, to divert aluminum from domestic markets. Despite depressed domestic demand, there were still shortages of the aluminum. For example, in Russia-the major exporter of aluminum from the FSU-producers of caviar (an attractive exportable) complained that there was not enough aluminum available for canning. Representatives of the Ukrainian fish canning industry voiced similar complaints. Surge in exports, mainly to the EU. Russian exports of aluminum increased from 300,000 tons in 1990 to 1.6 million tons in 1993. Despite falling world prices for aluminum, the value of exports increased by 65 percent and their share in world imports rose from 2.3 percent to 4 percent, with the EU absorbing most of this increase. The share of the EU in aluminum exports from the FSU rose from 29 percent to 59 percent. The main item in this export expansion was aluminum ingots (SITC. Rev.2. 6841). Between 1990 and 1992 the FSU's share in world imports more than doubled from, 3.8 percent to 8 percent and in EU imports it increased from 3 percent to almost 12 percent. The volume of EU imports of aluminum from the FSU increased almost fivefold-from about 125,000 tons in 1990 to 582,000 tons in 1992. As a result, the share of producers from the FSU rose from almost nothing in the 1980s to an estimated 9 per- cent of EU consumption of aluminum in 1992. This forced OECD producers, already affected by surplus capacity, to cut output by around 1 million tons, with the EU industry absorbing 80 percent of this cut. Source of competitiveness. Despite the technological obsolescence of most smelters in the FSU, their production costs were far below (around 42 percent) those of the most efficient producers in the world: Australia, Canada, and Venezuela. It is estimated that in mid-1992 the cost of producing 1 ton of aluminum in Russian smelters was about US$550. Since energy is the most important component of alu- minum production costs, as the gap between world and domestic energy prices in successor states is closing, many FSU producers have been losing their competitive advantage. It is estimated that the recently implemented 183 percent increase in energy prices in Russia (October, 1993) will raise costs to US$1,200 a tons, which is above the mid-October prices at the London Metal Exchange and on par with at the costs of many EU producers. Last March, according to Westem aluminum experts, the pro- duction cost in Russia moved above those in OECD countries. Thanks to the international agreement negotiated with Russian producers, the world price went up as well. Note that aluminum exports reflect the movement of prices of energy to world levels. The costs of aluminum producbon in Russia rose from 3 percent of the world average in 1990 to 60 percent by mid-1993. THE IMPACT OF TRANS1TION ON DEVELOPING COuNTREs' TRADE 53 FSU, oil prices would have been 10 to 15 per- around 3 million tons of annual trade was cent lower in 1991, if OPEC sales had bound to depress prices. Moreover, FSU remained unchanged. Indeed, Soviet exports exporters offered zinc at a discount of up to declined between 1990 (third quarter) and US$85 per ton to the average London Metal 1991 (third quarter) from 2.1 million barriers Exchange (LME) price of US$1,240 in 1992. per day to 800 thousand barriers per day,31 Although other zinc producers, especially from while those from OPEC increased by around 1 Australia and Canada, slashed their output, their million barriers per day in the same period. cutbacks were not sufficient to prevent prices Although Russian oil exports subsequently from falling until December 1993. At that time increased to about 1.3 million barriers per day, the trend was reversed in response to a meeting a shortfall of about 800 thousand barriers per of European producers scheduled to discuss pro- day relative to 1990 remained. This simple posals for coordinated output cuts. comparison does not necessarily imply that A similar story pertains to the aluminum changes in Soviet exports kept oil prices from glut and the financial difficulties faced by many falling-the comparison period included the aluminum producers in the world. With the price rise between the Gulf War and declines in decline in world import demand-triggered by oil stocks. It does, however, show that the tran- the economic slump in OECD economies- the sition had both positive and negative effects on surge in aluminum exports from the NIS has commodity prices. produced a huge world imbalance of supply and The transition had a more directly disrup- demand and brought the EU industry to the tive impact on markets for nonferrous metals, verge of collapse. EU producers demanded that supplementing cyclically weak world demand imports from the NIS be restricted to 80,000 with a simultaneous contraction in FSU imports tons-claimed to be the "normal" level. The and the emergence of export surpluses. These final quota has not yet been decided, but in the surpluses arose from the use of spare FSU meantime the European Commission has limit- smelter capacity to process imported concen- ed imports to 60,000 tons, angering U.S. alu- trates into metals, which were then shipped to minum producers fearful that exports from NIS world markets.32 Because some of these mar- would be diverted to their markets. The falling kets handle only residual trade after fixed-price prices prompted major producing countries to arrangements have covered most transactions, negotiate an unprecedented agreement limiting even small changes in net supply may have a output. The representatives of the Russian alu- huge price impact. An example is zinc, whose minum industry, threatened with antidumping London Metal Exchange price dropped by 20 measures and lower quotas and lured by a percent in the last quarter of 1992 compared promise of a financial package to assist them in with average prices during the first three quar- restructuring and diversifying the industry, ters. decided to participate in the agreement. The Although the FSU was the world's largest "memorandum of understanding" about pro- producer and the world's second largest con- duction cuts was approved in February 1994. sumer of zinc, it was a net importer until the late Developments leading to the post-Soviet "alu- 1980s. Net imports peaked in 1988 at around minum shock" in world markets shed some 32,000 tons, but by 1992 net exports reached extra light on links between export perfor- more than 100,000 tons. A net change of more mance and the transition from central planning than 132,000 tons in a stagnant market with (box 3). 54 FOREIGN TRADE IN THE TRANSITION In sum, the collapse of the CMEA and the ments originating in the FSU. The contraction Soviet Union had a more limited impact on in domestic supply-triggered by the plunder- developing country trade flows than might have ous exploitation of natural resources under been expected given past performance and communism and the collapse of interstate changes in competitiveness. Apart from zinc trade-weakened the growth in primary com- and aluminum, the redirection of Russian pri- modity net exports from the FSU. Exports of mary commodities from their previous man- primary commodities from the CEECs have aged markets to the OECD economies contracted mainly because output fell when depressed international prices very little in subsidies were cut. Thus, overall, while particu- 1990-92. Even in the cases of non-ferrous met- lar markets have been depressed by the reform als except aluminum the declines in price were process, the transition has not been a major mainly a result of recession in many OECD contributor to weak commodity prices in the countries rather than a large increase in ship- early 1990s. 7 Conclusion Under central planning foreign trade did not effectively transmit international efficiency standards into the transi- tional economies. Moreover, the soft payments arrange- ments in the CMEA contributed to the development of productive structures at variance with countries' factor endowments and comparative advantage. No matter how radical the reform measures were over the 1980s, foreign trade regimes remained a source of distortions and ineffi- ciencies throughout the CMEA region, with domestic pro- ducers isolated from changes in relative prices in world markets. While the reforms encouraged the development of some marketing expertise in some enterprises and provided incentives to boost exports, they failed to introduce the crit- ical discipline of the market and boost exports to OECD markets. Thus the export performance of the CMEA economies was very unimpressive in the 1980s, as shown by the decline in their competitive position in OECD mar- kets, and the shift in commodity composition of exports toward low value-added, resource-intensive products. The distortions in trade patterns inherited from central planning, the Soviet Union's inter-republic division of labor, and, to a lesser extent, the CMEA were very signifi- cant. Calculations based on the gravity model suggest that 56 FOREIGN TRADE IN THE TRANSITION the NIS and the CEECs overtraded with each more immediate priority than relaxing import other and significantly undertraded with OECD controls. This is because extensive export con- countries. These findings suggest that the ECA trols designed to keep subsidized products at transition economies may eventually become home induce undervaluation, which, in turn, important exporters and importers, and that a offers high levels of protection and effectively dramatic expansion of trade with OECD coun- insulates domestic producers and consumers tries is likely to take place in the long run. This from international markets. Exchange rate expansion will depend on removing distortions appreciation appears to have little impact on the to trade in the transition economies and on foreign trade performance of firms in the first nondiscriminatory access to OECD markets. stages of the transition and privatization is not Turning to the shorter run, ECA transition necessary if there is, instead, a significant hard- economies have already experienced dramatic ening of budget constraints for state-owned shifts in their foreign trade patterns in a very enterprises. In the long run, however, realistic short time and have already become relatively exchange rates and privatization are necessary to important partners for EU countries. Differences assure an efficient response to improved market in initial conditions provide some explanation signals and open trade. In brief, trade reforms for these differences in performance. The NIS are largely ineffective without macroeconomic (except Russia) lacked commercial contacts in stabilization, the liberalization of prices, and a OECD markets and have functioned under cen- suitable exchange rate in the short run. In the tral planning for about seventy years (fifty for the long run, enterprise reform is necessary. Baltic states); some of the CEECs, however, had It is also true that trade liberalization signif- experimented with market-oriented economic icantly aids price liberalization and stabilization. reforms before the collapse of central planning It injects competition, reduces the power of local and had been increasing their exports to OECD monopoly, establishes standards of comparison, countries since 1986. But initial conditions are and introduces new technology. Thus, to reiter- not a sufficient explanation: the Baltic states are ate, our real policy message is that reform must among the top performers in terms of export proceed on both domestic and foreign fronts performance despite their earlier dependence on simultaneously. Soviet markets and lack of easily marketable raw Although the collapse of central planning, materials. The removal of discriminatory measures the CMEA, and the Soviet Union triggered a in OECD markets also fails to explain the pace of realignment of trade patterns, fears of major dis- the trade reorientation because market access ruptions in world trade for primary commodi- improved for all countries, but trade performance ties and labor-intensive, lightly processed manu- did not. Thus, the explanation must lie elsewhere. factures have, so far, turned out to be exaggerat- This study suggests several factors that con- ed. Apart from trade in zinc and aluminum, the tribute to a successful foreign trade reorienta- demise of the FSU did not have very large effects tion. By far the most important to early success on primary commodity markets. Similarly, is that stabilization and liberalization reform examining EU imports of labor-intensive prod- tracks be pursued simultaneously-liberalizing ucts-particularly textiles and clothing and the foreign trade regime alone will not improve footwear, which are especially sensitive to trade performance unless domestic prices are changes in preferential access-we found little or decontrolled and inflation expectations lowered. no evidence of trade displacement triggered by Moreover, removing export controls appears a the normalization of East-West trade relations. Nors 57 Notes stantial shortfall in their trade witl market economies. The authors are grateful to Alan Gelb and the Transition Their products were mainly sold in either domestic (that is, Economics Division of the World Bank for proposing this Soviet) or developing-country markets. paper and to Stijn Claessens, Alan Gelb, Martha de Melo, Costas Michalopoulos, Martin Schrenk, David Tarr, 9. In the long term, one should expect a significant Ulrich Zachau and an anonymous referee for their very increase in trade, at least between the NlSs from Central helpful comments. We are also indebted to Minerva Patefia Asia and China. The Central Asian states are potentially for logistical support. important suppliers of energy and other raw materials. However, in the near term this trade potential can not be 1. Before its dissolution in 1991, the CMEA had eleven fully realized as both sides lack capital and transport net- full members: Albania, Bulgaria, Cuba, Czechoslovakia, works, and barter arrangements seem to have reached their German Democratic Republic, Hungary, Mongolia, potential. Poland, Romania, Vietnam, and the USSR Albania joined 10 the CMEA in 1949, but since 1961 had not participated in 0. Ths temporary improvement was mainly a result of a CMEA activity nor had it paid its dues (Staar 1988:292). one-time increase in Romanian exports across all major Yugoslavia had "limited participant" status. Three coun- product categories. However, subsequent years witnessed a tries-Finland (1973), Mexico (1975), and Iraq (1976)- dramatc contraction in Romanian exports. ratified cooperative status agreements with the CMEA. Since 1978 the observer status group induded: 11. Some countries fared slighty betterinsEComarkets: the Afghanistan, Angola, Kampuchea, Ethiopia, Laos, dedine in the EC import share of Czechoslovakia, Poland, Mozambique, Nicaragua, Democratic People's Republic of and Romania was lower than for other OECD markets. Korea, and Yemen Arab Republic. Not all of them For more details see Kaninski (1993b). remained observers until the dissolution of the organiza- 12. Kaninski (1993b), Landesmann (1995), and Padoan and Pericoli (1993) also suggest a shift toward labor-inten- 2. These include three Baltic states, Estonia, Larvia, and sive products, while Hoekman and Pohl (1995) have also Lithuania; five Central Asian states, Kazakhstan, Kyrghyz found evidence of composition changes in the CEECs Republic, Tajikistan, Turkmenistan, and Uzbekistan; four export bundles. European states, Belarus, Moldova, Russia, and Ukraine; 13 Given the collapse of East German industrial compet- and three Caucasian states, Armenia, Azerbaijan, and 13GietcolpofaGrmnduracme- Georgiae itiveness because of the sudden increase in labor costs after unification, producers from other former CMEA countries 3. Bulgaria, former Czechoslovakia, Hungary, Poland, and may have successfully increased their exports. Without a Romania. Albania should also strictly be included in this detailed data analysis (not readily available because the group,mbut weArarelydhave datafor it. German Democratic Republic enjoyed a de facto EC member status), it is impossible to assess the extent to 4. For an extensive discussion of similarities and differences which exports from CEECs substituted for East German between "socialist reforms" in Europe and China, see Gelb, products. Jefferson and Singh (1993). Jefferson and Rawski (1994) 14. Note that the value of earlier exports to the FSU was offer an explanation why China has been so successful even 14. overtate (and, hence, sorts the contras though institutional conditions regarded as necessary for probably overstated (and, hence, so was their contraction economic expansion were lacking. Sachs (1995) shows why in 1990) because of the use of different implict crossrates Chinese experience cannot be emulated by Eastern Europe between the TR and the US dollar in the period under and the former Soviet Union. consideration. In 1990 the crossrates used in both Hungary and Poland were increased significantly, thus depressing 5. Tarr (1992) observes that the dynamics of the bargain- the dollar value of exports to the FSU. For an extensive dis- ing process in the CMEA discouraged quality improve- cussion of statistical problems involved see Economic . , . . . , ~~~~~Bulletin for Europe 1993 and Winters and Wang (1994). ment, since one had difficulty obtaining a better price for r E quality in the negotiations. 15. Available data do not allow us to check whether this 6. Many argue that this had a significant impact on the increase was simply due to re-exports of chemicals orignat- pace of trade reorientation towards the EU. See, for exam- mg m othcr NIS. pIe, Baldwin (1 994) and Landesmann (1 995). 16. In an earlier draft of this paper we calculated the export 7. For more details about the model and its limitations, see performance index for 1993. It gave roughly similar results Winters and Wang 1994. to 1994, but placed Romania and Tajikistan in the "poor" group and Uzbekistan in the "good" one. Uzbekistan's 8. The high level of industrialization of the Baltics relative "success" was due to booming gold and cotton exports to tO other former Soviet republics accounted for a very sub- OECD, but it proved short-lived because it was not sup- ported by significant reform. (Connolly and Vatnik, 1994) 58 FOREIGN TRADE IN THE TRANSITION 17. One exception was Hungary, which was admitted to 26. See Balcerowicz and Gelb (1994) and Yotopoulos the GATT as a market economy despite its centrally (1995) for a discussion of the costs to development of planned economic system at the time of accession. chronic undervaluation. 18. Sensitive products indude agriculture, steel, textiles, 27. The discrepancy between the nominal and the pur- and dothing. chasing power parity exchange rate is indicative of trade barriers, transportation costs, distortions in prices due to 19. An important caveat is that many of these imports are monopolistic and oligopolistic practices, and different subject to GSP preferential rates up to a certain limit and prices on nontraded goods and services. Nominal and PPP above that to MFN rates. As a result, their significance is exchange rates usually diverge even for highly developed overstated. market economies. The difference tends to be greater for countries with a lower GDP per capita. 20. Because of its special political status in the FSU and its traditional status of a supplier of raw materials to other 28. In quite separate exercises Gelb, de Melo and Denitzer NIS and the CEECs, Russia was better off than the other (1995) calculate an index of liberalization, which correlates NIS. well with our export performance index, and Michalopoulos and Tarr (forthcoming) find that the 21. For a more comprehensive discussion, see growth of exports to OECD countries and the growth of Michalopoulos and Tarr (1994) and IMF (1994). total output (both indicative of good policies) are also strongly correlated across countries. 22. However, for fiscal reasons and in part to reduce illegal activities, the Estonian government retained trading 29. The Soviet Union was a major producer and exporter monopoly over lucrative metals exports. See Hansen and of several major primary commodities including oil, gas, Sorsa (1994:123). and ferrous and nonferrous metals. Several economies, not only those from the now defunct CMEA but also from the 23. For example, in early 1991 the Polish government, developing world (such as India), were heavily dependent faced with a rapidly deteriorating external position and on supplies of energy and raw materials from the FSU. On continued inflation, decided to strike a balance between the demand side, the Soviet Union was a major importer of the competing objectives of depressing imports and lower- grains, sugar, tea, coffee, and cocoa. It regularly purchased ing inflation by limited devaluation and increasing tariffs. around one-eighth or more of world grain exports. This policy of increasing protection of domestic producers, recendy "enriched" by introducing variable levies on agri- 30. See "Economic Trends," Business Week, 6 July 1992. cultural products, is not sustainable over a longer period (Wellisz 1994:22-24). 31. Data from Russian Statistical Office (Goskomstat) as quoted in Russian Economic Monitor, Vol. 2, No.1, 24. Pynzenyk (1994) quoted in Havrylyshyn, Miller, and January 1993. Perraudin (1994:356). 32. For more details, see Oxford Analytica, 22 February, 25. They are still reasonably liberal, however, relative not 1992 and Mooney (1993). only to their past but also relative to developing countries outside East Asia and Latin America. REFERENCES 59 References Hamilton, C., and L.A. Winters. 1992. "Opening-Up International Trade with Eastern Europe," Economic Balcerowicz, L. and A. Gelb. 1994. 'Macropolicies in Policy. Volume 14, April. Transition to a Market Economy: A Three-Year Perspective." paper presented at the Bank's Annual Hansen, J., and P. Sorsa. 1994. "Estonia: A Shining Star Conference on Development Economics, from the Baltics." in C. Michalopoulos and D. Tarr, Washington, D.C., April 28-29. eds. Trade in the New Independent States. World Bank/UNDP, Washington, D.C. Baldwin, R. 1994. Towards an Integrated Europe. Centre for Economic Policy Research, London. Harmsen, R. 1995. "The Uruguay Round: A Boon for the World Economy." Finance d Development. March. Bruno, M. 1994. "Stabilization and Reform in Eastern Europe: A Preliminary Evaluation." In O.J. Havrylyshyn, O., M. Miller, and W. Perraudin. 1994. Blanchard, K.A. Frost, and J.D. Sachs, eds., The "Deficit, Inflation and the Political Economy of Transition in Eastern Europe. Volume 1: Country Ukraine." Economic Policy, October. Studies (19-50). Havrylyshyn, 0. and Pritchett. 1991. "European Trade Cadot, O., and J. de Melo. 1994. "The Europe Agreements Patterns After the Transition." Policy, Research, and and EC-LDC Relations." CEPR Discussion Paper. External Affairs, Working Papers, Trade Policy, WPS 1001. London, August. 748. World Bank, Washington, D.C. Choksi, A.M., M. Michaely, and D. Papageorgiou. 1991. Havrylyshyn, O., and D. Tarr 1991. "Trade Liberalization "The Design of Successful Trade Liberalization and the Transition to a Market Economy." Policy, Policies." in A. Koves and P. Marer, eds., Foreign Research, and External Affairs, Working Papers, Economic Liberalization: Transfirmation in Socialist Trade Policy, WPS 700. World Bank, Washington, Economies. Boulder: Westview Press. D.C. Collins, S., and D. Rodrik. 1991. Eastern Europe and the Hewett, E.A. with C.G. Caddy 1992. Open for Business. Soviet Union in the World Economy. Institute for Russia's Return to the Global Economy. The Brookings International Economics. Washington, D.C. Institution, Washington, D.C. Connolly, M., and S. Vatnick. 1994. "Uzbekistan: Trade Hillman, A., and A. Schnytzer 1992. "Creating the Reform in a Cotton Based Economy." in C. Reform-Resistant Dependent Economy." in B. Michalopoulos and D. Tarr, eds. Trade in the New Milanovic, and A. Hillman, eds. The Transition from IndependentStates. World Bank/UNDP, Washington, Socialism in Eastern Europe: Domestic Restructuring IDpC. and Foreign Trade. World Bank, Washington D.C. Csaba, L. 1995. "The Political Economy of Trade Regimes Hoekman, B. and G. Pohl. 1995. "Enterprise Restructuring in Central Europe" in L.A. Winters, ed. Foundations in Eastern Europe: How Much? How Fast? Where? of an Open Economy: Trade Laws and Institutions for Preliminary Evidence from Trade Data". mimeo. The Eastern Europe. Centre for Economic Policy Research, World Bank, Washington, D.C. London. IMF. 1994. "Trade Policy Reform in the Countries of the Dadush, U. and D. He. 1995. "China: A New Power in Former Soviet Union." IMF Economic Reviews. World Trade," Finance &- Developmcnt, June. Volume 2. Washington, D.C. EBRD. 1993: Economic Outlook, European Bank for International Trade Reporter. Various issues. The Bureau Reconstruction and Development. September. of National Affairs, Inc., Washington, D.C. London. Jefferson, G.H., and T.G. Rawski. 1994. "How Industrial Gelb, A., G. Jefferson, and 1. Singh. 1993. "Can Reform Worked in China: The Role of Innovation, Communist Economies Transform Incrementally?" Competition, and Property Rights." Paper presented Policy Research Working Papers 1189. World Bank, at the Annual Bank Conference on Development Washington, D.C. Economics. World Bank, Washington D.C., April 28-29. Gelb, A., M. de Melo and C. Denitzer. 1995. 'From Plan to Market: The Patterns of Transition," Mimeo, Kaminski, B. 1994a. "The Significance of 'Europe World Bank, Washington, D.C. Agreements' for Central European Industrial Exports." Policy Research Working Paper 1314. World Bank, Washington, D.C. 60 FOREIGN TRADE IN THE TRANsITON Kaminski, B. 1994b. "Trade Performance and Access to Messerlin, P. 1995. "Central and East European Countries' OECD Markets." In C. Michalopoulos and D. Tarr, Trade Laws in the Light of International Experience." eds. Trade in the New Independent States. World In L.A. Winters, ed. Foundations of an Open Economy: Bank/UNDP, Washington, D.C. Trade Laws and Institutions for Eastern Europe. London: Centre for Economic Policy Research. Karninski, B. 1993a. "Emerging Patterns of Foreign Trade." In R.F. Staar, ed. Transition to Democracy in Messerlin, P. 1992. "The Association Agreements between Poland. New York and London: St. Martin's Press. the EC and Central Europe: Trade Liberalization vs. Constitutional Failure?" In J. Flemming and J.M.C. Kaminski, B. 1993b. "How the Market Transition Rollo, ed.. Trade, Payments andAdjustment in Central Affected Export Performance in the Central European and Eastern Europe. London: Royal Institute of Economies." Policy Research Working Paper 1179. International Affairs. World Bank, Washington, D.C. Michalopoulos, C., and D.G. Tarr 1994. "Summary and Kaminski, B. 1989. "Council for Mutual Economic Overview of Developments Since Independence." In Assistance: Divisions and Conflicts at its 40th C. Michalopoulos and D. Tarr, eds. Trade in the New Anniversary." in Richard F. Staar, ed. Yearbook on Independent States. World Bank/UNDP, Washington, International Communist Affairs. Stanford: The D.C. Hoover Institution Press. Michalopoulos, C., and D.G. Tarr. (forthcoming) 'Trade Kaminski, B., and A. Yeats 1993. "OECD Trade Barriers Performance and Policy in the New Independent Faced by the Successor States of the Soviet Union." States." Directions in Development Series. The Policy Research Working Papers WPS 1175. World World Bank: Washington, D.C. Bank, Washington, D.C. Mizsei, K. 1991. "The Institutional Framework of Foreign Konovalov, V. 1994. "Russian Trade Policy." in C. Trade Reforms in Hungary." In M. Kaser and A.M. Michalopoulos and D. Tarr, eds. Trade in the New Vacic, eds. Reforms in Foreign Economic Relations of Independent States. World Bank/UNDP, Washington, Eastern Europe and the Soviet Union. New York: D.C. United Nations Economic Commission for Europe. Krumm, K. 1994. "Kyrgyz Republic: External Trade for a Neal, C. 1992. "The CMEA Trade Collapse: Did It Small Country." in C. Michalopoulos and D. Tarr, Happen, When and Where?," Mimeo, World Bank, eds. Trade in the New Independent States. World October. BankWUNDP, Washington, D.C. Nuti, D.M. 1991. "Progress in Trade Reforms, Missing Kroll. 1991. "Monopoly and Transition to the Market." Links, Convertibility A Discussion." In M. Kaser and Soviet Economy. April-June,. 6: 143-174. A.M. Vacic, eds. Reforms in Foreign Economic Relations of Eastern Europe and the Soviet Union. New Landesmann, M.A. 1995. "The Pattern of East-West York: United Nations Economic Commission for European Integration: Catching Up or Falling Europe. Behind." Research Report No. 212, The Vienna Institute for Comparative Economic Studies, Vienna. Oblath, G. 1993. "Interpreting and Implementing Currency Convertibility in Central and Eastern Le Gall, F. 1994. "Ukraine: A Trade and Exchange System Europe: A Hungarian Perspective." Kopint-Datorg Still Seeking Direction." in C. Michalopoulos and D. Discussion Papers. 10, Budapest, December. Tarr, eds. Trade in the New Independent Statef. World Bank/UNDP, Washington, D.C. Oblath, G., and D. Tarr. 1992. "The Terms-of-Trade Effects from the Elimination of State Trading in Learner, E.E. 1984. Sources of International Comparative Soviet-Hungarian Trade." Journal of Comparative Advantage. Cambridge: The MIT Press. Economics. No. 16. Lista 500 Najwiekszych Polskich Przedsiebiorstw (The List Olechowski, A., and A. Yeats 1982. "The Influence of of 500 Largest Polish Enterprises). Gazeta Bankowa, Nontariff Barriers on Socialist Countries' Exports." Dodatek Specjalny, III Edycja, Warszawa. Economia Internazionale, Fall. McAuley, A. 1991. "The Economic Consequences of Padoan, P.C., and M. Pericoli 1993. "The Single Market Soviet Disintegration." Soviet Economy. 7 (3). and Eastern Europe. Specialization Patterns and Prospects for Integration." Economic Systemr. 4(17):279-99. REFERENCES 61 Panagariya, A. (1995) "What Can We Learn from China's Staar, R. F. 1988. Communist Regimes in Eastern Europe. Export Strategy?" Finance&-Development, June. 5th ed. Hoover Institution Press, Stanford. Panagariya, A. 1991. "Unravelling the Mysteries of China's Tarr, D.G. 1990. "Second-Best Foreign Exchange Policy Foreign Trade Regime. A View from Jiangsu in the Presence of Domestic Price Controls and Province." Policy Research Working Papers, WPS Export Subsidies." The World Bank Economic Review. 801. Country Economics Department, World Bank, 4(2). Washington D.C. Tarr, D.G. 1992. "Problems in the Transition from the Pynzenyk, V. 1994. "Ukrainian Economic Reforms: CMEA: Implications for Eastern Europe." Communist Reflections on the Past and the Future." Economic Fconomies and Economic Transformation. 4(1). Policy. October. Tarr, D.G. 1994. "The Terms-of-Trade Effects of Moving Rodrik, D. 1992. Foreign Trade in Eastern Europe's to World Prices on Countries of the Former Soviet Transition: Early Results. Mimeo. NBER. Union." Journal of Comparative Economics. (18):1-24. Roe, A., and J.Roy. 1989. "Trade Reform and External Toth, L.G. 1992. "The Liberalization of Foreign Trade in Adjustment. The Experience of Hungary, Poland, East-Central Europe: The Experiences of Poland and Portugal, Turkey, and Yugoslavia." EDI Policy the Czech and Slovak Federal Republic." Kopint- Seminar Report 16. World Bank, Washington, D.C. Datorg Discussion Papers. 10, Budapest. Rosati, D. 1992. "Problems of Post-CMEA Trade and Tovias, A., and S. Laird. 1991. "Whither Hungary and the Payments." Discussion Paper 650, Centre for European Communities?" Policy, Research, and Economic Policy Research, London. External Affairs. Working Papers, WPS 584, International Economics Department, World Bank, Sachs, J. 1995. "Reforms in Eastern Europe and the Washington, D.C. Former Soviet Union in Light of the East Asian Experience," paper presented at the NBER-TCER- UNECE. 1993a: Economic Commission for Europe. CEPR Conference on Transition from Socialist Economic Bulletin for Europe. Volume 45. United Economies, January 6-7, 1995, Tokyo. Nations: New York. Schroeder, G.E. 1993. "Post-Soviet Reforms in UNECE. 1993b: Economic Commission for Europe. Perspective." in The Former Soviet Union in Economic Survey of Europe in 1992-1993. United Transition, Vol.I. Joint Economic Committee, U.S. Nations: New York. Congress, US Government Printing Office, Washington, D.C. Webster, L. 1994. "Russia's Newly Privatized Firms." Finance and Private Sector Development. Note Schumacher, D., and U. Mobius. 1994. "Analysis of No.22. World Bank, Washington D.C. Community Trade Barriers Facing Central and East European Countries and Impact of the Europe Wellisz, S. 1994. Wolny handel czy ochrona rynku kra- Agreements." Mimeo, Deutches Institut fur jowego (Free trade or protectionism). CASE, Polish Wirtschaftsforschung, Berlin. Scientific Publishers, Warsaw. Sorsa, P. 1994a. "Larvia: Trade Issues in Ttansition." In C. Winiecki, E, and J. Winiecki. 1992. "The Structural Michalopoulos and D. Tarr, eds. Trade in the New Legacy of the Soviet-Type Economy." A Collection of Independent States. World Bank/UNDP, Washington, Papers. Centre for Research into Communist D.C. Economies, New Series 5, London. Sorsa, P. 1994b. "Lithuania: Trade Issues in Transition." Winters, L.A. 1994. "The Liberalization of European Steel in C. Michalopoulos and D. Tarr, eds. Trade in the Trade." Chapter 7 of R. Faini and R. Portes (eds). New Independent States. World BankIUNDP, European Union Trade with Eastern Europe: Washington, D.C. Adjustment and Opportunities, Centre for Economic Policy Research, London. Sorsa, P. 1994c. "Regional Integration and the Baltics-- which way?" Mimeo, International Trade Division, Winters, L.A. 1993. "The Europe Agreements: an assess- International Economics Department, World Bank, ment," in Patterns ofProduction and Trade in the New Washington, D.C. Europe. Geneva: European Free Trade Association. 62 FOREIGN TRADE IN THE TRANSMON Winters, LA., and Z.K. Wang 1994. Eastern Europe's Yeats, A. 1995. 'Are Partner Statistics Useful for rn'crnmaoad Trade. Manchester and New York: Estimating 'Missing' Trade Data?" Policy Research Manchester University Press. Working Paper 1501, August, International Economics Department, International Trade Writers, L.A, and Z.K Wang. 1993. 'EC Imports from Division, World Bank, Washington, D.C., Eastern Europe.' Centre for Economic Policy Research Discussion Paper 825. London. Yotopoulos, P-A. 1995. Exchange Rate Parity for Tradc and Development: Theory, Tests, and Case Studies. London and New York: Cambridge University Press. Distributors of tjalatt~j &Ecceont PoI, [aerator.AMlCE t RAE tEW ZEUND RUSSIAN FEDEMRIION Wennefgron-Wtllams AD Distributors of ~~~ChtnJ Ftn nct & Ercoommt World 6ank Pubkatc obs Yorrno Literature Ltt EBSCO NZ Ltd. Isdatelstvo