East Asia and Eastern Europe Trade Linkages and Issues

Transcription

1 247 PACIFIC ECONOMIC PAPERS NO. 261, NOVEMBER 1996 East Asia and Eastern Europe Trade Linkages and Issues Jocelyn Horne A USTRALIA JAPAN RESEARCH CENTRE

2

3 PACIFIC ECONOMIC PAPER NO. 261 NOVEMBER 1996 East Asia and Eastern Europe Trade Linkages and Issues Jocelyn Horne Macquarie University A USTRALIA JAPAN RESEARCH CENTRE

4 Australia Japan Research Centre 1996 This work is copyright. Apart from those uses which may be permitted under the Copyright Act 1968 as amended, no part may be reproduced by any process without written permission. Pacific Economic Papers are published under the direction of the Research Committee of the Australia Japan Research Centre. The opinions expressed are those of the author(s) and do not necessarily reflect the views of the Centre. The Australia Japan Research Centre is part of the Economics Division of the Research School of Pacific and Asian Studies, The Australian National University, Canberra. ISSN ISBN Australia Japan Research Centre Research School of Pacific and Asian Studies The Australian National University Canberra ACT 0200 Telephone: (61 6) Facsimile: (61 6) ajrcgen@ajrc.anu.edu.au Edited by Gary Anson Typeset by Minni Reis ii

5 CONTENTS List of tables and figures... iv Introduction... 1 Background and issues... 3 Stylised facts Concluding remarks Appendix Notes References iii

6 TABLES Table 1 Summary measure of merchandise trade... 5 Table 2a Merchandise trade shares, Table 2b Northeast Asia trade matrices, Table 3 Trade intensity, complementarity and bias Table 4 Factor composition of total trade Table 5 Export specialisation indexes Table 6 Revealed comparative advantage Table A1 Krause factor intensity classification Table A2 Murrell classification system Table A3 Classification of transition economies by income and region, FIGURES Figure 1 Relative endowments of natural resources, labour and capital, various economies, iv

7 EAST ASIA AND EASTERN EUROPE TRADE LINKAGES AND ISSUES This paper examines the pattern and structure of trade between Eastern Europe, the former Soviet Union and East Asia with a particular focus on the post-1991 reform period. The unexpected expansion in trade between Eastern Europe and East Asia has been accompanied by increased trade complementarity between East Asian and Eastern European transition economies. This trend is shown to reflect two concurrent developments; an intensification of pre-existing comparative advantage by Eastern Europe and changing comparative advantage by East Asian economies. Introduction Economic integration of Eastern Europe and the former Soviet Union (FSU) has attracted considerable interest in recent literature. 1 The main focus has been on the welfare implications of an enlarged European Union with little attention paid to evolving trade linkages between Eastern Europe and the Asia Pacific region. 2 The two issues, however, cannot be considered in isolation given the increased importance of the Asia Pacific region in the world economy and changing areas of comparative advantage as economies in these regions undergo economic development and reform. This study focuses on one aspect of these linkages or trade relations between core (advanced industrialised) and periphery (developing) countries in the trade blocs. In particular, it attempts to identify the main changes in merchandise trade pattern and structure between Eastern Europe, the FSU and East Asia in the aftermath of systemic reforms in the former group of countries. In examining these linkages, the scope of the study is restricted to merchandise trade and an aggregated treatment of the FSU. The central themes that emerge need to be interpreted within this context. Drysdale (1991), writing prior to the breakup of the Soviet Union, stated that The Soviet economy is not yet, nor for many years can it be, central to Asia Pacific interests. Despite the major upheavals that have since taken place, this statement remains true. The share of East 1

8 Asian trade by Eastern Europe and the FSU remains below 1 per cent. Rapid growth has occurred, however, in East Asia s share of Eastern Europe and FSU trade; that share has risen from 8.4 per cent in 1980 to 16.6 per cent in Nevertheless, the dramatic collapse in intratrade among former members of the Council for Mutual Economic Assistance (CMEA) has been mirrored largely in increased trade flows between Eastern and Western Europe. The motivation for the present analysis requires further justification. It rests upon three main grounds: the strategic importance of the FSU, the welfare effects of an enlarged European Union, and the likelihood of increased global trade shares of Eastern Europe and the FSU. First, the strategic influence of the Russian Federation in the world means that all countries have a stake in successful reform in this region. This factor, combined with the growing Asia Pacific share of global trade, means that trade and financial linkages bind the sustainability of growth of the regions to each other. In addition, just as it is argued that Eastern and Western Europe are natural trading partners because of geographical proximity, so the Central Asian republics are natural partners in trade to Northeast Asian economies, including South Korea, Japan and China. Trade linkages may be expected to strengthen within regions such as the Tumen River area. A second motivation for the study arises from the question of Eastern Europe s comparative advantage and the welfare impact of an enlarged European Union. Resolution of the question of comparative advantage in the transition and post-transition periods has a critical bearing on predicting the distribution of the gains from inter-industry trade; ceteris paribus, the greater the degree of complementarity in trade structures between Eastern and Western Europe and East Asia, the greater the gains to member countries and the smaller the losses to non-members. Conversely, insofar as East European economies compete in similar export markets to Asian trading partners, the greater the potential trade diversion losses to the latter that may arise from EU preferences to East European partners. As confirmed in the present study, a distinctive pattern is apparent in the trade structure of NIEs, with a shift in export specialisation away from unskilled labour and traditional manufacturing towards human capital and high technology exports. This trade pattern is consistent with recent theories of dynamic comparative advantage derived from models of technology-led endogenous growth (see Grossman and Helpman 1991, 1992). In the transition period, European central planning economies (CPEs) share certain similarities in economic structure with developing economies, including trade and financial repression, soft budget constraints and large macro imbalances. But there are also fundamental differences, such as a 2

9 higher share of industrial output and a more educated workforce. The question of their comparative advantage is by no means clearcut in terms of such analogies. Finally, as successful reform gathers momentum in Eastern Europe, the importance of this region in the global economy should increase. Thus far, only a small group of countries (the Czech Republic, Hungary, Poland and the Baltic states) have achieved positive output growth, with significant shares of output originating in the private sector and private capital flows exceeding 10 per cent of export earnings. Growth prospects for the remaining countries are contingent upon their continued reliance on external assistance, making any projections highly uncertain. Notwithstanding these uncertainties, trade interactions will be further strengthened as the world economy becomes more integrated with successful implementation of the Uruguay Round. The remainder of the paper is organised as follows. The next section discusses the background to recent trade reform initiatives in Eastern Europe and the associated issues. The third section presents an analysis of the structure and composition of trade flows between East Asia and Eastern Europe/FSU within a broad regional framework that includes the main trade blocs the European Union and APEC, as well as East Asia, North America and Australasia. The final section brings together the main findings of the study and suggestions for further research. Background and issues Trade reform issues in Eastern Europe and the FSU have already undergone a shift in emphasis away from the earlier focus on rapid trade liberalisation towards present concerns with market access, as reflected in the proliferation of regional trade agreements. A key question underpinning the economic rationale for trade strategy is the likely effect of the opening up of these economies on their trade volume, direction and structure. This question has been addressed in recent literature. See, for example, CEPR (1990), Collins and Rodrick (1991), Murrell (1990), Neven and Roller (1991), Wang and Winters (1993) and the Economic Commission for Europe (1993). As for trade volume, it is predicted that this will expand in the post-reform period due to the widespread distortions under central planning, resulting from bureaucratic coordination of production and trade as well as the asymmetric treatment of trade between CMEA and non- CMEA trading partners. 3 The dismantling of the CMEA in January 1991 was accompanied by a comprehensive set of trade liberalisation measures (see Kenen 1991; and IMF 1992). 3

10 Predictions of the magnitude of the expected expansion in trade vary according to the methodology employed and the definition of the post-reform period (transition or posttransition). Collins and Rodrick (1991) project a rise in Eastern Europe s share of world trade from 3.3 per cent in 1989 to 4.9 per cent (assuming full trade liberalisation and no growth catchup to industrialised economies) and to 12 per cent (assuming full catch-up). Comparable longrun rises in trade volume are projected in Wang and Winters (1993) based upon a gravity trade model. Summary data given in Table 1 show that these predictions have as yet failed to materialise. On the contrary, Eastern Europe s share of world trade has fallen to below 2 per cent. The unexpectedly large fall in output in the first few years of reform, itself related to the CMEA dismantling (principally the negative terms-of-trade shock for non-fsu members), helps explain the above outcome, given the above assumptions. Nevertheless, this finding serves as a warning of the sensitivity of projections about trade volume and, more generally, trade developments to assumptions about growth and successful reform. This issue becomes even more important in the post-1993 period in view of the diverse economic performance of former European CPEs. The second question concerns the predicted geographical composition of trade. The above studies project a reversal of the trade pattern under central planning in which trade flows were highly concentrated among CMEA members, accounting for per cent of trade, with dominance by the FSU. The Collins and Rodrick (1991) study adopts a 1928 trade matrix as its base period that shows high trade flows between Eastern and Western Europe but a very small share of FSU trade. They assume that European CPEs would have followed a similar trade pattern to a group of comparator countries in the absence of socialism (after adjusting for factors depressing FSU trade in the 1920s). 4 Wang and Winters (1993) also project a large reorientation of trade towards industrialised countries and especially to Western Europe based upon estimates of the gravity model of trade. Table 2a shows these predictions to be well supported. The share of the European Union in East European exports (excluding the FSU) has more than doubled since 1980 while intra- East European trade halved over the same period. There are, however, two features of these findings that are unexpected; first, the speed of redirection of trade and, second, the unexpected expansion of trade between Eastern Europe and East Asia. 4

11 Table 1 Summary measure of merchandise trade (in per cent) Share of exports that is intra-regional: Intra-East Asia a Intra-NIES b Intra- ASEAN c Intra-EEC Intra-East Europe (excl. FSU) d Intra-North America e Intra- Australasia f Intra-APEC g Share of world exports: East Asia Japan NIES ASEAN China EEC Eastern Europe (excl. FSU) FSU na na North America Australasia APEC Export share of GDP: East Asia Japan NIES ASEAN China EEC Eastern Europe (excl. FSU) North America Australasia APEC Trade intensity index: East Asia NIES ASEAN EEC Eastern Europe (excl. FSU) North America Australasia APEC Notes: a Country aggregate East Asia includes Japan, Korea, China, Taiwan, Hong Kong, ASEAN (as defined in c ), Laos and Cambodia. b Country aggregate NIES includes Korea, Taiwan, Hong Kong and Singapore. c Country aggregate ASEAN includes Thailand, Malaysia, the Philippines, Indonesia, Brunei, Vietnam and Singapore. d Eastern Europe excludes FSU and includes Albania, the German Democratic Republic (before 1990), Romania, the Czech Republic, Slovakia, Bulgaria, Poland and Hungary. 5

12 e Country aggregate North America includes the United States and Canada. f Country aggregate Australasia includes Australia, New Zealand and PNG. g Country aggregate APEC includes Australia, Brunei, Canada, Chile, China, Hong Kong, Indonesia, Japan, Korea, Malaysia, Mexico, New Zealand, Papua New Guinea, the Philippines, Singapore, Taiwan, Thailand and the United States. Export share of GDP is calculated by summing exports(us$). na Not available. Sources: IMF (1996); World Bank (1995); International Economic Databank, Australian National University. The speed of change is surprising given that both methodologies assume full functioning of markets and institutions as well as income catch-up with industrialised countries. One explanation lies in the smaller-than-expected import surge in Eastern Europe in response to price liberalisation and elimination of the monetary overhang. Undervalued (in other words, depreciated) real exchange rates is one factor behind increased exports, at least for Poland, Hungary and the Czech Republic. But, in any event, both approaches assume that trade takes place between similar industrialised economies, in which case the gains for trade derive primarily from intra-industry rather than inter-industry trade. Other evidence (see, for example, the Economic Commission for Europe [1993]) shows intra-east European industry trade to be low and to have fallen in the post-reform period. The point here is that neither of the above methodologies offers a satisfactory explanation of the observed transitional rather than longrun pattern of trade. A second feature of observed post-reform trade patterns is the unexpected growth in trade between Eastern Europe and East Asia. This development is not predicted in the above studies because of the low share of East Asia in world trade in 1928 and the gravity model based upon the hypothesis of natural trading partners, as determined by income, location and country size. Again, it is not difficult to offer ad hoc explanations for the observed trade pattern, especially based upon a stages-of-development approach in which middle and lower-income countries specialise in unskilled labour exports in return for imports of high-technology goods from more advanced economies. To explain this phenomenon, we need to turn to the next issue that of comparative advantage of Eastern Europe. The issue of East Europe s future comparative advantage has attracted considerable debate. This debate has arisen largely because of difficulties in drawing inferences about comparative advantage from pre-reform data on factor endowments and trade patterns as well as different assumptions about Eastern Europe s growth prospects after reform. Two main 6

13 approaches have been adopted; the first uses information on factor endowments to make inferences about comparative advantage using the Heckscher Ohlin theory of trade. The second draws inferences about comparative advantage from actual trade patterns. The specific question addressed here is whether a similar picture emerges from these different approaches. Misallocation of resources under socialism means that inferences drawn from raw data may be quite misleading. One way of addressing this problem is to examine historical trends before socialism. For example, the CEPR (1990) study shows that in the pre-1914 period, Russia specialised in exports of agricultural products and raw materials, a finding that is consistent with its relative abundance in natural resources. It is also observed that Eastern Europe (and especially the former Czechoslovakia) prior to the Second World War specialised in labour-intensive goods and thereby resembled Japan in the 1950s and 1960s. However, it is difficult to derive any clearcut inferences about post-reform trade patterns based on historical trends given the different stage of development of the Soviet economy and Eastern Europe after many decades of industrialisation under socialism. Some support for the view that pre-socialist data may provide a useful insight into postreform trade patterns comes from a different source. While measured capital-to-labour ratios and investment in terms of GDP in Eastern Europe and the FSU are very high, estimates by Borensztein and Montiel (1992) show unproductive investment to be per cent of total investment in the former Czechoslovakia, Hungary and Poland. Once capital to labour ratios are adjusted for excess investment, the ratios are very low. As a consequence, Eastern Europe matches the characteristics of the earlier historical data, resembling middle-income developing economies with low wages and an abundant labour force. The scarcity of capital and high marginal productivity of capital at the start of recent reform also implies that fairly low investment ratios may be sufficient to achieve growth convergence. 5 The need to correct resource endowment data for factor quality is also of relevance when using data on human capital skills. A broad range of human capital indicators (education expenditure in terms of GDP, school enrolment ratios and R & D ratios) are presented in the CEPR study cited above to show comparability between Eastern and Western Europe. On the basis of this evidence, the authors conclude these factor abundances suggest that among manufactures, it is high-technology goods rather than labour-intensive goods that represent Eastern Europe s area of comparative advantage (CEPR 1990, p. 13). This conclusion is at variance with other research using factor endowments and measures of revealed comparative advantage, as discussed below. 7

14 A simplified treatment of the problem of measuring factor endowments is suggested in an application by Anderson (1991) of the Leamer triangle examining comparative advantage of groups of countries, including Eastern Europe. 6 A three-factor model natural resources, unskilled labour and capital (human, physical and knowledge) is assumed in Figure 1, which is reproduced from Anderson. In Figure 1, per capita income and per capita agricultural land are used as proxies for capital-to-labour and resources-to-labour ratios, respectively. While measures of income per capita in CPEs vary greatly, Anderson s estimates (Eastern Europe s per worker capital endowment is 60 per cent of the world average) are broadly consistent with those of Borensztein and Montiel (1992). In the Leamer diagram, Eastern Europe (excluding the FSU) is located in the WLB zone that is, below the global average per worker endowments of capital with a resulting comparative disadvantage in skill and knowledge-intensive products. In contrast, West European economies lie in WCB, with a comparative advantage in capital-intensive goods but a comparative disadvantage in primary products and unskilled labour-intensive goods. Economic growth as modelled through an increase in the supply of capital relative to that of other factors shifts an economy s endowment point towards C and away from NL. As a result, countries in WLB (as well as the FSU in WAN) shift towards WBC and in the process strengthen their initial comparative advantage in labour-intensive products (Eastern Europe) and resource-intensive products (FSU). As discussed more fully in Anderson (1991), the direction and magnitude of predicted change in comparative advantage in response to growth depend critically upon assumptions about labour mobility, sufficient capital inflows to finance investment and enterprise restructuring in reforming countries. For example, large-scale labour emigration from Eastern Europe shifts the endowment point towards the N corner, thereby strengthening its comparative advantage in primary products. 7 Alternative approaches to identifying comparative advantage in Eastern Europe using measures of revealed comparative advantage (see Murrell 1990; and Collins and Rodrick 1991) reach similar conclusions in regard to the pre-reform period. For example, Collins and Rodrick show that while Eastern European economies were net importers of manufactures taken as a whole, they are similar to middle-income developing countries with a comparative advantage in standardised (low-skill items) in basic and miscellaneous manufactures. Based upon 1989 trade patterns, Collins and Rodrick (1991) conclude that Eastern Europe is likely to make its entrance into the world economy, at least where manufactures are concerned, mainly as a 8

15 Figure 1 Relative endowments of natural resources, labour and capital, various economies, 1991 (Natural resources) N ANZ 1.5 SSA SU D A 1.0 LA W 3.2 CH NA 10 OEA EET SA EC 32 EF JA (Labour time) L NIE C (Capital) B Notes: The distance along NL from N measures population per unit of agricultural land as a ratio of the world average (1.07 people per hectare). The distance along LC from L measures per capita income as a ratio of the world average (US$3,400). Both scales are in logs. Along any ray from C to NL the population per unit of agricultural land is constant, and similarly for rays from the other two corners of the triangle. W is the world s endowment point. Countries are represented as follows: ANZ Australia and New Zealand; CH China; EC the twelve EC member countries; EET Eastern Europe; JA Japan; LA Latin America; NA the United States and Canada; NIE the Asian NIEs; OEA other East Asian market economies; SA South Asia; SSA Sub-Saharan Africa; SU the Soviet Union. The estimates used for per capita income for Eastern Europe and the Soviet Union are US$1,800 and US$1,600 respectively, based on World Bank and other estimates reported in CEPR (1990, p. 33). Source: Anderson (1991). low-cost producer of relatively standardized commodities rather than as a producer of human capital-intensive goods (p. 61). Analysis of East West trade flows by Murrell (1990) using measures of revealed comparative advantage also shows that the comparative advantage of Eastern Europe under central planning lies in so-called Heckscher Ohlin goods (that is, in goods characterised by 9

16 standard manufacturing) while that of the FSU lies in Ricardo goods (namely, in resourceintensive goods). Stylised facts Methodology The methodology adopted in this paper is stylised fact analysis. The purpose is to identify issues for research that is, what needs to be explained. Presentation of the stylised facts is intended to answer the following questions: What is the strength of trade linkages between East Asia, Eastern Europe and the former Soviet Union? What is the structure of trade interactions among the above regions and how has this structure altered over the past decade, especially since the major economic reforms introduced by European CPEs in 1991? What is the comparative advantage of Eastern Europe based upon measures of export specialisation (or revealed comparative advantage)? Presentation of stylised facts is a highly selective process and this is particularly the case in analysing trade data on Eastern Europe. There are three main problems: non-comparability of trade data of former European CPEs and Western economies, classification of commodities, and definition of region and country. The first issue has already been noted and arises from three sources distortions under socialism, non-convertibility of trade within the former CMEA, and differences between UN and CMEA trade data classification. To minimise data problems, the focus of the analysis is on trade flows between Eastern Europe, the FSU and other trade regions rather than within Eastern Europe. A further solution to the above problem is to use mirror statistics (see, for example, Murrell 1990). Exports (imports) of Eastern European economies are derived by adding imports (exports) of major industrialised trading partners. Sensitivity tests using the latter show that the broad trends in trade patterns identified in this study do not appear to be altered appreciably by the use of mirror statistics. A second concern is the sensitivity of summary trade measures and interpretation of trade patterns to the method used to classify factor intensity. This paper follows Tyers and Phillips 10

17 (1984), who adopt Krause s methodology whereby production processes are assumed to involve multiple factors, each classified by its factor used most intensively and/or location of production. Trade of 187 commodities at the 3-digit SITC level are divided into five groups according to their intensities in five factors: agricultural resources, mineral resources, unskilled labour, technology and human capital (see Appendix Table A1.) The sensitivity of summary measures and, in particular, measures of export specialisation to the classification system are examined by comparing these results with those based upon an alternative classification system as presented in Murrell (1990), 8 details of which are given in Appendix Table A2. Third, there is the question of regional definition. There is considerable diversity in economic structure and development within each trade region and especially within East Asia and European CPEs. A further breakdown based upon core and periphery economies has therefore been adopted for East Asia. Within East Asia (excluding Japan), core economies are defined to include the four NIEs (Singapore, South Korea, Hong Kong and Taiwan) with the remainder ASEAN, Laos, Cambodia and China classified as periphery economies. The close proximity of Central Asian republics in the FSU to Northeast Asia is also of interest and some limited data on trade linkages between the FSU and Northeast Asian economies is presented. In regard to former European CPEs, there may be greater similarities between the more advanced Eastern European economies and the Baltic states than within either Eastern Europe or the former Soviet Union (see Appendix Table A3). The emphasis on European CPEs may also be misplaced given the similarity in economic structure of some Central Asian republics which specialise in cotton exports to developing Asian economies as well as complementarity (in terms of mineral resources) to more developed Asian economies. This type of disaggregation within the FSU would be highly desirable but is not possible using a consistent data set based upon UN sources. 9 Direction of trade The outstanding feature highlighted by the direction of trade data (Table 2a) is the small size of trade flows between East Asia and Eastern Europe, especially compared with the strength of trade linkages between Eastern and Western Europe in the aftermath of reforms. Since 1990, the share of the European Union in Eastern Europe s trade has doubled to over 50 per cent, rising from almost 20 per cent in In contrast, East Asia s share of total FSU and East European trade has risen from 8.4 to 16.6 per cent. The sharpest recorded growth has been between East 11

18 Table 2a Merchandise trade shares, (three-year averages, except ) (in per cent) Export share East Asia Japan NIES ASEAN China EEC-12 Eastern Europe FSU North Australasia APEC (excl. FSU) America East Asia Japan NIES ASEAN EEC

19 (Table 2a continued) Eastern Europe (excl. FSU) FSU North America Australasia APEC Sources: IIMF (1996); International Economic Databank, Australian National University. 13

20 Asia (excluding Japan) and the FSU, from less than 1 to 9.5 per cent, while the share of East Asia in East European trade has remained at between 2 to 4 per cent. The total share of Eastern Europe and FSU in exports of the major regional trading blocs has remained very small at less than 1 per cent (APEC and East Asia) and 3 per cent for the European Union. Decomposition of trade flows between developed and developing East Asia economies reveals some distinct patterns. First, within East Asia (excluding Japan) there has been a sharp reduction in trade among NIEs. At the same time, trade has strengthened between core and periphery East Asian countries and between core East Asian and Eastern European/FSU economies. The fastest growth has been the increased NIE share of FSU trade, rising from 0.2 to 0.5 per cent ( ) to 2.7 per cent in Somewhat slower growth (from per cent) is observed for ASEAN shares of FSU exports. A similar pattern is observed for NIE and ASEAN shares of East European exports. The above regional trade data do not capture the trade linkages associated with the geographical proximity of the FSU with Northeast Asia, specifically the Tumen River area. 10 Northeast Asia is defined to include South and North Korea, Japan, China, the FSU and Mongolia. Even within a relatively short period ( ), some changes have already occurred, reflecting FSU economic reforms and the increased importance of South Korea (Table 2b). 11 Trade within this region remains dominated by Japan. However, trade flows have shifted from Japan China (45.8 per cent) and Japan South Korea (28.4 per cent) in 1985 to dominance by Japan South Korea (47.0 per cent) in Some redirection of FSU trade away from former trading partners, Mongolia and North Korea, towards South Korea (from zero in 1985 to 6.6 per cent in 1991) is also observed and is consistent with the strengthening of trade linkages between the FSU and NIEs shown in Table 2a. Trade intensity, complementarity and bias Measures of trade intensity and their decomposition into trade complementarity and bias provide further insight into the structure of trade relationships between Eastern Europe and East Asia. The trade intensity index measures the share of trade of a given region with another region expressed as a proportion of that region s share of world trade. This measure has been further decomposed into the joint product of two indexes: trade complementarity (a measure of the degree of similarity between the commodity composition of one region s exports and imports of its trading partners) and trade bias (a measure of the relative strength of trade resistances)

21 Table 2b Northeast Asia trade matrices, Importer South Korea North Korea Japan China FSU Mongolia 1985 trade matrix (in per cent of total regional trade) Exporter South Korea North Korea na Japan China FSU Mongolia 0 na trade matrix (in per cent of total regional trade) Exporter South Korea North Korea na Japan China FSU Mongolia na trade matrix (in per cent of total regional trade) Exporter South Korea North Korea na Japan China FSU Mongolia na 0.3 Note: na: Not available. Source: Pomfret (1996, Table 10.1, p. 132). Movement in trade intensity indexes (i in Table 3) are broadly similar to trends in bilateral trade discussed above. 13 Trade intensity indexes for East Asia have risen over the past fifteen years while those for Eastern Europe have fallen sharply since Despite the high level of aggregation across countries and commodities, a pattern of changing regional trade structure is identifiable, especially if the NIEs are shown separately. The trend over the entire sample period has been towards greater regional complementarity between Eastern Europe and core East Asian economies, with an acceleration after For example, the trade complementarity index (Table 3) for trade flows between the NIEs and Eastern Europe (based upon exports by NIEs to Eastern Europe) doubled from 0.5 to 1.0 from 15

22 Table 3 Trade intensity, complementarity and bias Exports from: Exports to: APEC Eastern Europe FSU East Asia NIEs European (excl. FSU) (excl. Japan) Union i c b i c b i c b i c b i c b i c b APEC Eastern Europe (excl. FSU) FSU East Asia (excl. Japan) NIEs European Union Sources: International Economic Databank, Australian National University; UN trade data, June

23 1985 to Over this period, the complementarity index fell from 1.9 to 1.0 within the former CMEA (based upon FSU exports to Eastern Europe). In contrast, trade complementarity between Western and Eastern Europe remained stable, based upon EU exports to Eastern Europe. Trade complementarity among the NIEs has also remained strong despite a decline in the index since the mid-1980s. Commodity composition of trade Shifts in commodity composition of trade based upon the Krause factor classification scheme (see Appendix Table A1) are shown in Table 4. In terms of overall export structure, Eastern Europe presents the most diversified picture, especially compared to the FSU. FSU export structure is highly specialised in mineral-intensive products whose share of total exports has remained stable at around two-thirds. In contrast, export shares in Eastern Europe are fairly evenly distributed among agriculture, unskilled labour, technology and human-capital intensive goods. However, a noticeable shift has taken place away from technology-intensive goods towards agricultural and unskilled-labour intensive products. The export structure of East Asia lies in predominantly unskilled-labour and technology-intensive goods, with the latter showing the most dramatic rise in export share from 11 to 28 per cent. Examination of core and periphery economies also shows significant differences within East Asia. The NIEs specialise in technology-intensive and human capital exports while developing Asian economies specialise in unskilled-labour goods. APEC and EU export structures are dominated by human and technology-intensive goods in contrast to Australasia, where agriculture and resource-intensive goods dominate. Smaller differences in the regional commodity composition of trade arise in regard to imports. Import structure in all regions, including Eastern Europe and the FSU, is predominantly technological (with a small share of unskilled-labour imports). A similar but far more pronounced pattern is apparent for East Asia, whose share of technological imports (36 per cent) is well above that of the other regions (ranging around one-third). Export specialisation Trends in export specialisation are given in Table 5. Export specialisation is defined as the share of each commodity group in an economy s total exports relative to that commodity group s share of world exports

24 Table 4 Factor composition of total trade (per cent) Agriculture Human-capital Resource Technology Unskilled- Intensive intensive intensive intensive labour intensive X M X M X M X M X M APEC ASEAN China East Asia Eastern Europe EU Japan NIEs North America Australasia Rest of Asia Rest of the world FSU Western Europe World Sources: International Economic Databank, Australian National University; UN trade data. 18

25 Table 5 Export specialisation indexes APEC ASEAN China Eastern Eastern EEC-12 Japan North NIEs Australasia Rest of FSU Western Europe Asia America World Europe Agriculture intensive Human-capital intensive Mineral-resource intensive Technology intensive Unskilled-labour intensive Notes: See Appendix Table A1 for the commodity classification upon which the factors groups are based is derived. The export specialisation index is defined as the ratio of the share of a commodity group in total exports for a country or group of countries to that commodity group s share of world exports. Sources: International Economic Databank, Australian National University; UN trade data. 19

26 Summarising the results in Table 5, three trends are apparent: First, based upon export specialisation indexes, the present comparative advantage of Eastern Europe (excluding the FSU) appears to lie in agricultural and unskilled labour-intensive goods. The comparative advantage of the FSU lies unambiguously in mineral-intensive goods. In the post-reform period, the comparative advantage of Eastern Europe has intensified in the above areas and shifted away from technological goods. The comparative advantage of the FSU has remained fairly stable. Second, the comparative advantage of older, industrialised economies in human capital and technological goods has been stable over the past fifteen years. The most striking change is observed in advanced (core) East Asian economies (excluding Japan) with a shift away from labour-intensive towards technology-intensive goods. 15 Third, the export specialisation indexes for both pre-reform and post-reform data do not lend strong support to the hypothesis that the comparative advantage of Eastern Europe lies in human capital goods, at least in the transition period. However, specialisation indexes for human capital goods in Eastern Europe are comparable to those of NIEs and lie well above those of developing Asian economies including China. It is of interest to compare the above findings with RCAs derived using the data set and methodology adopted in Murrell (1990), which have been re-estimated and extended to include the post-reform period. Murrell s data set are based upon mirror statistics drawn from trade data of a sample of 44 countries representing 80 per cent of world trade. No separate category is included for Europe or East Asia and almost all older industrialised countries including the NIEs are classified within the region identified as market economies. Goods are classified according to whether they are Ricardo (natural resources) goods, Heckscher Ohlin (goods produced using a standard technology under constant-returns-to-scale) and product cycle (high technology) goods. 16 The results given in Table 5 match those obtained by Murrell (Table 6), which show that the comparative advantage of Eastern Europe in the pre-reform period lies in Heckscher Ohlin goods. (In the Soviet Union, comparative advantage lies in Ricardo goods.) Since 1990, the comparative advantage in Heckscher Ohlin goods has intensified while comparative disadvantage in technological goods has also increased. Comparative advantage of the FSU in resourceintensive goods has been retained. In contrast, market economies show little evidence of any change in RCAs. 20

27 Table 6 Revealed comparative advantage a Eastern Europe 6 b Ricardo goods Heckscher Ohlin goods Product-cycle goods Eastern Europe 9 c Ricardo goods Heckscher Ohlin goods Product-cycle goods Market economies d Ricardo goods Heckscher Ohlin goods Product-cycle goods Notes: a Based upon Murrell s (1990) classification system. See Appendix Table A2. b Eastern Europe 6 excludes the former FSU, the former Yugoslavia and Albania. c Eastern Europe 9 includes Bulgaria, Poland, Hungary, Romania, Czechoslavakia and East Germany (Till 1990). d Market economies include the OECD members plus South Korea, Singapore and Hong Kong. Source: UN trade data. Concluding remarks The purpose of this study has been to examine changes in trade flows and structure with a specific focus on Eastern Europe, the former Soviet Union and East Asia. It has sought to give a factual basis to explaining the unexpected expansion in trade between these regions in the aftermath of reform as well as to provide input into the present debate on Eastern Europe s comparative advantage. Increased trade flows between the above regions, viewed within the framework of changing regional trade structure, appear to reflect two concurrent developments: trade liberalisation within Eastern Europe and the FSU that has enabled former CPEs to fully exploit their existing areas of comparative advantage and changing comparative advantage of more advanced and developing Asian economies. The outcome of both forces is an observed acceleration of trade complementarity between the two regions. Behind these trends, similarity in trade structure and a shift towards technology-intensive exports by advanced East Asian 21