Structural adjustments in trade relations of the Visegrad group countries 1

Transcription

1 Structural adjustments in trade relations of the Visegrad group countries 1 Tomasz Brodzicki* Abstract: The goal of this article is to investigate structural adjustments in trade relations of the Visegrad countries resulting from their accession into the European Union and advanced stage of economic transition. Particular emphasis is placed on the similarity of adjustments in geographic trade patterns, the intensity and nature of intra-industry trade flows, comparative advantage indices as well as overall product variety in exports and imports. The empirical results for this particular group of small open-economies are compared with theoretical postulates of new trade theories. Work in progress, please do not quote. Paper to be presented at the 2011 Conference of the European Trade Study Group (ETSG). Keywords: intra-industry trade, product variety, comparative advantage, economic integration, Visegrad group JEL codes: F10, F14, F15 * Contact information: University of Gdansk Faculty of Economics Economics of European Integration Division Ul. Armii Krajowej 119/ Sopot, Poland mail: t.brodzicki@ug.edu.pl 1 This paper is part of a broader research project on competitive potential of new MS in the integration process with the EU ( Analiza porównawcza zmian konkurencyjności nowych krajów członkowskich w procesie integracji z Unia Europejską na przykładzie Polski, Węgier, Czech i Słowacji) ministerial research grant no 2898/B/H03/2010/39. 1

2 1 Introduction The Visegrad Group was formed in 1991 at a meeting of the presidents of Czechoslovak Republic, Hungary and Poland. There regional alliance consists nowadays of four countries due to division of Czechoslovakia and is frequently referred to as V4. In economic terms cooperation within V4 led to the creation of CEFTA. Since 1 May 2004 all V4 members have become Member States of the European Union and are nowadays functioning within its internal market. From economic perspective V4 is an interesting group of small open economies situated in Central Europe that have more or less successfully completed transition to fully-fledged market economy. Slovakia is the first of the group to have formally entered the eurozone (2009), the remaining countries have a temporary derogation as they do not fulfill the nominal convergence criteria of entry. From the beginning of economic transition roughly two decades ago the group has undergone tremendous structural adjustments which are clearly reflected in the geographical and product structure of their trade. The group seems to constitute an ideal natural experiment to analyze the consequences of economic transition intertwined with simultaneous economic integration. The goal of this article is to investigate these structural adjustments in trade relations of the Visegrad countries resulting from their accession into the European Union and advanced stage of economic transition. The analysis covers the period of (5 years preceding and following entry of V4 into the EU). Particular emphasis is placed on the similarity of adjustments in geographic trade patterns, the intensity and nature of intra-industry trade flows, comparative advantage indices as well as overall product variety in exports and imports. The empirical results for this particular group of small open-economies are compared with theoretical postulates of new trade theories. The scope of the paper is obviously limited by availability of data. The remainder of this paper is organized as follows. Section 2 presents theoretical considerations and discusses methods utilized in the subsequent analysis. The empirical analysis of trade relations of V4 Group countries is conducted in Section 3. Final section concludes and discusses some points requiring further more in depth research. 2 Theoretical considerations Since 1989 the V4 countries have been affected by two simultaneous processes rapid economic transition from inefficient, autarkic and overregulated centrally-planned economy towards fully fledged market-economy as well as gradual economic integration with the Western European structures. In the last 20 years they made a tremendous progress in most of 2

3 important indicators of economic and social indicators converging to EU mean. According to the latest edition of EBRD s Transition Report (2010) economic transition in the Czech Republic has been already completed while the other Visegrad countries are well advanced in the process. It is worth to note that V4 are converging not only towards the EU mean they are also converging towards each other (an evident club-convergence ). The initial discrepancies in income per capita, labor productivity as well as TFP levels among V4 have significantly diminished. From trade theory perspective trade relations of V4 countries at the beginning of transition could have been explained by traditional trade theories emphasizing differences in factor endowments and technological differences as their trade relations especially with the EU15 resembled typical North-South trade. With a high-share of interinustry trade, comparative advantage in low-skilled labor and resource intensive goods and specialization of low quality varieties of differentiated goods had a trade structure resembling developing states. Furthermore, most of the trade occurred within the COMECON 2. With time the structure evolved nowadays being much closer to a typical North-North trade due to geographical reorientation of trade relations as well as relatively fast progress in convergence to advanced type of an economy. The traditional theories of trade face also a problem when applied to trade relations within the V4 group itself a more or less coherent and homogenic group of similar economies in terms of the level of technological sophistication and factor endowments. Despite the asymmetry is in market sizes they all can be considered small open economies. From theoretical perspective it is also interesting whether the entry into EU led to a significant trade creation or trade diversion. Nowadays, most of V4 trade can be explained only within the framework of monopolistic competition with economies of scale and significant product differentiation. We have to stress, however, that pattern of international trade of a given group of countries typically reflects simultaneous interaction of several different causes. No single theoretical model is thus able to explain all patterns observed in reality. Traditional trade theory postulates that countries will specialize in production and export goods in which they have comparative advantage. In the present paper we utilize a related concept of revealed comparative advantage (RCA). RCA (Balassa 1965) concerns the relative trade performance of individual states in particular goods. Assuming that commodity pattern 2 COMECON formally seized to exist only in

4 of trade reflects the differences in relative costs, factor-endowments as well as other non-price factors, this is presumed to demonstrate comparative advantage of a given country. Traditional RCA indices are however frequently said to be biased. We utilize a different approach though a corrected revealed comparative advantage indicator (CRCA) which is not biased by significant imbalances in trade flows (Neven 1995, Cieślik 2001). The formula due to Neven (1995) is given below: [1] CRCA s = x i= 1 i, j x i, j m i= 1 i, j m i, j Positive value of the index is indicative of presence of comparative advantage in trade 3. Income convergence should with time lead to intensification of trade in varieties of differentiated products thus IIT should gain particular significance. IIT has been extensively modeled in the NTT literature. Interesting and significant early contributions include among others Krugman (1979, 1980), Helpman and Krugman (1985), Falvey and Kierzkowski (1987), Flam and Helpman (1987). We have to remember, however, that the pattern of intraindustry trade is highly unpredictable. Its intensity is however driven by similarity on supply and demand sides of trade partners. According to Helpman and Krugman (1985) the share of IIT in total trade between two countries is larger the smaller the difference in their relative factor endowments, given the constant relative country size. Thus, as the gap in relative factor endowments is slowly decreasing, the share of IIT in trade within V4 as well as in the trade with EU15 in particular should be steadily increasing. Furthermore, Helpman (1987) demonstrates that two countries of unequal size do not trade as much as two countries of similar size, holding constant the sum of both gross domestic products (GDP). IIT is frequently decomposed into its vertical and horizontal components. We have to stress here that determinants of vertical and horizontal IIT do differ (Greenaway 1994, 1995). Significant country-specific determinants of vertical IIT include both demand-side elements such as income distribution, market size, and similarity in preferences as well as supply-side determinants such as relative factor proportions and technological differences. Generally speaking according to Helpman and Krugman (1985) the larger the difference in factor endowments, the smaller (larger) should be extent of horizontal IIT (vertical IIT). (). In accordance with the demand overlapping model of Linder (1961) difference in income levels 3 The negative value of CRCA should be cautiously interpreted as a lack of comparative advantage as according to traditional Ricardian approach a given country will not produce a good in which it has comparative disadvantage (complete specialization occurs in bilateral trade). 4

5 should increase the share of vertical IIT (decrease the role of HIIT). Countries with lower per capita incomes should specialize and export low-quality varieties of a differentiated product (down-market specialization) whereas developed countries with higher per capita incomes should specialize and export high-quality varieties (up-market specialization). Bigger differences in market sizes between trade partners should increase the intensity of VIIT and lower intensity of HIIT in their mutual trade relations. Furthermore, similarity in factor endowments should increase intensity of HIIT and decrease the intensity of VIIT. Other country-specific determinants of IIT include geographical proximity, economic integration and barriers to trade. Intensity of IIT should be higher among closer partners, linked by economic integration agreements and having better access to their domestic markets. Empirical analyses show that sector-specific determinants (such as product differentiation, economies of scale, market structure, product life cycle, the role of multinational corporations) are on the whole less important in determining IIT flows. Furthermore, according to overlapping demands hypothesis of Linder (1961), nations with similar preference structures should develop similar industries and thus trade in similar, but differentiated goods. The intensity of IIT trade should increase in similarity of countries. We expect the Linder hypothesis to hold for the V4 group as they clearly converge in structural parameters. Moreover, according to Pugacewicz and Wincenciak (2007) an increase in the intensity of IIT in trade with advanced partners can be attributed mostly to higher rates of growth of income per capita (accelerated income convergence) as well as diminishing discrepancies in factor endowments (increase in K to L ratios). Consequently similarly to Kawecka-Wyrzykowska (2009) we expect vertical IIT to be relatively more pronounced between emerging and developed economies than between developed countries. At the same time IIT intensity within the V4 should be higher than in trade with EU15. Fontagne and Freudenberg (1997) pointed out sectoral aggregation bias of standard Grubel- Lloyd indices as a serious problem and thus following theoretical arguments proposed to empirically distinguish vertical IIT owing mainly to the international fragmentation of the production chain from horizontal intra-industry trade. In order to detect trade in similar products Fontagne and Freudenberg (1997) proposed that export and import unit values should differ by less than 15 per cent and that the lower trade flow amounts to at least 10 per cent of the higher trade flow 4. In the present study the intensity of intra-industry trade will be 4 The unit value approach is criticized for imperfections of trade statistics resulting from difficulties with right separation of products (unit values of two groups of products may also differ if the mix of products differs, so 5

6 measured by a modified version of the Grubel-Lloyed index (Grubel, Lloyd 1971) as proposed by Greenaway and Milner (1983) which adjusts the index for potential categorical aggregation bias. The adjusted GL index free of the potential distortion for sector s takes the following form: xsj msj j [2] GLs = 1 ( xsj + msj ) j, where x sj and m sj represent exports and imports of sector s and time t at a disaggregated level of individual product groups j. Higher values are indicative of higher intensity of intraindustry trade. A suggested by the literature of the subject we further break the intra-industry trade index into its two principal components of horizontal IIT and vertical IIT thus taking into account quality differences. We will further break the vertical IIT index into indices of VIIT in low quality products (the so-called down-market) and VIIT in high quality products (the so-called up-market) taking into account quality differences in trade in vertically differentiated products. In constructing the indices we follow the methodology utilized by Greenaway, D. et al (1995) in their empirical study of intra-industry trade in the UK which is based on relative unit values x a ratio of unit values in exports ( UV sj ) to UV in imports ( UV ) for a given product group j in sector s. Horizontally differentiated products are those that satisfy the following condition x UVsj [3] 1 α 1+ α m UV sj, where α is some dispersion factor. If relative unit value lies outside of this range we are dealing with vertically differentiated products. Taking into account quality differences we further differentiate between trade in low quality vertically differentiated product if relative UV is below of is above of α α and trade in high quality vertically differentiated product if relative UV Following Oulton (1991) we calculate unit values per ton. In line with the standards adopted in the empirical literature (please refer to Fontagne and Freudenberg 1997) we choose a m sj that one group may contain a higher proportion of high unit value items than the other one), consumers may buy a more expensive product for reasons other than quality and finally for setting an arbitrary threshold level for relative unit value (traditionally 15 per cent). 6

7 dispersion factor of 0.15, however, we test the robustness of the results assuming two other levels of the parameter 0.10 and The impact is significant results are highly vulnerable to the assumed level of the parameter the share of HIIT obviously increases in alpha. Lets assume that within sector s we have N disaggregated product groups j and that of them N H are horizontally differentiated, N LQV are low quality vertically differentiated products and N HQV are high quality vertically differentiated products. We can then obtain the following indices of horizontal IIT (HGL) as well as vertical IIT (VGL(LQ) and VGL(HQ) respectively): [4] HGL s NH ( xsj + msj ) j= 1 = N H ( xsj + msj ) j= 1 N j= 1 x sj m sj [5] [6] VGL ( LQ) s VGL ( LQ) s NLQV ( xsj + msj ) j= 1 = N LQV ( xsj + msj ) j= 1 HQV ( xsj + msj ) N j= 1 = N N j= 1 HQV x ( xsj + msj ) j= 1 N j= 1 x sj sj m m sj sj From basic calculus we should note that for a given sector s we should observe: [7] GL = HGL + VGL ( LQ) VGL ( HQ) s s s + s To some extent V4 countries and in particular Poland could be compared to Spain or Portugal. Martin-Montaner and Rios (2002) study the effects of differences in factor endowments on vertical intraindustry specialization as illustrated in Spanish IIT and find it to play a significant role. As Spain typically exports low-quality varieties to the northern countries of the OECD and high-quality varieties to the southern ones, V4 in their trade with EU15 should provide low-quality varieties (specialize in down-market varieties in exports). 7

8 Furthermore, Poland is a larger market than other V4. As such it thus creates more opportunities for product differentiation and exploitation of economies of scale. Assuming symmetric firms larger economy should be able to support a larger set of producers of differentiated products. At the same time the impact of home market effect should be more pronounced in larger economies. Consequently we should observe a higher number of goods or varieties of goods in Polish exports than in other V4 states. In existing bilateral trading relationship the trade may increase through time in commodities or varieties already present in the trade pattern, the so-called intensive margin of trade. It can otherwise be driven by introduction of new commodities or varieties, the so-called extensive margin of trade 5. The relative impact on countries export performance of each of the margins is rather inconclusive. Some authors have concluded the extensive margin to be a principal diver (Hummels and Klenow 2005) while others have found that the intensive margin plays the dominant role (e.g. Felbermayr and Kohler 2006, Helpman et al. 2008, Besedes and Prusa 2007). We expect the increase in trade potential of V4 to be mostly due to intensive margin of trade that is increase in trade volume in already existing trade relationships and established products and not in the in extensive margin of trade the increase in the number of new product lines. 3 Empirical analysis The study is based on own calculations using EUROSTAT trade data database COMEXT and conducted on the 4-digit level of CN (which gives 1630 potential product groups). The analysis covers the period , i.e. five years from accession of V4 countries and five years preceding their accession into the EU. In all cases, the nominal values of trade (in euro) were taken into account as well as quantity measured in tons. We analyze the trade pattern of each of the V4 group countries separately with other V4 MS, EU15 and the remaining MS of the EU-27 (EUN8 3 Baltic States, Slovenia, Cyprus, Malta, Romania and Bulgaria), internal and external trade of EU as well as all trade relations (global). We have to note that at first that V4 countries differ by their level of overall openness. The level of openness in 2008 varied from 87.6 for Poland to per cent for Czech Republic, for Slovakia and per cent in Hungary (Heston et al. 2011). The differences are 5 Amiti and Freund (2010) define the extensive margin as a number of products a country exports. 8

9 mostly due to the scale of individual economies with Poland the sixth biggest nation in the EU27 group. The adjustment to liberalized trade regime was rather painful, disruptive and far away from smooth. The V4 and in particular Poland suffered from structural deficit in trade with EU15 MS in the initial stage of economic transition which was also fueled by the significant inflow of FDI (later on this however led to improvement in trade deficit). In the analyzed period V4 exports increased significantly with an average speed of 14.1 per cent per annum. In accordance with our expectations the intra-eu exports increased at a much faster pace than extra-eu exports which resulted in the increase of share of intra-eu trade in total exports in all analyzed states. Within the EU V4 exports increased in particular in relations with other Visegrad states (16,4 rate of growth on average) and the EUN8 group (17,2 per cent on average). Exports to EU15 increased at a relatively slower but still high pace of 13,2 per cent on average. External exports had significantly lower average pace of 8,9 per cent (with Poland a clear exception 12,2 per cent). Overall there was significant trade creation with EU partners accompanied by overall trade expansion. There is little evidence for trade dispersion. Most of trade creation happened within the New Member States of the European Union despite of the prior formation of CEFTA and incorporation of V4 into the internal market of the European Union upon accession in May The extra-eu exports increased more than 3-fold in the case of Poland and 2,3-fold for Hungary, and nearly doubled in value terms in the case of Czech Republic and Slovakia. Overall, most of V4 exports and imports happen currently within the EU internal market in accordance with our expectations. The EU share varies from 68,1 for Hungary to 76,9 per cent for the Czech Republic in imports and from 76,9 for Hungary and 85,8 per cent for Slovakia in exports. In all countries EU share is higher in exports than in imports. It is interesting to note that overall increase in importance of intra-eu trade is due to two enlargements of the EU during the last decade. The share of the so-called old Europe the EU-15 decreased in all V4 countries the fall in exports for instance ranges from 9 per cent for Poland to 3,8 per cent for Slovakia. At the same time the role of intra-visegrad exports increased by 4,5 per cent for Hungary, 3,8 per cent for Poland, 3,0 per cent in the Czech Republic and diminished only in the case of Slovakia 1,6 per cent. Still in 2008 Slovakia was most dependent on intra-v4 exports 26,2 per cent with Poland being the least dependent of the four (10,8 per cent). The share of the EUN8 group increased in all Visegrad 9

10 countries and varied from 1,3 per cent in the case of Hungary to 5,3 per cent in the case of Poland. Central Europe is one of the most attractive regions for inflow of foreign direct investments. The impact of FDI is significant and clearly reflected in the structure trade of the V4 group. How similar is the trade pattern of Visegrad countries in product and geographical dimensions? We have calculated trade similarity index as a sum of squared deviation of shares in imports/exports treating Poland as a benchmark. The analysis has been conducted at the 2-digit level of disaggregation of CN for commodity dimension and total exports and imports with trade partners according to COMEXT classification for geographical dimension. The index similarly to classis Herfindahl index can take values from 0 to with lower values indicative of greater similarity. The structure of exports and imports of V4 countries are relatively similar to each other which could reveal structural similarity in export potential and its underlying base as well as relative homogeneity of preferences of consumers. This could include similarity in love for variety increasing in time with growing average incomes and their convergence to Western levels and within the group (club-convergence). It is worth to note that Czech and Slovakian trade pattern in commodity dimension ismore similar to Polish that Hungarian in both exports and imports (please refer to Figures 1 and 2). Overall similarity in exports (composition of goods) is lower than in imports which could be related to structural differences in comparative potential of countries. Surprisingly in geographical dimension, Hungary has a trade pattern most similar to Poland (please refer to Figures 3 and 4). Similarity of Slovakian trade pattern increased in exports and decreased in imports (with a clear change in trend after the entry into the EU). In accordance with our expectations Poland being the largest out of the V4 group has overall the highest product variety in exports with its principal trade partners (please refer to Table 3). The differences are however smaller that could have been expected. That could be a result of high aggregation of data analysis going to 6 or better 8-digit level could very well amplify the observed differences. In accordance with our expectations intensive margin of trade has a significant bearing on the trade of V4 states. Most of increase Polish exports (global dimension) in the period was due to higher intensity of trade in existing goods, with new goods responsible for only 0,2 per cent of value of exports in Only in the case of trade with EUN8 the role of extensive margin was significantly higher (10.6 per cent). In imports around 95,7 per cent of increase in the value of imports was due to intensive only 4,3 per cent due to extensive 10

11 margins respectively. Once again new goods were more pronounced in the case of EUN8. The role of the extensive margin is higher for imports than for exports. The pattern is quite similar for other V4 states with intensive margin playing a clearly dominant role. In these countries the role of extensive margin is the smallest in the extra-eu trade relations. It is however much higher for Slovakia in trade with EUN8 and EU15, and in the case of Hungary in trade with V4, EUN 8 and EU15. There is a clear heterogeneity in structural adjustments in this area. A high and growing proportion of trade of V4 takes place within rather between industries. We thus observe a growing intensity of intra-industry trade relative to inter-industry type (please refer to tables 4 to 8). Intra-industry specialization pushes out inter-industry specialization. Before the transformation started, the share of IIT in V4 and more general all across CEE was very low and horizontal IIT was almost non-existent (di Simone, 2007). The low share of IIT could be interpreted as an evidence of relative and absolute backwardness or underdevelopment of the region. As Michalek et al. (2009) correctly point out Visegrad group countries as well as other CEEs still trade more homogeneous (less differentiated) goods than more advanced EU economies which is reflected in much lower share of intra-industry trade in their trade flows. Factor proportions and technological differences as postulated by traditional trade theories still play a major role in determining their trade patterns and this is likely to continue for a considerable period of time. Nevertheless, rapidly rising IIT intensity is indicative of new and significant forces at play. Janda and Munich (2004) state that the Czech Republic had traditionally the highest IIT out of all transition economies of CEE. The structure of Czech IIT with respect to its horizontal and vertical elements is comparable with the structure of IIT in EU countries. In the case of Poland we observe an overall increase in intensity of intra-industry trade with all analyzed groups of partners. Overall it increased from 42.0 to 57.0 per cent between In comparison it went up by 4,8 per cent in Czech Republic, 5,2 per cent in Slovakia and 9,1 per cent in Hungary. It is worth to note that Poland and Slovakia had initially the lowest level of IIT in Nowadays its share exceeds 50 per cent benchmark. Initially and still it is the highest in the case of the Czech Republic. As could have been expected on the basis of theoretical postulates presented above the intensity of IIT is the highest in relations within the V4 group similarity of the Visegrad countries clearly counts. The intensity of IIT with the EU15 group could be considered as an 11

12 indirect measure of structural convergence of the V4 group. It noticeably increased in the last decade starting form rather moderate levels. Intensity of IIT is the lowest within extra-eu relations and interestingly enough it fell significantly in the analyzed period in the case of the Czech Republic and Slovakia and remained at roughly unchanged level in the case of Poland and Hungary. On average 2/3 of current extra-eu trade of V4 group is of inter-industry type. As we have mentioned before the internal structure of intra-industry trade is of particular importance. Poland has the highest share of HIIT (25,6) and Slovakia the lowest (14,0). Despite of changes VIIT clearly dominates within trade pattern of V4 however there is quite a a lot of heterogeneity in adjustments. The structure of Polish IIT changed the share of horizontal of IIT trade expanded significantly (from 7.1 to 25.6 per cent) while the significance of vertical IIT dropped. Similar adjustments happened in Hungary. Within IIT we see positive adjustments the share of lowquality goods decreased while the share of up-market goods increased. Poland is still here a laggard in comparison to other V4 states. The Czech Republic is the country with the highest intensity of VIIT with particularly strong shift from down-market to up-market specialization. It is particularly strong in the case of trade with EU15 which could be indicative of large competitive potential. In general, the most important structural adjustment in the trade pattern of V4 countries in the period is an increase in the intensity of intra-industry trade. Its intensity has been steadily increasing with major trade partners. The empirical results presented in this paper support the idea that the distinction between horizontal and vertical IIT is essential. Vertical IIT still plays a dominant role in trade pattern of V4 counties. However, the share of horizontal IIT is slowly but steadily increasing. Furthermore, the distinction between upmarket and down-market VIIT is significant. Traditionally V4 countries specialized in exports of down-market that is lower quality varieties of goods and importing up-market superiorquality varieties. The number of product groups with revealed comparative advantage in trade of V4 countries is decreasing overall (please refer to Table 9). They are more or less improving there position in comparison to extra-eu partners but loosing it with intra-eu partners and in particular with the EU15. Slovakia is here an exception as it made a significant qualitative improvement boosting its comparative potential especially towards other V4 and EUN8 states. 12

13 4 Conclusions and policy implications The trade pattern of Visegrad countries has undergone significant structural adjustments since the beginning of economic transition in The process continued in the second phase of transition analyzed in the present paper. EU Member States are most important trade partners of Visegrad countries. V4 are gradually converging to the structure of trade characteristic for more advanced economies. The intensity of intra-industry trade is steadily increasing with the Czech Republic being a clear leader. In comparison to advanced market economies the importance of relative factor endowments as the basis of comparative-advantage and as postulated by HOV theory is however still significant. IIT intensity clearly increases in similarity and is the highest within the V4 itself. As the group converges to EU mean this is likely to bring about further dramatic consequences in particular in trade with the EU15 group. At the same moment intensity of IIT is stagnating or even decreasing in trade relations with extra-eu trade partners. If this continue we are likely to converge to a typical North-South trade pattern. Furthermore, vertical IIT clearly dominates with horizontal IIT and up-market specialization in VIIT slowly gaining importance this could be considered a desired development. The geographical scope of V4 trade evolves in line with the traditional gravity approach and this pattern is unlikely to change in the foreseeable future. After accession to the EU trade creation was most evident within the group comprising V4 and EUN8 countries. The Rose effect related to euro-adoption is likely to be positive but not dramatic as the integration process is fairly advanced. The empirical results to a large extent support postulates of NTT literature emphasizing imperfect competition with representative (symmetric) firms. In accordance with new trade theory postulates increasing returns to scale matter for V4 Group. Absolute size of a country matters as well. The analysis should be conducted at higher level of commodity disaggregation in order to fully account for the role of product variety in trade and to fully acknowledge the role of extensive and intensive margin of trade. Further analysis will be required in order to estimate the size of potential Rose effect for V4 countries currently outside of the eurozone this time against the benchmark of Slovakia and its experiences preceding and following its entry into the eurozone in We lack resources to carry out comprehensive analysis of firm-level data on trade which could allow us to provide evidence for or arguments against theoretical postulates of the new 13

14 new trade theory concentrated on the presence of heterogeneous firms (attributed mainly to Melitz 2003). In a more elaborate model of Melitz and Ottaviano (2008) scale of a country matters - bigger markets exhibit higher levels of product variety and host more productive firms that set lower markups and thus are able to set lower prices (are more pricecompetitive). Larger market firms are bigger, more productive, earn higher profits (despite lower average markups), but face lower probability of survival at entry as they face tougher competition. A distinctive market selection mechanism is at work. Trade forces the least productive firms to exit and reallocates market shares towards more productive exporting firms (lower productivity firms only serve their domestic market or have to exit this market as well). Trade liberalization leads to pro-competitive effect which is welfare-improving at least in the short-run. Some empirical evidence available supports the presence of selection mechanisms in trade. The heterogeneity of firm potential could be even more evident than in the case of most advanced economies for which we have detailed results. 14

15 Literature Balassa, B. (1965), Trade Liberalization and Revealed Comparative Advantage, The Manchester School, Vol. 33, pp Besedes, T. (2010), Export Differentiation in Transition Economies, MPRA Paper Besedes T., Prusa T. (2007), The Role of Extensive and Intensive Margins and Export Growth, NBER Working Papers Cieślik, A. (2001), Nowa teoria handlu zagranicznego w świetle badań empirycznych, PWN, Warsaw. EBRD (2010) Recovery and Reform. Transition Report 2010, EBRD, London. Falvey R., Kierzkowski H. (1987), Product Quality, Intra-Industry Trade and (Im)Perfect Competition, in: Kierzkowski H. (ed.), Protection and Competition in International Trade. Essays in Honor of W.M. Corden, Basil Blackwell, pp Feenstra R. (2004), Advanced International Trade. Theory and Evidence, Princeton University Press, Princeton. Felbermayr G., Kohler W. (2006), Exploring the Intensive and Extensive Margins of World Trade, Review of World Economics, Vol. 142(4), pp Flam H., Helpman E. (1987), Vertical Product Differentiation and North-South Trade, American Economic Review, Vol. 77(5), pp Greenaway D., Milner C. (1983), On the Measurement of Intra-Industry Trade, The Economic Journal, Vol. 93, pp Greenaway, D. et al (1995), Vertical and Horizontal Intra-Industry Trade: A Cross Industry Analysis for the United Kingdom, Economic Journal, Vol. 105, pp Grubel H, Lloyd P.(1961), The Empirical Measurement of Intra-Industry Trade, The Economic Record, Vol. 47, pp Hummels D., Klenow P., (2005), The Variety and Quality of a Nation's Exports, American Economic Review, Vol. 95(3), pp Helpman E., Krugman P. (1985), Market Structure and International Trade. Increasing Returns, Imperfect Competition and the International Economy, MIT Press, Cambridge. Helpman E. et al. (2008), Estimating Trade Flows: Trading Partners and Trading Volumes, Quarterly Journal of Economics, Vol. 123(2), pp Heston A. et al. (2011), Penn World Table Version 7.0, Center for International Comparisons of Production, Income and Prices at the University of Pennsylvania, May Hine R.C. et al., Vertical and Horizontal Intra-Industry Trade: An Analysis of Country- and Industry- Specific Determinants, in: Brülhart M., Hine R.C. (1998), Intra-Industry Trade and Adjustment, Macmillan, London, pp Janda K., Munich D. (2004), The Intra-Industry Trade of the Czech Republic in the Economic Transition, Emerging Markets Finance and Trade, 2004, Vol. 40(2), p Kawecka-Wyrzykowska E., Evolving pattern of intra-industry trade specialization of the Visegrad countries, in: SGH (2009) Five years of the EU Eastward Enlargement Effects on Visegrad Countries: Lessons for the Future, SGH, Warsaw. 15

16 Kohler W., Felbermayr G. (2004), Exploring the Intensive and Extensive Margins of World Trade, CESifo Working Paper Series No Krugman P. (1979), Increasing Returns, Monopolistic Competition, and International Trade, Journal of International Economics, Vol. 9(4), pp Krugman P. (1980), Scale Economies, Product Differentiation, and the Pattern of Trade, American Economic Review, Vol. 70(5), pp Linder S. (1961), An Essay on Trade and Transformation, John Wiley, New York. Martín-Montaner J., Rios V. (2002), Vertical specialization and intra-industry trade: The role of factor endowments, Review of World Economics, Vol. 138 (2), pp Melitz M.J, Ottaviano GM (2005), Market Size, Trade, and Productivity, NBER Working Paper no Michałek J. et al. (2009), Trade effects of the euro adoption in Central and Eastern Europe, paper presented during ETSG Neven, D. (1994), Trade liberalization with Eastern nations: How Sensitive?, CEPR Discussion Paper no. 1000, London. Neven, D. (1995), Trade liberalization with Eastern Nations: Some Distribution Issues, European Economic Review, Vol. 39. Oulton, N. (1991), Quality Performance in UK Trade , NIESR Discussion Paper No 197. Pugacewicz A., Wincenciak L., Struktura handlu Polski w świetle teorii handlu in: Krugman P., Obstfeld M. (2007), Ekonomia międzynarodowa, PWN, Warsaw. 16

17 Table 1 Geographical structure of V4 exports Poland V4 7,1 7,2 7,5 7,6 8,1 8,7 9,3 10,7 10,6 10,8 EUN8 4,1 4,1 4,5 4,7 5,0 4,4 4,3 4,6 5,3 5,3 EU15 70,5 69,9 69,2 68,8 68,8 67,2 65,0 63,8 63,0 61,5 INTRA EU 70,5 69,9 69,2 68,8 68,8 75,5 77,2 77,6 79,0 77,6 EXTRA EU 29,5 30,1 30,8 31,2 31,2 24,5 22,8 22,6 21,1 22,6 GLOBAL 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 Hungary V4 11,3 11,4 11,0 11,2 11,3 12,4 13,7 14,9 15,6 15,8 EUN8 1,0 1,0 1,0 1,1 1,2 1,1 1,3 1,4 1,4 1,3 EU15 64,2 62,8 62,6 60,2 58,9 66,6 66,4 64,1 63,2 59,8 INTRA EU 64,2 62,8 62,6 60,2 58,9 76,4 80,9 80,0 80,1 76,9 EXTRA EU 35,8 37,2 37,4 39,8 41,1 23,5 19,0 20,0 19,9 23,1 GLOBAL 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 Czech Republic V4 15,5 15,0 15,1 14,9 15,0 16,0 16,8 17,2 17,8 18,5 EUN8 2,6 2,4 2,3 2,4 2,4 2,4 2,7 2,9 3,1 3,2 EU15 69,4 68,5 69,0 68,4 69,8 68,7 66,1 65,9 64,7 63,5 INTRA EU 69,4 68,5 69,0 68,4 69,8 81,2 84,2 84,3 85,6 85,2 EXTRA EU 30,6 31,5 31,0 31,6 30,2 18,7 15,9 16,0 14,8 15,1 GLOBAL 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 Slovakia V4 27,8 28,0 27,7 25,9 22,5 24,1 26,5 26,1 24,9 26,2 EUN8 2,2 2,5 2,8 2,9 2,7 2,9 3,5 3,3 3,7 3,9 EU15 59,5 59,2 60,1 60,7 60,8 59,7 57,4 57,7 58,6 55,7 INTRA EU 59,5 59,2 60,1 60,7 60,8 78,2 85,5 85,3 87,2 85,8 EXTRA EU 40,5 40,8 39,9 39,3 39,2 21,8 14,7 15,0 13,3 14,8 GLOBAL 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 Source: Own calculations based on COMEXT. 17

18 Table 2 Geographical structure of V4 imports Poland V4 5,8 6,3 6,6 6,4 6,8 7,4 7,5 7,8 8,0 7,9 EUN8 1,4 1,6 1,7 1,6 1,7 2,3 2,4 2,0 2,0 2,0 EU15 64,9 61,1 61,4 61,7 61,1 65,4 65,3 63,1 63,2 61,2 INTRA EU 64,9 61,1 61,4 61,7 61,1 71,9 74,7 72,3 73,2 71,2 EXTRA EU 35,1 38,9 38,6 38,3 38,9 27,8 25,2 27,6 26,7 28,7 GLOBAL 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 Hungary V4 5,7 5,8 6,2 6,6 7,1 8,0 8,9 10,0 10,5 11,3 EUN8 1,6 1,9 1,9 2,1 2,3 2,6 3,3 3,9 3,4 3,4 EU15 64,4 58,4 57,8 56,2 55,0 57,6 57,7 56,4 55,7 53,4 INTRA EU 64,4 58,4 57,8 56,2 55,0 64,1 67,7 67,3 69,7 68,1 EXTRA EU 35,6 41,6 42,2 43,8 45,0 35,6 32,3 32,8 30,6 32,2 GLOBAL 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 Czech Republic V4 11,3 11,4 11,0 11,2 11,3 12,4 13,7 14,9 15,6 15,8 EUN8 1,0 1,0 1,0 1,1 1,2 1,1 1,3 1,4 1,4 1,3 EU15 64,2 62,8 62,6 60,2 58,9 66,6 66,4 64,1 63,2 59,8 INTRA EU 64,2 62,8 62,6 60,2 58,9 76,4 80,9 80,0 80,1 76,9 EXTRA EU 35,8 37,2 37,4 39,8 41,1 23,5 19,0 20,0 19,9 23,1 GLOBAL 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 Slovakia V4 21,8 20,1 21,1 21,3 21,4 26,4 28,9 28,9 29,0 29,6 EUN8 1,0 1,0 1,0 1,3 1,6 1,4 1,3 1,4 1,5 2,0 EU15 51,7 49,1 49,9 50,5 51,5 50,7 47,5 44,8 44,0 41,4 INTRA EU 51,7 49,1 49,9 50,5 51,5 71,5 77,2 74,4 74,4 73,0 EXTRA EU 48,3 50,9 50,1 49,5 48,5 28,2 22,7 25,6 25,6 27,2 GLOBAL 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 100,0 Source: Own calculations based on COMEXT. 18

19 Table 3 Number of product varieties in exports of V4 Group Poland V EUN EU EXTRA EU Hungary V EUN EU EXTRA EU Czech Republic V EUN EU EXTRA EU Slovakia V EUN EU EXTRA EU Source: Own calculations based on COMEXT (4-digit CN). 19

20 Table 4 Intensity of horizontal IIT (HIIT) Poland V4 0,162 0,154 0,206 0,262 0,268 0,243 0,229 0,167 0,190 0,211 EUN8 0,056 0,061 0,062 0,051 0,094 0,105 0,113 0,145 0,149 0,117 EU15 0,050 0,066 0,090 0,117 0,187 0,215 0,150 0,161 0,224 0,164 INTRA EU 0,050 0,066 0,090 0,117 0,187 0,229 0,175 0,215 0,264 0,271 EXTRA EU 0,077 0,094 0,106 0,098 0,119 0,151 0,051 0,094 0,103 0,100 GLOBAL 0,071 0,091 0,124 0,141 0,193 0,242 0,183 0,223 0,276 0,256 Hungary V4 0,127 0,109 0,113 0,157 0,114 0,178 0,135 0,112 0,108 0,158 EUN8 0,071 0,086 0,059 0,104 0,053 0,092 0,039 0,071 0,091 0,106 EU15 0,082 0,073 0,162 0,131 0,124 0,111 0,137 0,163 0,174 0,186 INTRA EU 0,082 0,073 0,162 0,131 0,124 0,109 0,166 0,175 0,160 0,210 EXTRA EU 0,067 0,079 0,062 0,096 0,069 0,084 0,063 0,059 0,118 0,080 GLOBAL 0,072 0,131 0,156 0,144 0,108 0,133 0,150 0,167 0,173 0,186 Czech Republic V4 0,178 0,211 0,205 0,205 0,225 0,200 0,227 0,172 0,179 0,168 EUN8 0,039 0,042 0,060 0,067 0,068 0,056 0,077 0,045 0,081 0,079 EU15 0,128 0,182 0,146 0,074 0,091 0,121 0,136 0,103 0,120 0,108 INTRA EU 0,128 0,183 0,146 0,074 0,091 0,148 0,190 0,163 0,206 0,163 EXTRA EU 0,128 0,126 0,107 0,139 0,138 0,133 0,080 0,090 0,097 0,073 GLOBAL 0,181 0,182 0,184 0,132 0,113 0,130 0,192 0,166 0,202 0,175 Slovakia V4 0,124 0,119 0,145 0,203 0,163 0,121 0,097 0,135 0,121 0,104 EUN8 0,060 0,055 0,043 0,043 0,052 0,037 0,042 0,103 0,051 0,163 EU15 0,099 0,099 0,050 0,143 0,040 0,093 0,110 0,106 0,047 0,065 INTRA EU 0,099 0,099 0,050 0,143 0,040 0,100 0,116 0,120 0,075 0,092 EXTRA EU 0,097 0,109 0,131 0,157 0,141 0,096 0,036 0,063 0,074 0,057 GLOBAL 0,109 0,142 0,103 0,128 0,096 0,126 0,145 0,118 0,075 0,140 Source: Own calculations based on COMEXT (4-digit CN, adjusted GLI, alpha 0.15). 20

21 Table 5 Intensity of vertical IIT (VIIT) with down-market specialization (low-quality varieties) Poland V4 0,174 0,180 0,186 0,146 0,138 0,162 0,196 0,203 0,149 0,139 EUN8 0,111 0,115 0,118 0,125 0,119 0,118 0,087 0,059 0,108 0,117 EU15 0,258 0,248 0,227 0,215 0,225 0,189 0,195 0,226 0,186 0,254 INTRA EU 0,258 0,248 0,227 0,215 0,225 0,195 0,206 0,207 0,184 0,192 EXTRA EU 0,155 0,146 0,143 0,164 0,146 0,146 0,123 0,109 0,121 0,112 GLOBAL 0,248 0,231 0,212 0,205 0,209 0,192 0,187 0,188 0,159 0,180 Hungary V4 0,090 0,099 0,097 0,073 0,100 0,085 0,085 0,128 0,134 0,114 EUN8 0,095 0,101 0,109 0,101 0,110 0,131 0,155 0,125 0,086 0,088 EU15 0,192 0,229 0,190 0,198 0,168 0,169 0,176 0,183 0,141 0,113 INTRA EU 0,192 0,229 0,190 0,198 0,168 0,169 0,157 0,182 0,130 0,114 EXTRA EU 0,123 0,141 0,152 0,147 0,102 0,194 0,159 0,138 0,209 0,193 GLOBAL 0,179 0,197 0,195 0,182 0,151 0,171 0,150 0,148 0,179 0,163 Czech Republic V4 0,135 0,132 0,123 0,133 0,130 0,147 0,124 0,102 0,138 0,147 EUN8 0,152 0,112 0,112 0,113 0,131 0,117 0,106 0,185 0,084 0,113 EU15 0,365 0,307 0,344 0,318 0,294 0,205 0,240 0,167 0,161 0,170 INTRA EU 0,365 0,307 0,344 0,318 0,294 0,207 0,238 0,152 0,153 0,166 EXTRA EU 0,181 0,180 0,174 0,160 0,160 0,157 0,149 0,128 0,100 0,084 GLOBAL 0,306 0,296 0,291 0,269 0,258 0,195 0,218 0,206 0,186 0,191 Slovakia V4 0,156 0,123 0,140 0,155 0,160 0,209 0,174 0,192 0,223 0,249 EUN8 0,078 0,057 0,063 0,079 0,080 0,096 0,078 0,087 0,118 0,115 EU15 0,205 0,190 0,198 0,195 0,191 0,181 0,160 0,143 0,179 0,191 INTRA EU 0,205 0,190 0,198 0,195 0,191 0,192 0,170 0,156 0,181 0,220 EXTRA EU 0,143 0,108 0,129 0,123 0,128 0,120 0,094 0,065 0,046 0,047 GLOBAL 0,200 0,165 0,181 0,193 0,179 0,168 0,154 0,137 0,161 0,156 Source: Own calculations based on COMEXT (4-digit CN, adjusted GLI, alpha 0.15) 21

22 Table 6 Intensity of vertical IIT (VIIT) with up-market specialization (high-quality varieties) Poland V4 0,080 0,083 0,086 0,091 0,086 0,098 0,099 0,136 0,178 0,183 EUN8 0,085 0,089 0,086 0,121 0,091 0,108 0,156 0,168 0,133 0,169 EU15 0,086 0,114 0,112 0,117 0,044 0,075 0,131 0,099 0,092 0,107 INTRA EU 0,086 0,114 0,112 0,117 0,044 0,081 0,134 0,106 0,103 0,107 EXTRA EU 0,120 0,101 0,138 0,143 0,165 0,125 0,202 0,180 0,156 0,161 GLOBAL 0,100 0,118 0,131 0,143 0,097 0,093 0,153 0,121 0,116 0,133 Hungary V4 0,139 0,168 0,154 0,149 0,176 0,183 0,232 0,247 0,294 0,264 EUN8 0,135 0,124 0,139 0,151 0,156 0,178 0,218 0,215 0,195 0,186 EU15 0,163 0,148 0,104 0,121 0,139 0,201 0,176 0,173 0,206 0,222 INTRA EU 0,163 0,148 0,104 0,121 0,139 0,223 0,191 0,183 0,279 0,254 EXTRA EU 0,162 0,128 0,175 0,169 0,196 0,221 0,203 0,212 0,120 0,142 GLOBAL 0,232 0,182 0,154 0,191 0,231 0,250 0,262 0,246 0,221 0,226 Czech Republic V4 0,182 0,157 0,190 0,184 0,182 0,216 0,229 0,335 0,293 0,317 EUN8 0,088 0,110 0,115 0,121 0,125 0,179 0,173 0,140 0,196 0,166 EU15 0,080 0,074 0,092 0,188 0,189 0,273 0,205 0,307 0,306 0,326 INTRA EU 0,080 0,074 0,092 0,188 0,189 0,278 0,201 0,312 0,282 0,333 EXTRA EU 0,152 0,138 0,171 0,171 0,165 0,156 0,154 0,172 0,212 0,216 GLOBAL 0,103 0,097 0,120 0,218 0,254 0,308 0,200 0,236 0,242 0,273 Slovakia V4 0,146 0,181 0,184 0,141 0,180 0,157 0,228 0,215 0,213 0,226 EUN8 0,048 0,077 0,095 0,096 0,069 0,091 0,112 0,081 0,094 0,082 EU15 0,102 0,098 0,159 0,076 0,191 0,158 0,152 0,165 0,185 0,192 INTRA EU 0,102 0,098 0,159 0,076 0,191 0,190 0,211 0,222 0,243 0,222 EXTRA EU 0,148 0,160 0,158 0,147 0,135 0,115 0,089 0,076 0,088 0,112 GLOBAL 0,137 0,110 0,166 0,134 0,183 0,174 0,167 0,209 0,231 0,203 Source: Own calculations based on COMEXT (4-digit CN, adjusted GLI, alpha 0.15) 22

23 Table 7Overall intensity of IIT Poland V4 0,416 0,417 0,478 0,499 0,491 0,502 0,524 0,506 0,518 0,533 EUN8 0,252 0,265 0,266 0,298 0,303 0,331 0,356 0,372 0,390 0,403 EU15 0,394 0,428 0,429 0,449 0,456 0,479 0,476 0,487 0,503 0,525 EXTRA EU 0,352 0,341 0,387 0,404 0,430 0,423 0,376 0,383 0,380 0,372 GLOBAL 0,420 0,440 0,467 0,490 0,499 0,527 0,523 0,533 0,551 0,570 Hungary V4 0,356 0,376 0,364 0,379 0,390 0,446 0,452 0,487 0,537 0,536 EUN8 0,301 0,311 0,307 0,355 0,319 0,401 0,412 0,412 0,373 0,380 EU15 0,436 0,450 0,456 0,449 0,432 0,481 0,488 0,519 0,521 0,521 EXTRA EU 0,352 0,348 0,390 0,412 0,368 0,499 0,425 0,410 0,447 0,415 GLOBAL 0,484 0,510 0,505 0,517 0,490 0,555 0,562 0,562 0,574 0,574 Czech Republic V4 0,496 0,500 0,518 0,522 0,537 0,562 0,581 0,609 0,610 0,632 EUN8 0,280 0,265 0,287 0,300 0,325 0,352 0,356 0,370 0,361 0,358 EU15 0,572 0,564 0,582 0,580 0,574 0,599 0,581 0,577 0,587 0,604 INTRA EU 0,572 0,564 0,582 0,580 0,574 0,633 0,629 0,627 0,642 0,662 EXTRA EU 0,462 0,445 0,452 0,470 0,463 0,445 0,382 0,390 0,409 0,373 GLOBAL 0,591 0,575 0,595 0,620 0,625 0,633 0,610 0,608 0,630 0,639 Slovakia V4 0,426 0,424 0,469 0,499 0,502 0,486 0,499 0,541 0,557 0,579 EUN8 0,185 0,189 0,201 0,218 0,202 0,224 0,232 0,272 0,263 0,360 EU15 0,406 0,387 0,406 0,414 0,422 0,432 0,422 0,414 0,412 0,448 EXTRA EU 0,388 0,378 0,417 0,427 0,404 0,331 0,219 0,204 0,207 0,216 GLOBAL 0,446 0,417 0,450 0,455 0,458 0,468 0,466 0,464 0,466 0,498 Source: Own calculations based on COMEXT (4-digit CN, adjusted GLI). 23

24 Table 8 Share of vertical component in IIT (in per cent) Poland V4 61,0 63,2 56,9 47,5 45,6 51,7 56,3 67,0 63,3 60,5 EUN8 77,6 76,9 76,8 82,7 69,1 68,2 68,2 61,0 61,8 70,9 EU15 87,4 84,5 79,0 74,0 59,0 55,1 68,5 66,8 55,4 68,8 INTRA EU 87,4 84,5 79,0 74,0 59,0 54,7 66,0 59,3 52,1 52,4 EXTRA EU 78,1 72,5 72,7 75,9 72,3 64,2 86,4 75,4 73,0 73,2 GLOBAL 83,0 79,3 73,4 71,1 61,3 54,1 65,1 58,2 49,8 55,0 Hungary V4 64,4 71,1 68,9 58,5 70,8 60,0 70,1 77,0 79,8 70,6 EUN8 76,4 72,5 80,8 70,9 83,5 76,9 90,6 82,8 75,5 72,1 EU15 81,3 83,7 64,5 70,8 71,3 76,9 72,0 68,7 66,5 64,3 INTRA EU 81,3 83,7 64,5 70,8 71,3 78,2 67,7 67,5 71,9 63,7 EXTRA EU 81,1 77,3 84,0 76,8 81,2 83,2 85,1 85,6 73,6 80,7 GLOBAL 85,0 74,3 69,2 72,2 77,9 75,9 73,3 70,2 69,8 67,6 Czech Republic V4 64,1 57,7 60,4 60,7 58,2 64,5 60,8 71,7 70,7 73,4 EUN8 85,9 84,2 79,0 77,8 78,9 84,1 78,4 87,9 77,5 77,9 EU15 77,7 67,6 74,8 87,2 84,1 79,9 76,6 82,1 79,5 82,2 INTRA EU 77,7 67,6 74,8 87,2 84,1 76,6 69,8 74,0 67,8 75,3 EXTRA EU 72,3 71,6 76,4 70,3 70,2 70,2 79,2 77,0 76,2 80,5 GLOBAL 69,3 68,3 69,1 78,7 81,9 79,4 68,5 72,7 68,0 72,6 Slovakia V4 70,8 71,8 69,2 59,4 67,6 75,2 80,6 75,1 78,2 82,0 EUN8 67,8 71,0 78,8 80,1 74,2 83,4 82,1 62,0 80,8 54,7 EU15 75,6 74,5 87,8 65,4 90,5 78,4 73,9 74,3 88,5 85,5 INTRA EU 75,6 74,5 87,8 65,4 90,5 79,3 76,7 76,0 85,0 82,8 EXTRA EU 74,9 71,0 68,7 63,3 65,1 71,1 83,4 68,9 64,3 73,5 GLOBAL 75,6 65,9 77,2 71,8 79,1 73,0 68,9 74,6 84,0 71,9 Source: Own calculations based on COMEXT (4-digit CN, adjusted GLI). 24