Horizontal and Vertical Intra-Industry Trade between the Former cefta Countries and the European Union Stanislav Cernoša This paper analyses the production structure or intra-industry trade specialization (iit) of the Czech Republic, Hungary, Poland, Slovenia and Slovakia in foreign trade with eu member states from 1995 to 2001 at the five-digit level of the sitc. The results confirmed that former cefta countries in general showed iit specialization in the produc- tion of vertically differentiated products of lower quality either at the aggregate level of the manufacturing sectors (i.e. sitc 5-8) or at the level of the twenty chosen manufacturing activities (i. e. divisions 17-36 of the isic) and that differences also exist between iit specializations of these countries. Key Words: intra-industry trade, horizontal and vertical intra-industry trade, cefta countries jel Classification: F14 Introduction The fact is that the Czech Republic, Hungary, Poland, Slovenia and Slo- vakia as former cefta countries are less developed measured by gdp per capita1 than Austria, Belgium, France, Germany and Great Britain as relatively developed eu member states. If we use other economic indica- tors such as openness and gdp growth rate, than the position of former cefta countries is somewhat better in comparison with advanced coun- tries. But differences between gdp per capita income of former cefta countries and developed eu states are crucial for better understanding of the correlation between the stage of development of each observed country and its intra-industry specialization.2 It is also well known that the eu is the main foreign trade partner of the former cefta member countries.3 Therefore, the analysis of the foreign trade in European tran- sition countries provides a good opportunity for better understanding of the determinants of intra-industry trade specialization of these coun- tries. The purpose of this paper is to verify the production structure or Dr Stanislav (Cernoša is the Chief Salesman at Založba Aristej, d. o. o, Maribor, Slovenia. Managing Global Transitions 5 (2): 157-178 intra-industry specialization that former cefta countries revealed in trade with eu member states at the aggregate level of the manufactur- ing sectors (i. e. sitc 5-8) in 1995,1998 and 2001 and at the level of the twenty chosen manufacturing activities (i. e. divisions 17-36 of the isic) in 2001. In this way we test the hypothesis that five former cefta coun- tries showed predominant specialization in the production of vertically differentiated products of lower quality in their foreign trade with eu member states. The previous study (CCernoša 2005a) also analysed the production structure in Slovenia from 1994 to 2003 and revealed predominant spe- cialization of this country in the production of lower quality products. A similar study (CCernoša 2005b) verified intra-industry trade special- ization for Slovenia and Czech Republic in comparison with Germany and Austria in 1998 and showed that both Central European countries were specialized in the production of lower quality products, while Ger- many and Austria were simultaneously specialized in the production of vertically differentiated products of higher quality. The empirical study by Aturupane, Djankov and Hoekman (1999) also analysed the determi- nants of horizontal and vertical intra-industry trade of eight Central and Eastern European States in foreign trade with the European Union from 1990 to 1995. 4 Thus in the present study, the observed years 1995,1998 and 20015 were chosen in order to be able to compare the results of the measurements of horizontal and vertical intra-industry trade of this analysis with those obtained in the previous studies cited above. It is common knowledge that production structure or iit specialization is relatively stable and that methodology for the measurement of horizontal and vertical iit is based on unit value index (UV).6 Thus Greenaway, Hine, and Mil- ner (1994; 1995) in their fundamental studies use statistical data for only one observed year as basis for the measurement of horizontal and ver- tical iit.7 In this way, we suppose that the present study clearly repre- sents significant production structure or predominant iit specialization of the Czech Republic, Hungary, Poland, Slovenia and Slovakia as former cefta countries at the beginning of the twenty-first century. If the new theory of international trade supposes intra-industry trade in horizontally or vertically differentiated products between two coun- tries, then the theory of comparative advantage supposes inter-industry trade in homogeneous products between two countries. During the 1980s, different models of the new theory of international trade were developed. Thus, for example, Helpman and Krugman's book (1985) presented several theories that supposed trade with horizontally differ- entiated products. One of the basic messages of this book was that the traditional theory of comparative advantage is still alive and well, and that it had lost only its monopoly position. While during the last twenty years many other authors deepened the theoretical background of the new theory of international trade, the traditional theory of comparative advantages survived in different modifications. Thus one of the impor- tant messages of this paper is that it is still possible to explain trade between less developed cefta member countries and eu member states with the modified version of the Heckscher-Ohlin model, which is a significant representative of the theory of comparative advantage. The paper is structured as follows. The first section presents the the- oretical background for the new theories of international trade. The second section presents the methodology for the measurement of hor- izontal and vertical intra-industry trade, while the third section empiri- cally tests the production structure of each of the observed economies at the aggregated level. Similarly, the fourth section verifies intra-industry trade specialization for twenty manufacturing activities (divisions isic 17-36) of observed countries trading with the eu in 2001. The final part of this paper presents concluding comments. The Theoretical Background Research on two-way trade in similar products in the 1960s and the 1970s was mainly focused on the empirical estimation of the phenomenon of intra-industry trade (iit). Thus Grubel and Lloyd (1975) empirically confirmed that intra-industry trade is a real phenomenon and that the levels of intra-industry trade grow faster within the trade between devel- oped countries which are members of custom unions or other regional trading arrangements, than in the trade of the developed countries with other countries. One of the important distinctions made in theoretical literature is a distinction between horizontal and vertical product differ- entiation. The former arises when different varieties of the product are of a similar quality and the latter when varieties of the product are differ- entiated by quality. Thus vertical product differentiation is related more to the traditional theory8 of international trade and its modified version, while horizontal product differentiation is related to the new theories of international trade, which supposes horizontal product differentiation.9 All of this is well known, but it was empirically under researched due to difficulties connected with disentangling vertical and horizontal intra- industry trade.10 Therefore, the majority of studies investigated intra- industry trade exchange between highly developed economies. These works on iit, which estimated regression models for developed coun- tries, have generally found more support for the importance of country specific effects (i. e, gdp per capita) as opposed to industry specific fac- tors (Greenaway, Hine, and Milner 1995). Trade between former cefta countries and eu member states should be driven by differences in fac- tor endowments and also differences in technologies. Accordingly, the Aturupane, Djankov and Hoekman (1999) study estimated a regression model and empirically confirmed that vertical iit of the former Central European countries is positively associated with product differentiation, economy of scale, labour intensity of production and foreign direct in- vestment (fdi).u Since many analyses 12 confirmed that the share of vertical iit was larger than the share of horizontal iit for the observed cefta coun- tries, intra-industry trade in vertically differentiated products of these countries can be explained using the Falvey model (1981). The Falvey model is based on some presumptions of the Heckscher-Ohlin model, but it also includes a modification with the new presumptions. While the Heckscher-Ohlin model supposes trade in homogeneous products and inter-industry trade between two countries, the Falvey model supposes trade in vertically differentiated products and iit. The Falvey model also supposes that iit between two countries occurs at least in one industry, in which the home country produces and exports its own specific range of differentiated products and imports other types or a range of vertically differentiated products from a neighbour country. So the Falvey model supposes that the manufacturing industry of each country is best defined by capital^ and that each industry is no longer assumed to produce a sin- gle homogeneous output, but instead can produce a range of products using as inputs labour and its own industry specific capital. Therefore, the manufacturing industry of each country has different endowments of capital (K) and labour (L). The different endowments result in different factor prices in two coun- tries with the foreign price of labour being relatively low and the do- mestic price of capital being relatively low. It is assumed that the higher capital-labour ratio results in a higher quality of the vertically differenti- ated products. As a consequence of the assumption that a higher capital- labour ratio results in a higher quality, the capital abundant country will export relatively high quality products, while the labour-abundant coun- try will export relatively low quality products. The direction of trade is also determined by this model and provides an explanation of iit in vertically differentiated products between developed countries and less developed states. 14 Methodology Horizontal and vertical intra-industry trade of the five observed cefta countries was measured by using the Greenaway, Hine and Milner (1994; 1995) methodology. There also exists an alternative methodology for the measurement of horizontal and vertical intra-industry trade proposed by Fontagne and Freudenberg (1997), which is not useful for the measure- ment of multilateral tradei5 of the five observed countries. Nielsen and Lüthje (2002) also showed that the methodology introduced by Green- away, Hine and Milner is more appropriate for the measurement ofhori- zontal and vertical intra-industry trade than the alternative methodology mentioned above. ifi The method introduced by Greenaway, Hine and Milner also supposes the separation of total iit or, better said, disentangling of £>, on the be- longing shares of horizontal iit (HBi) and vertical iit (VBi): ~Bi = HBi + VBi. (1) Following this methodology, the unit value index (UV) is calculated for exports and imports of each manufacturing industry at the five-digit level of the sitc.i7 Horizontal iit is defined as a ratio between the unit value of exports UVX and the unit value of imports UVm for a particular industry i or, to put it differently, UV*/UVj. More specifically, horizon- tal iit is defined (HBi) when the unit value index (UV) was inside the range of ± 15%: UVx 0.85 1.15 and V2: -'- < 0.85, (3) 1 JJVm 2 JJVm ii where Vi represents the share of vertical iit when the ratio between the unit value of exports UVX and the unit value of imports (UV1") is greater than 1.15, and V2 represents the share of vertical iit when the ratio be- tween the unit value of exports UV* and the unit value of imports UV™ is smaller than 0.85. It is assumed that the relative quality of each prod- uct at the five-digit level of the sitc is best defined by the achieved rel- ative price for the same product and that the relative share of Vi repre- sents trade in vertically differentiated products of higher quality, which are sold at a higher average price, and that V2 represents trade in verti- cally differentiated products of lower quality, which are sold at a lower average price. (Cernoša (2005b) showed that the Greenaway, Hine, and Milner methodology (1994; 1995) is able to capture adjustment cost. In short, since the Greenaway, Hine and Milner methodology for measure- ment of horizontal and vertical intra-industry trade is grounded on unit value index (UV), it is able to capture adjustment costs. Results of the Measurements at the Aggregate Level It is widely recognized that the average levels of horizontal and verti- cal intra-industry trade show the production structure of the observed country or intra-industry trade specialisation of that country. The hor- izontal and vertical intra-industry trade is measured for each manufac- turing industry (of sections sitc 5-8) at the five-digit level of sitc18 for the Czech Republic, Hungary, Poland, Slovenia and Slovakia in 1995,1998 and 2001. The statistical data at the five-digit level of sitc were used in order to be able to compare the results with those obtained in the pre- vious studies.19 Statistical data were obtained from comext, Eurostat's trade database, where the eu was the reporter for both import and ex- port flows.20 Table 1 shows average levels of total, horizontal and vertical intra-industry trade for the Czech Republic's, Hungarian, Polish, Slove- nian and Slovakian trade with eu in 1995,1998 and 2001 at the aggregate level (sitc sectors 5-8), using the Greenaway, Hine, and Milner method- ology. Table 1 shows that the Czech Republic showed the highest levels of iit trade in the observed period and that Slovenia and Hungary simul- taneously revealed higher average levels of iit than Poland and Slo- vakia. These results of the measurement of total intra-industry trade for former cefta countries were similar to the results of iit levels in the trade between Eastern Europe and the eu from 1990 to 1995 (Aturupane, Djankov, and Hoekman 1999). The cited study also revealed the high- est level of total iit for the Czech Republic (42.5% in 1995), followed by iit level of Slovenia (37.0% in 1995) and Hungary (33.0% in 1995), while table 1 Average levels of total, horizontal and vertical intra-industry trade of observed countries in 1995,1998 and 2001 Country Year Total Horizontal Vertical Vi V2 Czech Republic 1995 43.21 5.88 37.34 6.78 30.56 1998 54.42 7.89 46.53 10.60 35.94 2001 53-75 8.26 45-49 9.89 35.60 Hungary 1995 31.82 4.64 27.18 8.10 19.08 1998 30.88 4.50 26.38 7.86 18.52 2001 37-33 5-44 31.89 9.50 22.39 Poland 1995 22.21 2.55 19.67 3.75 15.92 1998 25.78 3.64 22.14 5.18 16.96 2001 34.10 5-53 28.57 7-57 21.00 Slovenia 1995 33.27 5.16 28.11 7.99 20.12 1998 40.43 6.65 33.77 10.28 23.49 2001 36.92 5.87 31-05 10.03 21.02 Slovakia 1995 24.33 2.64 21.69 3.59 18.10 1998 26.58 3.78 22.8 4.47 18.33 2001 30.97 4-36 26.61 6.50 20.11 notes Vi - trade in vertically differentiated products of higher quality; V2 - trade in vertically differentiated products of lower quality. sources Eurostat 2002 and own calculations at the five-digit level of sitc. Poland and Slovakia both showed the lowest level of total intra-industry trade (approximately 28% in 1995). This study used Eurostat's statisti- cal database from 1990 to 1995 at the six-digit level of eu's Combined Nomenclature,21 while the present study used statistical data at the five- digit level of siTc from 1995 to 2001.22 While the summary statistics of the above mentioned study did not reveal significant changes in the average levels of total iiT for the chosen cefta countries in the period from 1990 to 1995, the present study has found that average levels of total, horizontal and vertical iiT of Hungary, Poland and Slovakia gradually grew from 1995 to 2001, while in the case of the Czech Republic and Slovenia these levels rapidly increased from 1995 to 1998, and then decreased or remained relatively stable from 1998 to 2001. In this way, the present study only captured the effects of the in- tegration process on the territory of Europe. Thus the levels of total iiT of the five observed cefta countries increased after year 1995 due to trade liberalization between the eu and these countries in the mid 1990's. The following paragraphs will show that the present study included complete methodology for the measurement of horizontal and vertical 11T.23 To be more precise, total intra-industry trade of each of the observed countries in table 1, was disentangled into horizontal and vertical com- ponents using relative unit values of exports and relative unit values of imports. Horizontal iit was defined as the simultaneous export and im- port of five-digit sitc products, where the unit value of exports relative to the unit value of imports was within the range of ±15%. Where unit values of exports relative to the unit value of imports were outside of that range, iit was considered to be vertical. Using the Greenaway, Hine and Milner methodology, vertical iit was separated into V1 and V2. Thus, V1 represents the share of vertical iit, where the ratio between the unit value of exports and the unit value of imports is greater than 1.15, and vice versa, V2 represents the share of vertical iit, where the ratio between the unit value of exports and the unit value of imports is smaller than 0.85. It is assumed that the unit value index as an indicator of quality gives us perfect information that the products sold at higher prices must be of higher quality and that products sold at lower prices must be of lower quality. It is further assumed that the relative share of V1 represents trade of the observed cefta countries in vertically differentiated products of higher quality, which are sold at a higher average price, and the relative share of V2 represents trade in vertically differentiated products of lower quality, which are sold at a lower average price. Table 2 shows the relative24 shares of horizontal iit and vertical iit, and relative shares of V1 and V2 in total intra-industry trade of the Czech Republic, Hungary, Poland, Slovenia and Slovakia in 1995, 1998 and 2001. The numbers confirmed that the shares of vertical iit are larger than the shares of horizontal iit of the observed cefta countries. Therefore, the share of vertical iit represented approximately five sixths of total iit of Slovenia and Poland in 2001, and approximately six sevenths of total iit of the Czech Republic, Hungary and Slovakia. The previous study ((Cernoša 2005b), which measured horizontal and vertical intra-industry trade (on multilateral basis) for Slovenia and the Czech Republic in com- parison with Germany and Austria in 1998, also showed that the share of vertical iit represented approximately three quarters of total iit of Ger- many, four fifths of total iit of Austria and five sixths of iit of Slovenia and Czech Republic in 1998. A similar empirical study by cepii (Fontagne, Freudenberg, and table 2 Relative shares of horizontal and vertical intra-industry trade in total intra-industry trade of observed countries in 1995,1998 and 2001 Country Year Horizontal Vertical Vi V2 Czech Republic 1995 13.60 86.40 15.69 70.71 1998 14.50 85.50 19.47 66.03 2001 15-37 84.63 18.41 66.22 Hungary 1995 14.58 85.42 25.45 59.97 1998 15.42 84.58 25.88 58.70 2001 14.67 85-33 27.84 57-49 Poland 1995 11.47 88.53 16.87 71.66 1998 14.12 85.88 20.09 65.79 2001 16.23 83-77 22.19 61.58 Slovenia 1995 15.52 84.48 24.00 60.48 1998 16.46 83.54 25.43 58.11 2001 15.89 84.11 27.17 56.94 Slovakia 1995 10.87 89.13 14.74 74.39 1998 14.24 85.76 16.80 68.96 2001 14.08 85.92 20.99 64-93 notes V1 - trade in vertically differentiated products of higher quality; V2 - trade in vertically differentiated products of lower quality. sources Eurostat 2002 and own calculations. Peridy 1997), which by contrast measured horizontal and vertical intra- industry trade for 12 eu member states on bilateral basis from 1980 to 1994 recorded that horizontal iit of Germany, France and Great Britain represented between one quarter and one fifth of total iit of these coun- tries from 1980 to 1994, that the less developed members of the eu si- multaneously showed lower levels of horizontal iit, that vertical intra- industry trade V1 of Germany, France and Great Britain was greater than vertical intra-industry trade V2 in the observed period, and that the less developed members of the eu simultaneously showed inverse propor- tion between vertical iit V1 and vertical iit V2. The authors of the cited study argued that in the case of eu member states two-way trade in ver- tically differentiated products is more significant than two-way trade in similar, or better said, horizontally differentiated products (Fontagne, Freudenberg, and Peridy 1997, 54). The relative shares of horizontal and vertical intra-industry trade in table 2 confirmed that vertical intra-industry trade also prevailed in to- tal intra-industry trade of five former cefta countries. Thus in Hun- gary, Poland, Slovenia and Slovakia the relative shares of vertical iiT Vi, which show specialization in the production of vertically differentiated products of higher quality, were increasing from 1995 to 2001, while in the Czech Republic the increase was recorded only from 1995 to 1998. The summary statistics also shows that Slovenia and Hungary revealed approximately the same proportion between the share of V1 and V2 and that during the observed period both countries expressed the lowermost proportion between the share of V1 and V2, while Poland, Slovakia and the Czech Republic simultaneously recorded a relatively lower share of V1. In this way, it was empirically confirmed that former cefta coun- tries due to the predominant share of vertical intra-industry trade of lower quality - V2 - showed predominant specialization in the produc- tion of vertically differentiated products of lower quality. The Results of the Measurements for Twenty Division of the isic While in the previous section we presented the results of the mea- surement of horizontal and vertical intra-industry trade of five former cefta member states at the aggregate level of the manufacturing sectors (sitc 5-8), in this section we present the results of the measurements at the two-digit level of isic for the same countries. While the Stan- dard International Trade Classification (sitc, rev. 3) is in concordance with International Standard Industrial Classification (isic, rev. 3; see http://unstats.un.org/unsd/cr/registry/) the same statistical data at the five-digit level of sitc were used as a basis for calculations in both anal- yses. The statistical data at the five-digit level of sitc were regrouped in concordance with the chosen divisions at the two-digit level of isic (see http://europa.eu.int/comm/eurostat/ramon/). In this way, twenty divisions25 were formed, which represent the following manufacturing activities at the two-digit level of the International Standard Industrial Classification: Code 17 - Manufacture of textiles Code 18 - Manufacture of wearing apparel; dressing and dyeing of fur Code 19 - Tanning and dressing of leather; manufacture of luggage, handbags, saddlery, harness and footwear Code 20 - Manufacture of wood and cork products, except furniture; manufacture of straw and plaiting materials Code 21 - Manufacture of paper and paper products Code 22 - Publishing, printing and reproduction of recorded media Code 23 - Manufacture of coke, refined petroleum and nuclear fuel Code 24 - Manufacture of chemicals and chemical products Code 25 - Manufacture of rubber and plastics products Code 26 - Manufacture of other non-metallic mineral products Code 27 - Manufacture of basic metals Code 28 - Manufacture of fabricated metal products, except machin- ery and equipment Code 29 - Manufacture of machinery and equipment nec Code 30 - Manufacture of office accounting and computing machin- ery Code 31 - Manufacture of electrical machinery and equipment nec Code 32 - Manufacture of radio, television and communication equip- ment and apparatus Code 33 - Manufacture of medical, precision and optical instru- ments, watches and clocks Code 34 - Manufacture of motor vehicles, trailers and semi trailers Code 35 - Manufacture of other transport equipment Code 30 - Manufacture of furniture; manufacturing nec Code 31 - Manufacture of electrical machinery and apparatus nec Code 32 - Manufacture of radio, television and communication equip- ment Code 33 - Manufacture of medical, precision and optical instruments and clocks Code 34 - Manufacture of motor vehicles, trailers and semi-trailers Code 35 - Manufacture of other transport equipment Code 36 - Manufacture of furniture, manufacturing nec The new aggregation of twenty activities included almost all con- corded manufacturing industries (sections sitc 5-8) and also a few con- corded industries of raw material (sections sitc 0-4) at the five-digit level of sitc. The main goal of this time-consuming work was to mea- sure horizontal and vertical intra-industry trade for each of the manufac- turing industries that formed the twenty above listed activities for Czech Republic, Hungary, Poland, Slovenia and Slovakia.26 While the previous section of this paper verified the production structure or intra-industry trade specialization of the five observed former cefta countries in 1995, 1998 and 2001 at the aggregate level of the sectors (sitc 5-8), this section tests intra-industry trade specialization of the observed countries at the level of the twenty chosen manufacturing activities (i. e. divisions 17-36 of the isic) in 2001. Table 3 shows average levels of total intra-industry trade for twenty manufacturing activities in the Czech Republic's, Hungarian, Polish, Slovenian and Slovakian trade with the eu in 2001, which were measured by using the Greenaway, Hine and Milner methodology. More precisely, the total intra-industry trade at the aggregate level of the selected twenty manufacturing activities was measured by using the Grubel and Lloyd index for weighted average.27 The calculated mean28 in the last row (at the bottom) of the table recorded the average total iit for each of the for- mer cefta countries in 2001 and also showed that the Czech Republic in comparison with Hungary, Poland and Slovenia achieved the highest level of total intra-industry trade and that Slovenia and Hungary simul- taneously revealed higher average levels of iit than Poland and Slovakia. The average levels of total intra-industry trade for twenty manufactur- ing activities of the former cefta countries are comparable with average levels of total intra-industry trade of these countries in 2001 at the ag- gregate level (see table 1). On the other side, the calculated mean29 in the last column of table 3 shows average levels of total intra-industry trade for twenty activities in 2001.3o It should be emphasized again that Eurostat's trade database was used, where the eu was the reporter for both import and export flows.3i Thus the average total intra-industry trade of twenty activities (at the two-digit level of isic) was measured in the Czech Republic's, Hungarian, Polish, Slovenian and Slovakian total trade with eu member states in 2001. If the particular activity recorded an average higher total intra-industry trade of former cefta countries, the calculated mean value of this activity (in the last column of table 3) is relatively higher. In this way, the achieved average intra-industry trade ranked twenty activities in 2001. This ranking of activities gave the following 'top eight' activities that revealed the highest average total intra-industry trade in the observed year: (1.) manufacturing of fabricated metal products, except machinery and equipment; (2.) manufacturing of electrical machinery and equip- ment; (3.) manufacturing of other transport equipment; (4.) publishing, printing and reproduction of recorded media; (5.) manufacturing of ma- chinery and equipment; (6.) manufacturing of furniture; manufacturing table 3 Average levels of total intra-industry trade for chosen manufacturing activities of observed countries in 2001 (in %) isic cz h pl si sk Mean 17 47.81 35.71 19.05 29.19 25.43 31.44 18 53.05 42.63 28.81 29.34 16.90 34.14 I9 60.07 25.94 53.18 57.67 22.18 43.81 20 36.75 45.50 42.56 34.24 25.33 36.88 21 35.72 20.40 11.04 42.24 19.84 25.85 22 83.65 41.35 40.55 42.82 42.24 50.12 23 56.67 46.07 15.17 — — 39.30 24 30.12 24.32 20.16 28.39 20.25 24.65 25 55.42 47.73 37.07 45.32 25.11 42.13 26 52.81 43.63 43.64 32.29 23.53 39.18 27 36.93 40.68 35.54 36.95 11.40 32.30 28 67.83 56.15 57.11 57.24 58.40 59.34 29 59.26 45.34 41.13 49.59 40.96 47.26 30 51.85 30.44 5.35 21.22 43.51 30.47 31 63.99 53.83 53.39 48.76 44.48 52.89 32 30.28 47.92 28.21 34.02 39.85 36.06 33 49.23 38.75 10.38 45.70 35.40 35.89 34 41.19 15.31 36.58 40.54 29.68 32.66 35 68.55 55.42 47.62 39.11 46.68 51.47 36 45.97 58.44 29.89 58.18 41.64 46.82 Mean 51-36 40.78 32.82 40.67 32.25 notes isic - selected divisions of the International Standard Industrial Classification signed by two-digit code; cz - Czech Republic; h - Hungary; pl - Poland; si - Slovenia; sk - Slovakia; mean - calculated arithmetic mean (by using equitation 6). sources Eurostat 2002 and own calculations at the two-digit level of sitc. of other non-mentioned products; (7.) tanning and dressing of leather; manufacturing of luggage, handbags, saddlery, harness and footwear and (8.) manufacturing of rubber and plastics products. It is important to note that the average total intra-industry trade" of listed manufacturing activities is not correlated with intra-industry trade specialization of the five former cefta countries. This ranking of eight manufacturing activities by average total intra-industry trade only confirmed that export and import flows of these activities are relatively balanced, while in the case of the remaining twelve manufacturing ac- tivities export and import flows are relatively lees balanced.33 In short, if the standard Grubel and Lloyd index per se shows the share of intra- industry trade of the particular industry in total trade in the same in- dustry, then only in combination with the Unit Value (uv) did the index represent a useful methodology for the measurement of intra-industry trade specialization^ In this way intra-industry trade specialization of twenty manufacturing activities was tested. The results of the measure- ments of horizontal and vertical iit for the Czech Republic, Hungary, Poland, Slovenia and Slovakia in 2001 confirmed that a great majority of the twenty chosen manufacturing activities (i. e. divisions 17-36 of the isic) showed predominant specialization in the production of lower quality products. At the same time, the present analysis also found a few manufacturing activities at the five-digit level of isic, which showed predominant spe- cialization in the production of higher quality products. The selection of these activities is based on the assumption (of the Greenaway, Hine, and Milner methodology) that the relative share of V1 represents trade in vertically differentiated products of higher quality, which are sold at a higher average price, and that relative share V2 represents trade in ver- tically differentiated products of lower quality, which are sold at a lower average price. Thus, if we assume that the unit value index gives us per- fect information that the products sold at higher prices must be of higher quality, then the manufacturing activities - which showed a greater share of vertically differentiated products of higher quality (V1 > V2) - simul- taneously reveal predominant specialization in higher quality products. Table 4 shows total, horizontal and vertical intra-industry trade for se- lected activities in observed cefta countries, which recorded predomi- nant specialization in the production of vertically differentiated products of higher quality V1. Thus the Czech Republic has one manufacturing ac- tivity, Slovenia and Slovakia have two activities that revealed specializa- tion in production of vertically differentiated products of higher quality in 2001, while Hungary and Poland recorded three activities that showed the predominant share of vertical intra-industry trade - V1 in 2001. Re- member that according to the Greenaway, Hine and Milner methodol- ogy, the relative share of V1 represents trade of the observed cefta coun- tries in vertically differentiated products of higher quality, which are sold at a higher average price, and that the relative share of V2 represents trade of these countries in vertically differentiated products of lower quality, which are sold at a lower average price. table 4 Total, horizontal and vertical intra-industry trade for selected manufacturing activities of former cefta countries in 2001 Country isic Total Horizontal Vertical Vi V2 Czech Republic 23 56.67 28.33 28.33 14.17 14.17 Hungary 18 42.63 7.19 35.44 22.60 12.84 21 20.40 6.22 14.17 6.91 7.26 34 15-31 2-45 12.86 7-35 5-51 Poland 23 15.17 3.03 12.14 9.10 3.03 32 28.21 3.94 24.28 12.47 11.81 34 36.58 8.83 27-75 13.88 13.88 Slovenia 18 29.34 5.08 24.26 15.80 8.46 19 57-67 8.24 49-43 28.84 20.60 Slovakia 18 16.90 3.82 13.09 6.54 6.54 30 43-51 3-96 39-55 27.69 11.87 notes isic - selected divisions of the International Standard Industrial Classification signed by two-digit code; V1 - trade in vertically differentiated products of higher qual- ity; V2 - trade in vertically differentiated products of lower quality. sources Eurostat 2002 and own calculation at the 2-digit level of sitc. The table also shows that former cefta countries revealed the simi- lar intra-industry trade specialization in the production of vertically dif- ferentiated products. Thus Hungary, Slovenia35 and Slovakia were spe- cialized in manufacturing wearing apparel, dressing and dyeing of fur products (isic 18), which represent so-called traditional manufacturing activity. Similarly, Czech Republic and Poland specialized in the produc- tion of coke, refined petroleum and nuclear fuel products (isic 23)^ while Hungary and Poland competed in manufacturing motor vehicles, trailers and semi-trailers (isic 33).37 We suppose that investments by foreign owned enterprises have caused Hungary and Poland to show predominant specialisation in the pro- duction of higher quality motor vehicles, trailers and semi trailers; that Poland reveals predominant specialisation in the production of higher quality radio, television and communication equipment and apparatus (isic 32); and that Poland shows the largest changes in the average levels of total iit in the period from 1995 to 2001. Concluding Comments The main goal of this paper to verify the production structure or intra- industry specialization of Czech Republic, Hungary, Poland, Slovenia and Slovakia as former members of cefta in trade with the eu member states. In this way the first analysis tested specialization of these coun- tries at the aggregate level of the manufacturing sectors (i.e. sitc 5-8) in 1995, 1998 and 2001. The results of the measurement of total, hori- zontal and vertical intra-industry trade at the aggregate level by using the Greenaway, Hine, and Milner (1994; 1995) methodology confirmed predominant specialization of the former cefta countries in produc- tion of vertically differentiated products of lower quality. The findings of this analysis are comparable with those of the previous studies (Cernoša 2005a; 2005b). The same statistical data at the five-digit level of sitc were also used as a basis for the calculation in the second analysis presented in this pa- per, which tested iit specialization at the level of twenty manufacturing activities (i. e. divisions 17-36 of the isic). This analysis also revealed the predominance of iit specialization of the majority of the chosen manufacturing activities in the production of lower quality products - V2 - and it also found a few activities in each of the five observed former cefta countries, which, by contrast, showed predominant specialization in the production of higher quality products. Thus, if former cefta countries in general showed a similar production structure either at the aggregate level of manufacturing sec- tors (sitc 5-8) or at the level of twenty chosen manufacturing activities (divisions isic 17-36), then in accordance with these results it is not pos- sible to conclude that these countries achieved the same intra-industry specialization. The important message of the present analysis is that the Czech Re- public, Hungary, Poland, Slovenia and Slovakia as former members of cefta showed predominant specialization in the production of verti- cally differentiated products of lower quality, while developed eu mem- ber states such as Austria, Belgium, France, Germany and Great Britain simultaneously showed predominant intra-industry specialization in the higher quality vertically differentiated products. The governments of former cefta countries will need to recognize that the existent production structure of the observed Central European countries at the beginning of the first decade of the 21st century is by no means comparable with predominant intra-industry trade special- ization of the advanced eu member states, which in comparison with former cefta countries also achieved higher gdp per capita income on average. Notes 1 For instance, Austria has approximately one time bigger gdp per capita than Slovenia and approximately one and a half times bigger gdp per capita than the Czech Republic. 2 The difference in gdp per capita income between the observed coun- try and the trading partner is frequently used in regression models as a proxy for differences in factor endowments. It is assumed that rela- tive capital abundance of an observed country is reflected in its relative gdp per capita income. 3 Trade with eu represented approximately two thirds of the total trade of these Central European countries. 4 It is important to note that the Aturupane, Djankov and Hoekman study estimated a regression model meaning that only disentangled total intra-industry trade of the Eastern and Central European coun- tries on the belonging share of horizontal and vertical intra-industry trade, while the present study disentangled total intra-industry trade of the observed former cefta countries on the belonging horizon- tal and vertical components and also separated the share of vertical intra-industry trade into two parts by using the Greenaway, Hine, and Milner methodology (1994; 1995). 5 The methodology for the measurement of the horizontal and vertical iiT also does not suppose use of statistical data for consecutive years. 6 Unit value index is defined as the ratio of the values (in national cur- rencies) and the quantities (in kilograms or tons) of the particular in- dustry i. 7 By contrast, Crespo and Fontoura (2004) use statistical data for the year 1994,1996,1998 and 2000. 8 The Heckscher-Ohlin model. 9 Krugman (1979), Lancaster (1980), Helpman (1981), Brander and Krugman (1983), Eaton-Kierzkowski (1984), Krugman and Venables (1990), Davis (1995), Markusen and Venables (1996) model. 10 Greenaway, Hine, and Milner (1994; 1995) developed a methodology which is based on the work of Abd-el-Rahman (1991) and which is able to identify vertical and horizontal intra-industry trade of the observed countries. 11 More precisely, the study by Aturupane, Djankov and Hoekman pri- marily tested an industry specific hypothesis and also found more sup- port for the industry specific factors. 12 Aturupane, Djankov and Hoekman (1999), Cernoša (2005a), Cernoša (2005b), and the present analysis. 13 In other words, the range of products which a certain type of capital equipment can produce defines an industry. 14 If Cernoša (2005b) showed that Germany and Austria as more de- veloped eu members' economies were specialized in the production of vertically differentiated products of higher quality, and Slovenia and Czech Republic as less developed Central European economies were specialized in the production of vertically differentiated prod- ucts of low quality at the end of the 1990s and at the beginning of the 2000s, then Germany and Austria, due to a higher capital-labour ra- tio, will produce and export vertically differentiated products of rela- tively higher quality, while Slovenia and Czech Republic, due to a lower capital-labour ratio, will produce and export vertically differentiated products of relatively lower quality. 15 The Fontagne and Freudenberg (1997) methodology is useful only for the observation of the bilateral trade. 16 The Greenaway, Hine and Milner methodology supposes calculation of the standard Grubel and Lloyd (1975) index. Thus the standard Grubel and Lloyd index is calculated: [(K, + Mi) - IX, - M,\] ■ 100 B, = —--——--- (0 < GL, < 100), (Xi+Mi) where B, represents the Grubel and Lloyd index for a particular indus- try i at the five-digit sitc level, K, represents exports of that particu- lar industry, while Mi represents imports of that particular industry. The intra-industry trade at the aggregate level (weighted average) was measured using the following index (Grubel and Lloyd 1975, 21): — B;(X; + Mi) n Bi = ——- = > w; ■ B;, where the weights are 2f=1(X! + M0 jrt 8 Xi + Mi ! ZUX. + M,)' where 5; represents Grubel and Lloyd's index for weighted average at the aggregate level, B, represents the standard Grubel and Lloyd index for a particular industry i at the five-digit sitc level, X, represents exports of that particular industry, while M, represents imports of that particular industry. 17 Unit value index (UV) is defined as a ratio of the values (in national currencies) and the quantities (in kilograms or tons) of the particular industry i. 18 The same level of aggregation was also used by Greenaway, Hine, and Milner (1994; 1995), Greenaway, Milner and Elliott (1999) and Nielsen and Lüthje (2002). It is important to note that alternative data clas- sification - Combine Nomenclature (cn) - in the case of Slovenia is available from 1996. 19 The results of the present analysis are comparable with the results of the measurement of horizontal and vertical intra-industry trade (on multilateral basis) for Slovenia from 1994 to 2003 at the five-digit level of sitc (Cernoša 2005a) and also comparable with the mea- sured horizontal and vertical intra-industry trade (on multilateral ba- sis) for Slovenia and the Czech Republic in comparison with Germany and Austria in 1998 at the five-digit level of sitc (Cernoša 2005b). In short, Slovenia and the Czech Republic in both studies cited above recorded relatively higher levels of total intra-industry trade in mul- tilateral trade and also in the present analysis both showed relatively lower levels of total intra-industry in trade with eu members states. 20 Concretely, we used Eurostat's statistical data inversely, where exports of eu member states (in 000 eur and tons) to a selected cefta coun- try were treated as imports of this country during the observed years, while imports of all eu member states from the same selected country (in 000 eur and tons) were treated as exports of this country during the observed years. 21 Where only 9 eu member states are included (Belgium, Luxembourg, Germany, France, The United Kingdom, Italy, The Netherlands, Den- mark and Ireland). 22 This study observed iit of each former cefta (5) member state in foreign trade with all eu member states in 1995,1998 and 2001. 23 While the present study following the Greenaway-Hine-Milner metho- dology disentangles total intra-industry trade on horizontal and ver- tical components and later also separates vertical intra-industry into two parts, Aturupane, Djankov, and Hoekman (1999) did only the first step. 24 Total intra-industry trade = 100. 25 Each particular division is signed by a two-digit code. 26 In short, we measured horizontal and vertical intra-industry trade of each particular manufacturing industry at the five-digit level of sitc and later repeated measurements at the aggregate level (at two-digit level of the isic). 27 Equitation 5;. 28 Calculated as the arithmetic mean vertically (by columns) in table 3. 29 Calculated as the arithmetic mean horizontally (by rows) in table 3. 30 This arithmetic mean is calculated using the following equitation (Greenaway and Milner 1986, 65): _ 1 " — Bi = — / B;, n t—i i=1 where 5; represents the arithmetic mean and 5; represents Grub el and Lloyd's index for weighted average at the two-digit level of isic. 31 We used Eurostats statistical data inversely, where exports of eu mem- ber states (in 000 eur and tons) to a selected cefta country were treated as imports of this country during the observed years, while imports of all eu member states from the same selected country (in 000 eur and tons) were treated as exports of this country during the observed years. 32 Measured by Grubel and Lloyd index. 33 Grubel and Lloyd defined intra-industry trade as the value of exports of an industry, which is exactly matched by the value of the imports of the same industry. (Grubel and Lloyd 1975, 20). Thus, the gl index showed the higher average total intra-industry trade at the aggregate level (see equitation (2) for weighted average) in the case where the re- lation between the exports and imports of the observed manufactur- ing industries is relatively balanced, and also revealed lower average total iit at the aggregate in all cases, when the proportion between the values of the exports and imports of the observed industries is on average relatively unbalanced. 34 See Greenaway, Hine, and Milner (1994; 1995) methodology for the measurement of horizontal and vertical iit. 35 The previous analysis (Cernoša 2002) also showed that Slovenia and the Czech Republic are competitors and partners in intra-industry ex- change of the same or similar products. 36 Both countries showed specialization in vertically differentiated prod- ucts of higher quality in the manufacturing of refined petroleum prod- ucts, which represent a sub-group of code 23 (of the isic). 37 This activity supposed the use of advanced technologies. References Aturupane, C., S. Djankov, and B. Hoekman 1999. Horizontal and ver- tical intra-industry trade between Eastern Europe and the European Union. Weltwirtschaftliches Archiv 135 (1): 63-81. Abd-el-Rahman, K. 1991. Firms competitive and national comparative ad- vantages as joint determinants of trade composition. Weltwirtschaftli- ches Archiv 127 (1): 83-97. Brander, J., and P. Krugman. 1983. A 'Reciprocal Dumping' model of in- ternational trade. Journal of International Economics 15 (2): 313-21. Crespo, N., and M. P. Fontoura. 2004. Intra-industry trade types: What can we learn from Portuguese data? Weltwirtschaftliches Archiv 140 (1): 53-79. Cernoša, S. 2002. Vnitroodvetvovy obchod a vliv rozdilü ve velikosti zemi: Na pripadu Slovinska a Ceske republiky. Finance a üver 52 (6): 355-70. -. 2005a. Horizontalna in vertikalna znotrajpanožna trgovina Slove- nije v obdobju od leta 1994 do leta 2003. ib revija 39 (1-2): 26-33. -. 2005b. Standard Grubel and Lloyd Index thirty years later. Paper presented at the 5th International Conference on Economic Integra- tions, Competition and Cooperation, April 22-23, Lovran, Croatia. Davis, R. D. 1995. Intra-industry trade: Heckscher-Ohlin-Ricardo ap- proach. Journal of International Economics 39 (3-4): 201-26. Eaton, J., and H. Kierzkowski. 1984. Oligopolistic competition, product variety and international trade. In Monopolistic competition and inter- national trade, ed. H. Kierzkowski, 69-83. Oxford: Clarendon. Eurostat. 2002. comext. od-rdm. Luxembourg: Eurostat. Falvey, E. R. 1981. Commercial policy and intra-industry trade. Journal of International Economics 11 (4): 495-511. Fontagne, L., and M. Freudenberg. 1997. Intra-industry trade: Method- ological issues reconsidered. cepii Working Paper 97/01. Fontagne, L., M. Freudenberg, and N. Peridy. 1997. Trade patterns inside the single market. cepii Working Paper 97/07. Greenaway, D., and C. Milner. 1986. Economics of intra-industry trade. Ox- ford: Blackwell. Greenaway, D., R. Hine, and C. Milner. 1994. Country-specific factors and the pattern of horizontal and vertical intra-industry trade in the uk. Weltwirtschaftliches Archiv 130 (1): 77-100. -. 1995. Vertical and horizontal intra-industry trade: A cross-industry analysis for the uk. Economic Journal 105:1505-19. Greenaway, D., C. Milner, and J. R. Elliott. 1999. uk intra-industry trade with eu north and south. Oxford Bulletin of Economics and Statistics 61 (3): 365-84. Grubel, H., and J. P. Lloyd. 1975. Intra-industry trade: The theory and measurement of international trade in differentiated products. London: Macmillan. Helpman, E. 1981. International trade in the presence of product dif- ferentiation, economies of scale and monopolistic competition: A Chamberlin-Hecksher-Ohlin approach. Journal of International Eco- nomics 11 (3): 305-40. Helpman, E., and R. P. Krugman. 1985. Market structure and foreign trade: Increasing returns, imperfect competition and the international economy. Brighton: Wheatsheaf. Krugman, R. P. 1979. Increasing returns, monopolistic competition and international trade. Journal of International Economics 9 (4): 469-79. Krugman, R. P., and J. A. Venables. 1990. Integration and the Competi- tiveness of Peripheral Industry. In Unity with diversity in the European economy, ed. C. Bliss and J. B. De Macedo, 56-77. Cambridge: Cam- bridge University Press. Lancaster, K. 1980. Intra-industry trade under perfect monopolistic com- petition. Journal of International Economics 10 (2): 151-75. Markusen, R. J., and J. A. Venables. 1996. Theory of endowment, intra- industry trade and multinational trade. cepr Discussion Paper 1341. Nielsen, J. U.-M., and T. Lüthje. 2002. Tests of the empirical classification of horizontal and vertical intra-industry trade. Weltwirtschaftliches Archiv 138 (4): 587-604.