3.5 VERTICAL PRODUCT DIFFERENTIATION IN EU MARKETS: THE RELATIVE POSITION OF EAST EUROPEAN PRODUCERS. Michael A. Landesmann, Johann Burgstaller (WIIW)

Transcription

1 3.5 VERTICAL PRODUCT DIFFERENTIATION IN EU MARKETS: THE RELATIVE POSITION OF EAST EUROPEAN PRODUCERS Michael A. Landesmann, Johann Burgstaller (WIIW) Introduction This study reports on a detailed examination of the price and quality positions of East European producers on EU markets comprising developments over the period 1988 to The raw material for this analysis was Eurostat s Detailed Trade Statistics which contain detailed information on trade values and volumes at the 8-digit CN (Combined Nomenclature, 6-digit NIMEXE before 1992) product level of trade to and from EU countries. We used this database to analyse the pattern of quality differentiation in intra-branch trade for some selected industries. In particular, the focus was to analyse the position of different Central and Eastern European (CEE) producers in the quality spectrum of European trade or, more precisely, trade with the EU (including intra-eu trade). This was done by means of comparisons with reference countries or country groups. First, there are the advanced Western European countries (EU-North) and ex-efta economies, as well as the USA, Japan and Canada, representing the more advanced economies; then there are the Southern European economies (EU-South without Italy, as well as Turkey), representing the countries which might be, in quality and technological terms, the countries with which the CEECs might be more immediately comparable; and, finally, there are groups of Asian reference countries (the NICs1, Hong Kong, Singapore, Taiwan and Korea; the NICs2, comprising Malaysia, Indonesia, Thailand and the Philippines; and, finally, China and India) which opened up a spectrum of economies which are at different stages of their (technological) catching-up and whose current position in world trade might again be a reference point for the CEE economies. All other trading partners of the EU were subsumed as RoW (rest of the world) 1. Methodologically, two basic types of exercises were carried out. Firstly, we simply calculated product prices (value per kg) at the detailed product level across the whole range of competitors in EU markets (more precisely, in total EU imports including intra-eu trade); we then compared these prices to the average price for that product in total EU imports and constructed a weighted "price/quality gap indicator" (Q c ) for a three- or two-digit NACE industry by aggregating the price j gaps (price ratios) for the individual products using their shares in the commodity structure of a particular country s exports to the EU as weights. Secondly, we compared the compositions of a country s exports to the EU within a particular (3- digit NACE) industry by ranking the products traded within that industry by their prices per kg in the EU as a whole. We could then see to which degree a country s exports were represented in the "high- 123

2 ", "medium-" and "low-quality" segments of the product range traded in this branch. The analysis of the location of different producers in the "quality segmented" structure of EU product markets (we again refer here only to total EU imports) was conducted for a range of (3-digit NACE) branches of the engineering sector, of the food, drink and tobacco sector and, finally, the textile, footwear and clothing sector. In the following section we report the methodology used for these two exercises in more detail (including some simple regression analysis), in section we discuss the results obtained and in section we report an additional exercise on outward processing (OP) trade Methodology Quality/price gaps The first exercise we wish to report is an analysis of price gaps in export sales. Given the very detailed product statistics available, price comparisons (price per kg) are more appropriate in revealing quality differences than at higher levels of aggregation where compositional differences would dominate the picture. Hence in order to obtain the price/quality gap information for a number of 2- and 3-digit NACE industries, the full product level information was used for each industry. The industry-level (weighted) price/quality gap indicator was arrived at as: Q c j = Σ (pc i /peu i ) sxc i i I(j) where p c i is the price (per kg) at which country c sells exports of the product item i on EU markets (refers here to the EU 12 market); p EU i is the average price of product item i in total EU 12 imports; sx c i is the share of product item i in country c s exports to the EU 12 market. We have Σ sx c i = 1 i I(j) where I(j) is the set of product items i belonging to NACE industry j. To minimise the effects of those errors in the data which show up in extreme outlier positions, we attempted to detect these and remove them. For the calculations with product data for the 3-digit industries we tried to solve the problem as follows: For each country and year the mean and standard deviation of the price ratios to the EU import price were calculated considering all products assigned to a certain industry. A product was then identified to be an extreme outlier if its price ratio to the EU import price exceeded the above mentioned mean plus three times the standard deviation. After the removal of outliers detected in this way the sum of the weights (export shares) had to be corrected to add up to unity again. Finally it has to be noted that the product level price comparisons were made using current ECU prices derived from EU import data. The results of our study will show that the price gap variables do not shift dramatically (with some exceptions) in spite of rather enormous exchange rate fluctuations 124

3 of CEECs currencies vis-à-vis the ECU. We take this as support for price-taking behaviour of CEE producers given the quality of their products. Product quality segmentation The next step in our analysis of quality differentiation is conducted only at the level of 3-digit industries. We first selected two groups of NACE engineering industries ( ; and 330, , ) and the groups of textile, clothing and footwear (436, , , 451, 453, ) and food, drink, and tobacco industries ( ). Within these industries we ranked products by their prices per kg which they fetch on the EU import market as a whole (including intra-eu trade). Once these products had been ranked in descending order, we demarcated three quality segments (Qual I comprising the more highly priced items, Qual II the medium priced items and Qual III the least priced items). Once such quality segments were defined within each 3-digit NACE industry we could compare the degrees to which the different national exporters product structures fell into these different quality segments. This defined their respective positions in the vertically (hierarchically) differentiated product market structure of EU markets. For EU total imports within each 3-digit NACE industry, each quality segment should in principle comprise one third of the total value of imports of the EU 12 (including intra-eu trade) in the respective year. This is not exactly the case because of the need to cut off before the product that causes the cumulative value to exceed a third of the total EU import value. Because of this problem (linked to the discrete number of products belonging to each quality segment) the demarcation of the segments therefore differs from year to year and industry to industry (i. e. they do not neatly lead to segments accounting each for exactly one third of total EU imports within each industry). But a comparison across countries is always possible. Some cross-industry regressions In a next step, the variables for a simple (cross-industry) regression analysis were prepared. Average price gaps were calculated for the periods 1988 to 1990 as well as 1992 to The countries export shares (exports to the EU 12) in the high- and low- quality segments in each 3-digit NACE industry were then divided by the values of the segments in total EU imports (which, as discussed above, did not precisely amount to one third of the total value of EU imports within each industry). This allowed regressions across industries. Then the logarithm of these variables was taken (leading to the variables LQ1, LQ3 for the representation in the high- and low-quality segment and LPG for the average price gaps) and we regressed these on simple country dummies as well as dummies for certain country groups such as EUN (Northern countries of the European Union except Ireland but including Italy), EUS (Southern EU countries Spain, Portugal and Greece), EFTA (with Switzerland, Austria, Sweden and Finland), EASTW (CSFR/Czech Republic, Hungary, Poland, Yugoslavia/Slovenia), EASTE (Bulgaria, Romania, Slovak Republic for the period and the Soviet Union/Russia) and the NICs 2. These regressions provide an overview across groups of 3-digit NACE industries concerning the significance of price gaps and the over-/under-representation of the different producers in high- and low-quality segments of the EU product markets, always relative to the average price and quality structure of total EU imports. Comparisons across the two defined periods ( and 125

4 ) show, furthermore, interesting shifts in the performances of the different countries and country groups. We examined further whether shifts in the price gap coefficients could be related to exchange rate movements (measured by the ratio of the nominal exchange rate to the PPP rate). Furthermore, we examined for the period whether reductions in the price/quality gaps of CEE producers were in any way a hindrance to expanding market shares in EU markets Discussion of the results The main results concerning the Eastern European producers can be summarised as follows: There is evidence for substantial price gaps (calculated in current ECU) between Eastern European producers exports to the EU and most other competitors (including the two groups of Asian NICs). This is also true for their under-representation in high quality segments in trade with the EU, especially with regard to their position in the engineering industries. Over the period 1989 to 1994 a remarkable process of differentiation has taken place across CEE economies regarding the price gaps revealed in their exports to the EU as well as with respect to their evolving positions in the vertically differentiated EU product markets (quality segments). The Western CEECs (Hungary, the Czech Republic, Poland and Slovenia) have "moved upstream" concerning their "product quality" and position in vertically differentiated product markets, while the Eastern CEECs (Bulgaria, Romania, Russia) have lagged very much behind. The persistence of price gaps and only gradual shifts in the location of national producers in quality segmented product markets can be taken as evidence of sustained hierarchies in vertically differentiated product markets in the EU. The CEECs operate in general at the low quality end of these hierarchies but there are interesting and differentiated movements by the different CEECs over the period from 1988 to Results for price/quality gaps Detailed figures for both 3- and 2-digit NACE industries 3 are reported in Landesmann - Burgstaller (1997, annex B, tables 14 and 10 respectively). In the following we report and discuss the statistics obtained for the weighted means of price gaps over all industries and for industry groups (mechanical engineering industries , electrical and instrument engineering industries 330, , as well as for food/drink/tobacco and textile/clothing/footwear industries; see table 3.19). 126

5 Table Price gap/quality measures by 3-digit NACE industries and 330, , ( ) (EU 12 = 1), Exports to EU NACE USA Japan Canada Germany France Belgium, Netherlands Italy UK Austria Switzerland Ireland Finland Sweden Denmark Spain Luxembourg Portugal Greece Turkey Hungary Poland CSFR Czech Republic Slovak Republic Bulgaria Romania Yugoslavia Slovenia Soviet Union Russia NICs 1 NICs 2 China India Rest of World

6 Table Price gap/quality measures by 3-digit NACE industries and 330, , ( ) (EU 12 = 1), Exports to EU (continued) USA Japan Canada Germany France Belgium, Netherlands Italy UK Austria Switzerland Ireland Finland Sweden Denmark Spain Luxembourg 330, , Portugal Greece Turkey Hungary Poland CSFR Czech Republic Slovak Republic Bulgaria Romania Yugoslavia Slovenia Soviet Union Russia NICs 1 NICs 2 China India Rest of World

7 Table Price gap/quality measures by 3-digit NACE industries: textiles, clothing and footwear 436, , , 451, , food, drinks and tobacco ( ) (EU 12 = 1), Exports to EU (continued) NACE USA Japan Canada Germany France Belgium, Netherlands Italy UK Austria Switzerland Ireland Finland Sweden Denmark Spain Luxembourg 436, , , 451, , Portugal Greece Turkey Hungary Poland CSFR Czech Republic Slovak Republic Bulgaria Romania Yugoslavia Slovenia Soviet Union Russia NICs 1 NICs 2 China India Rest of World

8 Table Price gap/quality measures by 3-digit NACE industries: textiles, clothing and footwear 436, , , 451, , food, drinks and tobacco ( ) (EU 12 = 1), Exports to EU (continued) USA Japan Canada Germany France Belgium, Netherlands Italy UK Austria Switzerland Ireland Finland Sweden Denmark Spain Luxembourg Portugal Greece Turkey Hungary Poland CSFR Czech Republic Slovak Republic Bulgaria Romania Yugoslavia Slovenia Soviet Union Russia NICs 1 NICs 2 China India Rest of World

9 The weighted price/quality gap indicators have been scaled so that they take the value of 1.0 for total EU imports; values below 1.0 signify a weighted sale of products by a particular producer on EU markets at prices below the average of total EU imports (including intra-eu trade); the opposite is true for values above 1.0. The indicators presented in table 3.19 have been calculated from detailed trade statistics at current ECU exchange rates. The following general results emerge from these tables: There seems to be something of an EU market integration effect, i. e. EU members sell broadly at lower prices on EU markets than comparable countries such as the (ex-)efta countries Austria, Switzerland and the Scandinavian countries. There might be some evidence of an impact of high/low values of exchange rates, such as the very high values for the price/quality variables for Japan and Switzerland indicate, but there is mostly remarkable stability of the indicators over the years, particularly in the case of most of the CEECs which experienced dramatic exchange rate movements. There is evidence for significant price/quality gaps between the CEECs and comparable countries of Southern EU (Spain, Portugal, less with Greece) and the groups of NICs in the engineering branches, but much less so in the textile/clothing/footwear and the food/drink groups of industries. The pattern of differentiation across the CEECs, particularly between the group of "Western CEECs" (the Czech Republic, Hungary, Poland, Slovenia) and the group of "Eastern CEECs" (Bulgaria, Romania, Russia) over the period 1988 to 1994 is remarkable, and also much more pronounced for the engineering industries than the textile/clothing/footwear and the food/drink groups of industries. We add some more detailed comments: The results for the engineering industries lead to the observation that the CEECs were more successful in closing somewhat the gap within the electrical and instrument engineering group of industries than for mechanical engineering. But the gap is still substantial between most of the CEECs and the EU, although Hungary and Slovenia, in particular, have been able to reduce it somewhat. For food, drink and tobacco industries the price gaps are less than for engineering industries (which means higher values for the price gap indicators). This indicates a worse position of the CEE exporters in industries which are intensive in capital and technology. Almost the same picture emerges for textile, clothing and footwear though the price gaps are smaller (i. e. the price levels relative to the EU import average higher) for the first years in the sample in comparison with the food, drink and tobacco industries. When evaluating the results at a detailed level there is a rather varying performance of Western European countries as well as of Japan, Canada and the USA across industrial branches. Amongst the CEECs, Hungary and Slovenia show the best positions in the engineering industries, with Bulgaria and Romania far behind. Hungary, Poland, the Czech Republic and Slovenia are the countries for which the upward movement of the price gap variables -- indicating the closure of the gap -- is most striking and persistent. This is also the case for the two other industry groups we examined. The upward trend for Bulgaria, Romania, the Slovak Republic and Russia is, if perceptible at all, much weaker. A detailed -- industrywise -- inspection of the specific position of the CEE exporters in individual 3-digit branches can be obtained from Landesmann - Burgstaller (1997, annex B, table 10). 131

10 The figures for food, drink and tobacco industries comprise many zero values which indicate that there are no products exported to EU 12 by the specific country in particular branches forming this industry group. This occurs in cases where the number of products imported by the EU 12 is small in the first place 1. Some figures, especially for Bulgaria, do not seem to be very reliable. At the 2-digit NACE industry level, price gap calculations were carried out for exports to the EU as well as for imports from EU countries across the whole range of NACE industries The full results are again shown in Landesmann - Burgstaller (1997, annex B, table 14). The results are less reliable (than those for 3-digit industries) because no outlier correction has been undertaken. Only the weighted price gap means over similar industries ("similar" in relation to an a priori classification of these industries by factor intensity) shall be presented here in table For these calculations, the 2-digit NACE industries were classified in the following way 2 : 24, 41, 42 and 46 in resource-, in labour- and in capital- and technology-intensive industries. The results show an obvious deficit of the CEECs in capital- and technology-intensive branches which exceeds that in the labour-intensive branches. Results for product quality segmentation The results of the location of the different producers in the product quality segments of EU imports are reported as averages for the two periods and Table 3.21 shows industries 322 (machine tools) and 342 (electrical machinery) as examples. The figures are quite revealing: CEE exporters in most cases have much higher shares of their exports to EU markets in the "medium-" and "low-quality" segments and, usually, deficits in the "high-quality" segment. But we find that there are some differences across the CEECs in this respect, with some (Hungary, Poland, the Czech Republic) occupying a higher position in the quality-segmented structure of EU imports than other CEE economies. The full set of the results is presented in Landesmann - Burgstaller (1997, annex B, table 11). These results are reported first with a table giving an overview of the representation of the different national producers in the product quality segments of the aggregate industry groupings (two groups of engineering industries and the textile, clothing and footwear industries group 4 ). Here, the original figures (such as the ones presented in table 3.21) were divided by the values of the EU 12 segments (i. e. figures such as those in column 1 of table 3.21) and were then summed up over industries using export shares as weights. These results also show generally a strong under-representation of CEECs' exports in the high-quality segments. Some exceptions emerge for textile, clothing and footwear industries: Russia for the period and Hungary and Romania for the second period. 132

11 Table Weighted price gaps for total manufacturing (NACE 24-49) and x-factor intensive branches (EU 12 = 1), Exports to EU USA Japan Canada Germany France Belgium, Netherlands Italy UK Austria Switzer- Ireland Finland Sweden Denmark Spain Luxembourg land Total manufacturing Resource intensive branches NACE 24, 41, 42 and Labour intensive branches NACE

12 Table Weighted price gaps for total manufacturing (NACE 24-49) and x-factor intensive branches (EU 12 = 1), Exports to EU (continued) USA Japan Canada Germany France Belgium, Netherlands Italy UK Austria Switzer- Ireland Finland Sweden Denmark Spain Luxembourg land Capital intensive branches NACE

13 Table Weighted price gaps for total manufacturing (NACE 24-49) and x-factor intensive branches (EU 12 = 1), Exports to EU (continued) Portugal Greece Turkey Hungary Poland CSFR Slovak Bulgaria Romania Yugoslavia Soviet Union NICs 1 NICs 2 China India Rest of Czech Republic Slovenia Russia World Republic Total manufacturing Resource intensive branches NACE 24, 41, 42 and Labour intensive branches NACE

14 Table Weighted price gaps for total manufacturing (NACE 24-49) and x-factor intensive branches (EU 12 = 1), Exports to EU (continued) Portugal Greece Turkey Hungary Poland CSFR Slovak Bulgaria Romania Yugoslavia Soviet Union NICs 1 NICs 2 China India Rest of Czech Republic Slovenia Russia World Republic Capital intensive branches NACE

15 Table Comparative export structure in different quality segments for 3-digit NACE engineering industries 322 (machine tools) and 342 (electrical machinery) Exports to EU, averages and NACE EU 12 USA Japan Canada Germany France Belgium, Netherlands Italy UK Austria Switzer- Ireland Finland Sweden Denmark Spain Luxembourg land Qual I Qual II Qual III Qual I Qual II Qual III Qual I Qual II Qual III Qual I Qual II Qual III

16 Table Comparative export structure in different quality segments for 3-digit NACE engineering industries 322 (machine tools) and 342 (electrical machinery) (continued) Exports to EU, averages and EU 12 Portugal Greece Turkey Hungary Poland CSFR Slovak Bulgaria Romania Yugoslavia Soviet Union NICs 1 NICs 2 China India Rest of Czech Republic Slovenia Russia World Republic Qual I Qual II Qual III Qual I Qual II Qual III Qual I Qual II Qual III Qual I Qual II Qual III

17 As mentioned above, detailed tables covering all examined 3-digit NACE industries can be consulted in Landesmann - Burgstaller (1997, annex B, table 11). Some of these show a much better position for almost all CEE exporters: e. g. 347 (electric lighting equipment) and 373 (optical instruments and photographic equipment) among the engineering industries and 438 (carpets, linoleum and other floor coverings) among the textile, clothing and footwear industries. For the latter industry group as a whole the export share in the high-quality segment is not bad at all (in comparison to the engineering industries). However, one should also keep in mind the still very small weight of the CEECs as trading partners of the EU 12. (Landesmann - Burgstaller, 1997, annex B, table 12, also provides detailed information concerning the evolution of market shares of the different suppliers to EU markets over the period 1988 to 1994 both by 2- and 3-digit NACE industries.) Results of cross-industry regressions The simple regression analysis based solely on country and country group dummies can "explain" (using the adjusted R-squared as indicator) only the price gap variables for the engineering industries to a reasonable extent (see table 3.22). However, examining significant coefficients for the country dummies, we can see that CEECs have highly significant deficits concerning their representation in high-quality segments of EU imports (LQ1); we can also observe negative coefficients for some Western European countries such as Belgium and Italy, although the coefficients are much smaller and less significant. The negative coefficients on the LQ1 variable are by far the highest for the CEECs compared to any other group of countries. The same can be said for the price gap variable (LPG). We find a negative (and significant) coefficient for this variable also for the group of Southern EU countries, but in this case we would interpret this as showing jointly the effects of quality and of the integrated EU market reducing via lower transaction and transport costs the prices of intra-eu trade generally compared to trade with non-eu member countries (this EU market integration effect is confirmed by the result that none of the Northern EU countries, including Germany, come anywhere near the positive and significant coefficients obtained for US or Japanese exports to the EU or those of some of the EFTA countries). The representation in low-quality segments (LQ3) is significantly higher than for EU average imports for Slovenia, the Czech and Slovak Republic and Poland. The USA, Switzerland, Japan and Canada stand out with low LQ3 and high LPG. For China, India and the NICs we do observe mostly negative coefficients for the LPG (price gap) and LQ1 (representation in the high-quality segments) variables, but they are generally of a much lower order of magnitude than the coefficients for the CEECs. Important and interesting are the developments of the CEECs positions over the periods compared with the situation in We can clearly see a bifurcation into two groups: the group of "Western" CEECs (comprising the Czech Republic, Hungary, Poland and Slovenia) and the group of "Eastern" CEECs (comprising Bulgaria, Romania, Russia and the Slovak Republic): the first group achieves distinctly lower (though still highly negative) values for the price gap variable than the second group of CEECs whose negative coefficients both on the LQ1 and LPG variables are extremely high (higher than any other group of importers). 139

18 Table Regressions of quality segment and price gap variables on country (country group) dummies coefficients and t-values for NACE engineering industries , 330, and Dependent variable LQ18890 LQ38890 LPG8890 LQ19294 LQ39294 LPG9294 R square adj USA Japan Canada Germany France Belgium / Lux Netherlands Italy UK Austria Switzerland Ireland Finland Sweden Denmark Spain Portugal Greece Turkey Hungary Poland CSFR / CR Slovak Rep Bulgaria Romania YU / Slovenia SU / Russia NICS NICS China

19 Table Regressions of quality segment and price gap variables on country (country group) dummies coefficients and t-values for NACE engineering industries , 330, and (continued) Dependent variable LQ18890 LQ38890 LPG8890 LQ19294 LQ39294 LPG9294 R square adj India RoW USA Japan Canada EUN EUS EFTA EASTW EASTE NICS China India RoW Notes: ( ) indicates significance at the 5 (10) % level 141

20 Figure 3.3. Shifts in country dummies (price gap) 1988/90 to 1992/94 and GDP per capita Estimates for 3-digit NACE engineering industries 25,000 USA CH GER CAN 20,000 JAP A 15,000 E GDP per capita IRL SU/RUS ROM BUL 10,000 YU/SLO GR P CSFR/CR H TR 5,000 PL NICs2 INDIA Price gap 0 142

21 Figure 3.4. Shifts in country dummies (representation in high quality segment) 1998/90 to 1992/94 and GDP per capita Estimates for 3-digit NACE engineering industries 25,000 USA CH GER 20,000 CAN JAP A E YU/SLO P GR CSFR/CR H PL TR BUL ROM SU/RUS NICs2 15,000 IRL 10,000 5,000 GDP per capita INDIA Representation in high quality segment 143

22 The same regressions for food, drink and tobacco industries (only price gap variables) result in a smaller adjusted R-squared. The existing price gap for products exported by CEECs is smaller than for the engineering industries (which is indicated by smaller negative values of the country dummy coefficients). In the first period the dummies for Bulgaria and the Soviet Union are not significant, however the same highly significant negative coefficients on the price gap variables are visible for the two groups of CEECs and, again, the bifurcation between "Eastern" and "Western" CEECs emerges very strongly if one compares the base period with the later positions in For the textile, clothing and footwear industries the results are similar, particularly what concerns the last point. The coefficients of the price gap variable are insignificant for Slovenia and Hungary in the second period revealing catching-up towards the EU import averages. In comparison with the regressions for the engineering industries the adjusted R-squared is even lower for the representation in high-quality segments but higher for the price gap variables regressions. Figures 3.3 and 3.4 show the country dummy intercept values -- significant or not -- of the above regressions for the two periods (for the price gap and the representation in the high-quality segments variables) plotted against real GDP per capita (measured in PPPs for the average of 1993/94) for each country except China, the Slovak Republic and the RoW. For the groups of NICs the real GDP data were summed up weighted by their degree of "openness" ((exports imports)/gdp). We can see that price/quality gaps and positions in the high-quality segments of EU product markets relate roughly to a country s position in terms of GDP per capita. The arrows in the figures reveal the movements in the estimated dummies for some of the countries over the period compared to In this regard the CEECs seem to have improved their position in the engineering industries (figures 3.3 and 3.4) with regard to the LQ1 variable and also with regard to the LPG variable with the exception of Bulgaria and Russia. For textile, clothing and footwear industries (see figures 3 and 4 in Landesmann - Burgstaller, 1997) an improvement of the CEECs is observable with the exceptions of Russia for LQ1 and Romania for LPG. Note that for the price gap variable the coefficients of Hungary s and Slovenia s dummies turn positive (though not significant) for the period The price gap regressions across food, drink and tobacco industries (figure 5, Landesmann - Burgstaller, 1997) show a different picture with worsening price gaps for all CEECs (except Hungary) over the period compared to Shifts in price/quality gaps and exchange rate movements The next figures (3.5 to 3.7) show the change of the coefficients in the above regressions for the price gap variable (vertical axis) plotted against the per centage change of the nominal to the PPP exchange rate for several countries (horizontal axis); depreciation of the currency relative to the PPP rate over the period as compared to would give a positive value, appreciation a negative value 1. The picture suggests a negative relationship between these two variables and econometric tests confirm that this influence of the above indicator of exchange rate appreciation/depreciation on the change in the regression coefficients is present for each group of industries. This suggests that an increase in the ratio nominal to PPP exchange rate caused e. g. by a revaluation of a country s currency, goes along with a reduction in price gaps. 144

23 Figure 3.5 Shifts in country dummies (price gap) and the percentage change of exchange rate/ppp (both 1988/90 to 1992/94) 3-digit NACE engineering industries 0.3 CR PL 0.2 ROM SLO H Price gap Appreciation Percentage change of exchange rate Depreciation 145

24 Figure 3.6. Shifts in country dummies (price gap) and the percentage change of exchange rate/ppp (both 1988/90 to 1992/94) 3-digit NACE textile, clothing and footwear industries 0.4 H PL 0.3 CR 0.2 SLO 0.1 Price gap ROM Appreciation Percentage change of exchange rate Depreciation 146

25 Figure 3.7. Exchange rate appreciations/depreciations, changes in price gaps and changes in market shares (,,-) over the period 1992/94 Mechanical engineering 0.8 H 0.6 BUL SLO CR ROM 0.4 SR - PL Changes in price gaps -0.4 Depreciation Percentage changes of exchange rates Appreciation 147

26 Figure 3.8. Exchange rate appreciations/depreciations, changes in price gaps and changes in market shares (,,-) over the period 1992/94 Electrical engineering ROM 0.4 SLO PL BUL CR 0.2 H SR Depreciation Percentage changes of exchange rates Appreciation 148

27 However, we also see that the group of CEECs do not fully conform to this general picture over the period to Take the case of the engineering industries (figure 3.5): Here we find that similar positive shifts in the price gap coefficients were experienced by all CEECs represented in this figure (Hungary, the Czech Republic, Poland, Slovenia, Romania); however, their experiences with respect to exchange rate appreciation/depreciation differed widely. We find countries such as Hungary and Poland, whose nominal exchange rate/ppp rate appreciated strongly, alongside the Czech Republic, Slovenia and Romania (the last of which experienced a substantial depreciation of their nominal exchange rate/ppp rate ratio) and all of them experienced similar closures of their price gaps in relation to the general price level in total EU imports. In the case of the textile/clothing/footwear industries group (figure 3.6) the CEECs do conform to the general pattern but they seem to lie on a line shifted upwards from the general regression line, i. e. they achieved greater closures in the price gaps compared to what the general exchange rate price gap shift relationship across the whole sample of countries would indicate. A further exercise along these lines was conducted for the most recent period 1992 to 1994 for which data were available. In figures 3.7 and 3.8, we plotted -- this time only for the CEECs -- the percentage change in the PPP rate/nominal exchange rate between 1992 and 1994 on the horizontal axis (note that this time the exchange rate variable is defined so that an appreciation of the currency relative to the PPP rate shows up as a positive value) against the changes in the price gap indicator over the same period on the horizontal axis for the different subindustries belonging to the two engineering industries (the mechanical engineering industries on figure 3.7 and the electrical cum instrument engineering industry group in figure 3.8). First of all, we can see that for the CEEC' industries there are mostly closures of the price gaps (positive values on the vertical axis) across most of the sub-branches of the two engineering groups. Secondly, the closures in the price gaps seem to take place irrespective of whether and to which degrees the countries appreciated or depreciated their currencies (relative to the PPP rate); this is the same result we found for the period to shifts before. Thirdly, we also indicated with plus () and minus (-) signs whether industries experienced an increase or decrease in their market share positions in EU total imports over that period and we indicated with an asterisk () those industries which experienced over this period an increase in their market share positions of over 50 per cent (!). From this we can see that for the great majority of industries, price gap reductions took place alongside improvements in their market shares in EU (import) markets. Furthermore, there are many "starred" performers and this is the case in countries with strongly appreciating currencies (relative to the PPP rate) -- such as the Czech Republic and Romania -- as well as in countries with only moderately appreciating currencies -- such as the Slovak Republic, Hungary and Slovenia -- or in Poland with a depreciating currency over the period 1992 to Outward processing (OP) Eurostat s Detailed Trade Statistics also contain information about exports for and imports after outward processing of EU products 1. The data for these trade flows are recorded separately from "normal" export and import data and one needs to sum both to obtain "total" exports respectively imports. For this study we chose to evaluate the outward processing data for the 3-digit NACE industries 436 (knitting industry), 451 (manufacture of mass-produced footwear) and 453 (manufacture of ready-made clothing and accessories). 149