Miklós Szanyi:
Structural changes of manufacturing and competitiveness Second draft of the research report for the Work Package 3 and deliverable 3.5 of the Competitiveness Project
June, 2004.
Paper prepared for the EU Fifth Framework Project: „Changes in Industrial Competitiveness as a Factor of Integration: Identifying Challenges of the Enlarged Single European Market” (Contract No.:HPSE-CT-2002-00148)
1
Table of contents page Introduction
3
1. The analytical framework
8
1.1. Statistical analysis
8
1.2. Other analytical tools
8
2. Results of the statistical analysis
11
2.1. background of structural changes
11
2.2. Creating structural change and performance measures
15
2.3. Winners and losers of structural change and performance race
15
3. Connecting the dimensions of performance and structural change: calculation of Spearman rank correlation indexes
26
4. Country specificities
29
4.1. Hungary
29
4.2. Poland
31
4.3. Czech Republic
32
4.4. Spain
33
4.5. Ireland
36
5. Attempts at refining the research results: group- and market share analysis
39
5.1. Trade intensity groups
39
5.2. Correlations in the groups
40
5.3. The BCG matrix: market analysis
42
6. Composite competitiveness measures
49
7. Conclusions
53
Literature
54
2
Introduction The major aim of Work Package 3 was to establish linkages between structural changes in the manufacturing sector and competitiveness. Transition in CEE countries resulted in fundamental changes in the structure of the economy. Though, Hungary experimented with the introduction of certain market economic elements in the economy already prior to the transition process, the basic rationale of the functioning remained strict state control and central planning like in the other too countries of this research, Poland and the Czech Republic. This logic determined the buildup of all kind of branches and industries, starting with the activities carried out and parameters of products produced, through the organizational structure down to the day-by-day practice of making “business”. All these circumstances had to be changed, either replaced or adjusted to the new environments, when the transition process started. The first phase of structural changes was therefore earmarked by the adjustment process of manufacturing firms to the requirements of the then established macroeconomic framework. This process was not identical in the transition economies. The then fierce debates of the gradual approach towards the introduction of market economic institutions and liberalization versus the shock therapy or „big bang” type of policies described clearly the dilemmas of economic policy. 10-12 years later, however, we can perhaps conclude, that despite of different sequencing and speed of transition process, the three Central-European economies show much more convergence in their economic development path, than differences. From the viewpoint of structural change, it is mainly the speed of the process, and different sequencing that resulted from the alternative policies of the economies. Today the transitionbound determinants of structural change seem to have lost importance in these economies. The second, current phase of structural developments is determined rather by traditional demand and supply factors and not by transition-specific issues. Sectors with quickly expanding and changing demand usually grow faster, than industries working on more mature and saturated markets. Factor endowments and relative prices on the other hand, also influence the pool of potential activities that can be carried out in the investigated countries with comparative (and competitive) advantage. If we insist on describing the process of uneven sectoral growth during the transition process and afterwards using the traditional 3
analytical framework of supply and demand, the difference between the two phases can be best grasped by the different speed of changes. Transition and structural changes occurred in the period of accelerating globalization of world markets and rapid changes of the technological paradigm in world economy. These interlinked global processes themselves ignited important structural changes in mature market economies, and other countries as well. Thus, transition economies’ structural changes were at least partially influenced also by global tendencies. The investigated three countries were exceptionally exposed to this process. Foreign direct investments played a crucial role in both the adjustment process (e.g. through privatization) and the integration into the new technological paradigm and global markets. Thus, it is rather difficult to separate the two processes, since they were very much interlinked and went on parallel. Again, it is perhaps the speed of structural changes that can be used in defining the borderline between the first development phase (transition dominated) and the second (globalization led). The third issue that deserves attention in influencing structural change is the accession process of the three countries with the European Union. This is again a factor the influence of which can be hardly separated from the other two. In the early 1990s it was the Association Agreement and liberalization of trade with the EU that had the most important influence. Increased competition from EU-area (and also other regions) fundamentally contributed to the separation of unviable under free trade circumstances activities, and these were soon eliminated or taken over by foreign investors (only few sensitive industries were protected). Instead, through the EU-Agreements, the three economies became more and more integrated into the EU economic area. Main carriers of the process were again foreign investments. Thus, we may state, that preparations for EU-membership also contributed to structural changes, and these changes occurred prior to the current actual entry of the three countries into the European Union. In fact, an important condition of the entry was the “preparedness” of these economies. Today, the three countries completed the most urging tasks of both transition and accession to the EU. The emphasis shifted to the challenges of globalization. The three accession economies face largely similar tasks, than what old members of the EU do. The Lisbon Agenda was launched in order to ignite those changes on the level of the integration that seem
4
to be necessary to improve competitiveness in the global competition. Thus, further structural changes can be expected out of this modernization process. The question emerges here, how structural changes can be linked with competitiveness? Obviously, specialization on activities with higher growth potential will boost the development of these branches. In the first phase of the transition process elimination of nonviable activities also played a fundamental role. Again, from a strict statistical viewpoint the dropout of badly performing industries alone caused an improvement of the overall performance (overall competitiveness) even without the creation or growth of more competitive activities. In a broader sense, of course, we can not state that destruction alone can improve overall competitiveness. But it enhances the allocation of resources towards more competitive activities. Thus, the Schumpeterian process of creative destruction is at the heart of the relationship between structural change and competitiveness. Now we can turn back to the original question of how to establish statistically testable linkage between structural change and competitiveness? Since at least three important factors determined the directions of structural change, one may expect that the outcome will not be always straightforward and consequent. This is exactly the task of this project: to test if structural changes (as outcomes of many interlinked factors) improved competitiveness? Did the three economies move towards more competitive structures? In order to evaluate research findings, we also use the results of similar analysis of Ireland and Spain. These countries serve as benchmark, since their structural change was not influenced by the special tasks of transformation. The basic relationship of competitiveness and structural change may be twofold. Entrepreneurs’ logic may indicate that investments in more competitive branches promise higher incomes and profits, hence, it is better to invest in more competitive branches. On the other hand, there may also be a reverse argument: more investment in a given branch may substantially improve the sector’s competitiveness. In this later case competitiveness potential is perhaps the ex ante incentive for investments, and thus, for above average increase of sectors. Nevertheless, the direction of the relationship is from realized or expected competitive advantage through investments to above average sectoral growth. Differences in either the absolute level of competitiveness or changes in the dynamics of competitiveness induce higher growth rates and changes in economic structure. 5
However, in the context of transition economies this relationship may be rather complicated, and several times also distorted. Most obviously, competitive positions of pre and post transition periods are hardly comparable. Due to this fact, the individual factors of competitiveness (labor, knowledge, capital equipment, marketing know-how, etc.) do not reveal exactly during the process, hence it is rather expected competitiveness that moves investors’ decisions. For example, in the Hungarian electronics industry, there was a quite large pool of accumulated knowledge, personnel and experience, also capital equipment. After the first period of transition (1989-1993) the value of this large amount of production factors was effectively annulled, especially when measured by any kind of business performance measure. The branch almost ceased to exist its products were wiped out of all markets. Nevertheless, this did not mean, that all of the means of potential competitive advantage were eliminated. Personnel, knowledge, sometimes also some general business frames (few companies) remained in place, though they did not produce effective, measurable output for some time. Hence, when Hungarian electronics industry became one of the major targets of foreign investments, there were practically no measurable positive competitive incentives in the sector in the preceding 1-2 years period! Another important circumstance that may seriously distort the main competitive factorsinvestments-high growth logic is economic policy. Economic policy is capable to influence competitive positions in financial terms despite of the obvious lack of competitiveness of companies’ activities. For example, the drastic devaluations which were applied in many transition economies in their stabilization policies, artificially increased incomes from export sales, and accommodated import competition, without any change in the real competitive position of firms. Continuous devaluations even sent a negative message to economic agents concerning improving competitiveness. Therefore, exchange rate policy and most importantly, state aid policy may have distorted the financial performance of firms from the values that they would have been able to achieve if there had not been state intervention. Obviously, such distortions also deteriorate chances of finding relationships between competitive strength (measured by any financial or other performance measure) and differences of growth patterns. When research was started in this work package, the extent of these distortions was not quite clear. The empirical evidence indicated that high growth was bound to high level of 6
investments, and also to high incomes, especially in the foreign-dominated sectors of Hungarian manufacturing industry. The question then was rather, which of the potential competitive factors is most frequently utilized by successful industries, which may be absent in slow growing branches. And the anecdotal evidence suggested the answer: it was relatively inexpensive, skilled and motivated labor. There have been numerous papers dealing with the different aspects of the topic. Structural changes were also analyzed at certain stages of the transition process. The basic direction of the research, however, has always been the micro-level adjustment, or the macroeconomic impacts. Relatively little attention has been paid to the mezzo level, especially in a comparative context. Individual industries were studied, but only very few studies tried to compare the development of individual manufacturing industries. We believe that the parallel stories of these industries can provide interesting new insight into the process of reorganization of the structure of transition economies’ manufacturing industry. For it is of course true, that major sectoral changes occurred on the macro level, providing booming services sectors with above average development momentum, meanwhile many of the manufacturing industries suffered heavy contractions. Nevertheless, if we consider manufacturing as a still important part of the economy, structural changes within manufacturing are by no means negligible. There has been a fundamental reorientation within, and between industries, a major reallocation of capital and human resources occurred. As a result, manufacturing industry became competitive in the international markets, production inputs were utilized in an efficient way, manufacturing production produced a lot more added value, than before the transition process. The aspects of structural development together with technological development and foreign investments were perhaps the three most important determinants of the revival of Hungarian, Czech and Polish manufacturing. The current stage of WP 3. tries to describe these fundamental structural changes, using a variety of statistical tools. The basic hypothesis is that the transition process opened up those determining factors of world markets that started reshaping manufacturing structure. Three major elements need special attention: general market economic principles and rules, the role of multinational companies as investors, and the opening up and liberalization of markets, both foreign markets and the domestic market. In this later context closing to the EU can be mentioned as a separate fourth dimension. This paper discusses mainly the first factor. It concentrates on the impacts of general market economic principles. The basic hypothesis is 7
that structural changes must have followed directions that have a kind of economic rationale. That is growing sectors of manufacturing are expected to be those, where there is more efficiency: more productivity, more added value that can increase returns on investments. On the other hand, those industries that are inferior from these aspects should try to allocate resources to expanding branches. WP 3 is mainly concerned with structural changes of production and trade and the relationship of these changes to competitiveness, whether the direction of changes is towards more competitive sectors and products or not. Also, the pace and mood of changes are important, whether there is a straightforward and speedy process with lots of new investments, or the contraction, downsizing is the more important process behind the changes. In order to make the comparison, first competitiveness has to be defined and measured. The competitiveness rank of industries is the basis of comparison we can relate structural change to this. As a first step, various factors and measures of competitiveness are introduced and related to measures of structural change. Later we make an attempt of creating a composite measure of competitiveness that could generate a single “final” list of competitiveness, instead of the several various lists. The main aim of this exercise is to test if structural change was favoring better performing, more competitive industrial branches or not. As a second thought, we also would like to compare the winners and losers of structural change in the various economies. The analysis of the figures may also shed light on many of the specific circumstances that explain the expected differences among the specialization patterns of the different countries. We also would like to draw some conclusions concerning the impact of joining the EU on competitiveness and structural changes in both the new member states and EU 15.
8
1. The analytical framework The design of the research evolved during work. For previous attempts only described the various aspects of structural change and some of the qualitative measures of development. The measurement of competitiveness itself is not quite clear cut. Therefore the research team experimented with various different measures of both aspects (structural change and competitiveness). Parallel with work package 1 we developed a series of competitiveness measures to be tested. We called them performance measures rather, since they were all intended to measure different components of competitiveness. The most important performance measures were productivity, profitability, income generating and investments. Structural change was measured in three dimensions: employment, output and value added. We calculated averages for the first and last two years of the period, and also used the individual years’ figures. As far as the quality of the database is concerned, we must be rather critical. Due to frequent changes in the subordination of large companies with differentiated product range significant changes occurred year by year due to changes in the composition of the single NACE 3 digit level manufacturing industries (we calculated figures only for manufacturing). Also, dynamics in manufacturing remained still high in the period, though somewhat slower than few years before, and new investments, new major capacities frequently started operation in one or another branch adding quite large new turnover and employment amounts to the previous figures from one to another year. A third reason of quick and radical changes in some of the figures was the relatively small size of especially the Hungarian and Czech economies, especially when compared with the size of some of the multinational companies operating in the country. The largest firms in automotive- and electrical equipment industry and in coke and petroleum products contributed to 3-5 % of industrial production and exports. Therefore, their sudden moves starting or ceasing operation resulted in a much bigger change in the figures, than in other larger countries. Due to these data problems and the sensitivity of the statistics we decided to stay with the analysis of simple statistical measures rather, than calculating more complicated ones. Nevertheless, we intended to go one step further, than previous analyses and tried to verify the relationship between structural change and competitiveness (or at least some of the competitive factors).
9
1.1. Statistical analysis As a first trial we compared structural changes to measures of competitiveness that the teams already had, partly in the accessible database, partly from other publications. This was the first relevant approach to quantify the issue. 1. We describe structural change with three measures: sales revenue, value added and employment. The measures can be used in the form of their change over time or change in their relative share. For a first approach we calculated with the simple share ranks, and the change over time that provided the basis for the Spearman rank correlation index. 2. We compared the above measures of structural change to four performance indicators: income generation, profitability, productivity and investments. The income generating power of branches was described by the division of added value and sales revenue. Proxy for profitability was gross profit of operations per sales revenue. Per capita sales revenue served as measure of productivity, for investment we used gross investments per gross stock of fixed capital (if available), or gross investments per subscribed capital (as second best). 3. Spearman rank correlation indices could be calculated in all six relations. 4. We could also separate growing and contracting branches according to sales revenue figures, and see if profit rates are significantly different in the two groups. This is basic information about the direction of structural changes.
1.2.. Other analytical tools for foreseen for the next stage of the research Another possibility is to see if structural changes are in line with the development path that theories consider beneficial. One option is to look at the technology intensity of the production structure using the categories of OECD (high- medium- and low-tech branches). Though, observers may be skeptical with this categorization. For the essence of comparative advantages is that they are different in various countries, but one can develop the counterargument too: let us see if specialization on high tech branches really means improvement in competitiveness or not.
10
The team also used methods that were introduced in a current competitiveness analysis produced by ECLAC. Since the first stage of WP 1 is about market performance anyway, we joined this string of thought creating the suggested matrix. This concept differentiates among four types of markets. Rising stars are those branches, where both the size of total market and the market share of the investigated country simultaneously increase. Missed opportunities are those branches, where market share declines, despite of market expansion. Declining stars are those, where market share increases despite of contracting markets. Retreat is observed in branches where both market size and market share declines. This is a very simple method, it can be calculated if we use the data gathered in WP 1, and it says something about the structural changes, if they follow overall market trends or not in the different branches. To make the method a little bit more complicated, we can make the same categorization for domestic market and main export market (EU market). Theoretically, in case the domestic market is part of the single European market, similar performance is expected in both domestic and EU market. It may be therefore very interesting to look at the pre-accession and post-accession performance of Spain and Ireland, since accession and liberalization went parallel there. For new accession countries liberalization already resulted in closing the two markets fairly tightly together. The other opportunity in this research is of course, looking at differences of branches and countries. Also the comparison with the two measures of structural change will be interesting. I do not think of advanced statistical measures here, since the method of grouping is very rough anyway, just to see if rising stars gain importance and also there is retreat, or the two other groups are more important, and if performance on the two markets is significantly different. A third grouping of branches is also interesting: branches vary greatly concerning their position in international labor division. A simple characterization possibility is provided by data of WP 1. We may calculate export/output and import/domestic demand ratios. Branches with below average export/output ratio are domestic oriented, above average export oriented. Branches with over average import/domestic demand ratio are import dependent. Branches with above average figures in both measures are internationalized. It is crucial to clearly separate these internationalized branches, since their competitive advantages stem from a
11
greater variety of sources (especially globalization advantages are unique for them). Then, we can see what group gains momentum. Using the results of the already calculated correlations of performance and structural measures we also moved on in creating the composite competitiveness measure. This measure was created as a balanced average of four performance indicators, that seemed to best describe the relationship of competitiveness and structural change. Demand, income generation, efficiency and investments was included in the measure, and the weights were calculated according the principle of best fit to structural changes. The use of the composite measure provided us a last “final” rank of branches, and we correlated this rank with the rank lists of the structural change measures..
12
2. Research findings until August 2004 2.1. Background of structural changes The statistical figures of structural change were reviewed in the project. The major findings were introduced in the summary of literature, but it is worthwhile to recall the main findings, so that we can relate our research results to them. The longer term reactions to liberalization and entry to the European Union (the major changes in the environment) obviously led to increased competitiveness in all 5 countries. The literature reviews prepared for Work Package 3 (O’Donnell, 2003; Éltető, 2003; Fonfria, 2003; Marczewski, 2003; and Jakubcek, 2003) reported that production structure changes followed shifts in market demand towards better sales opportunities and potentially at least higher returns. There were, however important deviations according to size of countries. The most profound structural changes occurred in small open economies, like Ireland, the Czech Republic and Hungary. Spain and Poland are less open economies, and due to the higher importance of the domestic market, their economic structure did not change so much. Nevertheless, this fact alone does not tell us anything negative, in case domestic markets are accessible. High degree of liberalization of domestic markets was achieved in Spain during the years following the accession to the EU. In case of Poland, liberalization took momentum after the Association Treaty was signed. Also, CEFTA regulations contributed to the process, though in certain markets effective market protection remained in place. All 5 countries underwent important structural changes. Light, industry, food processing and agriculture lost importance, meanwhile engineering, chemicals increased output and employment, together with services. There was solid evidence in the case of Ireland that this process accelerated after joining the EC (Croughan, 1984). Major structural changes also occurred in Spain, after joining the EU (Myro and Gandoy, 1999). However, in both countries liberalization of investments and trade was a parallel process. It is therefore difficult to separate the effects of liberalization, and the expansion of markets (Éltető, 2000). Liberalization of transition economies preceded joining the EU, thus, major economic restructuring took already place. The Hungarian experience emphasized export orientation, an increase in efficiency and productivity, as major components of structural change (IKIM, 1996, Borsi, 1998). Food, drink and tobacco, textile and clothing, chemicals, coke and petroleum lost significantly weight, engineering wood, paper and printing increased (Éltető, 13
2000). Éltető (2000) also stressed, that Hungarian production was still very much specialized on low demand and technology sectors. However, the share of high demand and technology sectors in value added increased spectacularly and during five years surpassed the achievement of Spain in ten years. Research findings concerning productivity changes vary, as far as national economies are concerned. It seems to be certain that during the overall previous history of transition Poland and Hungary both achieved labor productivity gains. The case of the Czech Republic is less obvious, Zeman (2000) states, that there was a very modest increase in the Czech Republic that did not reach the productivity increase of the EU average. Kaderabkova, (2002) also reports about stagnation of labor productivity in the period 1996-2000. Croughan (1984) reports steady productivity increase in Ireland throughout the pre- and post accession period, Fonfria reports an 5,1 % average labor productivity growth for the period 1966-1998 in Spanish manufacturing industry. The analysis of structural change on NACE 3 digit level proved to be more difficult than what we expected. A number of serious methodological and data handling problems occurred that could be eliminated only partially. The most serious problem was faced by the Czech partners. The Statistical Office declared, that the figures of NACE 3 digit level were estimations and suitable only for the orientation. It turned out, that the cautious approach of the Office was not exaggerated, and the quality of the data was quite dubious. But the team faced problems also in Poland and in Hungary. Polish colleagues drew the attention to the fact that due to high excise tax content of sales prices and revenues, a number of sectors unduly enjoy the image of efficiency in terms of sales and value added increase. In Hungary, the problem of shifting activities and shifting in branch codes by various (partly very big and important) enterprises caused troubles. This later problem is illustrated by the next two charts. They represent sales turnover figures for the years 1998 and 2001. There are important visible differences but a careful reading of the columns also discovers, that the single biggest manufacturing branch in both 1998 and 2001 representing over 10 % share was not identical in the two charts. In 1998 it was 341, in 2001 343. Very close, but not identical, due to shifts in activity and reporting codes. Such relatively unimportant changes in real terms may cause serious disturbances in the statistical analysis.
14
Sales shares 1998 % 12,00 10,00 8,00 6,00 4,00 2,00 0,00 1
6 11 16 21 26 31 36 41 46 51 56 61 66 71 76 81 86 91
Sales shares 2001 12 10 8 6 4 2 0 1
6 11 16 21 26 31 36 41 46 51 56 61 66 71 76 81 86 91
There are also other important observations concerning the two charts, and the data in the following table. First of all, it is very remarkable that Hungarian manufacturing is strongly concentrated on a few NACE 3 digit level branches. In 1998 almost one third of total sales turnover was produced by just 4 branches. In 2001 the firs 4 branches produced over one third, the first 6 over 45 %! Also interesting is that despite of the relatively short time period
15
Sectoral structure of Hungarian manufacturing in 1998 and in 2001 Sales shares 1998 151 152 153 154 155 156 157 158 159 160 171 172 174 175 176 177 181 182 183 191 192 193 201 202 203 204 205 211 212 221 222 232 241 242 243 244 245 246 247 251 252 261 262 263 264 265 266
6,04 0,03 1,78 0,78 2,38 1,35 1,68 3,58 2,31 0,64 0,44 0,20 0,53 0,49 0,11 0,08 0,13 1,39 0,01 0,17 0,15 0,56 0,36 0,41 0,43 0,11 0,14 0,46 1,61 1,53 1,47 7,56 3,72 0,11 0,38 3,05 0,57 0,23 0,18 0,73 2,86 0,59 0,30 0,09 0,37 0,41 0,56
Sales shares 2001 266 267 268 271 272 273 274 281 282 283 286 287 291 292 293 294 295 296 297 300 311 312 313 314 315 316 321 322 323 331 332 334 335 341 342 343 351 352 353 354 355 361 362 363 364 365 366
0,56 0,00 0,30 2,04 0,10 0,53 1,07 1,48 0,48 0,13 0,31 1,29 0,78 2,01 0,58 0,22 1,05 0,05 1,02 6,62 0,43 0,37 0,55 0,05 1,43 1,24 2,33 0,86 3,81 0,41 0,39 0,12 0,00 10,45 0,20 2,46 0,01 0,31 0,13 0,04 0,08 0,97 0,05 0,00 0,02 0,06 0,16
151 152 153 154 155 156 157 158 159 160 171 172 174 175 176 177 181 182 183 191 192 193 201 202 203 204 205 211 212 221 222 232 241 242 243 244 245 246 247 251 252 261 262 263 264 265
4,735713 0,012891 1,348213 0,341743 1,716841 1,117986 1,315826 2,854634 2,474891 1,723427 0,292714 0,174427 0,39281 0,395624 0,088222 0,086969 0,041319 1,480752 0,006254 0,056633 0,100902 0,486181 0,249512 0,326754 0,476934 0,120836 0,088176 0,83462 1,026886 1,361311 1,156937 10,05592 3,157324 0,104077 0,282346 1,873774 0,657747 0,134029 0,048292 0,813998 2,946564 0,396781 0,367568 0,071163 0,330168 0,424332
266 267 268 271 272 273 274 281 282 283 286 287 291 292 293 294 295 296 297 300 311 312 313 314 315 316 321 322 323 331 332 334 335 341 342 343 351 352 353 354 355 361 362 363 364 365 366
0,76956 0,055479 0,281135 1,370892 0,054168 0,292574 0,843796 1,719424 0,30408 0,070914 0,311242 0,899926 0,600251 1,254648 0,401767 0,273542 0,82333 0,016254 1,035929 5,483524 0,296415 0,367948 0,447121 0,28552 2,162382 5,056956 1,682817 1,843215 7,79454 0,2903 0,353625 0,025094 0,002916 1,610536 0,147152 10,64187 0,011951 0,333896 0,094275 0,03572 0,039029 0,781141 0,033491 0,004613 0,050833 0,03269 0,160499
of observation, changes are clearly visible. Thus, we may see, that even after the above described period of fast economic restructuring after liberalization and transition, the 16
Hungarian economy kept on moving, though true, at a slower pace. It seems to be therefore very much relevant to compare this structural movement with other countries. 2.2. Creating structural change and performance measures In order to measure the expected relationship between “performance” (competitive strength) and structural change, we decided to measure three dimensions of structural change and four dimensions of performance. The broad aspects of structural change to be measured were changes in output or sales turnover, changes in value added and change in employment. The performance aspects were the following: income generating power, efficiency (financial: profitability), productivity and investments. The measures to be used are introduced in the next table: Structural change
Income generation
efficiency
Change of gross sales
Value added per capita
Gross operating Sales turnover profit per sales per capita
Change in employment
Value added per sales turnover
Value added – labor cost per sales Cash-flow per sales
Change in value added
productivity
investments Gross cumulated investments per cumulated sales turnover
Value added per capita
For both structural and performance measures change in absolute figures (difference) and growth rates were calculated. In the firs approach starting years’ figures were compared to end-years’ figures. Later moving averages were introduced, since the exceptionally high (low) figures of the initial or ending years were to be smoothed. When the measures were calculated for all the NACE 3 digit level branches, the results were ranked, and the basic tool for further analysis were the various lists of rankings. We compared the lists with each-other, but especially the performance rankings with structural measure rankings. Using the lists we also summarized the results in a table of winners and losers of both the structural change and performance. The final step in the first phase was the comparison of the two lists of winners and losers of structural change and performance. 2.3. Winners and losers of structural change and performance race
17
The following analysis contains the results of the sectoral analysis done for the 5 countries. First we introduce structural figures (sales and employment) using shares and changes in shares to compare “winning” (that is growing) and “losing” (declining) industries. We produce also the rankings that can be used for Spearman rank-correlation calculations later on. The comparison of the various structural data series is also tested for “coherence”, that is we compare the various indicators concerning the branches if they move in the same direction over time. This is a kind of quality control of the database: if the indicators show by and large the same structural movements, we can rely on them more easily. Then, we compare the structures of the individual countries and search for similarities. The following step is a similar analysis of the performance indicators: value added per sales turnover or capita, profit per turnover, per capita turnover and investments. Here again, various indicators are used, checked for “coherence”, and branch-maps are prepared highlighting the “winning” and the “losing” branches. The results are then compared across the countries. This simple statistical analysis is then topped with the Spearman rankcorrelation calculations between indices of structural change and performance measures.
Structural characteristics: the largest industrial branches of the 5 countries at the beginning and at the end of the period of observation ( % share of branches in sales turnover) 1995-7 NACE 30 241 155 151 223 154- 159 244 321 6 % 14,8 10,7 7,6 6,1 6,1 5,2 4,8 3,9 2,6 Ireland 1998-00 NACE 30 241 223 154- 244 155 321 151 159 6 % 18,3 17,4 8,2 4,5 4,8 4,9 4,1 4,1 4,1 1993-1995 NACE 341 241 158 159 151 232 252 244 271 % Spain
Czech Rep.
Poland
8,8
4,5
4,3
4,0
3,9
3,5
2,6
2,6
2,6
1997-2000
NACE 341
241
158
151
159
252
343 232 361
1993-98
% 10,7 NACE 273
3,9 341
3,4 232
3,3 295
3,1 155
3,0 158
2,8 2,7 2,4 151 159 241
1998-2002
% 8,37 NACE 341
8,02 271
6,76 241
5,59 343
5,42 158
5,35 151
5,13 4,34 3,74 159 155 232
1995-1998
% 9,58 NACE 232
7,68 341
4,97 151
3,45 159
3,38 158
3,36 271
3,08 2,81 2,77 155 241 160
1998-2001
% 6,35 NACE 232
5,69 341
5,48 151
5,25 159
4,84 159
4,83 271
3,36 3,53 2,9 155 361 241
18
%
Hungary
8,12
5,85
5,18
4,64
4,55
3,37
3,24 2,96 2,89
1998
NACE 341
232
300
151
323
241
158 244 252
2001
% 10,54 7,56 NACE 343 232
6,62 323
6,04 300
3,81 316
3,72 151
3,58 3,05 2,86 241 252 158
10,64 10,06 7,79
5,48
5,06
4,74
3,16 2,95 2,85
%
Data in this table provides some basic information about the economic structure of the observed economies. It seems, that larger countries tend to have a more diversified, or at least less concentrated economic structure. At the extreme case of Hungary the single most important automotive industry provided over 10 % of manufacturing sales turnover at the end of the 1990’s. Together with further 3 branches, with electronics and telecommunication equipments as well as petroleum products the ratio increases to over 30 %. In other countries the structure is much less concentrated, but the first 4-5 biggest branches take 20-25 % in the other surveyed economies as well. This high level of concentration was a result of both interand intra-industry developments. For example, the Hungarian electronics industry used to have 5-6 % share in manufacturing before the transition as well, but this share declined to almost nil by 1992-3. Now if we compare industrial dynamics using this year as a comparative basis, we would get astonishing high growth rates. But this high growth is in fact the result of a deep intra-industry restructuring process, and does not necessarily mean quickly increasing specialization! We must be therefore very careful with the analysis of structural indexes. Another important lesson of the table is that the structure of the various economies differ. This by no means surprising, since the essence of specialization is international differences in the division of labor. However, from a development policy point of view, the desired specialization patterns are rather similar. Today’s high technology industries are the preferred targets of investment promotion. As many development theories pointed out specialization on branches with leading technologies may provide strong (technological) spillover effects for the economy as a whole. And to some extent the current fundamental wave of globalization points very much towards increasing intra-industry specialization. This means, that in our case parallel increase in the NACE 3 digit level industries is by no means surprising. In many cases global production networks establish intra-regional specialization affecting many neighboring economies. This is most obvious in the case of automotive industry, but there are
19
also many other examples. Thus, quick parallel growth in the three neighboring transition economies is also very likely in a number of industries. This is clearly shown in the table, where branch 341 (automotive) is on first or second place in four of the five countries. Also, 232 (petroleum refining) is on top everywhere, together with 241 (production of basic chemicals), 158 (miscellaneous food), 159 (beverages), 151 (meat products). Of course, these later industries are domestic market oriented, and internationally not so strongly competitive. Globalization of these branches occurs through international mergers and acquisitions. The pronounced similarities of the manufacturing structures cover important details of varying dynamics across branches. The next table contains structural dynamics: the most quickly growing branches of the 5 surveyed economies. Growth rates of sales and employment Sales NACE 241 321 95-00 % 421 375 Ireland Empl. NACE 171 333 95-00 % 266 163
171
333
223
284
363
30
311
331
350 284-
276 321
243 322
217 331
188 271
185 281
175 361
166 203
129 342
82 294
69 292
62 333
57 343
56 285
54 341
NACE
335
311
143 323
%
208
127
80
79
72
69
66
65
62
59
NACE
335
342
311
323
285
263
284
264
281
294
%
163
70
63
58
57
54
53
50
50
49
NACE
323
372
334
271
221
315
160
266
343
272
Czech Rep.
% Sales NACE 98-02 %
522,5 296,9 281,2 253,8 200,1 176,7 164 300 335 204 323 365 174 284
162 285
160 332
143,8 343
13127
221
252 263
225 316
199,2 312
Czech Rep.
Empl. NACE 93-98 % Empl. NACE 98-02 %
267 245 59,2 156
50,3 335
45,7 372
45,5 203
95-98 NACE
369,4 348,0 276,1 214,2 156,2 126,0 125 362 285 282 157 267 281 323
124 343
102 205
96,4 296
% 98-01 NACE
743 296
302 173
281 343
274 285
270 267
265 263
260 322
259 316
252 205
231 331
390
274
215
208
191
184
179
159
158
157
362
296
285
282
204
205
252
157
343
212
196
190
162
159
156
150
147
144
141
138
Spain
Poland sales growth
Sales 93/597/00 V.A. 93/597/00 Sales 93-98
% 95-98 NACE %
716,5 653,8 531,2 441,6 354 157 285 252 315 271
166,8 139,2 64,5 300 204 323
63,9 284
61,6 285
59,8 174
20
Poland 98-01 NACE employm % growth 93-97 NACE
296
173
285
267
281
221
223
252
343
204
246
194
166
130
117
113
112
109
108
108
335
352
242
247
251
372
362
156
313
364
% 98-01 NACE
185 267
102 314
98 343
81 316
75 160
56 364
55 322
54 323
53 363
51 211
3952
856
626
584
351
295
261
243
212
202
267
314
323
275
285
172
316
204
354
364
1088
152
148
84
64
57
56
55
52
50
Hungary sales
% 98-01 NACE
Hungary employm
%
The first important conclusion is methodological again. The very high growth rates e.g. in the case of the Hungarian top scorer branch 267 which is stone cutting and forming indicate in most cases either a small branch’s very low level of production at the base year, or some kind of inconsistency in the database. On the other hand, as concerns dynamism we see, that the highest growth rates in each country are very impressive, even if we take into consideration the different length of the captured periods. The quickest growth rates of Ireland are comparable with those of the transition economies. Spanish figures are somewhat lower, and also, figures of the earlier period of 1993-1997 of Hungary show slower growth, which is not surprising, this was just the time of deep recession of the Hungarian economy. Otherwise, the figures provide the impression of very dynamic growth and restructuring of the economies. The process of restructuring was strongest in branches which were not included in the list of the biggest manufacturing branches. Large industries usually grew slower than small or medium-sized ones. There were, however a few exceptions, mainly in the relatively smaller two economies of Ireland and Hungary. In Hungary branches 316 (other electrical apparatus), 323 (television and radio recorders), and 343 (engines of road vehicles), in Ireland 241 (basic chemicals) and 321 (electronic subassemblies) were the quickly growing large industries. In Spain branch 341 (motor vehicles) was big and quickly growing. All the above mentioned industries belong to the internationally most concentrated and globalized industries. The represented figures were shaped by the activity of large multinational corporations. On hand of the previous two tables we may draw some conclusions and make comparisons of the structural developments in the 5 countries. The next table contains the code numbers of those branches that appeared to be the winners and the losers of structural changes during the
21
second half of the 1990’s in the surveyed 5 countries. There is a considerable number of branches both winners and losers that appear in more than one countries’ list. The most important winners are 281 (metal structures), 316 (other electrical apparatus), 323 (television and radio receivers), 343 (engines and appliances of motor vehicles), 361 (furniture). This is a set of traditional material-intensive branches, and two globalized industries (automotive and electrical appliances). Among the losers we see 171 (textile fibers), 183 (fur), 191 (leather), 247 (man-made fibre), 271 (iron and steel), 273 (ferro-metal processing), 293 (agricultural machinery, 355 ((other transport equipment). That is, representants of the textile and leather industry, steel, and some machinery: in most cases the low end of the industries. It seems, that the basic direction of structural change is not so much between industries, but rather from the simplest low-end activities towards more complex processing, but not high technology either. Winners and losers of structural change according to the overall evaluation of the measures Ireland Spain Czech Rep. Poland Hungary win 223 241 244 267
lose 172 173175 176-
win 341 343 342 311
lose 365 296 154 193
win 174 204 223 247
lose 171 175 177 183
win 151 157 158 221
lose 171 172 181 183
win 232 251 267 314
lose 173 183 191 246
282-. 18
285
211
281
191
245
247
316
247
321
191-
294
205
284
192
252
271
323
273
333
193
353
181
285
193
281
293
351
355
246
244
182
300
265
343
355
364
365
251
246
175
316
271
361
353
295
155
172
323
273
297
335
293
351
343 361
315 366
The above described directions of structural change can be interpreted as a very logical process that indicates a smooth, albeit fairly quick shift, but no jumps from simple activities to skill and technology intensive production. This finding is in fact similar to what has been described in the literature reviews of the first phase of the research. It can be interpreted both as an important and welcome positive development process. In the transition economies this
22
means the dropout of many labor and material intensive activities of light and heavy industry, which played an important role in these economies’ strive for autarchy. On the other hand, in engineering industry the new activities sometimes seem to be more simple much less technology- and skill-intensive, than what once used to characterize these branches. A perhaps useful explanation of this is that this “technological degradation” is rather a consequence of globalization and not so much of transition. The concentration of production in the various engineering industries calls for selection and increased specialization. Full-scale production systems can hardly survive, and there is a chance for affiliates in transition economies to climb up on the technological ladder. In fact, the technological upgrading of production structure within the multinational affiliates is under way in many cases. There is sound empirical evidence on this process. Later on this process may become the driving force of structural changes. The second set of variables describe manufacturing branches’ performance in the 5 countries. Winners and losers are collected again in the next two tables in the four areas described in the previous section. There is very little overlapping among the 5 economies in terms of performance. It seems, that only in few exceptional cases coincide the best performing branches, and there is even less similarity in the group of losers. There were only 4 branches that were exceptionally efficient in more than two economies: 159 beverages, 160 tobacco, for income generation, but here excise tax also increases the primary measure of value added, also 311 electrical engine and turbines belonged to here. 223 reproduction of recorded media scored high for productivity. In case of the losers we found 152 fishing products, 154 vegetable and animal fat and oils, 193 shoe and apparel, 363 music instruments, at various performance indicators. Productivity (per capita sales) and income generation (value added per employee) differentials: winners and losers Product- win 241 30 321 244 155 159 331 221- 154 151 223 ivity loose 363 193 351 264 205 204 315 333 191 173 323 Ireland 1995-00 Income win 16 159 311 344 241 261 221 321 264 363 334 generat. loose 151 155 30 157 191 223 152 271 341 154 313
252 311 267 282 284 331
201 204 322 296 335 341 217 271 274 323 314 334 372 363 223
Product- win ivity loose 158 151 171 352 223 193 232 221 154 363 Spain
Income
win
264 296 271 274 311 211 314 335 273 354
23
1993-00 generat. loose 366 152 Product- win 300 355 ivity loose 351 364 Income win 365 300 Czech generat. loose 335 351 Rep.
333 223 355 193 154 183 351 363 365 335 322 332 353 201 268 174 321 251 232 372 156 272 203 293 362 314 274 315 158 154 353 322 160 323 191 354 321 342 245 212 192
232 372 314 283 223 263 272 203 316 313 294 341 323 362 232 264 322 352 175 343 160 354 247 297
Poland
Product- win ivity loose 223 154 231 204 183 283 313 221 296 364 30 284 203 Income win 159 160 182 183 221 265 283 294 331 332 333 334 353 generat. loose 151 152 155 156 157 173 191 231 271 273 300 323 341
Product- win ivity loose Hungary Income win generat. loose
154 155 157 160 223 232 241 243
244 245 300 323 341
174 177 181 182 183 192 193 204 205 351 363 365 366 154 159 160 211 232 241 244 245
246 265 266 311 323
152 174 176 177 181 182 183 193 201 204 205 363 366
The list of twice mentioned branches was somewhat longer. For productivity winners we can count branches 154,157, 159, 160, 217, 221 various branches of food industry. Losers here are 193, 204, 205, 351, 363 traditional crisis industries, like shipbuilding, and wood products. Interestingly, in terms of productivity gains, food industry was very much fragmented both among the various parts of the industry and also between countries. For example, 151 meat products and 154 vegetable and animal fat and oils were mentioned both as winners and losers in different countries. In the case of income generation the picture is more homogenous. Beside of the already mentioned branches, further winners in two countries were 211, 221, 241, 264, 265, 331, 334. These are branches like paper and printing, basic chemicals and construction materials, medical and optical equipment, usually material-intensive, low technology branches. The losers in the income generation competition were 151, 154, 155, 157, 183, 191, 193, various branches of food industry, fur, leather and shoe industries. Quite homogenous group. Profit/sales and total investments Profita Win 16 159 bility loose 151 271 Ireland Invest Win 321 311 ment loose 16 223
241 244 311 321 331 362 245 334 312 363 191 155 323 313 293 264 271 157 274 172 265 284 274 262 244 245 334 315 221 252
363 323 154 151 173 193 18 332 30 292 263 314 354 274 211 272 342 273 311 181 341 343
Profita win bility loose 333 282 232 334 323 322 352 355 363 296 316 333
24
Spain
Czech Rep. 19982002
Invest win ment loose Profita win bility loose
Invest win ment loose Profita win bility loose Poland Invest win ment loose Profita win bility loose Hungary cf/s Invest win ment loose
264 263 321 202 353 173 284 275 314 222 265 261 232 192 181 183 323 193 191 157 242 333 182 342 271 204 268 245 159 231 153 333 176 156 151 361 297 293 342 323 201 322 242 351 354 282 192 295 300 323 365 174 204 242 267 152 354 157 281 364 183 191 264 271 202 192 265 177 273 175 193 231 154 158 175 211 242 244 245 263 265 313 331 332 172 183 191 271 272 275 284 296 342 351 352 355 173 205 221 251 263 272 284 285 286 287 351 364 154 172 176 181 183 192 202 212 231 271 355 365 160
183
364
267
152
261
177
171
265
362
155
335
314 223 353 316 365 221 232 181 174 341 268 204 158
314
283
171
201
267
244
173
294
268
321
251
363 154 353 272 155 354 316 364 335 151 300 273
In the case of profitability multiple mentioning was even less frequent, and for investments it was practically absent. For profitability double mentioning was recorded for 160, 211, 244, 245, 263, 265, 311, 331, 362. These are tobacco again (because of the excise tax), paper and pulp, medicaments, drugs detergents and other fine chemicals, which are all suspect of enjoying some kind of patent or brand protection. There are also branches of the construction material industry, electrical engines and turbines, medical instruments and jewelry. Losers in terms of profitability were 172, 191, 232, 271, 296, 323, that is textile and leather industry, surprisingly petroleum products, iron and steel, weapons and ammunition, and radio and television receivers (another surprise). The most important finding of the investment performance lists was that they contradicted in many cases the other performance measures’ lists. It seemed, that high level of investments was carried out in branches of below average productivity and profitability. This can be interpreted easily as a sign of costly investment projects reducing profitability (despite of the delayed accounting of investment items through depreciation schemes). On the other hand, investments may have been necessary to be carried out in order to improve substandard performance. The last performance table summarizes the winners and losers of performance competition using the rankings of the above mentioned four measures. As it is seen, the most that we can observe is double mentioning (except for 221 – publishing, a winner in 3 countries, and 151 – meat processing, a loser in 3 countries). Likely winners in this table are 160 tobacco, 211
25
paper and pulp, 241 basic chemicals, 244 drugs and medicaments, 265 cement, lime and plaster, 311 electrical engines and turbines, 331 medical instruments. Losers were 154 vegetable and animal fat and oils, 155 milk and diary products, 157 animal feed, 183 fur, 191 leather, 271 iron and steel, 363 music instruments. Winners and losers in the performance competition Ireland Spain Czech Rep.
Poland
Hungary
win 159 16 221 241
loose 151 155 157 191
win 211 271 273 274
loose 154 193 223 232
win 172 173 176 211
loose 151 153 154 155
win 221 244 261 263
loose 172 183 191 231
Win 154 160 221 223
loose 151 152 177 183
244
204
296
333
247
156
265
271
232
205
311
264
311
352
265
157
331
341
241
334
321
271
314
363
351
158
332
342
243
361
331
313
335
172
352
334
323
354
175
343
206
285
211
294
194
363
355
363
The performance measures deteriorated in the five countries usually in food industry, leather and shoe industry, textile industry and steel industry. It seems, that the “usual” crisis industries’ position did not improve during the late 1990’s in neither of the 5 economies. Winners industries that could improve their performance measures were also traditional industries, many of them quite capital and material intensive. We could not detect a significant breakthrough of medium or high technology industries in the performance competition. The only exceptions were electrical engines and generators, as well as medical instruments. A preliminary conclusion here may state here therefore, that changes in the relative performance landscape discouraged investments in crisis industries, and seemed to encourage investments in industries enjoying some kind of protection (patents, established brands, etc.) or in a number of highly capital and material intensive branches. In this later case maybe the relatively cheaper access to some natural resources (including pollution rights) can improve financial conditions of investments.
26
As a concluding part of this section we may compare the list of performance and the list of structural change winners and losers. There is fairly little similarity between the two lists. For the individual countries we found that in the five countries there were only 7 branches that improved performance and also gained in weight. “Logical” winners tended to be more numerous in Ireland (3 out of the 7), meanwhile “logical losers concentrated to Poland (4 out of 6). There were only two loser branches that were mentioned in more than one country-lists in the performance competition, and were also included in the structural change losers’ “top list” of the five countries. They were 191 leather, and 271 steel. On this highest aggregation level we did not find any coinciding winners of both performance and growth competition.
27
3. Connecting the dimensions of performance and structural change: calculation of Spearman rank correlation indexes Comparisons of the lists of winners and losers of performance and structural change (growth) revealed the fact, that there was no general overlap between the two aspects. In a few cases the top and bottom performers were identical, but in most cases not. But they were not controversial either. Thus, the measurement of the strength of correlation between the individual indicators may further help us in defining details of the relationships. In this part the main results of the Spearman rank correlation calculations are presented. As already mentioned, we compared the rank lists of the various structural change and performance indicators. Since the research methodology was refined during the research process, not all indicators were used in all 5 countries, and in some cases different versions of the indicators were calculated. For example, per capita value added and value added divided by sales turnover both were used to describe income generating potential. Also, the same measures were used in different ways. For example, from figures of structural change rank lists were created using the absolute weight (share in total), change in the share, growth rate of both the absolute figure and the share. Wherever several versions of the measures were calculated, fit was compared. Usually, no major differences could be observed, but of course we considered the better results. If there was significant difference when using alternative versions of the measures, this will be described in the country chapters properly. The table on next page summarizes the most important results. As it is seen, we found significant correlation between a large number of performance and structural change indicators. It is also clear, however, that there were also numerous cases when the calculations did not prove the existence of correlation. The best, most numerous positive results were observed for the productivity measures. Sales/employee or output/employee measures showed correlations with both sales and value added, but usually not with employment. This is also logical, an increase in employment is usually considered as an extensive form of growth, meanwhile if production (measured by sales or turnover) grows without a proportional increase in employment, this by definition means an increase in productivity. This string of thought also suggests, that among the observed measures the strongest determining factor of intersectoral variation in growth was productivity differences.
28
Income generation and financial efficiency (profitability) showed correlation with growth measures in significantly less cases than productivity. Moreover, in the case of income generation both positive and negative correlation was observed in the relationship with sales and output. This suggests that expansion of manufacturing branches was followed by a decrease in the income generating potential. When for example sales grew, the share of value added in sales revenue was reduced. This may be a sign of intensive price competition. Expansion is only possible if prices go down, and this reduces the value added content of the sales turnover. This partly also corroborates with the efficiency measures, for out of the three cases financial efficiency did not show positive correlation with growth in two cases. In the third case we found contradicting efficiency figures. Now, if we look at the various growth measures, we can see that value added and sales turnover both “produced” significant results in roughly same number of cases (value added slightly more). Employment was much less sensitive. We can not draw clear conclusion as concerns the mode of calculation. In most cases the absolute figures or shares produced good results, but there were also numerous cases where growth rates proved to be the good measure instead of absolute figures.
Type of structural measure
Spain 1993-1995
Efficiency (cashflow/sales or operating profit/sales)
Productivity (sales/employee or output/employee
Value added
-
**
Income generation (value added/employee or value added/sales **
Investments (gross investments or cumulated investments/cumulated sales **
Sales Value added
**
** -
** **
-
Sales Value added
**
** **
** **
-
Sales Sales Value added
-
**
** ** (-) -
** **
Exports
-
-
**
Sales
**
-
-
Sales
**
** (-)
-
1998-2000 Growth 1993-5/199800 Czech Rep. Growth 199800/2000-02 Poland absolute change
29
Share Sales
-
**
**
Value added
**
-
**
Value added
**
-
-
Value added
**
**
**
Sales
**
**
-
Output
**
**
-
Value added
**
**
-
Employment
-
**
-
Sales
-
**
-
Output
-
**
**(-)
Value added
-
**
-
Employment
-
-
-
Sales
-
**
*
-
Employment
-
-
-
-
Value added
**
*
**
-
Sales
-
**
-
-
Employment
-
-
-
-
Value added
**
**
**
-
Change in share absolute change Share Change in share Ireland Absolute share Absolute share Absolute share Absolute share Growth Growth Growth Growth Hungary Absolute change Absolute change Absolute change Growth Growth Growth
30
4. Country specificities Now, in this section we discuss those details that seemed to be rather peculiar for one or the other country. The same set of variables will be reviewed at closer detail. 4.1. Hungary In the case of Hungary the basic measures were calculated in two basic versions: growth rates of the actual values of the measures and changes in the sectors’ share in total. The growth rates were calculated for both the starting and ending years’ values and for the averages of the first two years and the last two years. The available database covered the period 1998-2001, that is four subsequent years. For some measures data from another study were also used. These figures related to the years 1993-1997. Thus, in some cases the total time-span of the survey was 1993-2001. The comparisons of single years’ growth rates and average growth rates revealed the fact, that in some cases very large peaks and drops could be observed. These figures very much distorted the picture on growth dynamics They reflected either mistakes (or maybe changes in the statistical classification of some large industries) or exceptional market developments, but by no means steady and long term development. Such strange figures very much limited first of all the chances of successfully calculating robust statistical results, and also the scope of potential interpretations of the results. In terms of structural change the Hungarian study revealed the fact that a group of branches that increased turnover and also employment could do this expansion only at the cost of lower value added content. Branches of metal forming industry and electrical engineering expanded in such a way. On the other hand, automotive industries and other vehicle production simultaneously increased sales, employment and also value added. This segment was the most obvious winner of structural change in Hungary. However, the general lesson of the four structural change (or growth) indicators was that growth patterns varied, the various measures showed strong divergence, especially in the case of losers, weak performers. An explanation of this besides of weaknesses of the database could be that the three indicators that the survey used may reflect different type of growth. Especially value added and employment figures could reflect very different aspects: patterns of intensive and extensive growth. As concerns performance, the situation was somewhat better, since most performance measures supported each-other, and there was much less contradicting value. Using 5 31
performance indicators best performing branches seemed to be branches with some kind of monopoly or other type of protection (patents, trade marks, high level of concentration). This “security” was naturally reflected in increasing efficiency and higher level of value added, but also in productivity. But also branches of chemical industry, and construction materials industry belonged to the best performing branches. Investment did not seem to be a sensitive measure of performance. However, if we look at the lists of performance winners and the list of structural change winners, there is very few overlaps between the two. The next crosssection table contains the winners and losers of the two aspects. In the middle sections there are only two branches, these are the ones that were winners (314) and losers (183) of both structural change and performance. Cross section table of winners and losers of structural change and performance Structure winners + 251, 267, 316, 323, 351, 364 Performance win+ 314 154, 160, 221, 223, 232, 241, 243, 244, 264
++ -+ 183,
Performance lose151, 152, 155, 177, 205, 334, 353, 361, 363
+- -Structure losers – 191, 246, 247, 273, 355, 365
Nevertheless, the matrix does not exclude the possibility of finding some significant correlation between the two aspects, since there are no contradicting values that would be located at the second diagonal of the matrix (boxes -+ and +-). This was also tested by the calculation of Spearman rank-correlation indexes for the various lists of rankings based on the structural change and performance indicators. We also calculated indexes for the same type of measures, especially for the performance measures. Most of the highly significant correlations were found between performance measures and value added as growth indicator. Though value added as structural change measure is methodologically not fully independent from some of the performance measures (most obviously from per capita value added). Despite of this impact, changes in rankings of value added described the best the structural changes of the Hungarian manufacturing industry in the 1998-2001 period. Meanwhile earlier development was earmarked rather by rapid intersectoral changes, by the end of the 1990s the 32
main development pattern became intra-sectoral shifts towards more sophisticated and more value added producing activities. There is much empirical evidence gathered about this process. The calculations also proved a co-movement of value added and profitability. Sectoral change occurred in directions mainly where production resulted in higher level of cash-flow and profits. The co-movement of the production of value added and profitability obviously indicates improvements in competitiveness of growing sectors. We can also spot the sources of increased competitiveness. It is certainly local labor. The major source of improved competitiveness in the 1998-2001 period was a shifting towards activities producing higher level of value added. More added value can be divided among three major cost factors: labor cost, investment (depreciation), and profits. Our calculations show no significant correlation with investments, therefore, the shift was largely achieved using existing capacities. The higher level of added value was basically realized as profits, since until 2001 there was only a modest increase in average wage costs (in fact, wages started to rapidly increase only in 2001). 4.2. Poland Macroeconomic indicators of the Polish economy also suggested the differentiation of two development phases of the country. After the transformational recession a quick and deep restructuring process took place between 1995 and 1998. This quick restructuring was supported by advantageous internal and external conditions of the Polish economy. The 19982001 period was characterized by much worse conditions and a slowdown of growth, that also meant a limit to positive structural changes, “outgrowing” of one industry by another. Basic structural changes of the Polish economy were very similar to the Hungarian pattern. Most important losers were textile, leather and steel industries, meanwhile the quickly growing sector was a mix of new emerging branches like motor vehicles and publishing and traditional low tech industries like structural metal products or meat processing, and some other branches of the food industry. When looking at the performance indicators we can discover among the winners the same protected industries, than in Hungary. Also, various branches of construction material industry scored well. Apart of these there was also television and radio transmitters, measuring instruments, other transport equipment and some 33
metal forming among the winners of the performance race. They were mostly not identical with the winners of structural change (except one branch: 221 publishing). On the other hand, losers of performance were more likely to be also losers of structural change (183 fur, 271, steel, 363 musical instruments). All in all, the cross-section table of performance and structural change indicators did not bring much more encouraging picture than in Hungary. This was also reflected in the Spearman rank correlation indexes. Rather few and not very strong correlations were found. These were value added as structural measure and income generation, sales revenue and efficiency, value added and efficiency, sales revenue and value added with investment activity for absolute changes in the measures. 4.3. Czech Republic The Czech transition process can also be divided into phases according to the type and intensity of structural changes (and other factors). Between 1989 and 1993 a general decline of economic activity was observed due to the liberalization measures, but this affected branches very similarly, thus no large-scale structural changes occurred within manufacturing. During 1993-97 period macroeconomic conditions of growth and recovery were much more favorable. Certain branches could outgrow others. This period was characterized by rapid modernization of sectors too, thus, both intra- and inersectoral shifts occurred. The third, most current phase between 1998-2002 was earmarked by another recession. The major event of the period was the accomplishment of the privatization process. The sale of major banks that also owned much of the property of manufacturing industries, also affected the ownership patterns of industry. This gave a new impetus to industrial companies’ restructuring. On the other hand, fast growing real wages deteriorated competitiveness, since it was not offset by an equally quick increase in productivity. Structural change winners of the Czech manufacturing industry in the 1993-2002 period were a large number of engineering industries: electronic components, radio and television transmitters, radio and television receivers, measuring instruments, automotive parts, office machinery, aircraft and spacecraft. The long list of skill and technology intensive engineering branches makes the impression, that the Czech manufacturing industry changed its structure in the direction which is sought by so many politicians and observers. This was achieved at 34
the cost of losing weight in textile and clothing, footwear and steel industry. The direction of changes seem therefore very much in the line of more demanding, higher value-added producing engineering industries, which was observed also in Poland and in Hungary, but only in a few branches, and only second to some more traditional industries of low and medium level technology and skill intensity. Some of these lower end branches also gained weight in the Czech Republic, thus the picture is not completely different. Reproduction of recorded media (223), textile fibers (174, 247), metal products (281, 284, 285) are also among the winners. When looking at the performance measures we can observe not less difference: protected or monopolistic branches perform favorably, but there are also many others. Tobacco, paper and pulp, coke products are there on the list, but also apparel, pesticides, metal products, television receivers, medical equipment, various transport equipments. Some of them are suspicious to receive state subsidies. Among performance losers we can see a very mixed group of branches. Behind textile and leather there is also petroleum products (this was winner in Poland and Hungary), iron and steel, but also computers and aerospace. The latter two probably depend very much on the performance of very few companies. The correlation between the measures was not very strong in the Czech case either. The growth rate in the income generating power was negatively and significantly correlated to both growth in sales and growth in exports. Thus, the inverse correlation suggests also here an extensive growth pattern: increased sales are bound to lower prices and thus, shrinking value added potential. The Czech indicators showed a very strong significant correlation between the structural change measures and the growth rate of investments as well. Investment was not recorded as significant in Hungary, for example. 4.4. Spain The analysis of Spain and Ireland serves two purposes. For one they are economies of similar development level and structure than the surveyed three transition economies. Thus, a comparison of structural changes in these economies can serve as a kind of benchmark, an etalon of countries which were not to cope with tasks of transition, just with tasks of modernization. On the other hand, structural developments of the two countries before and after joining the European integration may indicate what can be expected in the three new 35
member countries. If there will be a change in development patterns? Or in general, will there be and what will be the effect of membership on the structure of manufacturing? Will it help restructuring, an increasing role of more competitive branches? Or will membership be rather a new factor that reshapes competitive environment of the single manufacturing branches? Unfortunately, until now, the Spanish paper could solve only the first task, since comparable NACE 3 digit level data for the pre- and post-accession years were not available. In the next phase of the research, even if data will not be fully comparable, the most important processes of pre- and post-accession will be described. The current contribution analyzed the period between 1993 and 2000. The primary concern of the Spanish authors was if structural change went in the direction of higher technology and skill level, which is also expressed in higher local value added. Though, competitiveness is not directly bound to such development, and it is also possible, that competitiveness of more traditional branches can also contribute to the general welfare, an up-grading in technological levels is widely believed to bring important general spillover effects for all segments of the economy. From this viewpoint the Spanish development is not regarded as a clear success story. High technology branches continuously play a marginal role, and the most robustly growing Spanish industries are rather traditional industries with low or medium tech requirements. Also, many of the traditional crisis industries play important role in the Spanish economy, and their restructuring is still not finished. The technology gap was not narrowed significantly. After Spain’s entry into the EU in 1986, there have been important changes in the productive structure of Spanish manufacturing basically as a result of: (1) the need for greater opening up to international markets, as a means to the Spanish economy’s full integration into the international context, (2) the restructuring of several sectors through an industrial modernization policy in the mid-80s, which at the end of the nineties was still taking place in industries such as shipping, mining or iron and steel; (3) the need for improvement in the productive efficiency of manufacturing in order to compete in a highly competitive market with no barriers to partners by increasing the size of firms in order to achieve economies of scale, and by improving technological capacities.
36
The Spanish industry has shown a growth model based on the positive evolution of productivity reducing the employment along time. This kind of growth model gives an idea of the maturity of the industry and the restructuring suffered specially from the mid 80´s until the end of the 90´s. The final situation of the Spanish industry after that process may be characterzsed by: 1.- The exploitation of scale economies by a number of manufacturing industries. This is based on the enlargement of the markets and on the restructuring of the manufactures mentioned. Industrial policy in these years was oriented to the financial and size reorganization of the industry in order to get competitive advantages. However the high tech industrial sectors does not show a good trade performance because of the lack of R&D efforts. 2.- The specialization of production in some industrial sectors as motor vehicles, paper products and printing and some branches of machinery and communication equipment. However, the lack of big firms reduces the possibility of exploitation of scale economies needed to get competitive advantages.
3.- The growth of labor productivity over 5,1% along the period 1966-1998 for the manufacturing while for the whole economy the productivity growth was 3,7%. The branches with outstanding growth were those of high and medium technological content.
4.- The growth of investment in capital goods which is higher than the growth of the GDP in the period 1985-1996 and accounts around the 12% of the GDP. However the evolution has been negative in the last years. 5.- The opening up of the industry to international markets which growths from 20% in 1985 to more than 50% in 1999, measured by the ratio exports over GDP. These processes have their reflections in the calculated structural and performance measures. The biggest industries were traditional industries like food industry, basic chemicals, pharmaceuticals, printing and publishing and automotive. The authors stressed the high degree of stability of the economic structure. This means that the quickest growing industries 37
are not the largest ones. Winners of the structural change are various engineering industries (335, 342, 311, 294), but also more traditional industries like steel and metal products and construction materials. The slowest growing losers of structural change were various other engineering branches including ship-building, textile and leather, and fishing: mostly traditional crisis industries. The restructuring process was quite similar to what we observed in the three transition countries. The performance measures were best in a substantially different group of industries, than in the case of structural change. Traditional industries of low and medium technology scored the best, again, many of them working on protected markets, like beverages, construction materials, paper and pulp, basic metals. Productivity was highest in chemical industries and in pharmaceuticals. Automotive industry also performed quite well, thus, at least some of the structural winners were also present among the performance winners as well. The group of losers was more overlapping with structural losers, with textile, leather, wood products, some food branches and repair of ships. As regards correlation of performance and structural change, the Spanish data showed significant positive correlation between productivity and income generating potential and size. Bigger industries tended to be more productive and produce also higher level of added value. The correlation was not so strong with financial efficiency, and profitability. or investments. When growth rates were compared, similarly to size, it was mainly productivity and not profitability that showed positive and significant correlation. 4.5. Ireland The Irish case was rather special due to several reasons. First of all, Ireland joined EU much earlier, in 1973, than Spain or the three transition economies. The era of the 1970s was completely different, with rather high level of protectionism even within the EU, much more different technological conditions (low level of use of electronics) which also affected the patterns of international specialization and trade relations. A further specific circumstance was Ireland’s FDI and export-oriented industrial policy launched in 1958 and the AIFTA1 episode of trade liberalization, both impacted on structural change prior to EC accession. The
1
Anglo Irish Free Trade Agreement between Ireland and the UK, it’s main trading partner.
38
trend evident prior to accession in terms of the increasing importance of these high-tech sectors, and the declining importance of the more traditional, labor intensive sectors appears to have continued unabated in the aftermath of accession. Under these circumstances, liberalization affected the Irish manufacturing industry much earlier, than EU accession, or the more current wave of liberalization with the lead of WTO. Ireland’s becoming an integrated part of international cooperation networks started at the earliest date among the observed economies. Therefore, periods of pre- and post-accession as well as more recent developments were not impacted by gradual or sudden changes in the level of market protection or other changes in the external institutional environment. Instead, Irish manufacturing developed at the influence of a strong industrial policy will equipped with various and mostly effective tools. During the pre- and post accession periods the Irish manufacturing showed therefore a quite high degree of stability in terms of size structure. No fundamental changes occurred in the rank list of the biggest and smallest manufacturing branches. However, rates of growth differed quite substantially, which meant a gradual shift among the branches. In the post accession period highest growth rates were achieved in office and data processing, instruments and also chemicals. Losers in weight were primary metals, motor vehicles, leather products and parts of the textile industry. This list was different in the pre-accession period, thus we may say, that there was some change in growth dynamics across branches. In the preaccession phase quickest growing branches were chemicals and drugs, textile goods, construction materials and industrial machinery. The losers than were canning, tobacco, linen and cotton, hosiery and shoe. As far as performance indicators are concerned, in the past-accession period best performance was achieved by non-metallic mineral products, chemicals, instrument engineering, metal forming, industrial machinery, office and data processing, electrical engineering and footwear-clothing. The only losers in terms of performance was leather, motor vehicles and oil refining. The difference between this list and the lists generated from data of the 1990s in the other countries is obviously different. Oil refining for example may be a loser here due to the ongoing oil crisis of the 1970s. On the other hand, there is some coincidence between a number of branches as concerns positions in the performance ranking and the structural change ranking. During the pre-accession period income generation and efficiency remained virtually unchanged in all sectors, and there was also very little change among the better or 39
poorly performing sectors’ lists. Better performing were beverages, textile fibers, printing and publishing, chemicals and drugs, construction materials. At the bottom we found several branches of food industry including tobacco, leather, motor vehicles. Both lists are slightly different in the pre- and post accession period. The Spearman rank correlations for the two periods showed fairly little significant contacts. None of the correlations were significant at the 1 % level. Sector size and performance showed no significant relationship. However, one relationship deserves special mentioning. The magnitude of changes in the growth of share in output was negatively related to the magnitude of change in income generation. The same negative relationship was also found in Poland, meaning that output and sales growth led also in Ireland to a decline of income generating potential. For the pre-accession period output and value added showed significant correlation with investments. For growth and performance, we find that the change in the share of gross investment and the rank of the change in share, are positively related to investment performance. The correlations also indicated that not only does the size of the change in share in VA and VA-labour costs matter for the size of changes in performance in income generation and efficiency, but the relative size of the change in share matters for the relative size of the change in performance in income generation and efficiency performance. Because of the use of different classifications for the pre and post accession periods, we are cautious about drawing comparisons between the two periods. However, there are a number of sectors which, on the basis of the performance measures calculated here, appear to have performed well in both periods, namely chemicals and drugs, electrical machinery2, brewing, mineral water/soft drinks and non-metallic mineral products. There is also some consistency in the sectors which performed relatively poorly before and after accession, namely meat, dairy, animal feed, motor vehicles and leather.
2
The UNISIC classification of manufacture and assembly of electrical machinery covers instrument engineering, and office and data processing machinery which are classified separately under NACE.
40
5. Attempts at refining the research results: group- and market share analysis Since correlations between the performance/competitiveness indicators and structural change measures were not very robust, an attempt was made at splitting the sample of very heterogeneous manufacturing branches into more homogeneous smaller groups. The creation of such sub-samples followed two major characteristics of the branches: ownership pattern, and trade affiliation (share of exports in turnover). The idea was to separate the impact of local affiliates of large multinational enterprises, for the logic of their functioning is based on global arena meanwhile most local firms’ room of action is either local markets, or deliveries to those multinationals, mainly based in their home country. We expected to see a better fit of our original hypothesis for local firms, than for affiliates of multinational companies. 5.1. Trade intensity groups First simple performance measures were observed for branches with above and below average export and import ratio in their sales and inputs. In the grouping NINT (not internationalized) means branches with below average shares of both exports and imports in total sales and total inputs. INT (internationalized) on the other hand represent branches that are highly integrated in international trade flows on both sides. EIX is the group of branches that export above average, but this export performance is based more on local sources. MIX are the branches that do not export but import above average share of their production inputs. The following to tables summarize the relative weights of these groups in the three economies, as well as their performance differences. Hungary
Poland
Czech Republic
Sales
Employment
Sales
Employment
Sales
Employment
Share Growth
Share Growth
Share Growth
Share Growth
Share Growth
Share Growth
(%)
1998=100 (%)
1998=100 (%)
1995=100 (%)
1995=100 (%)
1998=100 (%)
1998=100
NINT 15,5
121,9
22,0
99,2
43,8
212,7
28,1
83,8
27,3
160,7
31,6
118,9
INT
43,7
182,9
21,3
110,0
26,3
217,9
29,4
68,5
57,7
214,4
49,7
90,3
EIX
11,0
119,8
27,8
95,1
14,7
234,8
24,4
89,9
9,5
117,7
14,6
85,8
MIX
29,8
106,3
28,8
95,1
15,2
208,6
18,1
80,6
5,5
205,9
4,1
211,5
As it is seen, the various groups play different role in the observed economies, their weights in both sales and employment differ considerably. Especially striking is the difference in the
41
case of NINT branches. It is remarkable, however, that figures of employment laid much closer to each other, than sales. In fact, the share of Polish INT branches was higher in employment, than in sales. We do not try to explain the reasons of the differences here. Different weights can be explained by different size of the markets. It is also interesting, that the quicker growth was achieved by export-oriented EIX branches in Poland, meanwhile, quickest growing branches in Hungary were by far INT branches.
Hungary profit
Poland
Value Cum.
Profit-
profit
Czech Republic Value Cum.
Profit-
profit Value Cum.
Profit-
added investment ability
added investment ability
added investment ability
NINT 17,8
22,3
11,5
6,54
40,8
40,0
4,57
31,3
33,8
25,7
5,5
INT
41,4
33,7
34,1
5,39
26,8
30,3
3,63
48,6
46,5
57,9
4,9
EIX
18,6
17,5
8,7
9,64
14,8
14,4
4,78
4,9
9,9
7,6
3,4
MIX
22,2
26,5
45,8
4,23
17,6
15,3
6,14
15,2
9,8
8,8
9,8
When compared through performance measures, we can see once again important differences between countries and groups. The quick growth of INT branches in Hungary is underpinned by a high share in realized profits and also high share in value added, though, both shares are lower, than share in sales turnover. This means a below average rate of profits and value added. Accordingly, profitability is also below average, and this measure is highest in the case of EIX branches, which is interesting, since normally such import dependent and local market oriented branches like tobacco, beverages, various other food industries or petroleum products may earn well due to market regulations and oligopoly market structures. In the case of Poland the higher profitability of import-dependent branches was proved. Shares in investment and value added were very similar to weights in sales revenue. 5.2. Correlations in the groups Hungary We tested 8 measures and created 4 groups of companies. We tried to differentiate branches according to their relationship to major input and output markets: their positions in international labor division. The first group consisted of branches that imported and exported more than 30 % of their production inputs and sales, moreover they had over 30 % foreign
42
ownership in subscribed capital. This group consisted of 43 branches, and was called foreign assemblers. The second group had similar patterns of production inputs and sales, but were not dominated by foreign owners. This group incorporated 8 branches, and we called them subcontractors. The third group sold over 30 % of output on foreign markets, but had a lower level of imported inputs. 20 branches of this group were called export driven branches. Lastly, the fourth group included 28 branches with low levels of exports and imports. They were called domestic oriented manufacturing branches. The 8 measures were: 1. change in the share of sales revenues, 2. change in the share of employment, 3. relative investment effort (cumulated investment/cumulated sales), 4. change in per capita value added, 5. change in per capita cash-flow, 6. change in per capita sales, 7. growth rate in the share in domestic demand, 8. growth rate in domestic sales. Table 23 summarizes the results of the calculations.
Groups and pairs compared Gr.1 Average
Share in sales 2,94
Share Invest/ Val.add/ Cashflow/ Sales/ Growth in sales employ sales employ in DD employ share 4,88 0,071 0,82 -27,83 22,41 51,09
Growth in dom sales -58,31
Std. 69,45 30,48 Average -2,46 14,50 Std. 15,50 19,62
0,029 0,072
35,01 0,47
24,22 0,75
60,21 7,92
194,05 274,58 5,64 3,48
0,032
15,16
167,65
14,91
25,10
30,67
Average 12,49 2,48 Std. 89,92 56,67 F 20,07 2,41
0,083 0,142 1,18
17,29 41,29 5,33
-9,64 31,34 47,92
26,01 28,44 16,30
15,59 70,54 59,74
42,47 145,61 80,12
1-4
F 1,67 3,45 T estim 4,47* T critic 1,667
23,33
1,39 11,08* 1,667
1,67 14,41* 1,667
4,48
7,56
3,55 -2,01** 1,703
3-4
F 33,65 8,34 T estim T critic
19,62
7,41 1,97** 1,729
28,6
3,63 7,89 2,86** 1,729
Gr.3
Gr.4 1-3
22,53
*: Simple T test **: Welsh-type T values The main result of this exercise was the recognition, that there were significant differences between groups of branches in the Hungarian economy. However, these differences were registered only between domestic market oriented group, and the two export-oriented groups (including the assemblers). Since the subcontractor group consisted only 8 branches, we
43
omitted them. Not only differed domestic oriented branches from both other groups, but the tests did not show major differences between the assemblers and exporters. This means, that due to differing market orientation measures of trade-intensive branches are usually different than domestic sales oriented branches. The first conclusion from this result is that much of the obvious disturbances in the previous structural calculations can be traced back to inherently different business structures in trade oriented and in domestic market oriented manufacturing branches. This is, of course, not very much surprising. A little more analysis of the measures also discovers some interesting features of this split of manufacturing branches. The stars in Table 23 show those T values that were significantly lower than the critical T level, which means, that there is no extraordinary distortion in the two samples compared concerning the given measure, hence differences in the average values of the measures are significant and not random. There was one measure where domestic oriented branches differed from both the assemblers and the exporters: in per capita value added. The average was below 1 for the trade dependent branches and was over 17 for the domestic oriented firms. The difference was therefore quite large: branches producing for the domestic market produced much higher level of added value, than exporters and assemblers. per capita sales, cashflow/sales, share in sales and growth in domestic sales were the 4 further measures where domestic oriented branches performed significantly differently than others. And if we take a look at the averages, it is interesting that in all these cases their average values were higher, than those of the assemblers and exporters. It seems, that in the case of a variety of performance measures domestic oriented firms performed better. It is not the task of this paper to go beyond this phenomenon and search for reasons. It may be less competition, it may be differences in domestic and world market prices, or anything else.
Czech Republic Poland 5.3. The BCG matrix: market analysis The idea behind the creation of the BCG matrix was to measure competitive performance of branches on main markets. The method provided an opportunity to differentiate between 44
branches that could or could not expand on domestic markets and on main export market, the European Union. Moreover, we could also make a distinction of market dynamics: we observed changing market shares on growing stagnating and declining markets. When having a list of branches that performed good or poorly on the markets (meaning that the result of their competitiveness in terms of market share changes) we could compare these lists with the lists of performance and structural winners and losers. Hence, we had a further indirect analytical tool of checking the relationship between market performance (competitiveness) and structural change. On the following matrices stars are branches that increased market shares on growing markets, cash cows are branches with increasing market share on declining markets, question marks are branches performing on expanding markets with below average (missed opportunities) or above average (spoiled kids) growth rate of market shares. Dogs are branches that have declining market shares on declining markets. Hungary Table 21. Hungarian Domestic Market BCG Matrix of the Hungarian manufacturing on the domestic market versus structural indicators(average 2000-2001 vs. average 1998-1999) Red: double reds of Table 4. Blue: double blue of Table 4. QUESTION MARKS STARS 160, 203, 211, 251, 262, 266, 267, 275, 294, 300, 363
Missed Opportunities 182, 315, 316, 343, 365 Spoiled Kids 204, 245, 281, 282, 284, 297, 314, 322, 323, 351, 352, 364
CASH COWS
DOGS
Good Cash Cows 157, 158, 159, 172, 173, 192, 241, 243, 265, 268, 271, 283, 286, 312, 313, 321, 332, 335, 362, 366
151, 153, 154, 155, 171, 174, 175, 176, 177, 181, 183, 191, 193, 201, 202, 205, 212, 221, 222, 232, 242, 244, 246, 247, 252, 261, 263, 264, 272, 273, 274, 285, 287, 291, 292, 293, 295, 296, 311, 331, 333, 334, 342, 353, 354, 361
Bad Cash Cows 152, 156, 223, 341, 355
On domestic markets Hungarian branches performed relativel poorly. The many dogs and fairly few stars and question marks mean, that only few branches could increase market shares or operated on expanding markets at all. Branches with declining production operated on
45
declining markets, most of them could not maintain market share and became dogs. In contrast to this, booming industries were observed on expanding markets, rather evenly distributed among stars, spoiled kids and missed opportunities. Thus, there seems to be a strong correlation between changes in market size and growth patterns of an industry. Put it more general the demand pull of performance and structural change seems to have a strong effect in the case of domestic markets in Hungary. Table 22.
European import market BCG Matrix of the Hungarian export to the EU versus structural indicators (average 2000-2001 vs. average 1998-1999) Red: double reds of Table 4. Blue: double blue of Table 4. QUESTION MARKS STARS 212, 231, 265, 283, 286, 292, 297, 312, 313, 322, 332, 334, 341, 343, 351, 352, 361, 362
Missed Opportunities 159, 191, 244, 261, 267, 281, 315, 316, 321, 323, 331, 355 Spoiled Kids 232
CASH COWS Good Cash Cows 152, 154, 155, 157, 171, 175, 222, 241, 245, 252, 274, 291, 296, 311, 314, 353, 363 Bad Cash Cows 153, 156, 172, 176, 177, 183, 192, 203, 204, 233, 251, 268, 273, 293, 294, 295, 300, 335, 354, 366
DOGS 151, 158, 160, 174, 181, 182, 193, 201, 202, 205, 211, 221, 242, 243, 246, 247, 262, 263, 264, 266, 271, 272, 282, 287, 342, 364, 365
The picture is much more controversial in the case of the EU market. In this case growing and contracting branches had positions in the 4 major blocks of the matrix. Most important differences were observed in the case of dogs. Their number decreased and also, most of the contracting in size dogs changed place and became most typically cash cows. This means, their declining sales turnover still helped them to maintain market shares, or put it another way, the extent of their contraction measured by market size was smaller, than that of other competitors. The rather fuzzy position of growing and contracting industries on EU markets means, however, that despite of the importance and growing role of trade with the EU changes in the demand of the EU market does not have a determining influence on growth
46
patterns of the Hungarian manufacturing as a whole. Its impact concentrates on a number of highly internationalized branches (most importantly various branches of the engineering industry). Engineering stars of the EU-matrix were the most important ones among them.
Czech Republic In the case of the Czech Republic a much higher level of consistency of the two types of data was observed. For the same period, as in the case of Hungary (1998-2002) both on domestic and export (EU) markets structural winners and losers took their logical positions. Growing industries of the Czech manufacturing industry worked on growing markets. On domestic markets they usually were missed opportunities, meaning that the growth rate of the market was quicker, than their own growth rate. On export side most of them became stars. This result indicates an export-driven growth pattern of the period in the Czech Republic. The situation with contracting branches was rather similarly logical. Most of them were dogs in both the domestic and the EU-export matrix.
Table 1 BCG Matrix of the Czech manufacturing on the domestic market: position versus structural indicators (1998-2002) QUESTION MARKS Missed Opportunities 243, 282, 152, 267, 313, 203, 156, 314, 331, STARS 201, 242, 334, 205, 222, 353, 281, 247, 223, 321, 322, 268, 332, 285, 284, 174, 365, 157, 159, 212, 251, 252, 261, 266, 316, 343, 323, 204, 335, 300 361 Spoiled Kids 153, 173, 176, 182, 202, 231, 244, 245, 262, 263, 264, 283, 296, 297, 311, 333, 352, 355, 363, 371, 372 CASH COWS DOGS Good Cash Cows 151, 155, 158, 172, 211, 221, 232, 241, 274, 154, 171, 175, 177, 183, 191, 192, 193, 246, 275, 286, 287, 292, 294, 295, 312, 341 265, 272, 293, 315, 342, 351, 354, 362, 364, 366 Bad Cash Cows 160, 271, 273, 291 Source: Own calculation 47
Table 2 BCG Matrix of the Czech manufacturing on the domestic market: change of position versus structural indicators 1998-2002 QUESTION MARKS Missed Opportunities 151, 172, 173, 176, 202, 211, 231, 241, 263, 264, 274, 283, 296, 363 174, 204, 284, 285, 300, 323, 335, 343, 365 Spoiled Kids 153, 155, 158, 182, 221, 244, 245, 262, 295, 297, 312, 333, 352, 355, 371 STARS
CASH COWS DOGS Good Cash Cows 152, 156, 201, 203, 205, 212, 222, 223, 242, 154, 160, 171, 175, 177, 183, 191, 192, 193, 247, 251, 252, 261, 266, 267, 268, 281, 282, 246, 265, 271, 272, 273, 291, 293, 315, 342, 313, 314, 316, 321, 322, 331, 334, 353, 361 351, 354, 362, 364, 366 Bad Cash Cows 157, 159, 243 Source: Own calculation
Table3 BCG Matrix of the Czech export to the EU QUESTION MARKS Missed Opportunities 154, 155, 158, 171, 172, 175, 182, 202, 203, 211, 212, 221, 231, 241, 244, 245, 246, 261, 156, 174, 204, 242, 247, 268, 284, 300, 321, 262, 263, 266, 267, 274, 275, 286, 291, 294, 322, 323, 335, 343, 365 297, 334, 341, 352, 362, 363 Spoiled Kids 159, 173, 243, 272, 287, 292, 295, 311, 312, 371 STARS
CASH COWS DOGS Good Cash Cows 152, 201, 205, 222, 251, 252, 281, 282, 283, 151, 153, 160, 176, 177, 183, 191, 192, 193, 285, 313, 316, 331, 332, 333, 355, 361 264, 265, 271, 273, 293, 296, 315, 342, 351, 354, 364, 366, 372 Bad Cash Cows 157, 223, 232, 314, 353 Source : Own calculation
48
Poland The most important feature of the Polish domestic market share matrix is the strikingly high share of branches among dogs and question marks. This means, that on domestic markets only few branches could increase their market shares between 1995 and 2000. However, as in the case of the Czech Republic and also in Hungary, the strongest growing industries of the Polish economy were placed on expanding domestic markets, meanwhile also logically, the weakest branches were exclusively dogs. Thus, the strong correlation of market dynamics and structural change could be observed in Poland too.
Table 2. BCG Matrix of the Polish manufacturing on the domestic market vs structural indicators Number – strongest groups (see Table A) Number – weakest groups (see Table A) QUESTION MARKS STARS 151, 156, 157, 158, 160, 205, 212, 221, 222, 231, 252, 263, 266, 267, 281, 282, 296, 362
CASH COWS
Missed Opportunities 159, 201, 262, 283, 313, 315, 323, 352 Spoiled Kids 153, 155, 202, 203, 204, 232, 245, 261, 264, 265, 316, 343, 351, 361 DOGS
152, 154, 171, 172, 174, 175, 176, 177, 181, 182, 183, 191, 192, 193, 211, 241, 242, 243, 244, 246, 247, 251, 268, 271, 274, 286, 287, 291, 293, 294, 295, 297, 300, 311, 312, 314, Bad Cash Cows 321, 322, 331, 332, 334, 341, 353, 354, 355, 292 363, 364, 365 Source: Own calculation based on GUS (F-01, SAD) and Table A Good Cash Cows 342, 366
In the case of export markets the situation was largely different, just like in Hungary. The many dogs disappeared and became cash cows and even stars! Thus, contrary to the developments in domestic markets (with general decrease of market shares), there was a general expansion of Polish manufacturers on the markets of the European Union. Again, the reasons of this expansion may be various, and we do not try to go into deeper detail here. But we can draw the conclusion that growth patterns of Polish manufacturing in the second half of the 1990s were determined by demand factors of the markets in the European Union. The
49
very clear fit of growing branches in growing domestic markets and vice versa are less obvious in the case of the EU markets.
Table 5. BCG Matrix of the Polish export to the EU (average 2000-2001 vs. average 1995-1996) vs structural indicators Number – strongest groups (see Table A) Number – weakest groups (see Table A) QUESTION MARKS STARS 174, 176, 212, 232, 261, 292, 297, 311, 312, 322, 323, 331, 332, 334, 341, 342, 343, 352, 353
Missed Opportunities 177, 267, 300, 313, 315, 321, 351, 355 Spoiled Kids 361
CASH COWS Good Cash Cows 157, 171, 175, 211, 221, 222, 247, 251, 252, 268, 282, 283, 286, 291, 294, 295, 314, 316, 354, 366 Bad Cash Cows 151, 152, 153, 154, 181, 183, 191, 202, 203, 205, 231, 233, 262, 264, 281, 287, 293, 335, 364, 365,
DOGS 172, 182, 192, 193, 201, 204, 241, 265, 266, 271, 272, 273, 274, 362, 363
Source: Own calculation based on Commex Note: Strong groups 158 and 245 had to be left out Summing up the experiences of the market share analysis we can draw important conclusions. Firstly, it seems, that changes in demand which are expressed by both dynamics of markets and changing market performance of branches, have a crucial impact on the dynamics of industrial branches. This was most visible on the domestic market, but was not negligible on major export markets either. Secondly, there is a clear separation of branches that are more local market oriented and which depend rather on export markets. This difference is visibly reflected by the different position of branches on the two matrices (domestic market and export market), and especially the different content of the block “dogs”. Most dogs are traditional export-oriented industries, that are working on declining markets and can survive as cash cow of the markets in the EU. The other typical export-dependent group is highly
50
internationalized. They usually have declining market share on domestic markets (question mark), but an increasing share on export markets (stars). Their activity is dominated by large multinational companies. In terms of number they are in minority, but their share in sales turnover is fairly high.
6. Composite competitiveness measures Necessary basic data (home country data by industry (3 digit level) if otherwise not indicated): Total sales of domestic producers: S Total import: IM Total export: EX Total sales of EU: Se Total import of EU: IMe Total export of EU: EXe Home country export to EU: EXte Labour cost (total): LC Labour cost (total) in the EU: LCe Investment: I Amount of wages and salaries (without social contribution): W Employment: L Amount of wages and salaries int he EU (without social contribution): We Employment int he EU: Le Value added: VA Not compulsory: Jobs lost: JL Joibs created: JC
Supply side composite ranking: We decided to compose four indicators: 1.Average growth rate of market share(s) Domestic demand = DD = S-EX+IM Market share on domestic market = MS = (S-EX)/DD EU demand = DDe = Se-EXe+IMe Market share on EU market = MSe = EXte/DDe Yearly change is calculated as (in our example for 1999): YC99 =
MS 99 − MS 98 S 99 − EX 99 MSe99 − MSe98 EXte99 * + * MS 98 S 99 − EX 99 + EXte99 MSe98 S 99 − EX 99 + EXte99
51
If we have our yearly changes, we have to find the X which minimize the next expression: (X-(YC1+1))2+(X2-(YC1+1)*(YC2+1))2+…+(Xn-(YC1+1)*…*(YCn+1))2 where YC1 is the first yearly change we have, and certainly YCn is the last. (Excel can produce this X, using the logaritmic fitting function. The name of the function is maybe different (i do not have the english version of the Excel)). Thus (X-1)*100 will be a kind of average growth rate (in percentage points) of market share of the given industry. Using X, we can have a ranking of our industries. Industry with the highest X will get the rank 1.
2.Labour cost/sales relative to EU Yearly change is calculated as (in our example for 1999): LC 99 LCe99 YC99 =
S 99
LC 98 −
Se99 LC 98 LCe98
LCe98
S 98 Se98
S 98 S 98
If we have our yearly changes, we have to find the X which minimize the next expression: (X-(YC1+1))2+(X2-(YC1+1)*(YC2+1))2+…+(Xn-(YC1+1)*…*(YCn+1))2 where YC1 is the first yearly change we have, and certainly YCn is the last. Thus (X-1)*100 will be a kind of average growth rate (in percentage points) of labour cost/sales relative to EU of the given industry. Using X, we can have a ranking of our industries. Industry with the lowest X will get the rank 1.
3.Relative wage level relative to the EU
W Relative wage level = RWL =
ΣW
L
ΣL where Σ mean sum of all industries value. The same indicator for the EU is RWLe.
YC99 =
RWL98 RWLe98
During our consultation the Hungarian team decided to use this simple indicator instead of that mentioned in the previous version of the description of methodology, because we believe
52
this is more closely related to structural changes. If we have our yearly changes, we have to find the X which minimize the next expression: (X-(YC1+1))2+(X2-(YC1+1)*(YC2+1))2+…+(Xn-(YC1+1)*…*(YCn+1))2 where YC1 is the first yearly change we have, and certainly YCn is the last. Thus (X-1)*100 will be a kind of relative wage level relative to EU of the given industry. Using X, we can have a ranking of our industries. The industry with the highest X will get the rank 1. X-1 is the slope again.
4.Investment/sales (RI) RI = ΣI/ΣS where Σ means sum of yearly (invesment or sales) figures for the whole period. The industry with the highest RI will get the rank 1.
The composite ranking: For our WP3 work, we needed a composite rank indicator. Using our just calculated four ranks we can have a composite rank of supply side performance/competitiveness, which can be calculated as follows: For every industry we have four ranks, which mean four value between 1 and about 100 (it depends on how many industry in the individual country we can count). If we find (separately for every industry)Y which minimize the value of the next expression, using the value of Y for the different industries we can have our final ranking. The industry with the higest Y will get the rank 1. The expression is: (Y-R1)2+(Y-R2)2+(Y-R3)2+(Y-R4)2 where Ri is the ranking created using by the ith indicator we described above.
Demand side composite ranking: Growth rate of demand will be calculated as a composite of domestic demand growth and EU demand growth, where the weigths are similar that of indicator one of the supply side indicator, namely the importance of the given market for the given industry in the given year. YC99 =
DD99 − DD98 S 99 − EX 99 DDe99 − DDe98 EXte99 * + * DD98 S 99 − EX 99 + EXte99 DDe98 S 99 − EX 99 + EXte99
If we have our yearly changes, we have to find the X which minimize the next expression: (X-(YC1+1))2+(X2-(YC1+1)*(YC2+1))2+…+(Xn-(YC1+1)*…*(YCn+1))2 where YC1 is the first yearly change we have, and certainly YCn is the last.
53
Thus (X-1)*100 will be a kind of average growth rate of the demand for the products of the given industry. Using X, we can have a ranking of our industries. The industry with the highest X will get the rank 1.
Structural indicators: We decided to use 2 or 3 structural indicators:
1.Share in employment:
L99 YC99=
ΣL99 L98
−
L98
ΣL98
ΣL98
where ΣL is the total employment in manufacturing. If we have our yearly changes, we have to find the X which minimize the next expression: (X-(YC1+1))2+(X2-(YC1+1)*(YC2+1))2+…+(Xn-(YC1+1)*…*(YCn+1))2 where YC1 is the first yearly change we have, and certainly YCn is the last. Thus (X-1)*100 will be the average growth of weight (in percentage points) of the given industry in the total employment in manufacturing. Using X, we can have a ranking of our industries. The industry with the highest X will get the rank 1.
2.Share in Value Added:
VA99 YC99=
ΣVA99 VA98
−
VA98
ΣVA98
ΣVA98
where ΣVA is the total value added in manufacturing. If we have our yearly changes, we have to find the X which minimize the next expression: (X-(YC1+1))2+(X2-(YC1+1)*(YC2+1))2+…+(Xn-(YC1+1)*…*(YCn+1))2 where YC1 is the first yearly change we have, and certainly YCn is the last. Thus (X-1)*100 will be the average growth (in percentage points) of weight of the given industry in the total value added in manufacturing. Using X, we can have a ranking of our industries. The industry with the highest X will get the rank 1. X-1 is the slope again.
3.Jobs lost – jobs created This is a special indicator, which not necessarily can be calculated in every country. In the hungarian case, as we have firm-level data, we can separate the firms which increased the 54
number of employed persons, and the ones which decreased it. So we can have a sum of jobs lost and jobs created. Certainly it is not a perfect measure, but this is the best we can have. Using this data we can calculate an indicator of fluctuation of employment (FE): FE = (JL+JC)/L
JL + JC L Relative fluctuation of employment = RFE = ΣJL + ΣJC ΣL where Σ again means suming the values of all industry. Regarding this indicator we have to decide what form of it we want to use to generate a ranking. We can calculate the yearly change, or we can have a periodic sum. I propose to use yearly changes, which can be explained as the trend toward stabilization of employment of the given industry. Yearly changes and ranking can be calculated again as: YC99=(RFE99-RFE98)/RFE98 If we have our yearly changes, we have to find the X which minimize the next expression: (X-(YC1+1))2+(X2-(YC1+1)*(YC2+1))2+…+(Xn-(YC1+1)*…*(YCn+1))2 where YC1 is the first yearly change we have, and certainly YCn is the last. Thus (X-1)*100 will be the average pace toward stabilization of its empoyment portfolio of the given industry. Using X, we can have a ranking of our industries. The industry with the highest X will get the rank 1. 6.1. Hungary The previous body of the research tried to map the relationship of performance/ competitiveness and structural change. We used a variety of measures in order to see which indicators described best the relationship, provided it existed and was significant. We found that value-added-based measures responded most sensitively, and showed the strongest relationship. The various measures described different aspects of the relationships. Profitability, income generating power, investments were all measures that tried to characterize capital owners’ logic that makes them investing in one industry rather, than in another. Structural changes in manufacturing were regarded as a result of this logic: the weight of those branches should increase where there is some kind of advantage when compared with other branches, that makes certain industries more attractive for investments than others. Structural change itself was described by various measures to test which best describes the process.
55
Since we did not succeed in getting conclusive results as concerns the “best indicators” of structural change and of performance/competitiveness, we decided to try and combine the best performing few measures and create a composite measure of competitiveness. Competitiveness clearly influenced by both supply side abilities but also by demand side changes. For if changes in demand are such that are in favor of an existing set of capabilities, than it is easier to improve competitiveness. The two sides of demand and supply are of different nature, and also, their indicators are different, thus in fact two composite indicators were created. The supply side indicator included four aspects of competitiveness. Average growth rate of market share describes the flexibility of branches to improve positions on markets. Change in unit labor cost relative to EU average provides information about change in the competitive factor that is regarded as most important for the new member-states: the relative cost of labor. But it is not only the changes of labor cost that matter, but also differences in the relative wage levels of the various branches, compared to patterns in the EU 15. The third part of the composite is therefore relative per capita wage level. The fourth indicator used for the purpose of the composite measure was relative investment efforts. Structural change is linked to changing competitiveness if it is a result of new investments. For production structure may also change without having significant investments, but in this case competitiveness would hardly change. As concerns the demand side composite measure, we created an indicator that describes demand growth on both the domestic and the main export markets. Structural change was described by three indicators. The first was change in the branches’ shares in total manufacturing employment. The second was change in the branches’ shares in total value added. The third indicator was net job loss or creation in the branches. Spearman rank correlation indicators were created using the ranking lists of both composite competitiveness measures and the three structural change indicators. The next table contains the rankings that we achieved for the composite supply side and composite demand side competitiveness measures. It is rather clear when comparing the figures of the same industries in the two columns, that there were significant differences between the two types of rankings. There were only 24 branches where the difference
56
between the two rankings was less than 10 positions, that would perhaps mean a strong fit of the two types of competitiveness measures. Rankings generated using supply and demand composite competitiveness indicators 151 152 153 154 155 156 157 158 159 171 172 174 175 176 177 181 182 183 191 192 193 201 202 203 204 205 211 212 221 222 232 241 242 243 244 245 246 247 251 252 261 262 263 264 265
Supply rank 39 83 65 87 49 55 84 55 46 53 70 24 52 69 21 77 34 63 75 91 61 82 41 57 79 34 70 36 60 73 88 39 30 21 44 36 65 59 75 64 13 80 16 14 80
Demand rank 75 89 67 88 77 76 81 66 46 80 57 90 71 58 69 13 84 52 82 91 73 56 54 18 40 62 8 79 55 63 25 34 72 65 74 37 61 41 31 42 51 15 68 53 43
266 267 268 271 272 273 274 281 282 283 286 287 291 292 293 294 295 296 297 311 312 313 314 315 316 321 322 323 331 332 334 335 341 342 343 351 352 353 354 355 361 362 363 364 365 366
Supply rank 19 86 51 58 9 26 12 7 61 85 50 68 32 27 24 11 20 16 46 74 4 3 78 38 1 90 53 18 48 45 27 41 33 10 6 89 43 27 21 72 31 15 5 7 2 67
Demand rank 21 3 47 59 45 87 28 26 14 22 33 60 32 49 36 10 12 27 24 7 4 6 29 17 39 83 5 86 30 23 16 70 2 35 1 20 9 50 64 85 38 78 11 19 44 48
57
On the other hand, there were large differences, like 83 positions in the case of branch 267 or 69 positions in case of branch 351. These big differences emphasize the fact that the combination of the two types of competitiveness measures is hardly possible. Overall, the following branches showed parallel rankings for the two types of indicators: 152, 153, 154, 157, 159, 191, 192, 221, 222, 241, 245, 246, 266, 268, 271, 287, 291, 294, 295, 312, 313, 321, 343, 361, 363. In their cases branches’ responses on both demand and supply side were similarly strong or weak. Biggest differences (over 50 positions) were shown by the following branches: 174, 181, 182, 211, 232, 262, 263, 267, 273, 283, 311, 323, 351, 362. In these cases branches usually succeeded in one of the two competitiveness factors but not in both. It may also be the case, that the two sides were not equally relevant for these branches, and they responded quite well (or fairly badly) to the significant factor, but showed no reaction to the other. Calculations of the Spearman rank correlation indexes described in a more adequate way this contradicting nature of the two types of the competitiveness composite measures. The results are summarized in the following tables:
Spearmans Sside 1 Sside 1
Sside 2
Sside 3
Sside 4
-0,17041 0,186492
Supply side composite
Demand side
Structural: Labour
Jobs lostStructural: Jobs Value Added Created
-0,1444
-4,97877
-0,74544
-0,08745
0,751957
Sside 2
0,032003 0,090302
-5,54652
-2,48036
-0,36778
-3,37584
1,33473
Sside 3
-1,01213
-5,18285
-0,38736
-3,94229
0,1423
0,818603
-4,76437
-0,20468
-0,30862
0,348509
0,565943
1,563235
2,025934
0,836005
-1,87172
3,764845
2,987986
0,638957
7,124972
0,750137
Sside 4 Supply side composite Demand side compos Structural:Labour
0,274466
Structural:Value Added
-0,20514
Jobs lost-Jobs Created
Spearmans Sside 1 Sside 1 Sside 2 Sside 3 Sside 4 Supply side composite Demand side compos Structural:Labour Structural:Value Added
Supply side Demand Structural composite side :Labour
Sside 2
Sside 3
Sside 4
-0,02
0,02
-0,02
-0,47
0,01 -0,11
0,00
Jobs lostStructural Jobs :Value Added Created
-0,08
-0,01
0,08
0,03
-0,51
-0,25
-0,04
-0,34
0,14
-0,48
-0,04
-0,39
0,02
0,09
-0,45
-0,02
-0,03
0,04
0,06
0,16
0,21
0,09
-0,19
0,37
0,30
0,07
0,60
0,08 -0,02
Jobs lost-Jobs Created
58
Figures of the two tables can be interpreted as follows. From the three structural change indicators changes in the share of employment showed the better, not strong but significant correlation to both the demand side- and the supply side composite competitiveness measures. Changes in the share of value added proved to have significant correlation only with the demand side composite competitiveness measure. The third structural change measure, job creation and loss in the branches did not show significant correlation. Demand side composite competitiveness measure proved to have significant correlation with two of the here used three structural change measures, meanwhile the supply side measure had only one significant correlation. Hence, the demand side indicators performed better.
6.2. Poland Results presented below are based on the method proposed in the Hungarian paper entitled: „WP3 methodology for calculations of new composite competitiveness measures and ranks (supply side and demand side), and for structural indicators”. European data were extracted from Comext and New Cronos with addition of German figures for calculation of labour cost and wage indicators. Home data were compiled from the CSO (GUS) financial and foreign trade statistical reports. To preserve a correspondence between different sources of data we had to confine our analysis to 87 NACE-3 branches. In particular we had to omit following 12 branches: 160 Manufacture of tobacco products 172 Textile weaving 173 Finishing of textiles 223 Reproduction of recorded media 272 Manufacture of tubes 273 Other first processing of iron and steel and of non-ECSC ferroalloys 275 Casting of metals 284 Forging, pressing, stamping and roll forming of metal; etc. 285 Treatment and coating of metals; general mechanical engineering 300 Manufacture of office machinery and computers 333 Manufacture of industrial process control equipment 334 Manufacture of optical instruments,photographic equipement
59
Following seven variables (2 structural indicators and 5 performance measures) were taken into account: 1) Share in value added (VAD), 2) Share in employment (EMP), 3) Demand side composite (DCOM), 4) Average growth of market share (RMS), 5) Labour cost/sales relative to the EU (RLC), 6) Wage level relative to the EU (RWL), 7) Investment to sales ratio (INV). The last four variables were integrated (according to the 2-stage Hungarian method) into a composite supply side competitiveness indicator (SCOM). So, altogether we took into consideration 8 indicators. The analysis covered the 1996-2000 period for the first 6 variables. Only for calculation of investment to sales ratio we used cumulated data from the 1995-2000 period. Table 1 presents results of ranking of 87 branches according to 8 above mentioned measures. Table 1. Rankings of branches by composite competitiveness measures and structural indicators
NACE-3 151 152 153 154 155 156 157 158 159 171 174 175 176 177 181 182 183
VAD 3 23 2 84 34 74 4 45 67 73 31 66 77 75 64 65 81
EMP 30 21 45 76 44 59 14 39 26 82 35 64 72 80 69 49 74
DCOM 65 51 42 84 55 83 44 63 68 82 61 60 69 78 79 76 77
SCOM 46 56,5 15,5 71 40,5 37 53 18,5 38,5 60,5 42 38,5 70 67 86 85 78,5
RMS 10 30 23 55 26 14 18 17 21 71 42 59 75 73 68 48 2
RLC 24 66 10 69 42 55 84 44 75 14 25 15 41 33 83 68 79
RWL 59 34 46 61 38 35 7 32 41 73 62 47 74 53 37 72 82
INV 79 71 52 40 61 60 83 42 28 47 39 44 32 56 72 67 85
60
191 192 193 201 202 203 204 205 211 212 221 222 231 232 241 242 243 244 245 246 247 251 252 261 262 263 264 265 266 267 268 271 274 281 282 283 286 287 291 292 293 294 295 296 297 311 312 313 314 315 316 321
82 22 79 30 11 18 6 9 56 33 7 13 43 40 78 50 53 44 16 62 85 72 17 38 55 39 25 26 8 5 27 69 54 12 32 71 41 36 68 52 83 61 70 87 59 63 42 47 37 35 15 57
86 55 79 33 34 12 3 9 75 11 8 18 15 47 66 78 29 31 19 56 87 41 4 32 50 22 53 67 20 5 24 83 71 6 13 48 52 40 58 46 85 57 54 2 61 63 38 17 37 28 10 70
72 34 80 48 45 14 32 62 53 22 19 2 87 7 59 17 21 15 23 30 81 58 12 40 71 25 13 46 6 5 18 74 43 10 29 75 9 24 47 57 86 27 64 49 54 39 38 35 73 50 36 33
80,5 60,5 83,5 65,5 7,5 33 68 21 17 12,5 27 3 69 43 55 7,5 74 45 18,5 63 21 48 25 10 58 12,5 4 6 2 47 27 34 27 65,5 14 72 30 21 59 54 78,5 75 82 62 30 83,5 23,5 50 49 35 51 56,5
85 84 61 29 35 25 32 16 67 8 7 5 51 46 56 9 72 63 15 80 79 62 11 38 44 13 41 28 12 22 27 58 57 24 50 37 54 45 69 34 66 81 70 6 60 74 43 40 76 52 47 86
74 52 64 59 22 28 70 32 31 38 67 19 87 6 73 12 77 80 63 58 20 46 47 27 53 71 21 50 13 4 37 36 23 51 48 81 26 18 39 49 60 62 76 65 43 61 35 54 16 17 34 78
9 49 55 66 51 79 52 45 25 68 43 26 1 84 57 2 27 23 16 13 8 63 69 54 81 40 21 17 11 78 67 24 20 65 14 31 5 39 48 33 60 30 29 85 22 64 12 70 56 71 80 3
82 20 74 59 1 22 65 45 10 13 29 11 81 33 12 86 57 5 41 58 31 9 16 4 25 3 8 2 18 70 15 37 46 73 17 77 64 36 48 80 62 69 76 51 24 55 49 21 35 19 26 34
61
322 323 331 332 341 342 343 351 352 353 354 355 361 362 363 364 365 366
10 46 29 24 51 21 14 19 76 60 80 48 20 1 86 28 58 49
27 62 23 42 73 25 7 36 77 65 81 68 16 1 84 60 43 51
4 56 41 26 20 28 8 11 52 3 66 16 31 1 85 70 37 67
80,5 15,5 23,5 36 9 5 11 76 52 77 64 32 44 1 87 30 73 40,5
77 82 3 19 65 4 36 53 31 87 49 64 33 1 78 20 83 39
82 1 7 40 9 5 8 30 11 85 2 57 29 3 86 45 72 56
28 10 76 36 19 42 75 77 83 44 86 15 58 4 87 6 50 18
63 38 53 68 23 43 6 84 66 30 75 14 50 7 87 78 27 54
In the next step we calculated Spearman’s rank correlation coefficients. Results are presented in table 2. Correlations between composite competitiveness measures and share in value added are quite high. Structural indicator of share in employment is not so strongly influenced by competitiveness measures. Partial indicator of relative wage level shows no significant correlation with both structural indicators. Table 2. Spearman's rank correlation between competitiveness measures and structural indicators
VAD EMP DCOM SCOM RMS RLC RWL INV
VAD 1,0000
EMP 0,7424 1,0000
DCOM 0,6234 0,5185 1,0000
SCOM 0,4926 0,2926 0,3984 1,0000
RMS 0,5163 0,5820 0,2250 0,5050 1,0000
RLC RWL INV 0,3022 0,0948 0,2012 0,0455 -0,1266 0,2198 0,2357 0,1068 0,3554 0,6212 0,3327 0,5933 0,2289 -0,1117 0,0135 1,0000 -0,1260 0,2448 1,0000 0,0079 1,0000
Because of its additive nature Spearman’s rank correlation coefficient could be a subject of decomposition. Table 3 presents results of such decomposition for coefficients linking two structural indicators and two composite competitiveness measures. Among groups
62
of branches with the highest positive contribution to correlations between structural indicators and competitiveness measures we find following: 36 Manufacture of furniture, manufacturing n.e.c. 17 Manufacture of textiles 26 Manufacture of other non-metallic mineral products 29 Manufacture of machinery and equipment 15 Manufacture of food products and beverages 31 Manufacture of electrical machinery and apparatus On the other hand, among groups of branches with the lowest contribution to these correlations we find following: 23 Manufacture of coke and refined petroleum products 32 Manufacture of radio, television and communication equipment 27 Manufacture of basic metals Table 3. Decomposition of Spearman’s rank correlation coefficients by NACE-2 groups of branches (in %) 15 17 18 19 20 21 22 23 24 25 26 27 28 29 31 32 33 34 35 36
DCOM-EMP 9,17 9,85 5,18 5,53 4,05 3,37 3,77 -7,49 7,05 3,81 12,84 2,91 5,63 9,37 7,40 3,25 3,41 1,70 -3,02 12,22
DCOM-VAD 3,91 7,63 5,00 5,17 1,94 3,50 3,30 -0,73 6,47 3,36 13,16 3,47 7,47 8,32 7,12 4,49 3,47 4,00 1,75 7,19
EMP-SCOM 19,76 15,48 6,78 11,53 -0,82 -2,63 6,03 -1,27 -8,99 6,33 3,84 -5,65 4,19 9,51 12,38 -4,43 7,74 -2,27 7,33 15,15
VAD-SCOM 5,00 9,55 5,35 4,22 1,52 1,08 3,74 3,40 3,61 3,48 11,01 1,05 5,13 12,80 9,91 -3,92 4,34 3,25 3,11 12,38
Total
100,00
100,00
100,00
100,00
63
6.3. Czech Republic
7. Conclusions
64
8. Literature
Antalóczy, K., and M. Sass (1998), ‘A bérmunka szerepe a világgazdaságban és Magyarországon’ (The Role of Subcontracting in the World Economy and Hungary). Közgazdasági Szemle, Vol. XLV, pp. 747–70. Balcerowicz, L.Gray, C.W. and Hoshi, I. (eds.1998): Enterprise Exit Processes in Transition Economies. Central European University Press, Budapest, 1998. Barry F. and Hannan, A., (1996), Education, Industrial Change and Unemployment in Ireland, Working Paper WP96/18, Centre for Economic Research , University College, Dublin. Barry, F., (1999), FDI and Industrial Structure in Ireland, Spain, Portugal and the UK: Some Preliminary Results, Paper presented to the Annual Conference on the European Economy, ESEG, Lisbon, December 1999. Barry, F. Bradley, J. and O’Malley, E., (1999), Indigenous and Foreign Industry: Characteristics and Performance, in F. Barry (Ed.) Understanding Ireland’s Economic Growth, Macmillan Press, London. Borsi,B.- Molnár P.- Papanek G.- Viszt E.(1998): Structural Changes in the Hungarian Foreign Trade and Industry between 1993 and 1997. GKI Economic Research Co., Budapest Brülhart, M., McAleese, D. and O’Donnell, M., (1999), Ireland, in M. Brülhart and R.C. Hine (Eds.),Intra-Industry Trade and Adjustment: The European Experience, Macmillan, London. Buesa, M. and Molero, J. (1998) Economía Industrial de España. Organización, tecnología e internacionalización. Ed. Civitas, Madrid. Croughan, D., (1984), Irish Industry Today, in L.Ryan (Ed.), Irish Industry in the Eighties, Confederation of Irish Industry, Dublin Dobrinsky R. [1995], Economic transformation and the changing patterns of European East-West trade, in: Transforming Economies and European Integration, ed. R. Dobrinsky, M. Landesmann, Edward Elgar, Aldershot. Éltető Andrea (2000): Foreign direct investment in Spain and Hungary: main patterns and effects with a special regard to foreign trade. Ph.D. Dissertation, Budapest University of Economic Sciences Éltető Andrea (2001a): The competitiveness of Hungarian companies. A comparison of domestically owned firms and foreign investment enterprises in manufacturing. IWE Working Paper no.118. August. Éltető Andrea (2001b): A feldolgozóipar külkereskedelmében tükröződő szerkezeti változások nemzetközi összehasonlításban. February, IWE, Budapest. Fernández-Otheo, C.M. (2003): “Inversión directa extranjera y sistema productivo: una disociación creciente”, España 2003. Un balance, nº 95 (extra), marzo. Éltető, A.(2003): Structural changes int he Hungarian industry. Literature survey. WP3 background paper Farkas, P. (1997): The Effect of Foreign Direct Investment on Research, Development and Innovation in Hungary. IWE Working Paper, No. 81. July Farkas, P. (1999): A külföldi működőtőke hatása a K+F-re és az innovációra Magyarországon (Impact of FDI on R and D, and Innovation in Hungary) mimeo Fitzpatrick, J. and Kelly, J.H., (1985), Industry in Ireland: Policies, Performance and Problems, in J.Fitzpatrick and J. Kelly (Eds.), Perspectives on Irish Industry, Irish Management Institute, Dublin.
65
Fonfría, A. y Molero, J. (2000) “Innovation and internationalisation policies in Spain: Special consideration of less developed areas", Resegna Economica. Quaderni di Ricerca, nº 3, págs. 171-203. Abril. Fonfria, A.(2003): Structural changes and competitiveness int he Spanish industry. Review of literature. WP3 background paper Forfás, (1997), Survey of Product and Process Innovation in Irish Industry 1993-1995, Forfás, Dublin. Frischtak C. [1999], Manufacturing, Competitiveness: Concept, Measurement Policies, in: Competition and the World Economy UNIDO, ed. F. Sercovich, Ch.-Y. Ahn, C. Frischtak, M. Mrak, H. Huegge, W. Peres, S. Wangwe, Edward Elgar, Cheltenham. Hummels, D., and Levinsohn, J., (1995), Monopolistic Competition and International Trade: Reconsidering the Evidence, Quarterly Journal of Economics, Vol. 110, pp.799-836. Guerrieri Paolo (1998) Trade Patterns, FDI, and Industrial Restructuring of Central and Eastern Europe, Center for German and European Studies, Working paper 124, 21 pp. Greenaway, P. – Görg, H (2002): Havas, A. (2000): Technológiai Előretekintési Program. mimeo Havlik P. [1998], Export Competitiveness of Central and Eastern Europe, w: The Competitiveness of Transition Economies, OECD, Paris Havlík Peter, (2000) Trade and Cost Competitiveness in the Czech Republic, Hungary, Poland and Slovenia, World Bank Technical Paper, WTP482, 74 pp. Hunya Gábor (2000) International Competitiveness Impacts of FDI in CEECs, Vienna Institute for International Economic Studies, 32 pp. IKIM,1996: A feldolgozóipar növekedési térképe 1992-1995. IKIM, Budapest. Inzelt A.(1998): A külföldi befektetők kutatás-fejlesztési ráfordításainak szerepe az átalakuló gazdaságban. Elemzés statisztikai adatok alapján (The role of R&D expenditures of foreign investors in the transition economy). Külgazdaság vol.XLII no.6. p.59-75. Inzelt A. (2000): The FDI in R&D: Skin-deep and Soul-deep Co-operations. Science and Public Policy Journal of the International Science Policy Foundation. Vol.27, no. 4., August.p.241-251. Jakubcek, J. (2003): Literature review – Work Package 3. Kaminski, Bartolomiej (2000): Industrial Restructuring as Revealed in Hungary's Pattern of Integration into European Union Markets. Europe-Asia Studies, vol.52, no.3, p.457-487. Kaminski Bartlomiej (2001) Trade and production fragmentation: Central European Economies in European Union Networks of Production and Marketing, Research working paper series, WPS 2611, 80 pp. Kadeřábková Anna (2002) Structural and Technology Characteristics of the Czech Industry, Politická ekonomie, no.4, pp.487-504 Kaitila Ville (1999) Trade and Revealed Comparative Advantage: Hungary, the Czech Republic, and the European Union, BOFIT Discussion Papers, 8/1999, 60 pp. Kennedy, K., (1998), Industrial Development, in K. Kennedy (Ed.) From Famine to Feast: Economic and Social Change in Ireland 1847-1997, Institute of Public Administration, Dublin. Landesmann M. B. [2002], Trade Integration and Changing Trade Structures of Transition Economies, in: Completing Transition: The Main Challenges, ed. G. TumpelGugerell, L. Wolfe, P. Mooslechner, Springer, Verlag, Berlin. Marczewski K. (1999) “Foreign trade, industrial growth and structural changes in Poland”. Discussion Papers, no. 75, Foreign Trade Research Institute, Warsaw 66
Marczewski, K. (2003): Determinants of Structural Changes and Trade Specialisation in the Polish Industry. Review of the literature. WP3 background paper. Michałek J.J., Śledziewska-Kołodziejska K. (2000) “Analiza i perspektywy rozwoju handlu międzygałęziowego i wewnątrzgałęziowego oraz zmiany strukturalne handlu Polski z Unią Europejską”. In: J. Kotyński (ed.): Korzyści i koszty członkostwa Polski w Unii Europejskiej, tom 1, Instytut Koniunktur i Cen Handlu Zagranicznego, Warszawa McAleese, D., (1977) A Profile of Grant-Aided Industry in Ireland, Publication Series Paper 5, The Industrial Development Authority, Dublin. McAleese, D. and Foley, A., (1991), The Role of Overseas Industry in Industrial Development, in A. Foley and D. McAleese (Eds.), Overseas Industry in Ireland, Gill and MacMillan, Dublin. Molero, J. (2001) “Industrialisation and internationalisation in the Spanish economy”, One Europe or several. ESRC Working Paper, 28/01. Myro, R. and Gandoy, R. (1999) “Sector industrial”, Lecciones de Economía Española, 4ª O’Connor, T., (2001), Foreign Direct Investment and Indigenous Industry in Ireland: Review of Evidence, Working Paper No. 5, School of Slavonic and East European Studies University College London. O’Donnell, M., (1997), The Changing Commodity Structure and Destination of Irish Exports 1972-90, Technical Paper 97/5, Trinity Economic Paper Series, Trinity College, Dublin. O’Donnell, M.(2003): Literature review of structural change and trade specialization in Irish industry. WP 3 background paper O’Sullivan, M., (2000), Industrial Development in Ireland: A New Beginning? in J.W. O’Hagan (Ed.) The Economy of Ireland: Policy and Performance of a European Region, Gill and Macmillan, Dublin. Pellegrin, J. (2000), ‘German Production Networks in Central/Eastern Europe: Competitive Breakthroughs and Old Ghosts’. In: Lorentzen, J. (ed.), Globalization in Emerging Markets, Basingstoke: Macmillan. Raynauld A., Vidal J. P. [1998], Labour Standards and International Competitiveness. A Comparative Analysis of Developing and Industrialized Countries, Edward Elgar, Cheltenham. Ruane, F. and Görg, H., (1996), Aspects of Foreign Direct Investment in Irish Manufacturing since 1973: Policy and Performance, in Journal of the Statistical and Social Inquiry Society of Ireland, Col. XXVII, pp.37-87. Ruane, F., and McGibney, A., (1991), The Performance of Overseas Industry 1973-89 in A. Foley and D. McAleese (Eds.), Overseas Industry in Ireland, Gill and Macmillan, Dublin. Sereghyova, J. – Vesely, L. (1998): Progress Linking the Enterprise Sphere of Centraleuropean Countries in Transition into Westeuropean Corporate Networks. mimeo Sheehy J. [1995], Economic interpenetration between the European Union and Central and Eastern Europe, in: Transforming Economies and European Integration, ed. R. Dobrinsky, M. Landesmann, Edward Elgar, Aldershot. Sorenson, N.K., Dalum, K., Madsen, E.S. and Nielsen, J., (1991), Intra-Industry Trade in Denmark and Ireland: A Comparison, Memo1991-19, Institute of Economics, University of Aarhus. Soós Károly Attila (2000): Strukturális fejlődés a feldolgozóiparban a piacgazdasági átmenet idején. Külgazdaság, vol. XLIV, július-augusztus, p.15-51.
67
Szalavetz A. (1999): Technológia transzfer, innováció és modernizáció német ajdonban levő feldolgozóipari cégek példáján (Technology transfer, innovation and modernization on the example of manufacturing companies in German ownership). OMFB TAN 98-56-05, Budapest. Szalavetz, A. (2003): A hagyományos és a gyorsan növekedő ágazatok támogatására alkalmazott iparstratégiai eszközök. Mimeo. Szanyi, M.(2003): Science, Technology and Public Policy in Hungary: Challenges of the Emerging Knowledge-based Economy. mimeo Szanyi, M.(2002): Subcontracting and outward processing trade as a form of corporate networking in Hungary. Acta Oeconomica., vol.52. 2002. No. 3. Pp. 347-369 Szanyi, M. (2001): Privatization and greenfield FDI in the economic restructuring of Hungary. Transnational Corporations, vol. 10, no. 3. December 2001. pp 25-37. Szanyi, M.(1999a): The role of FDI in restructuring and modernization: an overview of literature in G. Hunya (ed.): Integration through Foreign Direct Investment - Making Central Europe Competitive (Cheltenham: Edward Elgar, 1999), pp. 50-80. Szanyi, M. (1999b): Adaptive Steps by Hungary's Industries During the Transition Crisis. In: J. Temesi - E. Zalai (eds.): Back to a Market Economy. Akadémiai Kiadó, Budapest, 1999. Szanyi, M. (1998): Ipari beruházások az átalakuló országokban. Egy empirikus felmérés eredményei. Közgazdasági Szemle XLV. évf. 1998. szeptember 851-868.o. Szanyi, M. (1997): Versenyképességi verseny. Figyelő, 1997 november 17. Tomšík Vladimír, Kubíček Jan, Srholec Martin (2002) Structural Adaptability of Manufacturing in Central European Countries during the Transition, Newton Holding, a.s., Macroeconomic Analyses and Research Department Tomšík Vladimír, Kubíček Jan, Srholec Martin (2002) Structural Changes in Central Europe during the Transition, Newton Holding, a.s., Macroeconomic Analyses and Research Department Walsh, P.P. and Whelan, C., (1999), The Importance of Structural Change in Industry for Growth, in Journal of the Statistical and Social Inquiry Society of Ireland, Vol. XXIX, pp.1-32. Wrynn, J., (1997), Foreign Direct Investment to a Peripheral Country: The Case of Ireland, in B. Fynes and S. Ennis (Eds.), Competing from the Periphery: Core Issues in International Business, Oak Tree Press, Dublin. Wziątek-Kubiak A. [2001a], Międzynarodowa specjalizacja a konkurencyjność Ekonomista, no 4. Wziątek-Kubiak A. [2003] Zmiany konkurenycjności polskiego przemysłu w latach 1994-2000, Bellona, INE-PAN, Warszawa. Wziątek-Kubiak A. [2003b]: Critical synthesis, review of the main findings, methodologies and current thought on competitiveness of accession countries. Mapping of competence. WP1 Synthesis paper Zeman Karel (2000) Competitiveness in Central and Eastern Europe, Working paper No. 19, Institute of Integration of the Czech Republic into European and World Economy, 46 pp.
68
69
