Measuring the knowledge base in Hungary: Triple Helix dynamics in a transition economy
Balázs Lengyel * and Loet Leydesdorff **
* Centre for Regional Studies Budapest Department, Hungarian Academy of Sciences, Teréz krt. 13, H- 1067 Budapest, Hungary; e-mail: lengyel@rkkmta.hu
** Amsterdam School of Communications Research (ASCoR), University of Amsterdam, Kloveniersburgwal 48, NL-1012 CX Amsterdam, The Netherlands; e-mail: loet@leydesdorff.net; http://www.leydesdorff.net .
How can the knowledge base of a transition economy be measured? Building on previous studies in the Netherlands and Germany, we combine the perspective of regional economics on the interrelationships among geography, technology, and organization with the triple-helix model of university-industry-government relations, and use the mutual information in three dimensions as an indicator of the configurations. Our data consists of firms categorized in terms of sub-regions (proxy for geography), industrial sector (proxy for technology), and firm size (proxy for organization). The results indicate that the knowledge base of Hungary is strongly differentiated in terms of regions. Budapest and its agglomeration are central to the country on every indicator. In the north-western part of the country, foreign-owned companies and FDI disturb an etastistic triple helix dynamics which is still dominant in the eastern part of the country. However, the national level seems no longer to add to the synergy among the regional innovation systems. Further analysis of the knowledge-intensive services and its high-tech components reveals that the transition from the planned economy to integration in the European common market is not yet completed.
Keywords: knowledge base, triple helix dynamics, mutual information, entropy statistics, transition economy.
1. Introduction
One of the main ideas behind the concept of innovation systems is that innovation takes place both within firms and across the interfaces among institutional agents like universities, industries, and government agencies. Innovation systems differ in terms of how the fluxes through the networks are integrated and whether the heterogeneous fluxes (economic exchange relations, novelty production, and organizational control) provide opportunities for synergy. The synergy between the industrial structure, geographical distributions, and academic traditions can be considered crucial for the strength of an innovation system (Fritsch, 2004).
The networks provide only the knowledge infrastructure, while the knowledge base of an innovation system is shaped by a division of innovative labour at the national and/or regional level given such an infrastructure. The Triple Helix of university-industry-government relations has hitherto been developed mainly as a neo-institutional model for studying the network arrangements among these agents (Etzkowitz et al., 2000; Powell and DiMaggio, 1991). A neo-evolutionary model should capture the relations among the different functions (organized knowledge production, diffusion, and control) that operate in and on these networks. The functions have to be carried by the agents at the nodes, but one can no longer expect a one-to-one correspondence between functions and institutions because the functions are also based on the arrangements among the institutions (Etzkowitz and Leydesdorff, 2000). More important than the mere presence of agencies is the quality of their relations in a given configuration. Since the functions are performed by different agents and relations, one expects an uncertainty which can be measured as probabilistic entropy.
Systemic effects may occur that cannot be directly traced back to specific exchanges, but emerge more indirectly (Burt, 1995). The different functions can be considered as sub-dynamics of the system. The sub-dynamics can be expected to interact to varying degrees. The mutual information flowing between these sub-dynamics will be used as an indicator of the synergy at the systems level.
The geographical distribution of industrial activities among regions is only one of the relevant dimensions of a configuration. Due to differences in the character of innovation processes, one can expect that geographical conditions have different effects on the various economic sectors such as manufacturing and knowledge-intensive services. We shall further distinguish between medium-tech and high-tech using the classifications of the OECD. The division of labour among corporations of various sizes (e.g., the number of SMEs in a region) can be considered as a third determining factor (Storper, 1997).
In summary, we use these three distributions of firms (geography, technology, and firm size) and analyze their mutual information at various levels of the Hungarian system (counties, regions, sub-regions) in order to test two hypotheses that were generated in previous studies about the German and Dutch economies, respectively (Leydesdorff et al., 2006; Leydesdorff and Fritsch, 2006):
1. medium-tech manufacturing can be considered as the drivers of the knowledge base of an economy more than high-tech;
2. knowledge-intensive services (KIS) tend to uncouple the knowledge base of an economy from its geographical location.
One can expect that mutual information among the three dimensions shows a less developed knowledge base in Hungary, as a transition country. We premise that the heritage of the etatistic model of Triple Helix configurations in terms of university-industry-government relations during the communist regime still has a significant effect on Triple Helix dynamics in transition economies (Etzkowitz and Leydesdorff, 2000; Inzelt, 2004). There are many other explicit signs of transition processes: the institutional background of local and regional innovation systems is just shaping up; the public decision-making is hardly decentralised, the role of central government is determining in economic development; the reform of universities and the healthcare system started just few years ago, etc.
The countries in Eastern Europe entered their transition period and faced the challenges of globalisation during the same period of time (Enyedi, 1995). Accession to the European Union became increasingly a crucial objective for facing both transition and globalization. Thus, one has to consider two processes when analyzing the knowledge base in Hungary: the mechanisms that prevailed in the “existing socialism” still remained dominant in the state-controlled services; the change in the relations was much faster in sectors where foreign-owned firms became heavily involved. Consequently, there are huge differences in terms of economic prosperity, share in foreign direct investment, and R&D spending among Hungarian regions: Budapest emerges as the center of the country in every sense (Barta, 2002; Varga, 2007), the rate of business R&D is higher in the Western parts while the big universities in the East are among the largest public R&D bodies (Grosz and Rechnitzer, 2005). Thus, two further hypotheses emerge:
3. foreign-owned firms have a restructuring effect on the synergy among the three dimensions (industrial organisation, technology, geographical spread);
4. the Hungarian regions are at different stages of the transition in terms of university-industry-government relations.
This paper introduces a way of assessing the quality of regional innovation systems by measuring the interaction and synergy between subsystems by means of an indicator based on entropy statistics (Jakulin and Bratko, 2004; Leydesdorff, 2003). The approach is applied to the 168 sub-regions in Hungary using micro-data of 660,290 firms as units of analysis. The following section presents the conceptual basis of the study. Section 3 outlines the spatial framework of the empirical analysis, and section 4 presents the data and methods. Following the general assessment of the quality of regional innovation systems (section 5); we compare results for different sub-sectors of the economy, in particular high- and medium-tech manufacturing and knowledge-intensive services (section 6). Conclusions and policy implications are specified in the final section (section 7).
2. From neo-institutional towards neo-evolutionary analysis
Because innovation processes involve the generation and application of knowledge the quality of innovation systems is dependent on how the knowledge base is related to the network among the interacting agents (Foray, 2004; Leydesdorff, 2006). The networks facilitate and constrain exchanges of knowledge and resources. For a number of reasons such as the costs and efforts involved in having face-to-face contact, a considerable part of these exchange relations is constrained geographically (Marshall, 1920); knowledge is accumulated in national and regional systems (Lundvall, 1992, Nelson, 1993). However, the distribution of the technologies in a system, the industrial organization, and the geographical spread can be considered as relatively independent sources of variation (Storper, 1997). One can expect that these three sources of heterogeneity are reflected in the division of innovative labour (Florida, 2002).
It is important to note that the organization of the division of innovative labour does not necessarily require direct relationship, but can also be ‘systemic’ in nature, steered by other market forces or externalities like knowledge spillovers (Anselin et al., 1997). For this reason, an analysis of the direct relations among actors in regional innovation systems such as market relations and R&D cooperation may not provide a sufficient basis for assessing the working of the system. The geographical dimension first positions the agents involved; secondly, economic exchange relations can be expected among the agents at the nodes; and thirdly, the dynamics of knowledge-based innovations can be expected to upset the tendency towards equilibrium prevailing in economic exchange relations (Schumpeter, [1939] 1964; Nelson and Winter, 1982).
Using the Triple Helix model, the knowledge base of an economy is considered as a trilateral interaction effect which operates on the bilateral interaction terms between each two of the subdynamics. However, this synergy at the level of the system has posed a problem for measurement (Carter, 1996; Hronszky, 2005). Is it possible to operationalize an indicator of the emerging and therefore ‘elusive’ order of a knowledge-based economy (Skolnikoff, 1993) and to measure this order as a reduction of the uncertainty which prevails at the systems level (David and Foray, 2002)? Is one able to distinguish innovation systems at different levels of geographical aggregation and in terms of technological and economic sectors?
In short, the knowledge infrastructure of institutional relations (e.g., among universities, industries, and governments) can be considered as a necessary but not a sufficient condition for developing a knowledge-based economy. The intensity and the quality of the interactions are decisive for the characteristics of such a system. Whether or not, and to what extent, a knowledge-based economy has emerged from a specific configuration of relations remains an empirical question (Storper, 1997).
Nations and regions can be expected to differ in combining the functional requirements of a knowledge-based economy (Lundvall, 1992; Nelson, 1993; Cooke, 2002). When a knowledge base is resulting from the synergy at the systems level, one can expect the system increasingly to ‘self-organize’ an additional feedback loop. This feedback may operate positively (that is, by reducing uncertainty in the relations) or negatively because, for example, it reinforces globalization in a previously more localized system. Etzkowitz and Leydesdorff (2000) called this additional feedback the operation of ‘a network overlay’ potentially emerging within a Triple Helix. In other words, the network of relations may turn into a configuration that can be productive, innovative, and flourishing, but not all networks can be expected to do so all the time.
For example, despite their productivity, innovativeness, and the density of relations networked in different dimensions (Biggiero, 1998), Italian industrial districts may suffer from deindustrialization because of the globalizing dynamics in the appropriation of the profits and the advantages of innovation (Beccatini et al., 2003). The neo-institutional perspective of social network analysis can provide us with a view of the (potentially changing) relations in the districts, but not on the dynamics. From this perspective, the emergence of a knowledge-based overlay to the system remains an unpredictable effect.
The neo-evolutionary model analyzes the Triple Helix dynamics in terms of how these relations operate: how much uncertainty is generated and/or reduced, at which level, and in which dimensions? Relations may function as lock-ins or be beneficial to the dynamics of the system. One can expect an additional reduction of the uncertainty in the configuration if the overlay feeds back on the generation of uncertainty in the institutional relations. This decrease of uncertainty results from the configuration of relations and can no longer be attributed to the individual agents at the nodes or to specific relations.
3. The transition features of Hungary: territorial structure, foreign-owned companies, research and development
Since the political changes in 1989 and 1990 Hungary has a small and relatively open economy. It joined the EU in 2004 after a transition period. The transition from planned into market economy first caused a significant economic fallback (Varga, 2007): many medium-tech state-owned companies went bankrupted, R&D expenditure declined.
Three regions (Central Hungary, Central Transdanubia, Western Transdanubia) actually began to reduce the gap between them and their Western-European counterparts with a growth of approximately 4-5% a year in the late ’90s. These three regions with dynamically expanding economies constitute one block situated in the northwest of Hungary between Budapest and the Austrian border. The Central Hungarian region (Budapest) almost reaches the level of the EU-25 average (96.5 %). The economic growth of the other four regions remained at a yearly 1.6-3% growth of GDP, which is more or less around the EU average or falling slightly below that (Kállay and Lengyel, 2007). These regions (Southern Transdanubia, Northern Hungary, Northern Great Plain, Southern Great Plain) are situated south and east of Budapest (Figure 1).
Figure 1: Regions (NUTS 2) and counties (NUTS 3) in Hungary
Source: http://en.wikipedia.org/wiki/Regions_of_Hungary
The employment situation in Hungary has improved parallel to the economic output; regional differences within the country are similar for the two indicators. In the three developed regions employment reached 60% in 2004, while this figure was 50% in the four less developed ones (Kállay and Lengyel, 2007). In 2003 the three developed regions produced 70% of Hungarian manufacturing exports, while the contribution of the Southern Great Plain region was only 6%. The basic figures of regional competitiveness show that the growth and competitiveness of Hungary’s economy depend on three regions and Budapest as a metropolitan area (Lengyel, 2004).
Weight of foreign-owned companies
In the Hungarian economy more than 50% of the registered capital of companies and partnerships is in the hands of foreign owners (Kállay and Lengyel, 2007). International stake is significant in all SME types, reaching at least 27-28 per cent. Similarly to large companies, half of the capital of medium-sized enterprises is in foreign hands[1].
The motives of foreign-owned companies for selecting specific locations (mainly in the energy-, banking and manufacturing sectors) were primarily the labour cost, accessibility, opportunities of privatisation in the beginning of the transition. However, the knowledge base of Budapest and the regional centres seems to be more and more attractive for multinational R&D (Inzelt, 2003); the growth of R&D spending by foreign affiliates was among the highests in Hungary in the period 1995-2003 (UNCTAD, 2005, p. 127). The share of large foreign-owned companies in manufacturing R&D is around 40 per cent (Eurostat, 2005), the share of foreign affiliates of business R&D is around 80% (EC, 2005).
Table 2: Distribution of foreign stake in foreign-owned companies, Hungary=100 (%)
Region, county |
2000 |
2001 |
2002 |
2003 |
2004 |
2005 |
Central Hungary |
68.3 |
65.1 |
64.4 |
63.0 |
65.6 |
67.2 |
of which Budapest |
58.8 |
54.0 |
52.9 |
47.3 |
50.2 |
53.4 |
|
Central Transdanubia |
7.1 |
8.3 |
8.4 |
10.0 |
10.1 |
10.2 |
|
Western Transdanubia |
10.8 |
12.4 |
11.9 |
11.9 |
11.6 |
11.1 |
|
Southern Transdanubia |
2.0 |
1.9 |
2.2 |
1.9 |
1.6 |
1.5 |
Northern Hungary |
4.7 |
4.0 |
4.7 |
5.7 |
4.0 |
4.1 |
|
Northern Great Plain |
3.8 |
4.1 |
5.4 |
5.1 |
4.8 |
3.9 |
Southern Great Plain |
3.3 |
3.2 |
3.0 |
2.4 |
2.3 |
1.9 |
Source: Hungarian Central Statistical Office (HCSO)
The territorial distribution of foreign investment in the Hungarian economy is uneven (Table 2). The majority of foreign stake in foreign-owned companies are in Budapest and Central Hungary, Central Transdanubia and Western Transdanubia.
Foreign-owned firms play a determining role in the Hungarian value production; they influence strongly the organisational structure of the economy. One can expect a significant effect of their unequal geographical distribution and high share of high- and medium-tech industry on the territorial differences among the three dimensions of our model. However, it is less clear to which extent this restructuring couples on the knowledge base of the Hungarian economy.
Knowledge intensity: R&D regional features
The R&D index in Hungary lags behind the European average: the total share of R&D expenditure of GDP was 0.89% in 2004; only one third of the R&D expenditures are spent by the business sector, the rest is provided by the state. Though, some regional centres have important universities which have taken an active role in the transition, the Central-Hungarian Region (CHR)—including Budapest—plays a determining role in the Hungarian R&D performance (Table 3).
Table 3: Number of R&D facilities and R&D employees in the Hungarian Regions
|
Region |
R&D places |
R&D employees |
||||
|
1996 |
2001 |
2004 |
1996 |
2001 |
2004 |
|
|
Central Hungary |
710 |
1,199 |
1,255 |
12,831 |
16,924 |
17,535 |
|
Middle Transdanubia |
64 |
158 |
158 |
732 |
1,513 |
1,712 |
|
Western Transdanubia |
109 |
150 |
194 |
830 |
1,411 |
1,500 |
|
Southern Transdanubia |
125 |
195 |
227 |
1,417 |
1,973 |
2,405 |
|
Northern Hungary |
101 |
118 |
145 |
1,160 |
1,326 |
1,571 |
|
Northern Great Plain |
162 |
250 |
280 |
2,213 |
2,489 |
2,873 |
|
Southern Great Plain |
190 |
267 |
282 |
2,126 |
2,715 |
2,824 |
|
Hungary |
1,461 |
2,337 |
2,541 |
20,859 |
28,351 |
30,420 |
Source: Hungarian Central Statistical Office (HCSO)
Several surveys concluded that university-industry relations are weak in Hungary (Inzelt, 2004; Papanek, 2000). As the majority of research and development in the respective centres of regions (for example in Northern Great Plain and Southern Great Plain) is publicly financed, one can expect that these Triple Helix configurations suffer from another distortion of the system. Local R&D demand hardly exists and the growing interest in international collaboration among researchers at the universities may have indirect effects on local and regional developments because publicly financed R&D has remained one of the main resources of these regions (Etzkowitz et al., 2000; Sun et al., 2007). In other words, the different sectors (R&D, state apparatuses, and industries) can be expected to enter the international and European arenas with different speeds.
In summary, the Hungarian economy has grown during the last fifteen years into three types of economy: Budapest as the metropole, the western part of the country which is integrated in structures of partners in the European Union, and the eastern part which has retained structural features of the old system. Paradoxically, when Hungary changed from a planned economy into a liberal one, it was in many respects too late for shaping a national system of innovation. The transition has been towards gradual accession to the common market of the European Union.
4. Data and methods
The dataset consists of 660,290 establishments and was collected by the Hungarian Central Statistical Office (HCSO). As it is obligatory for firms to supply data for the HCSO, our dataset can be considered as the entire population. The data applies to December 31, 2005. The use of this statistical register of enterprises provided us with the information at the company level: each company is classified into geographical, technological and organisational categories using the following systems:
The geographical dimension was investigated at the NUTS-4 level of sub-regions. Hungary as a whole is considered as a NUTS-1 unit according to the Eurostat classification.[2] There are seven regions (NUTS-2), 19 counties (NUTS-3) and 168 subregions (NUTS-4) in Hungary (see Figure 1 above). Since the data was collected at the NUTS-4 level, we are able to aggregate the information and define the uncertainty in geographical distributions at the NUTS-3 level of counties.
Table 4: Territorial units in Hungary
|
Level of territorial units |
Number of territorial units |
|
NUTS 2 = region |
7 |
|
NUTS 3 = county |
19 + Budapest (capital) |
|
NUTS 4 = subregion |
167 + Budapest (capital) |
Budapest is the only metropolitan district in Hungary and has hence to be considered as a special category in regional surveys. In this paper we have data from the Budapest enterprises as a NUTS-4 district. The Central Hungarian Region is specific to the effect that the NUTS-2 and NUTS-3 levels (including the Pest County) cover the same territory. However, we pursue the analysis at the NUTS-3 level.
In order to measure the technological dimension we use the NACE code[3] of industrial sectors developed by the OECD and Eurostat as in the Dutch and German studies (Table 5). Since various sectors of the economy can be expected to use different technologies, sector classifications can be used as a proxy for the technology (Pavitt, 1984). The OECD (2001, pp.137ff.) indicated the various sectors in terms of their knowledge intensity. Each enterprise is classified by its first activity at the two-digit level.
Table 5: Classification of high-tech and knowledge-intensive sectors
|
High-tech Manufacturing 30 Manufacturing of office machinery and computers 32 Manufacturing of radio, television and communication equipment and apparatus 33 Manufacturing of medical precision and optical instruments, watches and clocks Medium-high-tech Manufacturing 24 Manufacture of chemicals and chemical products 29 Manufacture of machinery and equipment n.e.c. 31 Manufacture of electrical machinery and apparatus n.e.c. 34 Manufacture of motor vehicles, trailers and semi-trailers 35 Manufacturing of other transport equipment |
Knowledge-intensive Sectors (KIS) 61 Water transport 62 Air transport 64 Post and telecommunications 65 Financial intermediation, except insurance and pension funding 66 Insurance and pension funding, except compulsory social security 67 Activities auxiliary to financial intermediation 70 Real estate activities 71 Renting of machinery and equipment without operator and of personal and household goods 72 Computer and related activities 73 Research and development 74 Other business activities 80 Education 85 Health and social work 92 Recreational, cultural and sporting activities
Of these sectors, 64, 72 and 73 are considered high-tech services. |
Source: Laafia, 2002: 7.
The organisational dimension can be operationalized by the size of enterprises and measured by the number of their employees (Pugh et al., 1969; Blau-Schoenherr, 1971). Like the Dutch data, the Hungarian enterprise register has a category with zero or unknown employees that includes the SMEs without employee or self-employing. This category contains, among others, spin-off companies that are already on the market but whose owners are employed by mother companies or universities (Table 6).
Table 6: Distribution of company data by size
|
Number of employees |
Number of firms included in this study
|
Number of registered firms – 31st Dec. 2005 |
|
0 or unknown |
275,202 |
365,861 |
|
1-9 |
369,869 |
805,209 |
|
10-19 |
5,976 |
20,870 |
|
20-49 |
4,921 |
11,046 |
|
50-249 |
3,733 |
4,860 |
|
250 or more |
589 |
944 |
|
Total |
660,290 |
1,228,999 |
Source: Hungarian Central Statistical Office (HCSO)
The main difference in the data mining among the three studies is that in this study, the data contain only the high- and medium-tech sectors and the knowledge-intensive services while in the Netherlands and Germany all firms were included in the databases. Surprisingly, a high percentage of firms (53.7%) is classified by these categories. However, we are not able to compare some of our results directly with those of the other two countries.
Methods
In the Hungarian analyses we follow precisely the same methodology as the previous two studies using the mutual information in three dimensions.
According to Shannon’s (1948) formula, the uncertainty in the distribution of a variable x (∑x px) can be measured as Hx = − ∑x px log2 px. Analogously, Hxy represents the uncertainty in the two-dimensional probability distribution of x and y (that is, Hxy = − ∑x ∑y pxy log2 pxy). In the case of two dimensions, the uncertainty in the two interacting dimensions (x and y) is reduced with the mutual information or transmission Txy. This can be formalized as follows:
Txy = (Hx + Hy) – Hxy (1)
In the limiting case that the distributions x and y are completely independent, Txy = 0 and Hxy = Hx + Hy. In all other cases Txy > 0, and therefore Hxy < Hx + Hy (Theil, 1972, at pp. 59f.). Because of the sigma in the formulae, all information terms can be fully decomposed; if base two is used for the logarithm, the values are expressed in bits of information. Note that these measures are formal (probability) measures and thus independent of size or any other reference to the empirical systems under study. In general, two interacting systems (or variables) determine each other in their mutual information (Txy) and condition each other in the remaining uncertainty on either side (Hx|y and Hy|x, respectively).
Abramson (1963, at p. 129) derived from the Shannon formulas that the mutual information in three dimensions is:
Txyz = Hx + Hy + Hz – Hxy – Hxz – Hyz + Hxyz (2)
While the bilateral relations between the variables reduce the uncertainty, the trilateral term in turn feeds back on this reduction, and therefore adds another term to the uncertainty. The layers alternate in terms of the sign. The configuration thus determines the net result in terms of the value of Txyz. The potential reduction of the uncertainty at the systems level cannot be attributed to individual nodes or links. McGill (1954) proposed to use the term ‘configurational information’ for this mutual information in three (or more) dimensions, but with the opposite sign (Jakulin and Bratko, 2004).
The mutual information can be generalized for more than three dimensions, but for reasons of parsimony we limit the discussion here to three dimensions.[4] As noted, the three dimensions under study in this case will be geography, technology, and organization, and the measure will accordingly be indicated as the TGTO. Similarly to formula 2 one can formulate as follows:
Tgto = Hg + Ht + Ho – Hgt – Hgo – Hto + Hgto (3)
The value of TGTO measures the interrelatedness of the three sources and the fit of the relations between and among them. Because it is a measure of the reduction of the uncertainty, a better fit will be indicated with a more negative value. This overall reduction of the uncertainty can be considered as a result of the intensity and the productivity of innovative labour division in a broad sense.
Assuming that a division of labor can yield efficiency gains, one would expect that regions with a distinctive profile would be more productive than regions with a lower level of division of labour. However, the indicator does not measure the innovative activity or economic output of a system. It measures only the structural conditions in the system for innovative activities, and thus specifies an expectation. Regions with a high potential for innovative activity can be expected to organize more innovative resources than regions with lower values of the indicator.
5. Results
As noted, the data allow us to disaggregate in terms of geographical regions (NUTS-2 and NUTS-3), and we are able to distinguish high-tech and medium-tech sectors versus knowledge-intensive services. The various dimensions can also be combined in order to compute the mutual information in the interaction among them in a next step.
Table 8 shows the probabilistic entropy values in the three dimensions (G = geography, T = technology, and O = organization) for Hungary as a whole and the decomposition at the NUTS-3 level of the counties. The counties are different in terms of the numbers of firms and their geographical distributions inside the counties. While according to NACE categories only 8,722 enterprises proved to be high-, medium-tech or knowledge-intensive in Nógrád county, Budapest contains 229,165 firms, and the Pest county 67,342 enterprises. The mean of analysed firms at the county level (without Budapest) is 22,690.
Table 8: Expected information contents (in bits) of the distributions in three dimensions and their combinations
|
Counties |
Hgeography |
Htechnology |
Horganisation |
Hgt |
Hgo |
H to |
H gto |
N |
Nr of subregions |
|
Hungary |
5.189 |
2.722 |
1.159 |
7.875 |
6.334 |
3.712 |
8.823 |
659,701 |
168 |
|
Budapest |
0.000 |
2.598 |
1.169 |
2.598 |
1.169 |
3.644 |
3.616 |
229,165 |
1 |
|
Baranya |
1.717 |
2.790 |
1.139 |
4.483 |
2.853 |
3.742 |