Factors explaining the changes in the rankings of Polish cities’ economic position (1992-2013)

With regard to economic theories on the impact of integration on regional development, good summaries are given by Niebuhr and Stiller (2004). According to them, the theoretical view that border regions may profit from integration due to the now closer distance to economic centers, highlights only one aspect. The authors show that such positive developments are scarce, and the process of integration remains open with regard to its results for particular regions. Bröcker (2009) analyzes the consequences of economic integration only for cities with the highest levels of urban hierarchies. The study on Leipzig and Wroclaw by Kunze and Lenk (2007) presents a broad approach for explaining differences in the long-term development of cities, but it was not aimed to present a quantitative analysis.

Cuaresma et al. (2014) tested the impact of 48 variables on the economic growth of 255 NUTS2 regions in the European Union between 1995 and 2005, using econometric models. The results of their research indicate the importance of education (10% increase in the share of workers with higher education is associated with a 0.4% higher growth rate of annual GDP per capita) and location of the capital city within the region. The latter factor is particularly important in CEE countries (regions with capital cities grew on average 1.8 percentage points faster than other regions).

Noseleit (2013), analyzing data for German regions (NUTS3) between 1975 and 2002, came to the conclusion that changes in the structure of the region’s economy are an important element in moving from entrepreneurial activity to real economic growth in the region. In regions with a low sectoral diversification of the economy, with the dominance of industry, the development of local entrepreneurship is not associated with economic growth as much as in regions with a more diversified economic structure.

Cullinan et al. (2013) investigated the determinants of relative economic performance for Irish towns (using a data set of 112 Irish towns with over 1000 inhabitants). Their results suggest that these determinants of success can act in different ways for different groups of towns. The dependent variable is a retail turnover per head of population. The group of independent variables consists of connectivity, capacity and competitiveness of the town. Authors construct models for different groups and also a model for all cities. In all cases, education has a positive association with relative retail performance.

There is extensive Polish literature on the organization of EU funds and its impact on regional economic development; however, these works are focused on regional rather than on urban issues (see, e.g. Ministerstwo Rozwoju Regionalnego 2010). In addition, the existing Polish literature usually concentrates on the absorption of funds, and not on a wider European integration context. The results of the latest work on the role of nationally managed programmes (EU funds) in financing local government projects show that the largest beneficiaries are especially large cities (Swianiewicz et al. 2013). The development of Polish cities is well documented with regard to demographic issues and economic key indicators (see, e.g. Węławowicz et al. 2010), state control functions (Śleszyński 2007), spatial distribution of large enterprises (Śleszyński 2008) and location decisions made by foreign investors in Polish municipalities (Dziemianowicz 2008) or issues of polycentricity (Gorzelak & Olechnicka 2009). One observation is that in general, the economic development of cities drops as one moves from west to east of Poland. Łukomska (2012) used regression models for all Polish municipalities to investigate factors of local economic development between 1995 and 2008. The results show that factors which may be directly influenced by local governments (like expenditures related to the improvement of local infrastructure, often with the contribution of EU funds) play a certain role in local development processes, although they do not dominate. However, the issue of European integration is hardly touched upon in these publications.

The main objective of the research is the identification of the changes in the economic position of large Polish cities in 1992-2013 and an attempt to answer the question: which cities had the chance to improve their economic position, and which cities lost population and economic potential since 1992? The second, but no less important objective of the research is the identification of factors which were the driving forces of these changes. In particular, what was the impact of factors related to European integration?

With reference to the research objectives hypotheses were formulated. Firstly, it was assumed that the changes in the economic position of Polish cities which took place during the last two decades would allow for the identification of the group of “success cities” and “loser cities”. The identification of the key drivers of changes in the “success cities” may be slightly more difficult than for “loser cities”. Factors that negatively influence the economic position of cities can be somewhat more group-oriented and it could be easier to identify them.

In relation to the second objective it is assumed that particular factors could have been crucial for explaining changes in the economic positions of cities:

The impact of the process of European integration.

Access to EU funds (including pre-accession funds) and economic integration, as well as the exchange of academic staff have contributed to the change in the level of the economic development of Polish cities. In the context of integration, the role of social capital (participation in the European Parliament elections) should not be insignificant to the economic development of cities.

Results of studies on European countries indicate the role of integration processes, particularly trade integration in the development of industry in the EU (Krugman & Venables 1996; Tsiapa 2014).

On the basis of the obtained values of HI, the five rankings of the economic position of cities for each of five points in time were created. The next step was the comparison of all five hierarchies, checking which groups of cities are ascending and which are descending the hierarchy.

Which of the analyzed cities made the biggest “leap” in the rankings in 1992-2013? The not very big cities (with less than 80 thousand inhabitants, except Koszalin) within the group of cities experienced the most significant positive changes. They are not concentrated in a given region, but are quite evenly distributed throughout the country. At first glance, it is hard to find distinctive features of this group. This is not the case for the group of cities that lost their economic position dramatically between 1992 and 2013. These are mono-industrial cities located in the Upper Silesia, usually with the dominance of the mining industry in the structure of their economy (see Table 1).

Table 1

Greatest gain and greatest loss of cities in the ranking 1992-2013

How did the rankings change over the years? The analysis has covered the changes in all possible combinations for the above five points over time: 1992-1998, 1992-2004, 1992-2008, 1992-2013, 1998-2004, 1998-2008, 1998-2013, 2004-2008, 2004-2013 and 2008-2013. It was assessed whether there is a relationship between the changes in these periods. Changes in almost all configurations are correlated with the 1992-2013 change. The only exception is the 2004-2008 change (Table 2). Changes in this period do not show a statistically significant relationship with the changes in the whole period of 1992-2013, and this may suggest that the economic changes in cities immediately after Poland’s accession to the EU were slightly different from the overall trend of changes throughout the analyzed period.

Table 2

Pearson correlation coefficients among changes of position in rankings

If we compare the cities in the 1992 ranking with the 2013 ranking, we can identify four groups of cities (see Figure 1). This diagram is a sort of summary of the first part of the empirical study.

Figure 1

The first group of cities (in the lower left corner of the diagram) – throughout the entire period this group kept its leading place in the ranking of economic development. These are the largest Polish cities. Their position in the hierarchy is stable.

The second group (cities in the upper left corner of the diagram) – their loss was the most dramatic in the rankings. This group comprises almost exclusively mining cities located in Upper Silesia.

The third group (cities in the bottom right corner of the diagram) – they have improved their positions in the rankings between 1992 and 2013. It is difficult to identify distinctive features of these cities. Some of them are very characteristic, e.g. with one large and very prosperous state company located there: Głogów (producer of copper and refined silver – KGHM Polska Miedź S.A.), Puławy (one of the largest melamine producers in the world and the biggest Polish company in the field of chemical synthesis – “Azoty Group Puławy” S.A.).

The fourth group are cities that have maintained a poor ranking for years.

The cities from the second and third group which were described above are the subject of quantitative analyses, as the dependent variables of the models tested are changes in the ranking (in the case of the first and fourth group of cities there were no changes, or the changes were insignificant).

Factors that can determine changes in the hierarchy of cities have been described and theoretically embedded in the section with the explanation of the research hypothesis. The next step was the operationalization of these factors. Finding the available data, which would be a good measure of the individual factors, was not an easy task, so in the case of selected factors some assumptions were needed. At this stage, in addition to data from the Main Statistical Office and data from the Ministry of Finance, databases specially prepared for this study were used, i.a. location in the context of communication accessibility, student and teacher mobility within the ERASMUS program, the quality of universities and the location of the biggest 500 companies on the basis of information available in “Wprost” and “Polityka” magazines. The data on the availability of EU funds for local governments in Poland in 1991-2006 prepared by K. Szmigiel (2007) and data on exports based on Komornicki’s work (2009) were also used.

The list of variables operationalizing the factors explaining changes in the economic development of cities in the groups is presented in Table 3.

Table 3

Operationalisation of factors explaining changes in economic hierarchies of Polish large cities

The results of the regression analysis with respect to the three explanatory variables (changes of the city’s ranking in 1992-2013, 1992-2004 and 2004-2013) are presented in Table 4.

Table 4

Results of OLS regression analyses (progressive selection method)

The explanatory power of the models tested is not very high (from about 20% in the case of explaining changes in 2004-2013 to about 60% in the model explaining the change in 1992-2004). Many of the independent variables were not statistically significant for explaining the changes in the economic position of cities

They were excluded from the explanatory models during the statistical procedure.

However, among the variables that are not irrelevant to the shifts in the economic hierarchy, the factors related to European integration can be found. These are the variables related to the availability of communication in relation to the core cities of the European Union:

-

location in relation to the airports in Hamburg and Munich – the closer to these cities, the more positive the change in the hierarchy of economic development;

The results show that the closer to vienna, the more negative the change in the economic development of the city. This result of regression analyses could be related to a group of cities located in Upper Silesia that can be characterized by negative economic changes and quite a short distance from the airports nearest to the airport in vienna (Katowice or Kraków).

Among variables that are statistically significant to explaining the changes in the rankings were also the variables related to European integration, such as the turnout in the accession referendum (the higher the turnout in the city, the more positive the change in the ranking of 1992-2004

However, the percentage of votes „in favor” as a percentage of potential votes in the 2003 referendum turned out to be a variable negatively correlated with economic changes in the period 1992-2004. This would indicate that the cities that experienced positive changes in the rankings in the 1990s and in the beginning of the 21st century were characterized by a relatively smaller group of Euro-Med participants in the accession referendum.

The relevance of “European integration” factors is confirmed by a hierarchical multiple regression analysis, but only for the model with the dependent variable related to the change of 1992-2013. Only for this period, the adding of a “European integration” block of variables to the model increased its explanatory power in a statistically significant way (at 5% level) (Table 5).

Table 5

Changes in the explanatory power of regression models after adding the variables related to EU integration (hierarchical multiple regression)

The variables related to human and social capital (% of population with higher education and electoral turnout in 1994 and 1998 local elections

Although the results regarding turnout are not clear, the turnout in the local elections of 1990 is negatively correlated with changes in the city’s economic hierarchy. But this election was exceptional in many respects, e.g. high percentage of invalid votes (Gendźwiłł 2014).

The loss of the administrative status of the regional capital turned out to be a variable explaining changes in the economic hierarchy of cities in 1992-2004. This relationship is positive, which means that cities holding the status of a regional capital in the period 1992-1998 have experienced positive changes in their economic positions between 1994 and 2004 (see Table 5).