Effective Cohesion Policy? Long-Term Economic and Social Convergence in Poland

Abstract

The aim of this article is to determine the relationship between the processes of economic and social convergence in Polish voivodeships. The correlation between these two processes is still unclear and the mechanisms behind them are not fully discovered. The goal of this article requires determining whether the changes in the level of disproportions in the GDP per capita translate into changes in social cohesion and what is the nature of this relationship. Those issues are still relevant and important due to their cognitive and practical value. In view of the above, using various analysis tools (i.e., the correlation coefficient and regression function), the authors determine whether economic convergence (both β and σ) and social convergence (with the use of the TOPSIS method) is observed in Polish regions in the years 2000–2022. The obtained results allow us to state that in the analyzed period, economic and social divergence processes occurred in Poland. The constantly deepening disproportions seem to be relatively permanent and caused by differences in the demographic potential (including urbanization potential) of individual regions that are difficult to eliminate, as well as by different development trajectories during the period of political transformation after 1989.

1. Introduction

The process of economic growth, its determinants, and its consequences have always been the key issues discussed in economic sciences. The work deemed to be the first modern economic book—“An Inquiry Into the Nature and Causes of the Wealth of Nations” by A. Smith—was in fact devoted to economic growth and the factors that influence the differentiation of its level in time and space. Ever since, a number of publications have been published on this issue [1,2,3,4], and the conclusions they present are still ambiguous despite the increasingly developed conceptual and methodological framework.

Some authors, mainly those associated with the classical and neoclassical approach to economics, emphasize that there are mechanisms that naturally push the economy to a state of equilibrium, also in spatial terms [5]. In other words, economies at different levels of aggregation (from the local to the macro level) are in pursuit of the same level of economic growth. Another branch of economic sciences emphasizes that differences in the endogenous potential (equipment and quality of production factors, capability for innovation and creating knowledge, production structure, etc.) cause economies to strive for different states of equilibrium, which will lead to the polarization of the growth level [6]. This debate, at the theoretical and empirical level, remains unresolved because there are theses and analytical results that support both these standpoints.

Another important thread of economic analyses related to the above is the issue of social cohesion. Economics as a social science examines the impact of various factors on the well-being of people [7], pointing to the level of economic development as its key determinant [8,9]. Here, too, the questions remain open as to the factors determining this well-being, to its changes over time between different economies, and to the verification of whether economic convergence (divergence) translates into social convergence (divergence).

In this context, social cohesion is based primarily on ensuring the long-term well-being of the population. Research on social cohesion is currently the result of two initially independent trends: research on the level and quality of life [10]. In the original assumptions, the standard of living should be operationalized using reliable quantitative indicators, which in turn, owing to their reliability, could underlie unbiased comparisons [11]. In turn, the study of the quality of life was supposed to be based on the use of qualitative elements, initially identified with sciences such as sociology [12]. Only the reflections of researchers from the end of the 20th century allowed the research apparatus to be focused on one trend—the well-being of the population, which combines both of these earlier concepts [13,14]. After all, Nobel Prizes in Economics were indeed awarded for research on well-being [15].

The issue of socio-economic development has also found recognition in strategic goals of international importance. The UN Sustainable Development Goals catalog includes those directly related described tasks (and indicators that enable their verification), i.e., to economic growth (SDG 8) and the well-being of the population (SDG 3). It is also present in the constituting documents of the economic associations of countries, the best example of which is the European Union. The European Commission implements cohesion policy, which assumes, in accordance with the definition indicated in the Single European Act of 1986 (and further emphasized in the Lisbon Treaty), reducing disproportions between regions and counteracting the backwardness of disadvantaged regions. This policy is being implemented by means of a tool called the Cohesion Fund, one of the five basic European Structural and Investment Funds. Poland is one of its beneficiaries.

Poland (the eighth country in terms of area and population in Europe) is an interesting example of a country following the path of economic transformation after decades of functioning under a communist regime [16]. Due to its difficult heritage, it is an interesting subject for socio-economic analyses [17]. The transformations made in the country in 1989 have triggered adjustment problems for its economy, but above all, for the population faced with the new reality [18]. The transition from a centrally planned economy and the primacy of state ownership over entrepreneurship and private property was not so convenient for every social group. Here, it is especially worth mentioning industry and agriculture [19]. The gap between Poland and other countries of Central and Eastern Europe and the mature economies of developed countries naturally placed Poland in the role of a long-term beneficiary of the above-mentioned Cohesion Fund [20]. After 35 years since the economic transformation and 20 years since joining the European Union, it seems advisable to empirically verify the effectiveness of the implementation of the Cohesion Fund, which should translate directly into economic and social convergence [21].

Taking into account the importance of the above issues, the aim of the presented analyses was to determine the relationship between the processes of economic and social convergence in Polish voivodeships as well as to answer the following questions: do changes in the level of disproportions in GDP per capita translate into changes in social cohesion, and what is the nature of this relationship?

This work is part of the research trend of empirically verifying cohesion policy [22,23,24,25,26,27,28,29,30,31,32,33,34,35]. Its advantage is, undoubtedly, a relatively long-time horizon and consideration given to both economic and social convergence, which make it distinguishable from the works of other authors. The conclusions drawn from it will also be important for increasing the understanding of the nature and determinants of economic and social processes and for a better selection of goals and tools of regional policy.

2. Economic and Social Convergence—Literature Review

The concept of convergence comes from the Latin word “convergere”, which originally meant moving in the same direction. Its meaning is similar in the case of economic sciences. Most often, it refers to a process whose essence is to equalize the level of economic growth between countries over time. However, as Gergics [36] notes, this term is often used in economics not so much to refer to economic systems that become similar to each other in certain aspects (usually in the level of GDP per capita) but rather to situations in which some economies are catching up or even surpassing others in terms of wealth levels. While the difference is subtle, the distinction is important.

A more complete understanding of the essence of economic convergence is possible by identifying and discussing its types. The literature usually distinguishes at least two types of convergence. The first is beta-type (β) convergence. It occurs when economies at a lower level of economic growth record higher growth dynamics than initially wealthier economies [37]. β-type convergence is further divided into absolute and conditional convergence [38].

The absolute convergence occurs when the growth rate of retarded economies is higher than that of the richer ones, assuming that the growth of the first is not limited by any other factors. Importantly, the growth dynamics of poorer areas are higher the lower the initial level of growth. It is also assumed that all economies strive for the same level of economic growth over time [39].

In turn, conditional convergence is observed when the level of economic growth is equalizing between countries or regions with similar structural characteristics (similar production structure, level of education of the population, and determinants of competitiveness). In other words, it is when economies with different structural parameters strive for different levels of income in the long term [40]. Comparing both types of β-type convergence, Liu et al. [41] note that absolute convergence can occur spontaneously, whereas conditional convergence can be modulated and catalyzed by the government when it identifies the factors determining it [41].

Sigma-type convergence (σ) occurs when differences in the level of economic growth between the analyzed economies decrease over time. The coefficient of variation or standard deviation of GDP per capita or population income per capita is most often used to measure this type of convergence [42]. The prior occurrence of β-type convergence is a necessary but insufficient prerequisite for the occurrence of σ-type convergence [43].

Also, club convergence is mentioned among the analyses of economic growth processes and convergence between economies. It refers to a situation when economic growth levels are approaching within groups (clubs) of studied areas, while at the same time, economic divergence is observed between clubs, and thus, an increase in development disproportions [39].

The theoretical foundations of the neoclassical approach to economic convergence are based on the Solow–Swan model. In this model, the long-term growth rate of GDP per capita is determined by the exogenous rate of technological progress. At the same time, the factor that makes it possible to achieve a steady state is the decreasing marginal productivity of capital. The steady state means a situation in which the value of the investments undertaken is equal to the investments needed to maintain constant relations of capital to technology and population. If economies are characterized by the same level of technology and the same rates of investment and population growth, their economic growth paths will converge over time. However, poorer economies, in which the capital stock is scarcer, will be characterized by a higher growth rate due to the higher marginal productivity of capital until the steady state is reached. The rate of economic growth is higher the smaller the capital stock per unit of labor. In other words, when certain initial conditions are met, the assumptions of this model include unconditional beta convergence [44].

This model also implies that in the case of economies with different initial states, the rates of economic growth in the process of reaching the steady state will also differ. Different investment rates and population growth rates mean that economies will tend to different equilibrium states, which is consistent with the conditional beta convergence hypothesis [ibid.].

The continuous exploration of the essence of economic growth processes, with particular emphasis put on the role of knowledge, innovation, and human capital within endogenous theories, has slightly modified the views on the universality and necessary nature of convergence processes. Hence, in addition to many theories pointing to the ability of economies to achieve economic convergence, there are ample theoretical concepts emphasizing objective economic mechanisms that push them in the opposite direction, i.e., toward increasing polarization of economic development. We can mention here the theory of polarization by F. Perroux, the theory of hierarchical diffusion by M. Lasuen, or the latest theories of endogenous development and the New Economic Geography [45]. Among these theories, the different endogenous potential of individual economies is indicated as the major cause of the emergence and aggravation of income inequalities. Its higher level not only allows for faster and more effective absorption of funds in the form of foreign direct investment (FDI) and research and development (R&D) expenditure but also, thanks to the emerging benefits of agglomeration, enables achieving increasing effects of the production scale. As a result, already better-developed economies are able to achieve higher growth rates in the future [36,45,46,47].

It follows from the above that there is no full agreement in theoretical considerations as to the possibility of achieving regional economic convergence and the conditions that could initiate this process. Likewise, there is no clear evidence at the level of empirical research that would indicate the inevitability of the processes of successive equalization of economic growth levels between selected economies.

For example, when it comes to the European Union, most analyses indicate the occurrence of slow convergence at the level of NUTS 2 regions over the last dozen or so years [36,47,48,49,50,51,52,53,54,55,56,57]. At the same time, the results of the cited and other studies show that this process is no longer so common in the case of greater disaggregation of data (in terms of time or territory). For example, López-Villuendas and del Campo [50] point out that this process is very poorly visible or that even its reverse trend can be observed, i.e., growing disproportions in the level of development, in the NUTS 3 subregions [36,52,58].

Divergence processes were also observed at the level of NUTS 2 regions, although only in selected periods [36,53,56]. The factors that could be responsible for the increase in development disparities include the following: geographical location of the studied regions [46,51,58,59], the degree of their urbanization and the resulting differences in the benefits of agglomeration, differences in institutional factors, different levels of support from European Union funds [57], and finally, disproportions in the course of demographic and social processes [54].

Also, analyses on economic convergence conducted for Polish regions do not always provide clear conclusions. Most of them [60,61,62,63] point to the ongoing processes of economic divergence of Polish voivodeships. There are also studies that indicate a slow equalization of development differences [64]. At the same time, it should be emphasized that practically all of these analyses point to the relatively dynamic divergence at the subregion level, where areas with above-average growth dynamics include large cities and metropolitan areas.

Poland, as a country with a socialist stage in its history, has been particularly struggling with regional divergence. Therefore, the expectation of the effects of accession to the European Union in 2004 also concerned socio-economic development. The European cohesion policy implemented in this country was systematically verified by measures of the level and quality of life. Theoretical aspects of achieving cohesion can be found in many Polish studies in this field. Ryszkiewicz [44] concluded, for example, that in the case of Poland, the aim of cohesion policy was not only to eliminate differences between regions and social groups but rather to provide assets for less developed regions and areas to encourage them to overcome disproportions. Korenik [65] also argued about the importance of these activities in conditions of irregularity. The first research works verifying this policy appeared a few years after accession to the EU. Already in 2008, Wójcik [40] pointed to the difficulties in achieving the intended effects, where a lack of unconditional convergence was observed, and only fragmentary club convergence occurred. In turn, Kusideł [66] indicated that the processes taking place from the perspective of the European Union, which actually improved social cohesion, were not able to prevent internal divergence in Poland. Dudek and Wrzochalska [28] and Busłowska and Marcinkiewicz [27] also confirmed this relationship. Walesiak and Dehnel [67] showed that the level of social cohesion was higher in regions that received more EU funds to support regional development. The positive relationship between EU funds and wages in Poland was also confirmed by Cieślik and Rokicki [68].

The regions of Eastern Poland were characterized by lower values of economic indicators, which resulted in lower endogenous potential and a greater number of socio-economic challenges. Therefore, actions within the framework of cohesion policy addressed to these regions were of strategic importance to them. The effectiveness of this policy was therefore also verified in these areas. Miś and Zając [32] and Busłowska and Marcinkiewicz [27] indicated that despite the improvement in living conditions and social situation noted by the residents, convergence and social cohesion in these regions occurred very slowly or did not occur at all. Moreover, Misiak [69] proved that the club convergence occurring in Poland did not affect the “catching up” of the eastern regions of Poland with relatively higher social development indicators.

More detailed analyses of convergence and cohesion were also ambiguous. For example, Jagódka and Snarska [26] asked about the justification for continuing the cohesion policy since the expected cohesion did not occur in relation to human capital and innovation. Cieślik and Misiak [70], in turn, showed that regional income inequalities can be reduced by promoting related differentiation in the eastern regions of the country. Dziembała [29] showed that high values of GDP per capita do not necessarily translate into equally high positions of regions in the rankings concerning human development. The previously cited Misiak [69] argued, however, that the research results were dependent on the adopted methodology, emphasizing ambiguity and the need for continuous validation.

In this context, an important cognitive and practical issue is the relationship between the processes of economic and social convergence. A common view found in the literature on the subject states that prior economic convergence is the prerequisite for reducing social polarization. However, as the conclusions of the World Development Report 2009 indicate, these two processes may occur independently of each other, i.e., social convergence may be accompanied by economic divergence. The need to verify this thesis has prompted analyses presented in this work.

3. Materials and Methods

The aim of the present research was to determine whether economic convergence processes took place in Polish regions in the years 2000–2022 and whether social convergence processes could also be observed against their background. The assumed goal of this research determined the sequence and methodology of this research.

First, the occurrence of β-type and σ-type economic convergence was examined. In the first case, analyses were based on the equation proposed by Sala and Martin [36]:

$${\displaystyle \frac{1}{T}}\ln \left({\displaystyle \frac{y\left(T\right)}{y\left(0\right)}}\right)={\alpha }_0+{\alpha }_1\mathrm{l}\mathrm{n}{y}_0$$

(1)

The expression y(T) represents the value of GDP per capita in the final year, y(0) is the value of GDP per capita in the initial year, and T stands for the number of years.

Knowing the specific form of the above function, based on the ${\alpha }_1$ value, one can also calculate the value of the β coefficient, which indicates the percentage of the distance from the established equilibrium state covered by the examined economy.

The occurrence of σ-type convergence was determined using its classic measures, i.e., the range (the difference between the maximum and minimum value of the studied variable in the group) and the coefficient of variation.

GDP per capita was adopted as a measure of the level of economic growth. Among the available measures of economic growth, it is by far the most accessible and transparent one, with a well-developed and harmonized calculation methodology [51]. All these features also make it relatively easy to interpret and universal for making spatial comparisons. At the same time, it should be borne in mind that it does not take into account external effects and processes of economic and social polarization within the studied units [51]. Moreover, it takes no account of the spatial dimension of interregional economic connections, which often result in interconnected levels of economic activity in neighboring regions [71].

Establishing diagnostic variables for social convergence was more difficult. Ultimately, a synthetic measure of social development was adopted for each voivodeship. To this end, twenty-one diagnostic variables were selected covering socio-economic aspects that significantly determine the long-term well-being of the population (

Table 1. Variables included in the social convergence analysis.

No.	Variable Name	The Nature of the Variable
1	Own income of the voivodeship/person	Stimulant
2	Number of entities registered in REGON per 10 thousand inhabitants	Stimulant
3	Disposable income in the household per capita	Stimulant
4	Gross salary	Stimulant
5	Population of non-working age per 100 inhabitants of working age	Destimulant
6	Average life expectancy for men (years)	Stimulant
7	Deaths due to cancer and cardiovascular system diseases per 10 thousand inhabitants	Destimulant
8	Beds in general hospitals per 10 thousand inhabitants	Stimulant
9	Doctors per 10 thousand inhabitants	Stimulant
10	Registered unemployment rate (%)	Destimulant
11	Percentage of households functioning below the statutory poverty line	Destimulant
12	Injured in fatal accidents per 100 thousand working inhabitants	Destimulant
13	Apartments per 1000 inhabitants	Stimulant
14	Average usable area of 1 apartment (m2)	Stimulant
15	Percentage of dwellings equipped with water supply in rural areas (%)	Stimulant
16	Percentage of population using sewage treatment plants (%)	Stimulant
17	Percentage of devastated and degraded land requiring recultivation in the total area (%)	Destimulant
18	Percentage of recycled waste in the amount of waste generated per year (%)	Stimulant
19	Children in kindergartens and other forms of pre-school education per 1000 children aged 3–5	Stimulant
20	Pass rate for secondary school leaving examinations (%)	Stimulant
21	University students per 10 thousand inhabitants	Stimulant

Source: Own study.

), which, as already mentioned, is an indicator of social cohesion. The diagnostic variables included in the study of social convergence are responsible for key areas of the functioning of the economy and households. Their nature and impact (positive or negative) significantly affect the well-being of the population and social cohesion. Therefore, variables were selected (based also on the completeness of data in this period) from the labor market, security, education, public finances, and household budgets, as well as demography and environmental sustainability indicators. It is noteworthy at this point that, although supported by the literature studies and an overview of similar research in this field, the selection of variables is of an original nature, and that different research outcomes could be achieved if other variables had been considered.

A synthetic social convergence indicator was calculated using TOPSIS [72], a method commonly used in economics research [73,74,75,76,77,78]. The method weighs decision alternatives by measuring their distance from two reference points—the PIS (positive ideal solution) and the NIS (negative ideal solution). The system identifies the best decision as the one closest to the PIS and furthest from the NIS. The procedure is performed in several steps. First, a decision matrix is built as follows:

X = [x_ij] and weight vector w = [w₁, …, w_n]

(2)

In the above equation, w₁ + …+ w_n = 1.

The next step is to build a normalized decision matrix N = [z_ij]_mxn, where z_ij is the value of the normalized decision alternative assessment, according to the following formula:

$$z_{ij}={\displaystyle \frac{x_{ij}}{\sqrt{{\sum }_{i=1}^m\left({x_{ij}}^2\right)}}}$$

(3)

In the formula, i = 1, …, m, j = 1, …, n.

The next step is to build a normalized weighted decision matrix:

W = [v_ij]_mxn,

(4)

The value of v_ij is calculated as w_j z_ij.

Then, the positive ideal solution A⁺ and negative ideal solution A⁻ are determined as follows:

$$A^{+}=\left[v_1^{+},\dots ,v_n^{+} \right] v_j^{+}=\left\{\begin{array}{c}\max v_{ij}, v_{ij}\in Z\\ min{ v}_{ij}, v_{ij }\in S\end{array}\right.$$

(5)

$$A^{-}=\left[v_1^{-}, \dots , v_n^{-}\right] v_1^{-}=\left\{\begin{array}{c}\min v_{ij}, v_{ij}\in Z\\ \max v_{ij}, v_{ij}\in S\end{array}\right.$$

(6)

The next step is to calculate the distance (d_i⁺) of the ith decision alternative from A⁺ and the distance (d_i⁻) of the ith decision alternative from A according to the following formula:

$$d_i^{+}=\sqrt{{\left(v_{ij}-v_j^{+}\right)}^2}, d_i^{-}=\sqrt{{\left(v_{ij}-v_j^{-}\right)}^2}$$

(7)

i = 1, …, m.

This allows the value of the synthetic assessment measure (the global score) to be calculated for the ith decision alternative using the formula below:

$$T_i={\displaystyle \frac{d_i^{-}}{d_i^{+}+d_i^{-}}}$$

(8)

i = 1, …, m. The value of T_i ∈ [0,1].

Finally, the decision alternatives are ranked in a decreasing order of the synthetic assessment score. The higher the score, the higher the alternative was ranked.

The data came from the Local Data Bank of the online statistical database of Statistics Poland (GUS) (https://bdl.stat.gov.pl/bdl/start, accesed on 14 August 2024), and the length of the research period depended on the availability of comparable statistical data.

4. Results and Discussion

Table 2. Value and dynamics of changes in GDP per capita of voivodeships in Poland in 2000–2022 (current prices).

Voivodeship	GDP per Capita Value (in PLN)												Change Dynamics
	2000	2002	2004	2006	2008	2010	2012	2014	2016	2018	2020	2022	Average Annual Growth Rate (in %)	2000 = 100%
DOLNOŚLĄSKIE	20,116	21,800	24,704	30,030	36,453	42,035	47,398	49,471	53,457	60,481	67,104	90,980	7.18	452.3
KUJAWSKO-POMORSKIE	17,530	19,084	21,502	24,231	28,911	30,893	34,019	36,058	39,343	44,964	50,246	66,547	6.31	379.6
LUBELSKIE	14,010	15,351	17,326	19,378	23,968	25,793	29,472	31,109	33,467	37,756	42,370	55,182	6.49	393.9
LUBUSKIE	17,495	18,690	21,667	25,167	29,285	31,636	34,874	37,561	40,554	45,398	50,209	66,313	6.31	379.0
ŁÓDZKIE	17,345	19,376	22,472	25,856	31,328	34,908	39,322	41,622	45,154	51,561	59,529	76,228	7.00	439.5
MAŁOPOLSKIE	17,552	18,834	21,644	25,295	30,232	32,689	36,935	39,258	43,634	50,873	55,138	72,004	6.69	410.2
MAZOWIECKIE	29,898	32,177	36,982	43,528	51,713	58,768	66,392	70,204	76,401	88,779	97,093	127,791	6.89	427.4
OPOLSKIE	16,294	17,235	20,852	22,831	29,004	30,516	33,878	36,064	38,511	43,907	48,834	64,383	6.56	395.1
PODKARPACKIE	14,225	15,758	17,746	20,017	24,164	25,961	29,278	31,384	34,055	39,021	42,501	55,125	6.40	387.5
PODLASKIE	14,387	16,207	18,038	20,388	24,598	27,345	30,167	32,357	34,434	39,770	45,345	59,818	6.76	415.8
POMORSKIE	19,328	21,327	23,988	27,667	32,241	35,810	40,949	42,108	46,754	53,813	57,680	81,149	6.84	419.9
ŚLĄSKIE	20,769	23,051	27,235	29,714	36,091	39,798	44,274	45,867	50,011	57,472	61,641	85,131	6.72	409.9
ŚWIĘTOKRZYSKIE	15,260	16,827	19,449	21,911	27,718	28,870	31,493	32,573	34,829	40,146	44,789	56,507	6.20	370.3
WARMIŃSKO-MAZURSKIE	15,167	16,013	18,490	20,904	24,855	27,160	30,068	31,973	34,556	38,271	43,662	56,368	6.19	371.6
WIELKOPOLSKIE	20,897	22,138	26,140	29,781	35,592	39,158	44,342	47,529	52,744	59,865	66,499	85,867	6.69	410.9
ZACHODNIOPOMORSKIE	19,361	20,298	22,062	25,231	30,187	31,981	35,288	37,443	40,673	46,455	51,790	66,443	5.82	343.2

Source: own study based on the Local Data Bank of the Central Statistical Office.

contains data on the value of GDP per capita of voivodeships in Poland in the years 2000–2022. It also provides selected information about the dynamics of changes in this indicator. Taking into account the period of analysis and care for the readability of data presentation, the data were presented in two-year intervals.

The presented data show that in the case of Poland, there is a clear and growing (in terms of absolute numbers) differentiation in the level of economic growth measured by GDP per capita. By far the highest level of GDP per capita was recorded in the Mazowieckie voivodeship. Over the entire study period, the average value of regional production per capita in this voivodeship was over 170% of the national average. This is, undoubtedly, due to the location of the largest Polish city, the capital of the country—Warsaw—in this voivodeship. As the largest urban center, with the highest administrative status, a large academic center, and an important and conveniently located transport hub of supra-national importance, it was and still is a natural place for the location of large, competitive enterprises, including those established as part of foreign direct investments.

Other regions with an above-average level of economic growth include, above all, the Dolnośląskie, Wielkopolskie, and Śląskie voivodeships. These are areas with a traditionally high level of economic activity and well-developed (especially during the socialist economy) industries and services. These are also regions with capitals in relatively large cities, which, along with their functional areas, display a number of features characteristic of metropolitan areas.

In turn, the least-developed voivodeships in the country included the Lubelskie, Podkarpackie, Warmińsko-Mazurskie, Podlaskie, and Świętokrzyskie voivodeships. They represent a conventional macro-region of Poland, traditionally called the Eastern Wall, characterized by a low level of economic growth, an uncompetitive production structure, a low level of urbanization, and a low level of human and social capital development. Throughout the entire study period, the average level of GDP per capita in this group of regions ranged from 76.8% (Lubelskie) to 83.1% (Świętokrzyskie) of the national average [79].

When considering changes in regional GDP per capita, several characteristic dependencies can also be noticed. The fastest developing voivodeships included Dolnośląskie, Łódzkie, Mazowieckie, and Pomorskie, i.e., regions characterized by a relatively high and average level of economic growth at the beginning of the analyzed period. Certainly, their common feature is the fact that their capitals are relatively large cities, with a highly developed sector of services (including those classified as the fourth sector), a favorable geographical location, and good transport accessibility.

In turn, the regions with the lowest dynamics of economic growth were the Lubuskie, Świętokrzyskie, Warmińsko-Mazurskie, and Zachodniopomorskie voivodeships, i.e., those with initially low and medium levels of economic growth. Compared to the national average, a relatively important part of their production structure is the agricultural sector, which in the years 1945–1989 was strongly dominated by large-scale, socialized but also ineffective and outdated State Agricultural Farms. Apart from the Kujawsko-Pomorskie voivodeship, these are also border voivodeships.

It follows from the above that there was no clear relationship between the initial level of economic growth and the dynamics of its growth in subsequent years. This is quite clearly visible in Figure 1, which shows the change in GDP per capita of the surveyed voivodeships (in percentage points) in relation to the national average between 2000 and 2022. The data presented in Figure 1 demonstrate that out of the 16 surveyed regions, exactly half improved the level of economic growth and the other half recorded a regression compared to the entire country. The regions that improved their position to the greatest extent include the Dolnośląskie, Mazowieckie, and Łódzkie voivodeships, which recorded an increase in their GDP per capita over the national average by 13.7 to 8.8 percentage points. In the case of the remaining voivodeships from the discussed group, this distance increased from 1.8 percentage points in the Małopolskie voivodeship to 4.6 percentage points in the Pomorskie voivodeship. This group included regions with initially relatively high and low levels of growth, with the richer areas prevailing. Similarly diverse, although with a slight dominance of less-developed voivodeships, was the composition of the group of regions that recorded a decline in GDP per capita compared to the national average. The largest regression (by 15.8 percentage points) and the smallest one (1.6 percentage points) were determined for the Zachodniopomorskie voivodeship and the Opolskie voivodeship, respectively, i.e., regions with a level of economic growth close to the national average in 2000.

Figure 2 presents data depicting in more detail the relationship between the initial level of economic growth (measured by the natural logarithm of the GDP per capita value in 2000) and the dynamics of its growth in the years 2000–2022 (as the natural logarithm of the GDP per capita value in 2022 to the value of GDP per capita in 2000). The figure also shows the average values of the mentioned variables for the entire studied group of voivodeships on the appropriate axes.

Based on data presented in Figure 2, the 16 surveyed voivodeships can be divided into four groups of regions:

(1)

Depressed regions—those that at the beginning of the study period had a lower level of economic growth and where the dynamics of its growth were lower than average. These were areas of divergence whose unfavorable economic situation further deteriorated during the period under study. These included the Warmińsko-Mazurskie, Świętokrzyskie, Lubuskie, and Kujawsko-Pomorskie voivodeships, i.e., regions with a relatively small population, non-competitive production structure, and most of them (except for the Kujawsko-Pomorskie voivodeship) being border regions.

(2)

Convergence regions—those that were initially less developed but in the years 2000–2022 recorded above-average GDP per capita growth rates. These included the Lubelskie, Podkarpackie, Podlaskie, Opolskie, and Łódzkie voivodeships; hence, a relatively heterogeneous group in terms of economic potential, demographic potential (including human capital development), and geographical location.

(3)

Crisis regions—voivodeships with above-average growth in 2000 and below-average growth dynamics in the following years. It is the least numerous group bringing together very diverse regions that includes the Zachodniopomorskie voivodeship and Mazowieckie voivodeship. The presence of the latter is, however, quite surprising because it has traditionally been considered the voivodeship with the highest level and dynamics of economic growth in Poland.

(4)

Expansion regions—areas with an above-average initial level and dynamics of economic growth, including the Pomorskie, Dolnośląskie, Śląskie, and Wielkopolskie voivodeships. These are voivodeships with highly developed industries during the years of the centrally planned economy, which, after 1989, quite quickly underwent transformation toward an economy dominated by services. Additionally, their administrative capitals are large cities representing important academic centers and transport hubs of supra-national importance.

In order to more formally determine whether β-type convergence occurred between the studied regions, a regression equation was estimated between the initial level of economic growth (measured by the natural logarithm of the GDP per capita value in 2000) and the average dynamics of its growth in the following years (as the natural logarithm of the ratio of the GDP value per capita in 2022 to its value in 2000). The estimated equation takes the following form:

$$y=1.1494-0.0008x$$

(p = 0.000) (p = 0.93)

R² = 0.02

Although the estimated regression function is characterized by a small, negative slope, which may suggest the occurrence of slow β-type convergence, neither the level of statistical significance of the estimated function parameters nor the value of the R² coefficient allowed for an unequivocal confirmation of this thesis.

Due to the impossibility of confirming the occurrence of β-type convergence, it was checked whether disproportions in the level of economic growth between Polish regions diminished in the years 2000–2022—in other words, whether σ-type convergence occurred. To this end, the values of the range and the coefficient of variation of GDP per capita of voivodeships in subsequent study years were calculated and presented in

Table 3. The value of the range and coefficient of variation of GDP per capita of voivodeships in Poland in the years 2000–2022.

Years	Range (in PLN)	Coefficient of Variation
2000	15,888	20.87
2001	16,473	20.84
2002	16,826	20.41
2003	17,546	20.43
2004	19,656	20.84
2005	21,938	21.79
2006	24,150	22.19
2007	26,682	22.13
2008	27,745	21.41
2009	31,297	23.02
2010	32,975	23.57
2011	35,379	23.73
2012	37,114	24.18
2013	37,612	24.17
2014	39,095	23.91
2015	42,010	24.35
2016	42,934	24.35
2017	46,273	24.68
2018	51,023	25.00
2019	55,727	25.54
2020	54,723	24.17
2021	60,100	23.62
2022	72,666	24.93

Source: own study based on the Local Data Bank of Statistics Poland.

.

The presented data show that disproportions in the level of economic growth increased in the analyzed period and, therefore, σ-type convergence did not occur. In the case of the range, its value increased more than 4.5 times, while the average value of GDP per capita in the entire country increased only slightly more than 4 times. The only year when a decrease was recorded in the range value was 2020, i.e., the initial period of the COVID-19 pandemic, when the most severe restrictions were imposed against the population and economic entities, which, as the above data show, had the strongest impact on the most developed areas.

The analysis of changes in the coefficient of variation leads to similar conclusions. In this case, its increase was also observed between the extreme years (by 4.06 percentage points). In the case of this measure, however, the increase in disproportions was not so dynamic and stable. Throughout the entire research period, there were four sub-periods when the value of the coefficient of variation of regional GDP per capita decreased. These were 2001–2002 (the period preceding Poland’s accession to the European Union), 2008 (the onset of the global economic crisis caused by the bursting of the speculative bubble on the US real estate market), 2013 (the debt crisis in the EU), and 2020–2021 (the onset of the coronavirus pandemic). It follows that the factors that stimulated the decline in development disparities in Poland were the periods of economic crises. Under normal economic conditions, the disproportions were observed to increase.

To recapitulate, a lack of economic convergence of both the β-type and the σ-type could be observed in the Polish regions in the first two decades of the 21st century. Given the adopted research goal, the question should be asked how this situation translated into the processes of social convergence estimated by changes in the synthetic measure of social development.

The research procedure for evaluating social convergence was the same as for economic convergence. It was therefore based, first of all, on the calculation of synthetic measures for each of the sixteen Polish voivodeships in the years 2000–2022 and their dynamics. The highest values of the synthetic measure of human development were recorded for the Mazowieckie, Śląskie, and Dolnośląskie voivodeships, which basically correlates with the positions of these voivodeships in the rank of the highest GDP per capita values. The reason behind this situation also seems to be identical to the previous one, i.e., strong urban centers having the nature of agglomerations significantly influence the diagnostic variables determined for voivodeships. While in the case of the Śląskie and Dolnośląskie voivodeships, their wealth largely translates into the socio-economic capital of the region, in the case of the Mazowieckie voivodeship, there are certainly significant development disproportions between the capital and several nearby cities surrounding it and the other areas of the voivodeship. In many aspects, especially the areas located east of Warsaw face socio-economic problems similar to those of the eastern voivodeships. Authorities of these regions have been struggling with a multitude of problems and challenges for years, whereas their inhabitants are forced to face the consequences of such deficits. The lowest values of the synthetic measure of social development were recorded in the areas located in the east of Poland, i.e., in the Podkarpackie, Świętokrzyskie, Lubelskie, and Warmińsko-Mazurskie voivodeships. Moreover, a somewhat positive phenomenon observed (especially in the context of convergence processes) was the reduction in the spatial dispersion of this phenomenon, measured by the range value. It was not permanently positive (it is difficult to indicate a distinct trend in the examined period); however, the range decreased by approximately 0.03 points and the average value of this synthetic measure increased by 0.03 during the examined period. This can quite simply suggest trends of changes in social convergence processes.

Next, the relationship between the initial value of the synthetic measure of human development and the dynamics of its growth in 2000–2022 was verified taking into account the natural logarithms of these values. The regions examined were also assigned to four groups, as previously based on the initial level of economic growth and the dynamics of its growth. However, their distribution was different and was characterized by smaller spatial dispersion:

(1)

Seven voivodeships were included in the group of depressed regions that were at a lower level of social development and where the dynamics of its growth were lower than average (areas of divergence): Świętokrzyskie, Warmińsko-Mazurskie, Lubelskie, Śląskie, Kujawsko-Pomorskie, and Lubuskie. Their catalog has therefore expanded compared to their division based on economic divergence.

(2)

The regions of social convergence characterized by an above-average growth rate of social development measured by the growth rate of a synthetic measure included only the Dolnośląskie voivodeship.

(3)

Podkarpacie turned out to be the region that recorded social regression with growth dynamics below the country’s average.

(4)

The expansion regions that, in turn, maintained their above-average social development while recording an above-average pace of growth turned out to be the remaining seven voivodeships: Podlaskie, Opolskie, Zachodniopomorskie, Wielkopolskie, Pomorskie, Małopolskie, and Mazowieckie.

Then, the occurrence of β-type social convergence was verified using the estimated regression equation between the initial level of the synthetic measure of human development (measured by the natural logarithm of its value in 2002) and the average dynamics of its growth (as the natural logarithm of the ratio of the value of this measure in 2022 to its value in 2000). The estimated equation was

$$y=-1.3586+0.5543x$$

R² = 0.3364

The parameter values did not allow for confirming the occurrence of β-type social convergence due to the positive slope of the function. As a consequence, no attempt was undertaken to verify the occurrence of σ-type convergence because, as stated earlier, the occurrence of β-type convergence is a necessary but not sufficient prerequisite for the occurrence of σ-type convergence.

5. Summary and Conclusions

The aim of the presented study was to check whether, in relation to Polish regions in the years 2000–2022, it was possible to observe the processes of economic convergence, i.e., the approximation of the examined units in terms of the dynamics and level of economic growth, and social convergence, i.e., the equalization of spatial disproportions in the standard of living. Each of these issues separately but also together (in the context of the occurrence of possible connections and dependencies between them) is important for the implementation of the cognitive goals of economics as well as for practical purposes. One of the overarching directions of the economic and social policy of the European Union, in which Poland celebrated its 20th anniversary in 2024, is the pursuit of economic and social cohesion.

The results of analyses conducted within this study enable concluding that the processes of both economic and social divergence occurred in Poland in the years 2000–2022. The basic assumptions of the cohesion policy implemented primarily by the EU Cohesion Fund failed to bring the expected results. Thus, the research results correspond with the findings from other studies indicating the lack of economic and social convergence in Polish regions (e.g., [25,26,33]). The dynamics of changes in GDP and its range were increasing, and there was no convergence of the well-being of the population measured by the synthetic measure of human development created using the TOPSIS method.

The results obtained are partly supported by the works of other authors, especially in the context of economic cohesion in other East and Central European countries. Most of these studies find that at the national level, one can observe quite clear (consistent with the “iron law of convergence”) economic convergence between different member states of the EU. However, in the case of new member states that joined the EU in 2004 and later, there is a visible divergence between the regions and subregions of these countries. The process of spatial polarization of economic growth is most evident between the capital and peripheral regions. Economic divergence is usually accompanied by social divergence, see [77,78,79,80,81,82,83,84,85]. However, one can also find studies in which the thesis of social convergence finds its support in the empirical analysis, see [54,85].

The added value of this study is the evaluation of the situation of Polish regions over the last two decades, taking into account economic and social indicators. Their comprehensive summary is presented in Figure 3 (left side—economic development index, right side—synthetic measure of social development).

Understandably, the Polish specificity of regional development, both in economic and social terms, highlighted the similarities between Polish voivodeships. High positions of the Pomorskie and Wielkopolskie voivodeships emphasized sustainable regional development, taking into account rural environments. In both rankings, these regions were included in the group with above-average growth rates of the already high measures. Disproportions in the urban–rural system, in turn, prevented high positions of those parts of the country whose capitals can be classified as agglomerations (Warsaw, Cracow, Wrocław). There is also a certain regularity in the context of the regions affected by the greatest socio-economic challenges, which certainly include the voivodeships located in the eastern part of the country (Świętokrzyskie, Warmińsko-Mazurskie).

Finally, it is worth recalling the limitations of this study, especially in its social dimension. The synthetic measure of social development, although supported by an analysis of the literature and an overview of other research in this area, is still an original approach because there is no objective measure that completely defines social development, and therefore social cohesion and social convergence, in an empirical dimension.

However, this does not change the fact that it is much easier to find publications pointing to the difficulty of achieving both economic and social convergence. Such conclusions can also be drawn from this study. This calls into question the assumptions of cohesion policy, although, on the other hand, it is difficult to imagine basing sustainable development on the exclusive support of strong socio-economic centers at the expense of peripheral regions with a lower or negligible growth rate. Dilemmas of this type remain the responsibility of decision-makers and creators of dedicated policies. This work is, therefore, another voice in the discussion on equal opportunities and care for the well-being of every citizen, with negative empirical verification of the outcomes of this policy.

If social cohesion requires trust, social relations, civic and social participation, integration, and attitudes toward social services, as well as constructive dispute resolution, then the practical conclusion from this study is the undertakings aimed at the constant strengthening of social capital. Despite the still greater scale of their economic problems than the mature EU member states, the standard of living in Poland has improved significantly. Therefore, investing in social capital can definitely accelerate the process of social convergence. And, building trust between the state and the citizen seems to be an important task to be performed. In many countries of Central and Eastern Europe, this element requires significant improvement. Such activities should be planned in the long term because qualitative changes in societies rarely occur quickly.