Factors of Divergence between Peripheral and Central Subregions: The Case of Mazovia Province, Poland

Based on the outlined premises, the following research questions arise: –

Are there any correlations between features of the analysed subregions that derive from the peripheral location in a province and the adopted development indicators?

In order to answer these questions, features of the analysed subregions related and not related to the peripheral location in a province have been identified as independent variables (IndVar). Then, their correlations with the adopted development indicators (dependent variables [DepVar]) have been measured.

The following features of the analysed subregions (IndVar), which are related to the peripheral location in a province, have been considered: –

provincial capital status of the subregion's largest city (IndVar 1),

The following features that are not related to the subregions’ peripheral location in a province have been considered: –

the size of the subregion's largest city (IndVar 3),

Subregions’ development levels in the years 2005 and 2015 (in some cases 2008 and 2016) have been taken into account because of data availability. The following features, deemed as development factors, have been considered: –

fixed assets in companies per capita,

They have been aggregated into the following development indicators (DepVar): –

development levels for the years 2005 (2008) and 2015 (2016) (DepVar 1) (see Spychała 2018),

All subregions in Mazovia Province have been analysed (see below). As a reference, all subregions located in provinces neighbouring the Radom subregion have been considered, i.e. the provinces of Łódź, Świętokrzyskie and Lublin. Such a delimitation of reference area accounts for the subregions that differ in features deemed as possible development factors (IndVar); however, in many terms, they may be considered as similar to minimise the third variable problem (León, Montero 1997) (Fig. 2).

Fig. 2

To maximise subregions’ comparability, statistical NUTS 3 subregions were modified in a form close to functional city regions, perceived as a core city with its hinterland that exceeds boundaries of its urban agglomeration. It was needed because among statistical subregions, some are similar to functional city regions (e.g. subregions of Radom, Lublin and Kielce), some are doughnut shaped (e.g. the subregion of Łódź) and some are identical to big cities (Warsaw and Łódź).

In 2008, joint subregions of Ostrołęka and Siedlce (ostrołęcko–siedlecki) as well as Ciechanów and Płock (ciechanowsko–płocki) were split. To reach comparability, they have been treated as joint subregions, as they were in 2005. All units developed in such a way that they have been uniformly called ‘subregions’ in this research. Figure 3 shows all the subregions considered.

Fig. 3

As a method of assessment, The Pearson correlation coefficient r_xy was applied (see Spychała 2018): rxy=n∑xiyi−∑xiyin∑xi2−(∑xi)2n∑yi2−(∑yi)2{r_{xy}} = {{n\sum {x_i}{y_i} - \sum {x_i}{y_i}} \over {\sqrt {n\sum x_i^2 - {{\left( {\sum {x_i}} \right)}^2}\sqrt {n\sum y_i^2 - {{\left( {\sum {y_i}} \right)}^2}} } }} where n is the number of subregions analysed (n = 14), x is the value of each development factor (1–10), y is the value of each independent variable (IndVar 1–5) and i is the respective subregion.

To measure the desired correlation between independent (IndVar 1–5) and dependent (DepVar 1–3) variables, firstly the Pearson correlation coefficients between the values of each independent variable (IndVar 1–5) and the values of each development factor relevant to each of the dependent variables (DepVar 1–3) were calculated (r_{IndVar1DepVar1xy} … r_{IndVar5DepVar3xy}). Secondly, mean values (M_{rIndVar1DepVar1} ... M_{rIndVar5DepVar3}) of the Pearson correlation coefficients obtained in this way were calculated for each pair of independent (IndVar 1–5) and dependent (DepVar 1–3) variables.

The following detailed research hypotheses (HD) have been formulated:

HD 1–3: The provincial capital status of the subregion's capital (IndVar 1) as well as subregion's level of centrality within the province (IndVar 2) are the positive factors of the following: –

development levels for the years 2005 (2006) and 2015 (2016) (DepVar 1) (HD 1),

Table 1 depicts the calculated mean values of the Pearson correlation coefficients Mr_{IndVar1DepVar1} ... Mr_{IndVar5DepVar3} that represent correlations between each of the independent (IndVar 1–5) and dependent (DepVar 1–3) variables.

As MrIndVar1DepVar1⋯MrIndVar2DepVar3>0{M_{{r_{IndVar1DepVar1}}}} \ldots {M_{{r_{IndVar2DepVar3}}}} > 0

As detailed hypotheses HD 1, HD 2 and HD 3 have been positively verified.

In order to better illustrate ranking of significance of subregions’ features (IndVar) in terms of their correlation with respective dependent variables (DepVar), several graphs have been used (Figs 4–6). As can be seen in Fig. 4, it is the size of the subregion's largest city that mostly correlates with subregions’ development levels for the years 2005 (2006) and 2015 (2016).

Fig. 4

Fig. 5

Fig. 6

The research conducted here has shown that both the independent variables related to the subregions’ peripheral location in a province, i.e. lack of the provincial capital status of the subregion's largest city (IndVar 1) and low level of centrality within the province (IndVar 2), are not as significant factors of subregions’ development levels as the size (IndVar 3) and administrative status of their largest city (IndVar 4) (Fig. 4): MrIndVar1DepVar1<MrIndVar3DepVar1MrIndVar2DepVar1<MrIndVar3DepVar1\matrix{ {{M_{{r_{IndVar1DepVar1}}}} < {M_{{r_{IndVar3DepVar1}}}}} \cr {{M_{{r_{IndVar2DepVar1}}}} < {M_{{r_{IndVar3DepVar1}}}}} \cr }

This means that the detailed hypotheses HD 4 and HD 7 have been negatively verified.

As far as the absolute difference of development is concerned, Fig. 5 illustrates its correlation with the respective independent variables.

The ranking of correlations shows that the two most important factors that correlated with the absolute pace of subregions’ development are the provincial capital status of the subregion's largest city (IndVar 1) and its level of centrality within the province (IndVar 2): MrIndVar1DepVar2<MrIndVar3DepVar2⋯MrIndVar5DepVar2MrIndVar2DepVar1<MrIndVar3DepVar2⋯MrIndVar5DepVar2\matrix{ {{M_{{r_{IndVar1DepVar2}}}} < {M_{{r_{IndVar3DepVar2}}}} \ldots {M_{{r_{IndVar5DepVar2}}}}} \cr {{M_{{r_{IndVar2DepVar1}}}} < {M_{{r_{IndVar3DepVar2}}}} \ldots {M_{{r_{IndVar5DepVar2}}}}} \cr }

This denotes that the detailed hypotheses HD 5 and HD 8 have been positively verified.

Other results can be seen in Figure 6 in the case of the correlation between independent variables and the relative difference in subregions’ development levels in the years analysed.

The most important factor of alleviating existing disparities is the size of the subregion's largest city (IndVar 3): MrIndVar1DepVar3<MrIndVar3DepVar3MrIndVar2DepVar3<MrIndVar3DepVar3\matrix{ {{M_{{r_{IndVar1DepVar3}}}} < {M_{{r_{IndVar3DepVar3}}}}} \cr {{M_{{r_{IndVar2DepVar3}}}} < {M_{{r_{IndVar3DepVar3}}}}} \cr }

It implies negative verification of the detailed hypotheses HD 6 and HD 9.

The following conclusions may be drawn from the conducted research.

The provincial capital status of the subregion's largest city, as well as its high level of centrality within the province positively impact the performance of the subregion in terms of the development level and its dynamics (Table 1). In the case of peripheral subregions of Mazovia Province, it means that their location in this huge province and their capital cities’ loss of the former provincial capital status are negative development factors. This conclusion supports the thesis, that above-mentioned two circumstances may be one of the reasons for the degradation of the Radom subregion that has been observed since its inclusion in Mazovia Province.

Figure 4 shows that all the surveyed features of the subregions, i.e. the size of a subregion's largest city, the administrative status of its largest city, its provincial capital status, the level of centrality within the country and the level of centrality within the province, are significant factors of the past and current subregions’ level of development in recent years. However, the absolute level of subregions’ development is mostly related to factors historically developed over time, such as the size or the overall administrative status of their largest cities.

Figure 5 shows the factors for the absolute rate of subregions’ development in recent years. It indicates that the provincial capital status as well as the central location of the subregion within its province are the most significant development factors among all surveyed ones.

The interpretation of Figure 6 is more complicated. In terms of the relative rate of surveyed subregions’ development in recent years, the most important factor is once again the size of their largest city. It means that the poor subregions which have big central cities are catching up quicker than the poor subregions with smaller cities. The second important and the only other significant factor in this regard is the provincial capital status of the subregion's largest city. It implies that the poor subregions of provincial capitals are catching up quicker than the poor subregions the largest cities of which are deprived of the provincial capital status.

Additionally, the development distance is reduced faster by the subregions located peripherally in the whole country, which may be explained by the ‘Development of Eastern Poland’ programme launched under the EU Cohesion Policy (Spychała 2018). This programme aims to support peripheral and poor eastern Poland regions (Czudec et al. 2019). What does it signify for the Radom subregion?

As the capital city of Radom is a considerably large subregional centre, the Radom subregion is privileged in terms of the relative rate of development. On the other hand, Radom's lack of provincial capital status as well as its central location in Poland inhibit the catch-up pace. Other Mazovian subregions share these two regrettable circumstances too.

Combining the results presented in Figures 5 and 6, one can conclude that in the case of considerably affluent subregions, their further development is conditioned mainly by the provincial capital status of their largest cities as well as their central location within the province. Both features, by definition, are present in the subregions of provincial capitals. This fact detrimentally affects the pace of development of all peripheral Mazovian subregions in comparison to the provincial capital ones.

The reasoning presented allows answering the general research questions posed. Firstly, there are correlations between features of analysed subregions that derive from their peripheral location in a province and adopted development indicators in all terms, i.e. the development level and its absolute and relative pace. These correlations are stronger than all other presumable ones in the case of absolute development pace only. However, in other two measures, these two features are the second and third most important development factors, outdone by the feature of the size of the subregion's largest city only.

This research has supported the thesis that both the provincial capital status of the subregion's largest city as well as the subregion's central location within its province are important growth factors in contemporary Poland. In the case of the Radom subregion, it implies that Radom's loss of the provincial capital status in 1999 and its location in Poland's largest Mazovia Province, the capital city of which is also the capital of the country, may be deemed an important factor of its relative downfall ever since. This regrettable situation is shared by other former provincial capitals located in the analysed area too, especially those located in Mazovia Province.

The main objective of The Strategy of Development of Mazowieckie Voivodeship until 2030 has been built on a paradox: ‘territorial cohesion, perceived as lowering development disparities in Mazovia Province as well as growth of Warsaw Metropolitan Area's importance in Europe (…)’ (SRWM 2013, 1). So, such an inconsistent definition of ‘territorial cohesion’ inevitably implies a dilemma: what should be the priority – growth of backward subregions or the prosperous Warsaw Metropolitan Area? A ‘strategic goal’ was formulated as follows: ‘increase of region's competitiveness through growth of economic activity (...)’ (SRWM 2013, 4). It should be implemented e.g. through the following ways: ‘Warsaw as the capital city – growth and supplementing of metropolitan functions, supporting growth of regional and subregional centres’ (SRWM 2013, 4). Unfortunately, such a provision does not grant a higher priority to activities aimed at developing economically backward subregions, but treats them equally or even worse in comparison to the further strengthening of the Warsaw city region. The authors of the strategy explain such a choice by arguing that the Warsaw Metropolitan Area has ‘the biggest potential of socio-economic development that has a vital impact on the growth of the whole region’ (SRWM 2013, 6.2). This claim is disputable. Firstly, it seems rather that places with the greatest development potential are those that suffer economic downfall while having a demographic potential, unemployed labour force and material assets, e.g. the city of Radom. Secondly, significant disadvantage of peripheral Mazovian subregions towards the Warsaw Metropolitan Area denotes that promoted growth of the Warsaw city region does not pass on to province peripheries. As this research has proved, the grim reality of statistics clearly indicates that development diffusion model does not work in practice. The main dilemma of Mazovia Province is thus how to bind two types of areas, so that the economic success of Warsaw could be beneficial to the whole province (Śleszyński 2015).

The lack of desirable development priority of disadvantaged subregions discussed above reflects a delimitation of ‘strategic intervention areas’ set up in the strategy (SRWM 2013, 6). Because intervention measures are to be taken in order to change a poor condition, it could be assumed that a ‘strategic intervention’ would apply only to problem areas. Nonetheless, the strategy defines practically the whole province as ‘strategic intervention areas’. It encompasses all peripheral subregions as well as the Warsaw Metropolitan Area, causing any ‘strategic intervention’ to be a fiction (Zaborowski T. 2013).

The current regional administrative division of Poland has been frequently criticised by scientists (Sokołowski 2014; Miszczuk 2003; Zaborowski Ł. 2009). It has been pointed out that the number and the form of regional territorial entities is an effect not only of a rational debate, but also of the pressure put by local communities, lobbies and political horse-trading (Nelicki 2001; Kowalczyk 2000; Wendt 2001; Miszczuk 2003; Piotrowski 2004; Bober et al. 2013; Zaborowski Ł. 2013). In effect, present administrative division does not suit any of the proposals made by scientists, being a combination of big and middle-sized regions’ concepts (Miszczuk 2003; Zaborowski Ł. 2013, 2016). Therefore, many experts advocate a reform of the controversial division (Habuda, Habuda 2014; Honka 2014; Sokołowski 2014; Zaborowski Ł. 2014; Śleszyński 2015a).