Zastosowanie metody przestrzenno-czasowych składowych głównych (STPCA) do identyfikacji przestrzennego zróżnicowania społeczno-gospodarczego powiatów

Abstract

The aim of the article is to identify the spatial socio-economic differentiation of counties in the years 2005–2024. A set of 30 variables was used, including demographic and economic variables, the state of infrastructure, and socio-environmental variables. A spatio-temporal principal component method for functional data was applied to reduce the dimensionality of the observations. First, the variables were transformed into a functional form, and a matrix representing the spatial interaction structure was determined. Then, 4 positive and 4 negative principal components were extracted. Their interpretation allowed them to state that, in the case of positive components, infrastructural variables make the most important contribution. County differentiation is determined by variables describing access to hospital care and county finances.

The classification of county types was carried out using the cluster analysis method, conducted separately for the positive and negative principal components. As a result, two divisions into 3 county types each and one into 9 types were obtained. The resulting typology confirms the persistent differentiation of counties based on their location within the territories of former partitioning states, the dichotomy between urban and rural counties, and their location relative to metropolitan areas. The obtained results confirm the usefulness of the applied method for constructing a typology of counties. The types distinguished based on the cluster analysis of the components take into account both agglomeration effects, which are more significant than processes that differentiate counties. Moreover, the county typology lends itself to straightforward interpretation, and its spatial distribution can be explained.