Spatial econometric approach to the EU regional employment process

This paper deals with the estimation of spatial econometric models of employment rate across 259 NUTS 2 (Nomenclature of Units for Territorial Statistics) regions of the European Union in 2018 regarding different region-specific factors. Since, spatial autocorrelation and spatial heterogeneity often occur jointly, the paper is oriented at verification of two hypotheses. Hypothesis 1 related to the existence of the spatial autocorrelation, i.e., that the regional employment process is not a spatially isolated process, was confirmed. Based on the estimation of Spatial Durbin Model, direct, indirect and total spatial impacts were quantified and verified. The results proved the significant impact of neighbouring regions for GDP and compensation of employees variables in explaining regional employment rate. Significant influence of factors like educational attainment level and population density seems to be limited only to the particular region. Hypothesis 2 reflected the existence of the spatial heterogeneity. Based on the geographically weighted regression the assumption of spatial variability of the model parameters was also verified. The regional employment in the EU seems to be affected by both spatial effects and the presented approaches thus represent two different insights into the complex spatial character of the modelled process.