Spatial distribution of income inequality and its determinants in Italy and Germany

Abstract

The study is a comparative analysis of spatial and socioeconomic determinants of income inequality in municipalities in Italy and Germany. The general aim is determining the accuracy in estimating unknown inequality values by learning from different types of municipality subsamples, with a particular focus on regional differences and applications for actionable policy implications. For this purpose, we combine demographics and socioeconomic data, with natural and built environment data, at multiple spatial scales. We use machine learning for estimating variable importance for the prediction of income inequality variables extracted from tabulated income tax declarations. Results show that selected variables explain 52% of the variability in income inequality in Germany, and 59% of the variability in income inequality in Italy. Employment and housing factors associated with income inequality are largely dissimlar between East and West Germany. Education and mobility factors are largely dissimilar between South, Central and North Italy.