Kernel Density Techniques as a Tool for Estimating and Comparing Income Distributions: A Cross European–Country Study

Abstract

This paper compares and assesses the income inequality between five European countries in the mid 1990’s, employing the non-parametric technique of kernel density estimation. The countries used in this inequality exercise were Germany, Hungary, Luxembourg, Poland and the United Kingdom, and the analysis was based on comparative data and variables provided by the PACO project. Kernel density estimates were found particularly revealing for comparing the shape of income distributions between populations, and for mapping the impact that differences in income polarization and concentration in various subgroups have on the overall income distribution of a country.