The Use of Hazard Models for the Analysis of Income Inequalities in Poland

Abstract

The aim of the paper is to examine the income inequalities of men and women in Poland. We estimate conditional cumulative distribution functions for incomes in both groups of people using a flexible hazard-function based estimator in the presence of covariates. The conditional piecewise-constant exponential hazard models are applied. Then, we decompose the estimated income differences along the whole income distribution. For this purpose, we construct the counterfactual distribution, which is the distribution of incomes that would prevail for women if they had the distribution of men’s characteristics. The method allowed to investigate the structure of inequalities in the entire range of income values for the two analyzed groups of people. The empirical investigation is based on the data collected within the EU-SILC project.