Dekompozycja nierówności płacowych w Polsce za pomocą metod bazujących na konstrukcji rozkładów warunkowych
Decomposition of income inequalities in Poland using methods based on the construction of conditional distributions

Summary/Abstract: In the paper, we consider various estimation methods of conditional distribution functions for income distribution using regression models such as binomial logit, multinomial probit and hazard models. These concepts allow to model the probability that a person’s income will not exceed a certain level. Estimated distribution functions and counterfactual distributions were used to achieve the goal of decomposition of income inequalities among men and women along the whole income distribution. The total differences were decomposed into the explained and the unexplained part. The high share of the positive unexplained part indicates a higher “valuation” by the market of men’s characteristics than of women’s (the effect of discrimination), while the negative values of the explained component show a reduction of inequality due to “better” characteristics of women. Empirical data from the EU-SILC study for Poland in 2014 were used.

• Details
• Contents