Two-part models of income distributions in Poland

Research background: studies of structures of incomes distributions have been performer for about 15 years. They indicate that there is no one model which describes the distributions in their whole range. This effect is explained by the existence of different mechanisms yielding to low-medium and high incomes. While more than 97% of the distributions can be described by models with two or three parameter, high incomes (about 3% or less) is in agreement with power law. Purpose of the article: the aim of this paper is an analysis of the structure of distributions of households’ incomes in Poland. By using various models we verify the hypothesis about two-part structure of those distributions. Methodology/methods: the studies are based on the households’ budgets microdata for years 2004 – 2012. The two-component models were used to describe the incomes distributions. The major parts of the distributions have been described by the two or three parametric models: lognormal, Dagum, and Singha-Madalla. The highest incomes were described by the Pareto model. Findings: one has showed that two or three parametric models explain from about 95% to more than 99% of ranges of income distributions. The poorest agreement with data is for lognormal model, while the best agreement has been obtained for Dagum model. Regarding the highest incomes the Pareto model describe the data very well only for the selected years. For the remaining years the results are not so obvious. The tails of the income distributions seems to have more complex structure