Differences in income distribution given by the structure of household

Incomes of households reflect the socioeconomic development of population and simultaneously influence the state of society because the incomes affect the behavior of households as the basic socioeconomic subjects. A direct relation exists between incomes and purchasing power which to a significant extent determines the consumption behavior of households. The presented paper is focused on specification of differences in the income distribution according to their structure. Since the different types of households mutually differ not only in their size and number economically active members but also in their demographic structure (age, education, etc.) we can expect differences also in the character of their income distributions. The goal of the paper is a construction of appropriate income distribution models and identification of differences. For this purpose, we can use both basic sample characteristics and suitable models of frequency distribution. The modeling of frequency distribution can be performed using non-parametric methods (kernel estimates with Gaussian kernel) and parametric models (finite mixtures of densities). The finite mixtures of densities are usually considered as more suitable for modeling of distributions of random variables in case of inhomogeneous populations than simple lognormal models with two or three parameters.