مدل احتمالاتی توزیع درآمد

The results showing relationships between distributions of individual incomes and incomes of two-earners households are presented. It is shown that individual incomes are very well described by Dagum's distributions. Income distributions in two quite different countries (USA and Poland) are studied for comparison. Obtained results show very striking and interesting differences.

The preliminary results of income distributions of households in two different countries (USA and Poland) are presented. We focus on individual incomes (incomes of one-earner households) and on incomes of two-earners households. Obviously the income of a two-earners household is just the sum of individual incomes. The aim of this work is to examine if there is any relationship between a distribution of individual incomes and a distribution of incomes in two-earners households. We present result that concern two countries: USA and Poland. In the case of the USA such a relationship has recently been investigated in Ref. [1]. For a description of individual incomes the authors of that paper used the exponential model. In this work we use another model of income distributions which seems to provide a better fit to experimental data, especially for Polish households. Full results of our investigations relating individual distributions of incomes in the USA can be found at an internet site [2]. They include all the years 1975–2002. Due to the lack of space, in this paper we restrict ourselves to the year 2000. Results concerning distributions of incomes of Polish households for the year 2000 are an extension of our previous paper [3].

We have shown that Dagum's distribution describes very well the distributions of individual incomes in both the USA and Poland. We confirm, however, the results of Ref. [1] that the American incomes are well-approximated by an exponential distribution function. The distribution of incomes of two-earners household in the USA can be constructed as a convolution of the individual income distributions. This suggests that incomes of members of households in the USA are statistically independent. It is not, however, the case of the Polish households—the result are completely different. The incomes of members of households in Poland seems to be strongly statistically dependent. A further analysis of this dependence should deliver more information about distributions of incomes in Polish families. To check the generality of the obtained results, in a forthcoming paper we plan to apply our analysis to other European and non-European countries.