Description:
Decomposition methods for income inequality measures, such as the Gini index and the members of the Generalised Entropy family, are widely applied. Most methods decompose income inequality into a between (explained) and a within (unexplained) part, according to two or more population subgroups or income sources. In this article, we use a regression analysis for a lognormal distribution of personal income, modelling both the mean and the variance, decomposing the variance as a measure of income inequality, and apply the method to survey data from Russia spanning the first decade of market transition (1992-2002). For the first years of the transition, only a small part of the income inequality could be explained. Thereafter, between 1996 and 1999, a larger part (up to 40%) could be explained, and ‘winner’ and ‘loser’ categories of the transition could be spotted. Moving to the upper end of the income distribution, the self-employed won from the transition. The unemployed were among the losers.