Estimation of the income distribution in latin america using limited information

The estimation of the income distribution is one of the main research aspects in growth theory and poverty dynamics. However, for many countries the information is partial and only a few pieces of information are available. The objective of this paper is to estimate the income distribution in Latin America, which is traditionally considered one of the most unequal regions of the world. We combine a measure of inequality which is taken from UNU-WIDER database and the average income provided by the World Bank database. To ensure comparability across years and countries, Gini coefficients are homogenized using a regression based adjustment. The methodology employed has two steps. In a first step, we estimate the within country distribution using a version of the Singh Maddala distribution, and in a second step we estimate the regional distribution using finite mixtures of this family. Countries distributions are derived using national accounts GDP per capita and empirical Gini estimates. In order to study the robustness of the results with respect to the considered parametric form, we estimate the within country distribution using a version of the Dagum distribution. Finally, we present some additive decomposable inequality measures and poverty rates, concluding that the countries have converged in the last decade.