Estimating and Combining National Income Distributions Using Limited Data

Abstract

A major problem encountered in studies of income inequality at regional and global levels is estimating income distributions from data that are in summary form. In this article we estimate national and regional income distributions within a general framework that relaxes the assumption of constant income within groups. We propose a technique to estimate the parameters of a beta-2 distribution using grouped data. We model regional income distribution using a mixture of country-specific distributions and examine its properties. We use the techniques to analyze national and regional inequality trends for eight East Asian countries and two benchmark years, 1988 and 1993.