In recent years a large number of cross-national studies have examined the causes and consequences of income inequality within nations. Unfortunately, few of these studies have attended very carefully to problems of measurement and definitional consistency that can seriously undermine the comparative use of currently available data on income shares. This article offers a discussion of the major theoretical and practical problems that can arise in measuring and comparing patterns of income distribution across nations, focusing on the completeness of income coverage, the unit of analysis, the time period over which income is measured, the scope of population coverage, the underreporting of income, and the effect of public sector fiscal policies. It then assesses major published sources of cross-national data on the size distribution of income in light of these problems. Finally, the article offers several suggestions for minimizing the negative consequences of measurement problems that remain in even the best available data on income shares.
