The problem of statistical inference in inequality research has persisted despite the solutions long suggested by asymptotic theory. This article provides a thorough analysis of the theoretical and practical aspects of conducting statistical inference in the context of inequality data. Using the income surveys of the Luxembourg Income Study, the article outlines in detail the new bootstrap procedures, assesses their performance, and demonstrates their effectiveness in evaluating levels and trends in cross-national inequality. It provides an analytical evaluation of current advancements in inequality measurement and for the first time applies these methods in the context of comparative research, but it also can be used as practical guide for handling the problems of statistical inference in more general social scientific settings.
Key Words: inequality • statistical inference • bootstrap confidence intervals