ABSTRACT

The empirical application of measures of inequality raises various issues which set this branch of statistics apart from others. A first and minor problem is the proper inference from a sample to a larger population. As samples are typically large, or nonrandom, or populations rather small as in the context of industrial concentration, relatively little work has been done on this.*

Much more important, in particular in the context of income inequality, is the incompleteness of the data: Typically, income figures are available only for certain quantiles of the population. and the rather voluminous literature on inequality measurement with incomplete data can be classified according to the amount of arlditional information available.