Correcting the Concentration Index for Binary Variables

This paper discusses measurement of socioeconomic inequalities in prevalence of a health condition. As its point of departure, it uses the recent exchange between Guido Erreygers and Adam Wagstaff in this journal, where they discuss merits of their own corrections of the frequently used concentration index. The paper first reconciles the debate between Erreygers and Wagstaff and discusses the value judgments that are hidden behind their indices. Secondly, it discusses when the property level independence is desirable. Thirdly, it empirically illustrates whether and when the choice of index affects comparisons between contexts using binary health indicators from the European Survey of Health, Ageing and Retirement (SHARE). The results show that choice of index matters and that it not only affects the magnitude of measured inequalities but also internal ranking between countries.