Abstract
Researchers have investigated the effects of ethnic heterogeneity on a range of socioeconomic and political outcomes. However, approaches to measuring ethnic diversity vary not only across fields of study but even within subfields. In this review, we systematically dissect the computational approaches of prominent measures of diversity, including polarization, and discuss where and how differences emerge in their relationships with outcomes of interest to sociologists (social capital and trust, economic growth and redistribution, conflict, and crime). There are substantial similarities across computations, which are often generalizations or specializations of one another. Differences in how racial and ethnic groupings are constructed and in level of geographic analysis explain many divergences in empirical findings. We conclude by summarizing the type of measurement technique preferred by outcome, when relevant, and provide considerations for future researchers contemplating how best to operationalize diversity. Finally, we highlight two less widely used yet promising measures of diversity.