Why is our understanding of segregation contradictory?

Abstract

Investigating segregation presupposes a wide range of decisions. These concern, first, what type of segregation is to be examined (for example residential segregation or school segregation) and which kinds of attributes (such as ethnicity or socioeconomic status) are to be placed at the centre of the analysis. In addition, one must decide which measure of segregation to use, as well as make delimitations in time and space. All of these choices have consequences for the results of a study. They are therefore of interest both to those who are about to conduct a study and to those who engage with the work of others. Here it is demonstrated that in Sweden the imaginations of segregation are contradictory. An account of how interest in residential segregation developed over the course of the twentieth century is therefore broadened to include other forms of segregation and to explain why segregation is perceived as a social problem. The conditions for studying segregation depending on the chosen attributes are then discussed. Against this background, it is to be expected that the picture of segregation is multifaceted. The choice of segregation measure, as well as the ways in which data are aggregated over time and space, also affects the results. Yet too often the properties of the chosen measure are not considered, and data aggregation is taken for granted. As a result, segregation is able, to varying degrees, to “hide” in studies that are intended to shed light on it. Consequently, the overall picture of segregation becomes unnecessarily distorted and contradictory.