An asymmetrical approach to measuring residential segregation

The index of dissimilarity is the most widely used measure of residential segregation employed by sociologists. Since a single index value is used to describe the segregation between a pair of groups or between one group and the remainder of the population, its usage without taking the population composition into account implies a certain symmetry. A different way of viewing residential segregation is suggested here, in which the isolation of a pair of groups from one another is not assumed to be equivalent. This occurs when the relative numbers of each group is taken into account a matter intentionally avoided with the index of dissimilarity. This paper describes a set of basic asymmetrical measures for determining the isolation of a group from either the remainder of the population or another specific group. The segregation of blacks and whites in a number of cities in the United States is then examined for several decades. The results generated by these asymmetrical indexes are quite different from the conclusions that would otherwise be obtained through the index of dissimilarity. In turn this leads to a novel interpretation of shifts in black-white residential segregation, one which could not possibly be observed through a symmetrical measure