More on the Index of Dissimilarity

In my presentation on racial integration, I talked about something called the index of dissimilarity as a way of measuring segregation and there seemed to be some confusion about what exactly it represented. I’ll do my best here to explain it a bit better.

At first, the index of dissimilarity can seem like a confusing concept. One simple way to define it is as “a measure of the evenness with which two groups are distributed across the component geographic areas that make up a larger area.” Usually, this means neighborhoods or census tracts within a city. The index is measured between two different groups as the X-Y index of dissimilarity and is given as percentage of people from one ethnic group (either X or Y) that would need to move to other areas in order to create a distribution of people that matches the actual population demographics. For example, a black-white index of dissimilarity of 100 would represent complete apartheid because it would mean that every black person or every white person would need to move to a different neighborhood to create an even distribution of blacks and whites among the neighborhoods in the city. Usually, an index of dissimilarity of 60 is considered the threshold for severe segregation.

Of course the index of dissimilarity isn’t the only way to measure segregation, however. Another measure of segregation that might be easier to grasp is the index of exposure. The X-Y index of exposure represents for the average individual of group X how many people of group Y live in the same neighborhood. For example, a white-white index of exposure of 73% would indicate that in a given city, the average white person lives in a neighborhood that is 73% white. This would be a pretty significant degree of segregation. Oftentimes, the index of exposure and index of dissimilarity are positively correlated, and either one can be used as a fairly reliable indicator of segregation. Personally, I like the index of exposure because its meaning is more obvious, but both essentially measure the same thing.
Hopefully, this clears up some of the confusion. But if you have a question, feel free to ask!