Quantifying Neighborhood Racial and Ethnic Transition Clusters in Multiethnic Cities

Dynamic racial and ethnic transitions are a critically important aspect of urban neighborhood social geography and demography, which receive little attention from urban analysts compared to static racial and ethnic patterns. To determine whether cluster analysis could bring analytic power to the study of neighborhood racial and ethnic change in multigroup context, we performed a case study of Los Angeles County during the period 1990 to 2000, using the prediction strength technique to determine empirically, rather than arbitrarily, how many clusters fit the data. Clustering identifies which combinations of the direction and magnitude of groups' local changes drive local trends across the region and, equally important, those that do not. The case study supports the conclusion that cluster analysis serves as a powerful data-mining technique for local racial and ethnic trends, and will yield satisfactory results for any region at any scale.