The intensity of ethnic residential clustering: Exploring scale effects using local indicators of spatial association

Most analyses of ethnic residential segregation in cities rely on single-number indices that pay no attention to the degree of spatial clustering of the areas in which a group is either underrepresented or overrepresented. Recently, local statistical measures have been proposed as a set of approaches to overcome this deficiency. One such method-the Getis-Ord G *-is used to portray patterns of segregation in Auckland, New Zealand. That procedure requires analysts to make a number of judgments about the parameters of the statistics deployed, but also offers greater insights into scale effects in the measurement and delineation of segregation. We examine the information that can be derived from changing two of those parameters-the distance band used to define spatial proximity and the statistical significance of the local statistic-using data on two ethnic groups (Asians and Pacific Islanders) in Auckland. The procedure is also combined with a recently developed approach to measuring absolute segregation levels, thereby giving a fuller picture of both the extent of residential clustering and its intensity for those two groups.