European Regional Science Association
|
European Regional Science Association
|
|
|
|
|
The abstract for paper number 411:
Lawrence Brown, Su-Yeul Chung, Ohio State University, Columbus, Ohio, USA
Spatial Segregation, Segregation Indices, and the Geographic Perspective
What could be more inherently geographic than segregation. Yet, segregation indices are notoriously aspatial. This paper first illustrates this with regard to the most commonly used Dissimilarity Index. Using a simulation approach, it is shown that the index is incapable of distinguishing highly dissimilar patterns of segregation. Nevertheless, it is arguable that the index is useful as a summary measure by which different places could be compared. To illustrate, the dissimilarity indices for more than 300 metropolitan areas in the United States are subject to regression analysis, using variables that might seen to be related to different levels of segregation. While the results are successful statistically, they provide virtually no insight concerning the process(es) of segregation. Attention then turns to local area statistics such as the location quotient and Moran?s I applied to census tracts within a city. While these also are descriptive indices, they better provide measures of where segregation occurs, where it does not, and the nature of segregation. The result is a roadmap for a more in-depth investigation of the ground level reality of segregation. In this way, the Geography is returned to spatial segregation research.
Unfortunately full paper has not been submitted.