《ANewApproachtoMeasuringSocio-SpatialEconomic》-课件.pptVIP

A New Approach to Measuring Socio-Spatial Economic Segregation Sean F. Reardon Stanford University Glenn Firebaugh Pennsylvania State University David O’Sullivan University of Auckland Stephen Matthews Pennsylvania State University Goal of the paper To develop an approach to measuring spatial economic segregation To develop different measures of segregation along an ordinal dimension since income data are often reported using ordered categories To show how these measures can be adapted to take into account the spatial or social proximity of individuals Existing Measures of Economic Segregation Category-Based Measures of Economic Segregation Most common way of measuring income segregation has been to divide the population into two categories, based on some chosen income threshold Segregation between the two groups is then computed using some segregation measure (e.g. the dissimilarity index) that assumes the existence of only two groups Category-Based Measures of Economic Segregation Problems : - Dichotomizing the income distribution amounts to discarding a lot of information - The results of such an approach may depend on the choice of threshold Possible Solutions : - Compute the two-group segregation for all possible pairs of income categories, and then construct some summary measure of the pairwise indices - Compute segregation among multiple income category groups using for example the Theil information theory index of segregation Category-Based Measures of Economic Segregation But still unsolved issues: The Theil index measures segregation among a set of unordered groups (such as racial groups) and hence is insensitive to the ordinal nature of income segregation Such an approach remains sensitive to the number and location of the thresholds used to define income categories Existing Measures of Economic Segregation Variation-Ratio Based Measures of Economic Segregation An alternative approach defines segregation as a ratio of the between