Social Area Analysis: Some Theoretical and Methodological Comments Illustrated with Australian Data

One criticism of social area analysis has been that, while its underlying theory is mainly concerned with understanding social differentiation in modern industrial society, the main empirical concern among social area analysts has been with a problem of a different, although related, order: the residential differentiation of socially differentiated groups. The first part of this paper is concerned with a discussion of the general processes through which the three constructs of social rank, urbanization, and segregation come to have a spatial referent. The second part of the paper reports an empirical test of the Shevky-Bell schema using 1961 census data for Melbourne, Australia. A two-stage component analysis is used to test for the existence of the three hypothesized dimensions of residential differentiation. A more efficient method of classification is used to derive a typology of residential areas, using two component scores for each area. Although some theoretical criticisms and methodological modifications are suggested, the findings tend generally to support the Shevky-Bell approach.