Measuring Segregation and Controlling for Independent Variables
This paper suggests a procedure to control for independent variables in the measurement of segregation by linking the well known Index of Dissimilarity and the Multinomial Logit Model. While the first one may be considered as a standard macro measure of inequality, the latter one is a typical tool to analyze the determinants of individual behavior and attainment. Combining both enables a judgement or a comparison of inequality structures taking into account respective distributions of relevant influential variables. After a short review of the debates on segregation indexes, the technique is described in detail. It is further illustrated using an example dealing with the assimilation by family types of foreigners in the Federal Republic of Germany. It is demonstrated that the method may be a helpful tool to come to more adequate judgements concerning the development of inequality structures.