Measuring urban segregation based on individuals' daily activity patterns: A multidimensional approach

This paper develops a methodology to measure urban segregation based on individuals' sociospatial experience of daily life. Since segregation can be considered as the isolation of people from those unlike themselves, its degree increases with the similarity in ethnicity, economic status, or other sociodemographic dimensions of interest between individuals and people who they are exposed to in their daily usage of urban space. Based on this perspective, we propose a regression estimator that measures segregation by assessing similarity or likeness between people and the social environments they experience in daily activity spaces. Compared to traditional segregation measures, the proposed estimator is not restricted to measuring residential segregation, but recognizes and assesses segregation as a dynamic process that unfolds in the daily life routines of individuals in a society and depends on the different ways individuals or social groups use urban space. It can be applied to various segregation factors, categorical or continuous, as well as to examine their interactions in a society. An empirical study in Hong Kong is used to demonstrate the proposed approach.