Dynamic Models of Neighborhood Change

Elizabeth E. Bruch, University of California, Los Angeles

This paper uses an agent-based (microsimulation) model to examine the relationships among economic inequality, residential mobility, and race and income segregation. I estimate models of residential choice using data from the Los Angeles Family and Neighborhood Survey (LA FANS). These models describe the probability that a household with a given race and income moves into a housing unit, conditional on the unit's price and neighborhood composition. I then use a realistic agent-based model to relate the residential choice behaviors estimated from the LA FANS to patterns of neighborhood turnover. The agent-based model includes a market mechanism that allows housing prices to respond to changes in aggregate demand for neighborhoods. I simulate mobility behavior in Los Angeles County from 1990-2000 using the residential choice functions estimated from the LA FANS, and examine segregation outcomes under varying assumptions about economic inequality within and among race groups.