An evolutionary approach to residential segregation

Residential segregation between blacks and whites remains high even after many years of fair housing legislation. This dissertation combines recently developed techniques in evolutionary game theory and agent-based simulation to study the dynamics of residential segregation.; After an introduction in Chapter 1, Chapter 2 gives a brief literature review. We then, in Chapter 3, argue that racial income disparity is not able to account for the observed high level of segregation. First, blacks and whites who earn the same income are found to be highly segregated. Second, the level of segregation that can be sustained solely by income disparity is significantly below the actual level of observed segregation.; Chapter 4 introduces a variation of the Schelling model that incorporates economic variables, and whose dynamics can be rigorously analyzed. Extensive survey data has shown that most whites prefer to live close to other whites, while blacks are not as demanding for like-color neighbors. We show that generally this is enough to generate residential segregation as individuals move in response to their environments. Agent-based simulation is employed to illustrate the evolutionary dynamics of such a process. The asymmetric setting of the model has an interesting implication for housing prices and vacancies, for which there is some empirical evidence.; Chapter 5 extends Schelling's original “Checkerboard model.” In studying housing segregation, researchers find that many individuals do not mind living in mixed-race neighborhoods. In fact, as the General Social Survey revealed, a majority of blacks and a fair number of whites prefer half-black, half-white neighborhoods. Such neighborhoods are, however, rarely found in urban America. Following Schelling's pioneering work, we construct an evolutionary model of residential choice and characterize its long-run dynamic properties. It shows that, even when every agent prefers to live in a half black, half-white neighborhood, the society will end up in a segregated pattern in the long run. The model puts on a rigorous footing Schelling's insight that segregation is persistent not necessarily because people like it, but due to the fact that people dislike to be isolated in a neighborhood of different color. The last chapter summarizes the main findings.