Mark Fossett's new research, published here in the Journal of Mathematical Sociology, is arguably the most important advance in studies of residential segregation in the past decade. While this study of the role of preferences in creating the patterns of residential separation does not answer all the questions about how preferences create separation in the residential mosaic, it provides a major extension of Schelling's seminal work of three decades ago. The paper shows clearly that preferences do matter and that the set of simulations leave little doubt that residential preferences and their underlying social dynamics have the capacity to generate high levels of ethnic segregation. The agent-based modeling technique, on which the results are based, is a major advance on previouswork using agent-based modeling and will set the standard for further studies of the underlying processes that create residential separation in U.S. cities.
Keywords:
Notes
1Fossett provides a brief history of the project and the documentation of validation tests by programmers funded through NIH.
2The code for a simple formulation (useful for classroom teaching) and a more complicated formulation of SimSeg is available on the Web at http://www.simseg.com