Prejudice versus Preference: What do we Really Know about Housing Market Discrimination?

A persistent feature of urban regions in the United States has been the marked degree of racial segregation in residential areas which effectively encloses nonwhite households in central city ghettos. It is widely believed that only a fraction of this segregation can be attributed to either racial economic differentials or to voluntary actions, although the evidence on the latter point is far from conclusive. A few authors have tried to provide alternative explanations for the observed interracial housing price differentials which revolve around either disequilibrium or higher cost of operation arguments. This paper, on the other hand, will provide a new skeptical counterpoint to the claim of discrimination by attempting to demonstrate that the existing econometric specifications cannot conclusively identify housing discrimination without recourse to arbitrary and often implausible assumptions concerning households' preferences for neighborhood racial composition and other components of the housing package. While it is not claimed that existing studies yield no inferences about discrimination, the potentially-biasing flaws of the specifications cannot be overlooked in any comprehensive consideration of the topic.