Racial Segregation and Discrimination: Evidence from the Rental Housing Market
Paul E. Carrillo
George Washington University - Department of Economics
Dirk W. Early
Southwestern University - Department of Economics and Business
Edgar O. Olsen
University of Virginia - Department of Economics
November 29, 2010
46th Annual AREUEA Conference Paper
Rental housing discrimination can take many forms. The most visible potential result of discrimination is the segregation of neighborhoods by race and ethnicity. The more subtle form of discrimination examined in this study is to charge different prices to different groups. This form of discrimination is possible if, among other things, owners of housing units believe unobserved factors important to the return on their investment are correlated with race or ethnicity, or if landlords have a taste for discrimination as described by Becker (1957). Regardless of the reason why landlords desire to charge different prices to different groups, theory suggests that these differences will be more pronounced in tighter housing markets, where the cost of discrimination to the landlord is lower. The study proposed here is an extension of the use of hedonic methods to detect discrimination using a much richer set of data than what has been available to previous authors (Follain and Malpezzi (1981), Kiel and Zabel (1995), and Myers (2004) for example) to answer three policy relevant questions related to the variations in access to housing across race and ethnicity: 1) Do minorities pay more for equal quality housing to live in majority neighborhoods? 2) Do minorities pay more for equal quality housing to live in areas with less concentrations of poverty? and 3) Does the tightness of the housing market effect the ability of landlords to charge different rents for equal quality housing based on race and ethnicity? The last question appears to have been largely ignored in previous empirical studies of housing discrimination. We make a considerable effort to include a comprehensive set of covariates in the hedonic equation that, due to data constraints, were not considered in previous studies. First, our primary source of data includes over 450,000 observations on rental housing units with more than 70 detailed questions about the units type, condition, and neighborhood attributes. Second, these data can be linked with data from the 2000 Decennial Census providing additional detailed characteristics about the neighborhood. Finally, information about credit history (at the census tract level) has been provided by Equifax, one of the three national credit reporting agencies. Although it is aggregated at the census tract level, the credit data is notably rich.
JEL Classification: J1
Date posted: November 29, 2010
© 2015 Social Science Electronic Publishing, Inc. All Rights Reserved.
This page was processed by apollo1 in 0.312 seconds