ÿÈÕ³Ô¹Ï

This paper estimates a discrete choice model of housing product demand to study the causes of black urbanization. Our estimation strategy incorporates that there are unobserved product attributes which are correlated with observed product attributes. We bound racial differences in household willingness to pay for product attributes without implementing an instrumental variables strategy. Thus, we relax a number of assumptions implicit in "hedonic two step" housing research. Our primary explanation for excess black urbanization focuses on the disutility from commuting and the bundling of housing and labor markets.