Neighborhood Information and Home Mortgage Lending

In this paper, we empirically examine how information about a neighborhood affects the level of lending activity in it. Specifically, do lenders deny mortgage applications at higher rates in neighborhoods where they have little experience in evaluating applications, and/or in neighborhoods where the lending community in general has little experience? The analysis uses data collected under the Home Mortgage Disclosure Act (HMDA) for 1990 and 1991 to construct denial rates for each lender in each census tract, controlling for applicant characteristics observed in the HMDA data. We then estimate the relationship between these lender-tract denial rates and both the number of applications processed by the lender in that neighborhood and the number of applications processed by all lenders in that neighborhood, controlling for other characteristics of the census tract and for the lender.

We find that the more applications a lender processes from a given neighborhood, the lower the neighborhood-lender denial rates -- both statistically and economically. Furthermore, we find that the low number of applications taken by individual lenders from specific low-income and minority neighborhoods does contribute to the relatively high denial rates in these neighborhoods. These findings are consistent with recent theoretical models linking redlining to incomplete information.