A Neighborhood-Level View of Riots, Property Values, and Population Loss: Cleveland 1950-1980

Abstract

We undertake a case study of riots in the context of Cleveland's economic decline between 1950 and 1980. Our empirical perspective emphasizes differential changes in property values and population levels across census tracts depending on their proximity to the riots' epicenter. We find patterns that are consistent with concentrated, negative, and long-lasting effects from the 1960s riots. These estimates do not depend on whether we use a narrow or a broad categorization for "riot tracts", whether we use simple difference-in-difference measures or detailed information on the distance of each tract from the riot center, or whether we use ordinary least squares or matching estimation techniques. Moreover, the negative relationship between riots and property value trends is not merely a reflection of the pre-existing trend in value, the pre-riot racial composition of the neighborhoods, the pre-riot proportion of neighborhood residents holding manufacturing jobs, the neighborhood crime rate, nor changes in the observable characteristics of the housing stock. Cleveland¬πs economic difficulties did not start with the riots. Rather, we suggest that the impact of the riots was compounded by long-run forces that were already eroding Cleveland¬πs economic base.