Poverty, Place or Race: Causes of the Retail Gap in Smaller U.S. Cities

Abstract

In this study we explore the possibility that any or all of three dimensions of the U.S. inner city, central location, relatively low income, and a relatively high percentage of minority residents, correlate with a relative lack of neighborhood retail services, the so-called retail gap. Our empirical analysis is based on zip code level data for 39 U.S. cities with populations from 50 to 225 thousand people. After specifying an empirical model based on hypotheses drawn from the urban economics, urban studies, and urban development literature, we test access to retail services using both geographical density and per capita retail measures. While several results from this study are worthy of discussion, our primary empirical finding is that neighborhoods with high percentages of African-Americans are systematically under-served by retail, all else equal, while Latino, low income, or otherwise centrally located neighborhoods are not. We then test a selection of product categories for retail density, finding some differing results by product category but verifying our more general conclusions for most products. Finally, we find that grocery stores have significantly smaller scale in African-American communities, a finding that provides partial support for the urban “food desert” hypothesis. These findings suggest that the inner city retail gap may be more of a racial than geographic issue, and that increased emphasis on racial composition is warranted in the retail development policies applied to smaller U.S. cities.