During the past several years, academic research and popular media have drawn attention to the limited access to retail in many low-income urban neighborhoods, often referred to as “food deserts”.  Retail establishments in low-income areas are predominantly small, independently-owned “mom-and-pop” stores, while middle- and high-income neighborhoods tend to have fewer but larger stores, which are more likely to be part of regional or national chains.  The welfare implications of these differences in retail industry structure by neighborhood income are somewhat ambiguous.  Traditional economic development policy has often favored mom-and-pop businesses as a means of wealth building for business owners, while critics argue that these stores typically offer a narrower range of goods and lack the economies of scale – and potentially lower prices – of larger firms.

However, a recent trend is emerging that may alter the retail structure in urban neighborhoods.  Several notable Big Box chains – including Target, Wal-Mart, and Home Depot – have begun opening stores in central urban locations.  This spatial shift marks a fundamental change in the business strategy of Big Box stores, which have traditionally been designed around a suburban development model: large footprint single-story buildings surrounded by ample surface parking in close proximity to highways.  The particular design features of Big Box stores, and the necessity of large (and relatively cheap) land parcels would appear to make central cities undesirable locations, raising the question of why Big Boxes are moving downtown.  One hypothesis is that, although development costs are likely to be higher in central cities, Big Box stores perceive fewer competitors in the so-called “retail deserts” of central cities than in suburbs.  Alternatively, urban locations may offer access to more affluent consumers, who are willing to purchase higher-margin products.

In this paper, I will test these two hypotheses using detailed data on locations of Big Box and competitor stores in designated retail segments, as well as neighborhood demographic and economic characteristics, for large metropolitan areas in California.  Using the California National Establishment Time Series (NETS) data from 1992-2010, I can identify the exact location of specific Big Box firms and non-Big Box stores in competing retail categories, to test whether urban locations offer fewer direct competitors than suburban ones.  I also merge these data with Census data on market-area demographic and economic characteristics, such as population density, income, educational attainment and household structure, to determine whether urban locations offer access to different types of consumers.  A trend of urbanization among Big Box stores has important implications for the physical landscape of cities, for downtown retailers and for urban consumers.  Thus understanding the extent of and reasons behind this trend is important for policymakers and urban scholars.