Substitutes or complements? The effect of opening a food store on customer visits to neighborhood food stores

Abstract

This study examines the effects of new food store openings on customer traffic to incumbent stores using a modified difference-in-differences approach and cellphone location data from the contiguous United States in 2019. Analysis of both the intensive margin (total visit counts) and the extensive margin (unique visitors) reveals substantial spatial, temporal, and demographic heterogeneity, highlighting that ignoring these variations may lead to misleading conclusions about customer behavior. Grocery store openings, for example, reduce visits to food-at-home incumbents in close proximity but affect food-away-from-home incumbents over broader distances. The magnitude of these effects varies with demographics. Entrant stores offering fresh food options, such as grocery stores, reduce traffic to outlets that sell more processed foods—such as fast-food and convenience stores. However, the reductions remain modest—generally below 5% of monthly visits—and are statistically significant for White and high-income visitors when disaggregated by demographics. Complementarity effects emerge when entrants and incumbents are not direct competitors, particularly within a one-mile range. For instance, grocery entry does not reduce traffic to fast-food stores nearby but does so at greater distances. Conversely, competition within similar store formats intensifies business-stealing effects, such as dollar store openings reducing traffic to other smaller store formats while leaving grocery stores unaffected. These findings underscore the competitive dynamics shaping the U.S. food environment.