Searching under the influence? Explaining beer price dispersion in New York City

Abstract. We use a data set comprising 12,565 price observations of 2,130 beers sold at 462 venues in Manhattan and Brooklyn to investigate the extent of beer price dispersion across the bars of New York City. We combine these with data about bar characteristics, geographic concentration and ?touristiness? to explain both residual and observed price variation. While a substantial fraction of price variation is explained by differences in bar and beer characteristics, beer price dispersion in New York City persists even after accounting for these drivers of product differentiation. Neighborhoods with a higher geographic concentration of bars feature higher rather than lower average residual price dispersion, although variations in beer-specific local competitive density do explain some of these price differences. Beers served at bars that cater primarily to tourists have higher residual price dispersion than those served in ?neighborhood bars? whose clientele is primarily local. These findings present new evidence of price dispersion in a highly competitive and understudied market, and suggest important subtleties in the interaction between consumer search, product differentiation, geographic concentration and market segmentation that are not present in theoretical models of price variation for homogeneous products or in existing empirical studies of price dispersion with differentiated sellers.