A previous version of this working paper was originally published in February 2016.

Using a large-scale dataset on prices in the US retail market, we document that relative price dispersion accounts for about 30% of the variance of prices for the same good, in the same market, during the same week. Using a search-theoretic model of the retail market, we show that relative price dispersion can be rationalized as the equilibrium consequence of a pricing strategy used by sellers to discriminate between high-valuation buyers who need to make all of their purchases in one store, and low-valuation buyers who are able to purchase different items in different stores.