The Bigger They Are, The Harder They Fall: Retail Price Differences across U.S. Cities

Recent theoretical and empirical work has made headway in examining the nonlinear dynamics of relative prices. This paper seeks to extend this work by analyzing the behavior of retail goods prices across the U.S. First, a simple continuous-time model is used to explore the types of behavior that might characterize retail prices in different locations. The model allows for the existence of both fixed and variable frictions that cause similar goods to be spatially differentiated. These frictions could be interpreted narrowly as transport costs. However we prefer a broader interpretation in which the frictions operate at the level of technology and preferences. Second, we gather data from 24 U.S. cities to examine the actual behavior of retail goods prices. Our data are disaggregated to the level of individual goods, such as orange juice and toothpaste. Thus we avoid the problem of aggregation bias that afflicts price indices. The empirical analysis reveals that: (a) price discrepancies between U.S. cities are stationary, and indeed often behave in a manner consistent with the absolute law of one price; and (b) there is evidence of nonlinear reversion in the price discrepancies: large price disparities decay faster than small disparities.