"The Bigger They Are, The Harder They Fall": How Price Differences Across U.S. Cities Are Arbitraged

Recent empirical work has made headway in exploring the non-linear dynamics of deviations from the law of one price and" purchasing power parity that are apt to arise from transaction costs. However, there are two important facets of this work that need improvement. First, the choice of empirical specification is arbitrary. Second, the data used are typically composite price indices which are subject to potentially serious aggregation biases. This paper examines the evidence for transport-cost-induced nonlinear price behavior within the U.S. We address both of the above shortcomings. First, we use a simple continuous-time model to inform the choice of empirical specification. The model indicates that the behavior of deviations from price parity depends on the relative importance of fixed and variable transport costs. Second, we employ data on disaggregated commodity prices, yielding a pure' measure of the deviations from price parity. We find strong evidence of nonlinear reversion in these deviations. The nature of this reversion suggests that fixed costs of transportation are integral to an understanding of law-of-one-price deviations.