Border, Border, Wide and Far, How We Wonder What You Are

This paper exploits a three-dimensional panel data set of prices on 27 traded goods, over 88 quarters, across 96 cities in the U.S. and Japan. We present evidence that the distribution of intra-national real exchange rates is substantially less volatile and on average closer to zero, than the comparable distribution for international relative prices. We also show that an equally-weighted average of good-level real exchange rates tracks the nominal exchange rate well, suggesting strong evidence of sticky prices. We turn next to economic explanations for the dynamics of this so-called "Border" effect. Focusing on dispersion in prices between city pairs, we confirm previous findings that crossing national borders adds significantly to price dispersion. Using our point estimates crossing the U.S.-Japan "Border" is equivalent to adding between 2.5 and 13 million miles to the cross-country volatility of relative prices. We make a direct and explicit inference on the influence of shipping costs, distance, exchange rate and relative wage variability on the "Border" effect. In our calculations, the "Border" effect disappears after controlling for these additional variables.