Money and Price Dispersion in the United States

Abstract

This paper reports an empirical test of a price dispersion equation, using data on the U.S. after World War II. The equation, derived elsewhere from aversion of the partial information-localized market models, relates price dispersion to the magnitude of changes in the aggregate disturbances. In order to test the model a series on price dispersion is computed using annual wholesale price indexes for the period 1948-76. The data on money shocks are the unanticipated money growth series estimated by Barro. The tests also include a measure of aggregate-real disturbances. From the theoretical point of view the results are negative. They reject the hypothesis that unexpected money shocks, as measured by Barro, affect price dispersion in the way predicted by the model. In a previous paper, a similar model was tested with data from the German hyperinflation and found supported to a considerable extent. The difference in the results may be related to the extreme magnitude of the monetary disturbances during that period, and to the apparently important effect of unincluded relative disturbances in the United States.