Money and the Dispersion of Relative Prices
NBER Working Paper No. 431
A price dispersion equation is tested with data from the German hyper-inflation. The equation is derived from a version of Lucas' (1973) and Barro's (1976) partial information-localized market models. In this extension, different excess demand elasticities across commodities imply a testable dispersion equation, in which the explanatory variable is the magnitude of the unperceived money growth. The testing of this hypothesis requires two preliminary steps. First, a price dispersion series is computed using an interesting set of data. It consists of monthly average wholesale prices of 68 commodities ranging from foods to metals, for the period of January, 1921 to July, 1923. The next step is the delicate one of measuring unperceived money growth. This estimation implies the postulation of an available information set and also a function relating the variables in this set to money creation. The function used was based on considerations related to government demand for revenue. The model receives support from the empirical analysis although it is evident that unincluded variables have important effects on price dispersion.
Published: Hercowitz, Zvi, 1981. "Money and the Dispersion of Relative Prices," Journal of Political Economy, University of Chicago Press, vol. 89(2), pages 328-56, April.