Price Rigidity and Price Dispersion: Evidence from Micro Data

We use large unpublished data set about the prices by store of 381 products collected by the Israeli bureau of statistics during 1991-92 in the process of computing the CPI. On average 24% of the stores changed their price where the average is over products and months. Using the standard calculation this would imply that on average prices remain unchanged for 4.1 months. We argue that the standard calculation suffers from a large aggregation bias due to Jensen's inequality and our best estimate suggests that prices remain unchanged on average for more than 7.5 months. We then assess the importance of price rigidity in generating price dispersion. We find no evidence that price rigidity as measured by the frequency of nominal price changes is related to price dispersion. We also find no evidence that a shock to the inflation rate increases price dispersion. These findings are not consistent with standard versions of the staggered price setting model but are roughly consistent with a simple version of the uncertain and sequential trade model.