Abstract:
This paper examines the lag distribution relating wholesale to consumer price changes. Sims' causality test indicates a one-sided lag structure. Following Hatanaka and Wallace, parameters of the lag distribution which can be estimated with relatively high precision are emphasized. Thus, our concern is with the sum of coefficients and the first four moments of the distribution. Short-run effects are estimated from lag moments using Pearson's method for equating moments. As a smoothness prior, a Beta distribution in the lag weights is suggested. Tests for bias due to missing components in the wholesale price index indicate little is lost because of misspecification.
