U.S. wage and price dynamics: a limited information approach

This paper analyzes the dynamics of prices and wages using a limited information approach to estimation. I estimate a two-equation model for the determination of prices and wages derived from an optimization-based dynamic model in which both goods and labor markets are monopolistically competitive; prices and wages can be reoptimized only at random intervals; and, when prices and wages are not reoptimized, they can be partially adjusted to previous-period aggregate inflation. The estimation procedure is a two-step minimum distance estimation that exploits the restrictions imposed by the model on a time-series representation of the data. In the first step, I estimate an unrestricted autoregressive representation of the variables of interest. In the second, I express the model solution as a constrained autoregressive representation of the data and define the distance between unconstrained and constrained representations as a function of the structural parameters that characterize the joint dynamics of inflation and labor share. This function summarizes the cross-equation restrictions between the model and the time-series representations of the data. I then estimate the parameters of interest by minimizing a quadratic function of that distance. I find that the estimated dynamics of prices and wages track actual dynamics quite well and that the estimated parameters are consistent with the observed length of nominal contracts.