THE FREQUENCY OF PRICE ADJUSTMENT AND NEW KEYNESIAN BUSINESS CYCLE DYNAMICS

The Calvo pricing model that lies at the heart of many New Keynesian business cycle models has been roundly criticized for being inconsistent both with time series data on inflation and with micro-data on the frequency of price changes. In this paper I develop a new pricing model whose structure can be interpreted in terms of menu costs and infor- mation gathering/processing costs, that usefully addresses both criticisms. The resulting Phillips curve encompasses the partial-indexation model, the full-indexation model, and the Calvo model, and can speak to micro-data in ways that these models cannot. Taking the Phillips curve to the data, I ?nd that the share of ?rms that change prices each quar- ter is about 60 percent and, reflecting the importance of information gathering/processing costs, that most ?rms that change prices use indexation. Exploiting an isomorphism result, I show that these values are consistent with estimates implied by the partial-indexation model.