How New Keynesian Is the US Phillips Curve?

Abstract

I provide a generalization of Calvo price setting, to include non-overlapping contracts as a special case and embed this in a small DSGE model. The resulting Generalized Phillips Curve (GPC) nests New-Keynesian and Neoclassical versions. I linearize the model around a potentially non-zero trend inflation rate, and estimate it on US data using Bayesian methods, allowing for Markov switching in the variances of structural shocks. I find that the Phillips curve is 100% New Keynesian. There is no evidence of either forward or backward indexation. I illustrate that trend inflation affects the estimation of the Phillips curve.