Heterogeneous beliefs and the Phillips curve

Heterogeneous beliefs modify the New Keynesian Phillips curve by introducing a term in the cross-section distribution of expectations. To take that model to the data, we develop a novel functional data approach to estimation and inference that accounts for variation in distributions of expectations. We find that this variation may be summarized using a handful of functional factors, and demonstrate their statistical and economic relevance for inflation dynamics. Our results are among the first to highlight the potential benefits to be gained in empirical work from a rigorous treatment of diverse beliefs in the study of macroeconomic outcomes.