The authors use a Bayesian Markov chain Monte Carlo algorithm to estimate a model that allows temporary gaps between a true expectational Phillips curve and the monetary authority’s approximating nonexpectational Phillips curve. A dynamic programming problem implies that the monetary...