How optimal is US monetary policy?

Abstract

Using a small-scale microfounded DSGE model with Markov switching in shock variances and policy parameters, we show that the data-preferred description of US monetary policy is a time-consistent targeting rule with a marked increase in conservatism after the 1970s. However, the Fed lost its conservatism temporarily in the aftermath of the 1987 stock market crash, and again following the 2000 dot-com crash and has not subsequently regained it. The high inflation of the 1970s would have been avoided had the Fed been able to commit, even without the appointment of Paul Volcker or the reduction in shock volatilities.