Optimal versus realized policy rules in a regime-switching framework

In this paper we compare a deterministic model and a Markov switching model to analyze the behavior of the US economy and the Federal Reserve. We examine both optimal and empirical monetary policies for the US Federal Reserve between 1960 and 2008. We compare the optimal monetary policy to the actual interest rates and to the empirical reaction function. We also evaluate the sensitivity of the results to the preferences assigned to each objective. We find that there is no unique optimal solution that fits the Federal Reserve behavior over the entire period. The best fit to the actual interest rates is obtained by an optimal policy with preference switches following the rule: a high-volatility regime coincides with a priority on inflation alone while in a low-volatility regime there is equal policy priority on output stabilization and inflation.