Some Issues in the Empirical Specification of Federal Reserve Reaction Rules

Abstract

Simple estimated monetary policy reaction rules of the Taylor rule type have recently become a popular way to describe and interpret monetary policy. Empirical implementation of such rules always involves a partial adjustment component which captures inertia in the Federal funds rate. Looking at the period since 1980, in many ways a single monetary policy regime, I show that the inertial component of monetary policy is larger in the 1990s than in the 1980s. This means that the estimated short-run response of the Fed to inflation is quite small, raising questions about the credibility with the public of such a rule. Furthermore, the adjustment parameter is quite close to its limiting value of 1, creating instability in the estimated long-run responses. The increased inertia may be related to the Fed's concern about the impact of its decisions on financial markets.