The Taylor Rule with real-time data and an analysis of its impact on historical policy analysis

The Federal Reserve system (the Fed) is the United States monetary policy authority and is mandated by Congress to pursue two goals: 1) maximum sustainable output and employment and 2) stable prices. Among the actions the Fed can take to achieve these goals is the ability to set the target federal funds rate. Simple policy rules that set the level of funds rate in response to changes in economic variables have gained attention as a means to more effective monetary policy. One of the most researched and cited rules is that proposed by economist John Taylor in 1993. Taylor’s rule is simple and intuitive, and it was found to be surprisingly accurate during the period from 1987 to 1992. However, this analysis is based on impractical assumptions about the amount of accurate data available to a policymaker at the moment of his or her decision. I explore the differences in policy rules using real-time data – that is, data available to a policymaker at the moment of policy decision, versus ex-post data – that is, data that has been fully revised and is accepted as the most accurate representation of an economic variable. Within this analysis, I also evaluate differences in policy rules between Fed Chairmen. While analysis using ex-post data by Judd and Rudebusch (1998) has found statistically significant differences in reaction functions between Fed Chairmen, I find that not only does a real-time Taylor rule recommend different levels of the federal funds rate than an ex-post Taylor rule, but also Fed Chairman is not necessarily a determinant of structural change in policy formation. I conclude that monetary policy rules in real-time seem to describe “eras” of economic events and recommendations generated within these “eras” do not differ from actual observations as much as Taylor found using his rule.