A new application of Taylor rules: Model evaluation

Taylor rules posit a linear relationship between the output gap, inflation, and short-term nominal interest rates. Previous work has shown that the relationship between these key economic variables as captured by the Taylor rule is quite robust both across countries and monetary policy regimes. Consequently, the Taylor rule has become a useful characterization of monetary policy with much recent work focussed on the optimal formulation of the Taylor rule and the properties of equilibrium. Our interest in the Taylor rule is from a quite different perspective: we ask whether a calibrated monetary model can produce Taylor rule behavior similar to that seen in the data. That is, since the Taylor rule is a useful summary of the characteristics of a monetary economy, it seems reasonable to ask whether a monetary model, when calibrated to the data, produces a similar relationship. For our analysis, we employ a version of the limited participation model of Christiano, Eichenbaum, and Evans (1997) that permits both technology and money shocks. We find that this model, when the shock process is calibrated to US data, is able to replicate qualtitatively the correlation of interest rates with inflation implied by the Taylor rule but fails dramatically to match that between nominal interest rates and output.