Performance of Operational Policy Rules in an Estimated Semi-Classical Structural Model

Abstract

This paper reports results of simulation exercises that explore several questions relating to the design of rules for monetary policy. Emphasis is given to issues raised by the concept of rule operationality, i.e., reliance on feasible instrument variables and information sets. Many of the results pertain to rules of the Taylor type -- i.e., with an interest rate instrument set in response to inflation and output-gap measures -- but some are reported for rules using a nominal income target and/or a monetary base instrument. The macroeconomic model utilized is small in scale but features a specification designed to represent rational dynamic optimizing choices by the economy's private agents. Saving and portfolio-balance behavior are expressed by optimizing versions of exceptional IS and LM functions, with gradual price adjustments specified differently in two variants of the model. One variant uses the well-known Calvo-Rotemberg price adjustment relation, whereas the second employs a newly-rationalized version of the Mussa-McCallum-Barro-Grossman P-bar model. Parameter values are estimated by instrumental variables on U.S. quarterly data for 1995-1996.