Monetary Policy under Model and Data-Parameter Uncertainty

Policy-makers in the United States over the past 15 to 20 years seem to have been cautious in setting policy: empirical estimates of monetary policy rules such as Taylor's (1993) rule are much less aggressive than those derived from optimizing models. The author analyzes the effect of an aversion to model and data-parameter uncertainty on monetary policy. Model uncertainty arises because a central bank finds three competing models of the economy to be plausible. Data uncertainty arises because real-time data are noisy estimates of the true data. The central bank explicitly models the measurement-error processes for both inflation and the output gap, and it acknowledges that it may not know the parameters of those processes precisely (which leads to data-parameter uncertainty). The central bank chooses policy according to a Taylor rule in a framework that allows an aversion to the distinct risk associated with multiple models and dataparameter configurations. The author finds that, if the central bank cares strongly enough about stabilizing the output gap, this aversion generates significant declines in the coefficients of the Taylor rule, even if the bank's loss function assigns little weight to reducing interest rate variability. He also finds that an aversion to model and data-parameter uncertainty can yield an optimal Taylor rule that matches the empirical Taylor rule. Under some conditions, a small degree of aversion is enough to match the historical rule.