Implications of the Zero Bound on Interest Rates for the Design of Monetary Policy Rules

In the presence of nominal rigidities, monetary policy can potentially improve welfare by reducing the magnitude of short-run fluctuations in inflation and resource utilization. According to a standard view of the monetary transmission mechanism, monetary policy stimulates economic activity by reducing the short-term real rate of interest. The zero lower bound on nominal interest rates, however, can constrain the ability to lower real rates. In this paper, we use the FRB/US macroeconometric model to quantify the effects of the zero bound on macroeconomic stabilization and to explore formulations of policy that minimize these effects. We draw three broad conclusions from this research. First, during particularly severe contractions, monetary policy alone is insufficient to restore macroeconomic equilibrium; fiscal policy or some other autonomous stimulus is also needed. Second, the one-sided nature of the zero bound implies that a standard linear policy rule, such as the Taylor rule, yields an upward bias to interest rates and hence a downward bias in the rate of inflation. We find that this bias can be easily overcome by introducing an offsetting downward adjustment to the policy rule. Finally, in a low-inflation environment, where policy follows a Taylor rule, the constraint on policy from the zero bound leads to a significant increase in the volatility of resource utilization and, to a lesser extent, inflation. Augmenting the Taylor rule to incorporate a response to past and expected future constraints on policy dramatically reduces the detrimental effects of the zero bound. Interestingly, policy rules that are efficient in the absence of the zero bound, implicitly incorporate such behavior; hence, for such rules, the zero bound generates only relatively small stabilization costs.