Evaluating Rules in the Fed’s Report and Measuring Discretion

In recent Monetary Policy Reports the Federal Reserve has been reporting and discussing specific monetary policy rules. This paper evaluates the monetary policy rules reported by Fed in these Reports. We define the rules in Section 2—three from the Report and one alternative. In Section 3 we calculate the performance of these rules in seven well-known macroeconomic models—a small new Keynesian model, a small old Keynesian model, a larger policy-oriented model, and four other models from the Macro Model Data Base. Each model has an optimal rule, that by construction outperforms any of these simple modelindependent rules. But good rules should not produce outcomes substantially worse than modelspecific optimal rules. In Section 4 we compute optimal policies and the inflation-output variance tradeoff curves for two models, and we compare the economic performance under model-specific optimal policies with performance under the 4 policy rules. Central banks value discretion, which we model as residuals or deviations from policy rules. Such deviations can in principle improve performance, by reacting to variables and shocks excluded from simple rules. If they do, then the actual volatility of inflation and output, produced by rule plus deviation, will be less than the volatility produced by the rule without deviations. To this end, in Section 5 we compare the interest rate prescriptions that result from these rules with the actual path of the federal funds rate. This paper was prepared for presentation at the conference “Strategies for Monetary Policy” held at the Hoover Institution at Stanford University on May 3, 2019. We thank Livio Cuzzi Maya, Lazar Milivojevic and Balint Tatar for research assistance. The research was supported by a grant from the Sloan Foundation.