New York Fed and San Francisco Fed Researchers: Substantial Improvements in U.S. Monetary Policy Occurred Only in the Last Thirty Years
Evaluating whether monetary policy has genuinely improved faces a fundamental difficulty: policymakers cannot choose the economic environment they face. A central banker may confront a severe banking panic, while another presides over a long period of stability. Even if the former makes higher-quality decisions, she may still experience higher inflation and unemployment simply because the shocks were larger and harder to stabilize. This confounding between policy quality and economic environment has long plagued cross-historical comparisons—was the Great Moderation the result of better policy or smaller, milder shocks? Should the early Federal Reserve be judged by the severity of the Great Depression, or by its effectiveness in handling unconventional challenges?
In a paper forthcoming in Econometrica, Regis Barnichon (Senior Research Advisor at the Federal Reserve Bank of San Francisco) and Geert Mesters (Economic Research Advisor at the Federal Reserve Bank of New York) propose a framework to disentangle these two factors. Their method does not require a fully specified structural model but instead uses empirically estimated impulse responses to compare the gap between actual policy paths and the optimal feasible response given specific shocks. This yields two complementary metrics: policy-path correction (the percentage adjustment required in the interest rate path to reach the optimum) and distance to minimum loss (the amount of macroeconomic loss—measured by expected squared deviations in inflation and unemployment—that could have been avoided under optimal policy).
Applying this framework to four periods—the Gold Standard era (1879–1912), the early Federal Reserve (1913–1941), the postwar period (1951–1984), and the post-Volcker era (1990–2019)—yields a core conclusion: meaningful policy improvements occurred only in the last thirty years. The first three periods exhibited large, statistically significant policy-path corrections, indicating systematic reaction functions far from optimal. Only in the post-Volcker era did these corrections shrink substantially and become statistically insignificant, with an average absolute correction of around 30 percent.
Improvements concentrated in two specific channels. Regarding responses to financial shocks: during the early Federal Reserve era, the Fed raised policy rates during banking panics. Because panics simultaneously suppressed inflation, real policy rates rose sharply, making policy contractionary precisely when the economy needed support; optimal policy would have lowered rates. In the post-Volcker era, the Fed already lowers rates following financial shocks, and estimated corrections are small and insignificant, resulting in negligible additional macroeconomic losses compared to the early Fed. Regarding responses to supply shocks: the postwar Fed reacted too weakly to energy price shocks, productivity shocks, and inflation-expectation shocks. Following an inflation-expectation shock, inflation rose steadily while policy rates responded only weakly, causing real interest rates to fall and reinforcing inflationary pressures; the Great Inflation was a reflection of this weak response.
Sources:Liberty Street Economics