Regis Barnichon, Federal Reserve Bank of San Francisco
How should we evaluate and compare the performances of policy institutions? An evaluation based on realized outcomes is flawed, since different institutions may face different economic environments and different economic or geopolitical shocks. In this work, we show that it is possible to quantitatively evaluate and rank policy performance using only two estimable statistics: (i) the impulse responses of the policy objectives to policy shocks, and (ii) the impulse responses of the same policy objectives to non-policy shocks, e.g., aggregate demand shocks, energy price shocks, productivity shocks, or war shocks. For a large class of models, the correlation between these two sets of impulse responses directly captures the performance of the policy institution: A correlation of zero indicates best performance ---the policy institution could not have reacted any better to the shocks that affected the economy---, while a correlation of one (in absolute value) indicates worst performance ---the institution could have (but did not) perfectly met its objectives by undoing the effects of the non-policy shocks---. We use our methodology to evaluate US monetary policy over the past 150 years; from the Gold standard period, the early Fed years and the Great Depression, to the post World War II period, and the post-Volcker regime.