Inferring monetary policy objectives with a partially observed state

Accounting for the uncertainty inherent in real-time perceptions of the state of the economy is believed to be critical for the analysis of historical monetary policy. We investigate this claim through the lens of a small-scale new-Keynesian model with optimal discretionary policy and partial information about the state. The model is estimated using maximum likelihood on US data over the Volcker-Greenspan-Bernanke regime. A comparison of our estimates to those from a version of the model with complete information reveals that under partial information: (i) the Federal Reserve demonstrates a significant concern for stabilizing fluctuations in the output gap, and (ii) the discrepancy between optimal and observed policy behavior is smaller.