Is the "information effect" of monetary policy quantitatively important? We first use a simple model to show that under asymmetric information, monetary policy surprises are correlated with the unobserved state of the economy. This correlation implies that monetary policy surprises provide information about the state of the economy, and at the same time, explains why the estimation of the information effect may be biased. We then develop a New Keynesian DSGE model under asymmetric information and calibrate model parameters to match macroeconomic dynamics in the US and forecasting accuracy in the Greenbook. Under our calibration, both the central bank and the private sector initially have noisy information. Over time, the information effect of monetary policy mitigates information frictions by enhancing the two-way learning between the central bank and the private sector.
