Survey Expectations, Adaptive Learning and Inflation Dynamics

The use of survey information on inflation expectations as an observable in a DSGE model can substantially refine identification of the shocks that drive inflation. Optimal integration of the survey information improves the model forecast for inflation and for other macroeconomic variables. Models with expectations based on an Adaptive Learning setup can exploit survey information more efficiently than their Rational Expectations counterparts. The resulting time-variation in the perceived inflation target, in inflation persistence and in the sensitivity of inflation to various shocks provide a rich and consistent description of the joint dynamics of realized and expected inflation. Our framework produces a reasonable interpretation of the post-Covid inflation dynamics. Our learning model successfully identifies the more persistent nature of the recent inflation surge.