Uncertainty of Household Inflation Expectations: Reconciling Point and Density Forecasts

We examine the uncertainty of household inflation expectations using matched point and density forecasts from the New York Fed's Survey of Consumer Expectations. We argue that using in- formation from both types of forecasts allows for better estimates of uncertainty. Since the two types of forecasts may be inconsistent, we propose to reconcile them by matching the mean (or the median) of individual density forecasts and the corresponding point forecasts using exponential tilting. The reconciled densities provide uncertainty measures that are strictly consistent with the point forecasts by construction. We compare the uncertainty of inflation expectations derived from the reconciled densities with that derived from the original densities. Our results suggest that, at the micro-level, the uncertainty of consistent forecasts tends to be lower after reconciliation, while that of inconsistent forecasts tends to be higher. Aggregate uncertainty measured by averaging individual uncertainty is likely underestimated when using the survey responses directly, without reconciliation. This study contributes to the literature on the measurement of uncertainty and provides insights into the interplay of matched point and density forecasts in this context.