Can we estimate macroforecasters’ mis-behavior?

We answer positively to this question by using Maximum Lq-Likelihood (or Deformed Likelihood) estimator. This is based on a parameter which measures the aggregate quote of judgment in the forecasting (game-based) system formed by three players—Forecaster, Policy Maker and Reality. For the first time in econometric literature, we apply this estimator to a dynamic system and derive a robust version of the Kalman Filter—the Deformed Kalman Filter (DKF). The evidence from U.S. data suggests that the judgmental dynamics exists and is correlated (but not coincident) with the phases of the Business Cycle. Furthermore its knowledge improves in-sample as well as out-of-sample estimation.