Incentivizing Inflation Expectations

Abstract: Inflation expectations are central to macroeconomics, yet surveys elicit them without marginal incentives for accuracy. In a multi-wave representative survey experiment, incentives reshape reported expectations: they reduce upward bias by 2.3–3.4 percentage points, disagreement by one-third, close the gender gap, and double learning rates in an RCT – all without increasing participant remuneration. Incentives raise effort and attention, not search or strategic behavior, and make expectations more informative about consumption plans. Because these effects are larger when attention to inflation is low, unincentivized elicitation generates state-dependent measurement error, with important implications for empirical research, model calibration, and monetary policy.