Abstract

Most standard economics models assume that financial incentives improve performance, while leading theories in psychology allow for the opposite. Experimental results are mixed and have not yet been corrected for publication bias and model uncertainty. We collect 2,193 estimates from 88 economics studies, along with 48 variables capturing the context in which the estimates were obtained. We apply recently developed techniques to correct for publication bias and use Bayesian and frequentist model averaging to account for model uncertainty. The corrected mean estimates are close to zero across most field experiment contexts, including differences in performance measurement, task definition, reward size, motivation beyond money, subject pool, and estimation technique. Framing as a loss and laboratory experiments produce statistically significant but small effects on average even after correcting for publication bias.

Fig: Heterogeneity in the effects of incentives

Bayesian model averaging


Reference: Cala Petr, Havranek Tomas, Irsova Zuzana, Luskova Martina, Matousek Jindrich, and Jiri Novak (2025), "Financial Incentives and Performance: A Meta-Analysis of Experiments in Economics." Charles University, Prague.