题目：Belief Disagreement, Instability, and Uncertainty under Stochastic Recall

主讲人：Pei Kuang，University of Macau

摘要：Using preregistered field studies, we show that even when individuals observe identical information, recall is highly heterogeneous across people and varies within the same person over time. Beliefs closely track recalled information and change even in the absence of new signals. Motivated by these findings, we develop a parsimonious model of stochastic recall in which agents store all observed signals but recall each past signal only probabilistically when forming beliefs, while updating optimally conditional on what they remember. This single friction generates cross-sectional disagreement, within-person belief instability, and heterogeneity in subjective uncertainty among ex ante identical individuals. The model yields sharp predictions for how disagreement, instability, and uncertainty vary with informational ambivalence, recall probability, and contextual cues. Large-scale preregistered experiments support these predictions, and survey evidence on macroeconomic and financial expectations is consistent with the model’s implications. Our results identify stochastic recall as a foundational informational friction in expectation formation.

主讲人简介：Pei Kuang is an Associate Professor at the University of Macau with a PhD in Economics from Goethe University Frankfurt. He previously served as Associate Professor and Co-lead of the Macro and Finance Research Group at the University of Birmingham. His research covers macroeconomics, monetary economics, behavioral and experimental economics, and political economics, with recent work focusing on beliefs, expectations, and their implications for behavior and policy. His work has appeared in the NBER Macroeconomics Annual, Journal of the European Economic Association, Journal of Monetary Economics, Journal of Public Economics, and other journals.