Reject & resubmit at Quarterly Journal of Economics
This paper explores the impact of the learning environment on how people react to information. We develop a model of belief-updating where people respond to complexity by forming a simplified representation of the environment via salience-driven channeled attention, then process information using Bayes’ rule subject to cognitive imprecision. The model predicts overreaction when environments are complex, signals are noisy, information is surprising, or priors are concentrated on less salient states; it predicts underreaction when environments are simple, signals are precise, information is expected, or priors are concentrated on salient states. Results from a series of pre-registered experiments provide support for these predictions and evidence for the proposed cognitive mechanisms. Our model is highly complete in capturing explainable variation in belief-updating; the interaction between the two psychological mechanisms is critical to explaining belief data. These results connect disparate findings in prior work: underreaction is typically found in laboratory studies, which feature simple learning settings, while overreaction is more prevalent in financial markets with more complex environments.