Abstract

We develop and estimate a model of policy-making that incorporates social learning through sequential deliberation. In the model, committee members offer policy recommendations one after another, weighing their own information and biases against the views expressed by prior speakers. We apply this framework to quantify social learning in the U.S. Federal Open Market Committee (FOMC) between 1970 and 2008. The results show that earlier recommendations significantly influence the decisions of later speakers and, ultimately, aggregate monetary policy. Contrary to standard models of social learning, the FOMC’s sequential process enhances the quality of policy decisions relative to a benchmark in which members rely solely on their private information. Counterfactual simulations further indicate that the observed deliberation order within the FOMC leads to more effective monetary policy than alternative sequential decision-making mechanisms.
Download Paper (Last Version: August, 2025)
Download Appendix