| 讲座简介： | Agents arriving sequentially may adopt a new technology. Early adoption reveals information about its value, but individuals tend to wait and free-ride on information generated by others. Facing this challenge, we study how a designer—who controls the public release of past information—can improve social learning efficiency. The optimal design follows a threshold stopping rule, which keeps recommending adoption until the designer’s belief becomes more pessimistic than a time-varying threshold. This threshold typically starts above the first-best threshold, drops below it in some later periods, and eventually returns to it. Meanwhile, public information flow is restrained by over-recommending adoption, but not by under-recommending. We also examine special cases of conclusive-news learning. In good-news environments, the optimal design terminates exploration in a random manner; in bad-news environments, adoption may continue even after bad news arrives. |
