Sequential Learning

In this Department of Economics Seminar, Yair Antler, from Tel Aviv discusses his paper on Sequential Learning.

 Abstract

Two players examine the quality of a joint project sequentially and privately. Launching the project requires mutual consent; however, the players disagree on the value of the project, with the first mover having more at stake.

The induced conflict of interests inhibits efficient collaboration. We characterise the Pareto-efficient equilibria of this game, and show that for a wide range of parameters, the first mover occasionally makes false claims about positive findings but over time becomes trustworthy and starts reporting honestly.

In these equilibria, the players’ relevant beliefs diverge from each other over time and the second mover will occasionally rubber-stamp a project even though the players have invested in learning and obtained only bad news.