Abstract
In this paper, we consider a principal who gathers information from a sequence of privately informed agents. The agents, in turn, observe and learn from previous agents’ actions, but the principal can compel them to take actions uninformatively. We show that the principal’s optimal strategy involves destroying information from early agents in order to create more information from subsequent agents. This way, the principal is able to mitigate herding behavior. Asymptotically, the principal keeps agents from learning until he gathers enough information, so, although agents observe their predecessors’ actions, they believe them to be manipulated with high probability.
Biography
David Rahman is Associate Professor in the Department of Economics at the University of Minnesota. He specializes in economic theory, specifically mechanism design and repeated games.
