| dc.description.abstract | This dissertation is a study in simple information design. In the first chapter, I examine dynamic information design under constrained communication rules, motivated by record keeping regulations and targeted transparency policies. Constrained rules require that messages cannot be conditioned on private information. A long-run and uninformed designer wishes to persuade shortlived agents to invest in a project of fixed, but unknown, quality. The restriction on the possible communication rules hinders the designer without benefiting the agents. It unambiguously decreases welfare if the designer has altruistic motives. The optimal policy depends on the conditional payoff distribution and if failures are very informative, constrained rules are less restrictive: simple policies approximate the designer’s first-best payoff. In the second chapter, I consider dynamic information design with bad reputational concerns. Customers increasingly rely on rating systems when hiring experts. If a rating system is ill designed, the reputational gain from taking a certain action might be excessively high, inducing the experts to over-choose it. In such a case, markets could even cease to exist altogether. I show how to design simple, optimal rating systems for both customers and experts. Such rating systems overcome market failures and improve upon both the full memory case and the case with no memory at all. Customers benefit from a higher number of ratings, while binary systems are sufficient to achieve the expert’s highest value. In the third chapter, I study information design in an observational learning environment. A sequence of short-lived agents must choose which action to take under a fixed, but unknown, state of the world. Prior to the realization of the state, the long-lived principal designs and commits to a dynamic information policy to persuade agents toward his most preferred action. The principal’s persuasion power is potentially limited by the existence of conditionally independent and identically distributed private signals for the agents as well as their ability to observe the history of past actions. I characterize the problem for the principal in terms of a dynamic belief manipulation mechanism and analyze its implications for social learning. For a class of private information structure - the logconcave class, I derive conditions under which the principal should encourage some social learning and when he should induce herd behavior from the start (single disclosure). I also show that social learning is less valuable to a more patient principal: as his discount factor converges to one, the value of any optimal policy converges to the value of the single disclosure policy. | eng |
