Robust Regulation of a Monopolist

Advisor: Vinicius Nascimento Carrasco

This work studies the problem of a regulator who faces a monopolist with unknown costs. Different to previous works, we leave the strong assumption that the regulator knows the true probability distribution of monopolist cost, and instead we assume that regulator holds only a prior belief which gives him exact information about the mean. Regulator maximizes the expected social welfare under the worst distribution in the set of mean preserving spread distribution of his prior. The solution entails a linear social welfare which induces a state-dependent optimal quantity, in contrast to the two point quantity which solves the problem under perfect knowledge of the probability distribution of monopolist cost.