Research

Abstract: This paper analyzes optimal insurance design when the insurer internalizes the effect of coverage on third-party service prices. A monopolistic insurer contracts with risk-averse agents who have sequential two-dimensional private information and preferences represented by Yaari’s dual utility. Insurance contracts shape service demand and, through a market-clearing condition, determine equilibrium third-party prices. We characterize the structure of optimal contracts and show they take simple forms---either full coverage after a deductible is paid or limited coverage with an out-of-pocket maximum---closely mirroring real-world insurance plans. Technically, we formulate the problem as a sequential screening model and solve it using tools from optimal transport theory.

Abstract: I study how to regulate firms' access to consumer data when a regulator faces non-Bayesian uncertainty about how firms will exploit the consumer's information to segment the market and set prices. I fully characterize all worst-case optimal policies when the regulator maximizes consumer surplus: the regulator allows a firm to access data only if the firm cannot use the database to identify a small group of consumers. 

Abstract: We study information disclosure in competitive markets with adverse selection. Sellers privately observe product quality, with higher quality entailing higher production costs, while buyers trade at the market-clearing price after observing a public signal. Because sellers’ participation in trade conveys information, feasible posteriors are endogenously constrained. We reformulate the designer’s problem as a martingale optimal transport exercise with an additional obedience condition that rules out further information transmission through participation decisions. We characterize optimal signals. When the designer maximizes trade volume, the solution features negative-assortative matching of inefficient and efficient sellers. When the objective is a weighted combination of price and surplus, optimal signals either preserve this structure or fully reveal low-quality types while pooling middle types with some efficient sellers. We discuss applications to credit scoring, grading systems, and online labor platforms.