When Rules Make Markets Fragile: Regulation- and Privacy-Induced Adverse Selection

Use policy variation such as medical underwriting prohibitions, community rating changes, privacy regulations (e.g., GDPR, U.S. state laws), or bans on using certain data to test the claim that adverse selection often arises because insurers are compelled to ignore available risk information. Employ difference-in-differences with granular plan-level enrollment and claims data, studying heterogeneity such as stronger selection among younger, self-rated healthy individuals. This reframes adverse selection as often policy-induced—an equilibrium consequence of constrained information use—contrasting with the standard narrative emphasizing unobservable risk. The research anticipates new frictions around AI-based underwriting and fairness constraints. It extends disentangling literature by making the information set endogenous to regulation and reconciles conflicting field results by showing context-dependent information constraints. It also draws on community-based schemes’ struggles with persistent selection. The project provides concrete guidance on regulatory levers—risk adjustment refinements, permissible signal sets, transparency—that reduce harmful selection without enabling discrimination. It sheds light on puzzling heterogeneity in selection patterns and can inform smarter regulation that pairs privacy and fairness goals with improved risk adjustment or dynamic information tools rather than blunt bans that destabilize insurance pools.

References:

1. NBER WORKING PAPER SERIES DISENTANGLING MORAL HAZARD AND ADVERSE SELECTION IN PRIVATE HEALTH INSURANCE. David Powell, Dana (2015).
2. DESIGN OF INCENTIVES IN COMMUNITY BASED HEALTH INSURANCE SCHEMES. R. Ahuja, J. Jütting (2003).