Essays on health insurance markets: Asymmetric information and multiple periods.

Abstract

Adverse selection and moral hazard are two effects of incomplete information in the market for health insurance. Theoretical articles have dealt with these two issues separately, but little past research has examined their combined effects. This dissertation demonstrates that under reasonable conditions in single period models having both effects of incomplete information can result in the high risks imposing no externality on the low risk individuals. In cases in which the high risk individuals impose an externality, any willing provider laws that restrict the ability to construct managed care plans will limit the utility for low risk individuals and may not increase utility for high risk individuals. Using an anticipatory equilibrium concept, the type of equilibria before and after the imposition of the any willing provider regulations will not necessarily be the same. Using an anticipatory equilibrium concept, low risk individuals will be able to obtain a higher utility in markets in which insurers subsidize across products than in markets in which this does not occur. However, high risk individuals may have a lower utility with cross-subsidization. Cross-subsidization will lead to lower premiums for fee-for-service products but will increase health maintenance organization premiums and may increase premiums and medical care expenditures at the population level. Multiple period models of health insurance choice are a recent consideration in the theoretical health insurance literature. Previous models are extended by adding constraints on the consumer capital market and subjective rates of time discounting. Constraints on the capital market alter the optimal sequence of policies linked over time, although both high and low risk individuals still face decreasing premiums over time. Consumers who sufficiently discount future consumption will not want to participate in sequences of policies linked over time. The public policies that could be used to limit the utility of low risk individuals in single period models can now be used to increase lifetime expected utility. The public policies must be applied to all sequences of policies except those lasting a lifetime. This suggests a selective role for government regulation that does not appear in the single period models.