Trading Off Flexibility in Health Insurance Choice

Abstract

In developed countries, consumers have little choice about health insurance. Options are restricted either by national health insurance programs or employer-provided group plans. An exception was the large number of households in the U.S. who weree not covered by group plans, but this changed with the introduction of ObamaCare. The most controversial provision in the on-going U.S. health care reforms is the requirement that such individuals purchase insurance. At the same time, there are some proposal in the U.S. for group plans to offer more flexibility. This paper provides a theoretical model of the trade-offs in allowing for flexibility. The main market failures that insurance restrictions address are the under-insurance that results from adverse selection and the reclassification risk that results from the lack of binding (on both sides) longterm health insurance. For both sources of market failure, the loss in welfare comes from the fact that individuals choose health insurance based on their health risk—hence the benefit of restricting choice. However, there are legitimate taste-based reasons that different individuals should have, in a first-best world, different health plans. Because these tastes are typically unobservable, designing the optimal portfolio of health plans allowed in the market (or offered by a state insurer) is a multidimensional screening problem. The design problem trades off the benefit of taste separation and the benefit of risk pooling. We develop this idea for a specific form of taste heterogeneity: individuals have different preferences for the quality of health care (such heterogeneity could come, for example, from differences in income, but we use a reduced-form representation that avoids a budget constraint). The extra dimension of hidden information is not accompanied by an extra dimension in the contract space, distinguishing this from the literature on multidimensional screening (e.g., non-linear pricing) by a multiproduct monopolist, but in common with several other papers on multidimensional screening. Furthermore, another distinction with much of the screening literature is that the welfare perspective is ex ante with respect to the realization of health risks, which is the right approach for studying reclassification risk rather than a cheap way to sneak in utilitarianism. Though we always take a social planner perspective with respect to the allowable contracts in the market (other than to consider unfettered competition as a benchmark), we consider the cases in which prices are set by the planner and in which prices are set competitively. A main conclusion is that, when the taste heterogeneity is weak enough compared to the risk heterogeneity, substantial or complete inflexibility is optimal.