Modeling Health Insurance Choice Using the Heterogeneous Logit Model

Recent advances in "simulation based inference" have made it feasible to estimate discrete choice models with several alternatives and rich patterns of consumer taste heterogeneity. These new methods have important potential application in health economics. One important application is the analysis of consumer choice behavior in insurance markets characterized by competition among several alternative insurance plans. Analysis of consumer choice behavior in insurance markets is of great interest in health economics for a number of reasons. For example, the longstanding interest in optimal design of insurance markets stems from the inefficiency of competitive equilibrium in these markets. But a deep understanding of the structure of consumer taste heterogeneity is necessary before one can hope to achieve an optimal design of insurance markets. New methods of choice modeling (like the heterogeneous logit) offer hope of providing such an understanding.