LAUSR

Abstract This paper develops a model of insurance choice and derives the implications for insurance plan selection when consumers are loss averse. Previous work has shown that health insurance plan switching costs can increase welfare by reducing adverse selection. Loss aversion provides a natural mechanism for modeling non-pecuniary switching costs and thus gives a basis for modeling insurance choice in environments where explicit switching costs might be minimal. I also derive a test for the presence of both adverse selection and switching costs based on differential spending of switchers and non-switchers. I test this model using data from a benefits firm providing employment-based health insurance. Employees choose between low and high coverage plans from the same insurer during the annual open enrollment period, and explicit switching costs are likely to be low. I assume employees take their current plan as the reference point for choosing the next year’s plan. I find evidence of adverse selection and significant switching costs.