From Ben Handel and Jonathan Kolstad:
Traditional models of insurance choice are predicated on fully informed and rational consumers protecting themselves from exposure to financial risk. In practice, choosing an insurance plan from a set of complex non-linear contracts is a complicated decision often made without full information on several potentially important
dimensions. In this paper we combine new administrative data on health plan choices and claims with unique survey data on consumer information and other typically unobserved preference factors in order to separately identify risk preferences, information frictions,
and perceived plan hassle costs. The administrative and survey data are linked at the individual level, allowing in-depth investigations of the links between these micro- foundations in both descriptive and choice-model based analyses. We find that consumers lack information
on many important dimensions that they are typically assumed to understand, perceive high plan hassle costs, and make choices that depend on these frictions. Moreover, in the context of an expected utility model, including the additional frictions that we measure has
direct implications for risk preference estimates, which are typically assumed to be the only source of persistent unobserved preference heterogeneity in such models. In our setting, we show that incorporating measures of these frictions leads to meaningful reductions in estimated consumer risk aversion. This result has both
positive and normative implications since risk aversion generally has different welfare implications than information frictions. We assess the welfare impact of a counterfactual menu design and find that the welfare loss from risk exposure when additional frictions are not
taken into account is more than double that when they are, illustrating the potential importance of our analysis for policy decisions.