A model of health plan choice:: Inferring preferences and perceptions from a combination of revealed preference and attitudinal data

In this paper, we model the elderlyʹs choice among health plans using data from a 1988 study of the Minneapolis–St. Paul Medicare health plan market. We show how attitudinal data can be combined with revealed preference data to provide more reliable estimates of consumersʹ preferences for and perceptions of the attributes of choice alternatives. We develop an extended heterogeneous logit model that incorporates information about attribute importance and allows for heterogeneity in tastes for observed and unobserved attributes. Our results indicate that the inclusion of attitudinal information produces a substantial improvement in model fit and in the substantive interpretation of estimated parameters.