作者: Paul Gertler , Roland Sturm , Bruce Davidson
DOI: 10.3386/W4700
关键词: Economics 、 Actuarial science 、 Information system 、 Variance (accounting) 、 Discrete choice 、 Panel data 、 Probit 、 Goodness of fit 、 Random effects model 、 Perfect information
摘要: While the critical role of imperfect information has become axiomatic in explaining health care market failure theory is backed by little empirical evidence. In this paper we use a unique panel data set with explicit measures and an educational intervention to investigate uncertain about insurance benefits on demand for supplemental Medicare insurance. We estimate structural discrete choice model that allows affect both mean variance expected distribution. The specification multinornial probit unrestricted variance-covariance including heteroskedasticity random effects control unobserved heterogeneity. computationally complex estimated simulated maximum likelihood. results indicate affects supplement increasing distribution rather than systematically shifting Since majority people already purchase increase due reduces demand. if everyone perfect then proportion individuals not purchasing would fall 23% from .096 .074. also found controlling heterogeneity was important. goodness fit increased 25% precision effect dramatically. (authors)