Essays on medical care using Semiparametric and structural econometrics

This dissertation consists of an empirical chapter, an econometrics chapter, and a theoretical chapter, all of which advance the study of the price elasticity of expenditure on medical care. In Chapter 1, I estimate the price elasticity of expenditure on medical care across the quantiles of the expenditure distribution. My identification strategy relies on family cost sharing provisions that generate differences in marginal prices between individuals who have injured family members and individuals who do not. I use a new censored quantile instrumental variables (CQIV) estimator, which allows me to examine variations in price responsiveness across the skewed distribution of medical expenditure. The CQIV estimator does not require any parametric assumptions to account for individuals who consume zero medical care. Using CQIV, as well as traditional estimators, I find elasticities that are an order of magnitude larger than those in the literature. My CQIV estimates suggest strong price responsiveness among people who spend the most. I find that the price elasticity of expenditure is approximately -2.3, which is stable across the .65 to .95 quantiles of the expenditure distribution. In Chapter 2, Chernozhukov and Kowalski (2008), we develop a censored quantile instrumental variables (CQIV) estimator. The CQIV estimator handles censoring nonparametrically in the tradition of Powell (1986), and it generalizes standard censored quantile regression (CQR) methods to incorporate endogeneity. Our computational algorithm combines a control function approach with the Chernozhukov and Hong (2002) CQR algorithm. Through Monte-Carlo simulation, we show that CQIV performs well relative to Tobit IV in terms of median bias and interquartile range.