Effects of Thai healthcare policy on household demand, hospital efficiency and household earnings

This dissertation evaluates how healthcare and the health insurance policy in Thailand affect both supply and demand sides. This research applies different econometrics methods such as a nonparametric boostrapping Data Envelopment Analysis method, a Quadratic Almost Ideal Demand (QUAIDS) model, a random-effect Tobit regression, etc. The first chapter investigates the short-term impact of the new national health insurance program or Universal Coverage (UC) in Thailand on technical efficiency in provincial public hospitals. By measuring efficiency before and after the reform, the study applies a two-stage analysis with bootstrapping Data Envelopment Analysis (DEA) and a censored Tobit model. The results indicated that UC improved efficiency in larger public hospitals across the country. The effect differed by region around the country, and hospitals in provinces with more wealth not only started with greater efficiency, but improved their relative position after UC was implemented.; The second chapter investigates the effects of health status and healthcare utilization on agricultural household earnings in Thailand. A utility-maximization production model in which health status and education affect household resource allocation is formulated. Using the Box-Cox transformation, the 2SLS and OLS estimations are applied. A key finding indicated that education appeared to increase farm household income, while the effect from health was unclear. However, the disaggregated analysis showed that health may be a determinant to income for rice farming household in which farm income appeared to increase by 0.3% with a 10% increase in health investment.; The third chapter explores how national health insurance affects the allocation of household expenditures on consumption goods (i.e., housing, food, etc.) changed by comparing expenditure patterns before and after the health insurance reform. A Quadratic Almost Ideal Demand System (QUAIDS) model developed by Banks, Blundell, and Lewbel (1997) is used incorporating with a two-step approach introduced by Shonkwiler and Yen (1999). The programming was done on GAUSS 7.0 in order to solve the nonlinear least squares problems applying the Gauss-Newton optimization algorithm