Text

theses

This dissertation focuses on examining the impact of public spending in health insurance and health care markets. Health care subsidies account for a fast-growing share of public expenditures in many developed and developing countries, making them an ever more important component of fiscal policy discussions. Two principle projects constitute my dissertation research. In the first project, I examine the heterogeneity in the impact of subsidized health insurance coverage on individual welfare, in the context of a Chinese public health insurance program. In the course of this research, I have also developed new econometric methods to address the empirical challenges of studying the effects of health insurance. These methods have broad applications beyond topics in health economics. In the second project, I look at the role of tax subsidies in the supply of health care. In particular, I exploit variations in state and federal level tax policies in the U.S to estimate the impact of government subsidies on ownership choice, provision of public services and the quality of hospitals. The first chapter of the dissertation mainly assesses the effect of public health insurance on program beneficiaries' welfare, by evaluating a new national public medical insurance program in China, Urban Resident Basic Insurance (URBMI). This program, introduced in 2007 and having an annual fiscal expenditure of 30 billion RMB, aims to provide coverage to more than 200 million urban residents including elderly, children, college students and unemployed adults. I exploit the city-variation in policy generosity as an exogenous determinant of URBMI enrollment. Using data from the Chinese Health and Nutrition Survey (CHNS), I find that URBMI increases welfare on several margins. Having insurance coverage increases health care spending while decreasing the out-of-pocket payments, providing protection from the financial risk. It also increases efficiency in medical spending by inducing the use of preventative care and reducing the probability of hospitalization. In terms of health outcomes, insurance coverage has a significant impact on subjective self-ratings in health and happiness. I also extend my examination to consider the labor market effects of URBMI. Since this program provides insurance coverage outside of employment status, it will potentially increase an individual's mobility between jobs and impact the retirement decision. In Chapter 2, building on the results of the first chapter, I explore the heterogeneity in the impact of health insurance through a semiparametric model. Since URBMI is a national program covering a wide range of subpopulations, observed and unobserved individual characteristics may play an important role in determining the response of an individual to insurance coverage. This chapter builds a panel data model with endogenous treatment, which incorporates unobserved individual heterogeneity non-additively into the outcome. The model is estimated in the context of a semiparametric setting. I first propose a two-stage semiparametric least square (SLS) method to consistently estimate the model parameters and then conduct a localized 2SLS procedure to recover the quantile treatment effect. Identification, consistency, and root-N asymptotic normality of estimators for parameters and marginal effects are proved. The estimation results reveal substantial variation in the impact of URBMI by age, income and gender. Children, the elderly above the age of 70, and females ages 25-40 benefit the most from the program. Adult males and individuals with incomes below the median level do not respond significantly to insurance coverage. The findings of heterogeneous insurance effects have important policy implications for the cost-effectiveness of URBMI across population groups, suggesting the need for differentiated insurance programs. In the third chapter, another form of subsidy in health care markets is studied. This chapter focuses on assessing the effect of government subsidies on the supply side of the health care market in the U.S. An important form of government subsidies to health care providers is the tax exemption for non-profit organizations. The validity and efficiency of such practice has long been under debate. Recently, many state and federal laws have been enacted that mandate the reporting of benefits provided to the community by non-profit providers. This chapter studies the hospital sector. Given the preferential tax treatment for nonprofit hospitals, the tax rate, in conjunction with community benefit reporting requirement (CRR), determine the net subsidy provided to a nonprofit hospital compared to its for-profit counterpart. I exploit the variation in tax policy across states and over time to identify the effect of tax subsidy on the ownership choice of hospitals. I further differentiate behavior between nonprofit versus for-profit hospitals, including cost, provision of undercompensated care as well as quality. Using Center for Medicare and Medicaid Services(CMS) hospital cost report data from 1996 to 2015, I estimate a 4-6 percent increase in the probability of non-profit conversion into for-profit hospitals due to the enactment of CRR. Moreover, the effect of CRR diminishes with the tax rate. My results further show that hospitals divert community benefit spending to teaching to meet the requirement of CRR, rather than increasing provision of uncompensated care.

ETD_8077

Ph.D.

Includes bibliographical references

by Shuyang Yang

doi:10.7282/T3KK9FQF

ETD doctoral

The author owns the copyright to this work.