Three Essays on Economics of Health Behavior in China

This dissertation consists of three essays, each focusing on one topic in economics of health behaviors in China. The first essay attempts to examine the determinants of alcohol demand with concentration on impact of alcohol price among Chinese adult population. Although literature in developed countries context has indicated that average alcohol demand is price responsive, evidences from developing countries are limited and few work studied the heterogeneity in price elasticities. This paper estimates the price elasticity of demand for alcohol consumption using China Health and Nutrition Survey (CNHS), a representative panel data interviewed during 1993-2006. Specifically, three sources of heterogeneities in alcohol use are explored: differential responses across socio-economic characteristics, drinking levels, and types of drinks. Focusing on public health perspective, we define alcohol consumption in this study as milliliters of total pure alcohol usage in past week. Results of average price elasticity for pure alcohol consumption in China yield estimates ranging between -.07 to -.11 for males and -.07 for females in pooled data, and getting smaller (-.03) or even becoming insignificantly positive in panel data. Examination in heterogeneous demand suggests that alcohol price responsiveness varies substantially across residence, age groups, gender, income level, and labor force status; people who drink alcohol above median level are as sensitive as median level drinkers, and more sensitive than drinkers who drink below median level; beer and liquor are substitutes, drinkers are significantly responsive to liquor price but not beer price. This study suggests that individual unobserved heterogeneity plays an important role in the relationship between alcohol demand and alcohol price. The effect of alcohol price on alcohol consumption is much smaller than western countries. Taxation on alcohol may be not an efficient instrument to improve alcohol related health, although it is efficient in increasing government revenues. Non-monetary policy levers should be considered in China. The second essay estimates healthcare expenditure in China and evaluates the performance of econometric models. The heavy skewness and non-normal distribution have made ordinary least squares (OLS) biased and inefficient in estimating individual healthcare expenditure. I compare a set of alternative methods including log-normal density maximum likelihood, generalized linear model, quantile regression and finite mixture model using CNHS individual healthcare expenditure data among Chinese adult healthcare users. The comparisons of statistical model selection criteria suggest that finite mixture model is preferred overall in terms of information criteria, goodness of fit, and cross sample validation. The in-sample prediction indicates that finite mixture model overestimates mean value of expenditure, but gives the most precise median prediction, and performs the best in both lower and upper tail predictions. The objective of the third essay is to examine the time trend of obesity disparities across sociodemographic groups in school-aged youth population from 1991 to 2006 in mainland China. 1 estimate mean body mass index age- and gender-specific percentiles for sociodemographic subgroups defined by gender, nationality, residence, relative household per capita income, and parental education, using multivariate regression to adjust for changes in sample/population composition. Six waves of CNHS were analyzed, including 14 204 school-aged youth (age 6-18) observations between 1991 and 2006. The findings suggest that all sociodemographic subgroups experienced substantial weight gain during study period, but the gap in mean BMI between population subgroups is widening in many cases. Boys gained more than girls; Han Chinese gained more than minorities; youth in high income families gained more than youth in low-income families; youth whose parents (for both mother and father) had little education (only primary school d ploma or less) gained less weight than youth whose parents had more education. In contrast, there was no strong evidence for a widening BMI gap between urban and rural residents. There appears to be a widening in BMI (and obesity rates) across sociodemographic groups in mainland China, yet it is the wealthier, more educated, ethnic majority that becomes more obese. In contrast, it is poorer, less educated, and minority groups in the United States and other developed countries that are at higher risk for obesity.

• http://libproxy.lib.unc.edu/login?url=http://search.proquest.com/docview/921240969?accountid=14244
• http://vb3lk7eb4t.search.serialssolutions.com/?ctx_ver=Z39.88-2004&ctx;_enc=info:ofi/enc:UTF-8𝔯_id=info:sid/ProQuest+Dissertations+%26+Theses+Full+Text&rft;_val_fmt=info:ofi/fmt:kev:mtx:dissertation&rft.genre;=dissertations+%26+theses&rft.jtitle;=&rft.atitle;=&rft.au;=Shi%2C+Yuyan&rft.aulast;=Shi&rft.aufirst;=Yuyan&rft.date;=2011-01-01&rft.volume;=&rft.issue;=&rft.spage;=&rft.isbn;=9781267165015&rft.btitle;=&rft.title;=Three+Essays+on+Economics+of+Health+Behavior+in+China&rft.issn;=