Health demand and health determinants in China

Abstract

This paper identifies health determinants in rural and urban China. Using the 2000 wave of the China Health and Nutrition Survey, we find that education has an important positive effect on health. We also find that regional location is an important determinant of health. Our results indicate that the self-reported health status is not significantly different between the urban and the rural population. Our study suggests that Chinese males have better health than females, and married persons have better health than single persons. We also find that rural residents who live in suburbs have worse health than those who live in remote villages.

Keywords:

• self-reported health status
• health determinants
• ordered probit
• China

Acknowledgements

The author would like to thank the participants in the 2005 Annual Conference of the Chinese Economists Association for helpful comments, and Hongchun Zhao for excellent research assistance. Financial support from the China Medical Board of New York (CMB), the China Academy of Health Policy (CAHP), and a 211 project at Peking University is acknowledged. All errors are the author's sole responsibility.

Notes

Notes

1. Liu et al. (2004) is one of the few exceptions. They study the relationship between economic growth and health capital.

2. There are two reasons for adopting a linear model instead of a double-logarithm model derived from the Grossman model. One is that the study of Wagstaff (1993) finds that the assumption is unconvincing and that the linear model is more consistent with the data. The other is that we use an ordered probit model to analyze the ordered categorical health status variable instead of continuous variables.

3. The surveys of 1989, 1991, and 1993 include eight provinces: Guangxi, Guizhou, Henan, Hubei, Hunan, Jiangshu, Liaoning, and Shangdong. In 1997, Heilongjiang replaces Liaoning. In 2000, both Liaoning and Heilongjiang are included in the survey along with the other provinces.

4. Studies such as Kaplan and Camacho (1983) find that this categorical health variable contains important information on an individual's health.

5. The sample average earnings per week are 131 RMB yuan.

6. We use the community cost instead of the individual cost to avoid the problem that individual cost is only observed for the people who have the flu.

7. To interpret the result on education, it is necessary to point out that in our analysis we cannot model unobservable factors such as ability. If the correlations between ability and education and between ability and health are both positive, our result on education will be biased upward due to omitted-variable bias (see Grossman, 2000).