| This dissertation consists of three essays on health, aging and the family in contemporary China. The first essay addresses socioeconomic differentials in mortality among the oldest old Chinese. The other two essays examine intergenerational transfers between parents and children.;The first essay explores how socioeconomic status affects mortality among the oldest old Chinese. Previous literature suggests that socioeconomic differentials might disappear at very old ages. To delve more deeply into this issue, I use data from the 1998, 2000, and 2002 waves of "The Chinese Longitudinal Healthy Longevity Survey" to examine whether SES differentials still persist even among the oldest old Chinese (80+). Findings show the continuing prevalence of SES differentials in mortality--higher SES is significantly associated with lower mortality risks--among the oldest old Chinese. Further analyses show that the relationship holds regardless of whether the cutpoint for the oldest old category is set at 80+, 90+, or 100+ years old.;The second essay concerns the patterns of intergenerational financial support in urban China. It examines whether children with high socioeconomic status buy themselves out of fulfilling their traditional obligation to live with their parents, by instead providing their parents with increased financial support. In particular, this study treats coresidence and financial transfer as joint outcomes and use endogenous switching models to take into account the selection bias associated with coresidence. The results show that, based only on observable factors, children opt to buy their way out of their obligation to live with their parents; however, after jointly considering coresidence and financial transfer by controlling unobservable factors, the buy-out pattern disappears. This indicates that the buy-out pattern is driven by the selection into co-residence/non-coresidence.;The third essay explores how the characteristics of adult children and their siblings affect their financial support for their parents, with particular attention given to gender and birth order differences (traditional social norm hypothesis), educational differences (long-term exchange hypothesis), and redistribution of resources within the family (resource redistribution hypothesis). This study takes unobserved differences across families into account by estimating fixed-effects models. Results show that the long-term exchange and resource redistribution hypotheses are supported. |
