Document Type



Doctor of Philosophy



First Adviser

Yang, Muzhe

Other advisers/committee members

Deily, Mary E.; Richards-Shubik, Seth; Wang, Yang


This dissertation studies the short-term infant health effects of manipulation of birth timing, the short-term elderly health effects of retirement and long-run associations between childhood socioeconomic status and health outcomes later in life. In the first chapter, I assess the impact of the small change of birth timing on infant health at birth, by exploiting a phenomenon in which some parents avoid holiday babies, especially during the holiday season at the end of the year. I focus on all full-term singleton births delivered by mothers without medical problems during pregnancy because newborns’ health outcomes are believed to be similar in the traditional full term pregnant period (37 weeks to 41 weeks). Using New Jersey birth certificate data with a baby’s exact date of birth and the exact date of the mother’s last menstrual period during the years 1989 to 2011, it was estimated that about 207 singleton births per year are moved to an early time from the Christmas break and the New Year’s break in the holiday season at the end of the year. Among different delivery methods, I find that induced vaginal and induced C-section deliveries are the two main methods used for the birth timing manipulation. I also find the group of mothers whose babies were born shortly before the holiday break appear to be slightly older, more educated, more likely to be white, more likely to be married, having more prenatal visits and fewer risky behaviors during pregnancy than mothers whose babies born during the holiday break. In addition, birth timing manipulation causes babies to be born before their expected delivery dates, which carries greater health risks, including lower Apgar scores, a lower probability of getting normal Apgar scores (i.e., scores > 7), and a higher probability of having respiratory problems. In the second chapter, I investigate the short-term effects of retirement on health related outcomes and the mechanisms behind the effects by using a comprehensive, nationally representative sample from the China Health and Retirement Longitudinal Study (CHARLS). The mandatory retirement policy in China provides a quasi-experimental setting for the identification of the effects of retirement. By focusing on a sharp change in retirement status among males just below and just above the mandatory retirement age 60, and using regression discontinuity models, I show that retirement improves the mental health and well-beings of retirees. Although the effects are not statistically significant, I find retirement still improve the subjective health status and objective physical health of retirees. I further find that paying more attention to one’s own health, increased social activities, as well as more frequent physical exercises, might be the key mechanisms through which retirement improve health and well-beings. In the third chapter, I also use the CHARLS to estimate the long-term effect of childhood social-economics status (CSES) on several health outcomes of the elderly (people aged 45 to 80) in China. I find that for both men and women, unfavorable childhood life situation is associated with adverse health status in later life. Although those effects are partially mediated through education and adult income status, those effects remain statistically significant when I control for the education and income. I also find long-term CSES effects are stronger for women than for men.

Included in

Economics Commons