Julia Dennett

Julia Dennett

Postdoctoral Associate, Department of Epidemiology of Microbial Diseases, Yale School of Public Health

Dissertation Title:  "Essays on the Determinants of Health and Human Capital"Economic, environmental, and social factors influence population health and human capital development. Often these outcomes reflect past exposures that occurred earlier in life, or the complex interplay between multiple conditions and events. This dissertation evaluates the effects and policy implications of three determinants of health in the United States: prenatal exposure to the seasonal flu, life course cigarette interventions, and the interaction between Medicaid coverage and neighborhood characteristics.

Chapter 1 analyzes the effects of prenatal exposure to seasonal influenza on life course outcomes in the United States. The seasonal flu is a common infectious disease that jeopardizes the health of pregnant women. Prenatal exposure to the flu likely disturbs fetal development and harms health at birth, but long run effects have been difficult to identify. I investigate the impact of in utero exposure to seasonal influenza over the life course in the U.S. by exploiting state and time variation in influenza-related mortality, a proxy for disease severity in the local environment. I first show adverse effects on birth weight and an increased risk of heart malformations, and then evaluate impacts on long-term outcomes. Exposure to seasonal influenza while in utero increases disability and decreases childhood school attendance, adult high school completion, and labor force participation. I examine implications for influenza vaccination as a policy intervention and find that historical vaccine uptake accounted for economically meaningful improvements in life course outcomes. Furthermore, my estimates suggest substantial returns to future reductions in flu exposure due to vaccination expansions, including 10,000 fewer infants born each year with low birth weight, 54,000 more workers in the labor force, and 34,000 more adults without a disability.

Chapter 2 conducts a life course evaluation of cigarette taxes and smoking. Cigarette-related diseases develop from the cumulative impacts of tobacco use over time. As a result, interventions are often targeted towards young people. The impacts of life course anti-smoking policies at older ages are less certain. This paper analyzes how cigarette policies experienced over a lifetime affect long-term smoking behaviors. I primarily focus on state cigarette excise taxes. I consider how early- and mid-life tax exposures impact the smoking behaviors of young, middle-aged, and older adults using 1984-2010 pooled Behavioral Risk Factor Surveillance System data. I find that policies should be directed at the entire age distribution in order to reduce lifetime smoking. Early life cigarette taxes decrease young adult smoking, but these effects dissipate with age, while mid-life cigarette taxes decrease contemporaneous smoking and smoking amongst older adults. Other results highlight the role of anti-smoking interventions to minimize smoking in subgroups of particular policy importance, including young adults, less educated individuals, and women of childbearing age. I explore policy implications by simulating the effects of cigarette tax changes at different stages of adulthood. I show that increases in mid-life taxes reduce the probability of smoking by age 65 nearly as much as increases in the teenage cigarette tax.

Chapter 3 (with Katherine Baicker) evaluates the extent to which living in neighborhoods with different characteristics moderates the impacts of Medicaid coverage on health. Policies aimed at improving population health in the United States often focus on expanding access to health insurance. However, neighborhood characteristics and other determinants of health beyond the health care system also affect outcomes. This study evaluates how neighborhood features moderate the causal effects of health insurance on physical, mental, and health-related quality of life outcomes. We undertake this analysis using data from the in-person survey of the Oregon Health Insurance Experiment, which randomized access to Medicaid, combined with information on multiple domains of neighborhood features in the greater Portland area. We find that factors associated with active living attributes enhance the ability of Medicaid coverage to improve body mass index, but also exacerbate increases in cardiovascular risk. These results suggest that the interaction between Medicaid coverage and certain neighborhood features are complex and varied. We find limited moderating effects for other neighborhood characteristics and outcomes. Understanding how neighborhood attributes shape the impacts of Medicaid coverage may improve policy design and enhance the effectiveness of future health insurance expansions.



Graduation Year