A substantial amount of evidence documents the correlation between socio-economic status (SES) and health outcomes. People with higher incomes, more education or better jobs also experience improvements in general health and lower disease incidence. But the correlation between SES and health does not imply causation, and the past decade has seen an increased effort to discern causal relationships using natural experiments. One of the health variables of particular interest is excess body weight, which has received much attention over the last 30 years as many countries have experienced a surge in obesity prevalence. Obesity is commonly measured by the body mass index (BMI) and is closely associated with SES, especially among women.
This dissertation examines three topics related to SES and health: First, how does BMI vary with SES, as measured by education and income, and have these relationships changed over time? How do these relationships differ by gender and race, and across three major U.S. national health data sets (National Health and Nutrition Examination Survey, National Health Interview Survey, Behavioral Risk Factor Surveillance System)? Second, does education have a causal effect on obesity? I examine this association using compulsory schooling laws as a source of exogenous variation in schooling. Special attention is paid to differences by gender and estimation and testing procedures that are robust to weak instruments. Third, is there a causal connection between education and other health behaviors and outcomes, such as self-rated health, hypertension, smoking, and alcohol consumption?
Results presented in Chapter 1 suggest that disparities between low- and medium-SES groups have almost disappeared over the last 30 years, while high-SES groups continue to enjoy an advantage. In terms of BMI trends, Hispanic males are comparable to white and black males, while Hispanic females are in-between white and black females. The three datasets examined in this work yield different prevalence rates but similar trends.
The second chapter, using data from the National Health and Nutrition Examination Survey (NHANES) from 1971 to 1980, finds a strong and statistically significant negative effect of additional schooling on BMI for women and no effect for men (an extra year of schooling is estimated to reduce BMI by about one unit for women). These results are robust to weak instruments and various other validity checks, and suggest that policies designed to increase years of schooling for at-risk populations might lead to substantial improvements in obesity rates.
In the third chapter, using the same methodology and data set as in Chapter 2, I find strong correlations between educational attainment and various other health outcomes. Weak instruments do not permit any conclusions about causality in this analysis.
In sum, this work contributes to the literature in three ways: First, it provides the most credible estimate to date of the causal impact of schooling on obesity. Second, it documents trends in the association between SES and obesity over the last 30 years using three major nationally representative data sets from the U.S., paying special attention to differences across gender and race. Third, it examines the usefulness of compulsory schooling laws in estimating causal effects of education on other health outcomes in the early NHANES. Weak instrument problems suggest that alternative approaches should be pursued in future research.
|Advisor:||Miller, Douglas L.|
|Commitee:||Cameron, A. Colin, Leigh, J. Paul|
|School:||University of California, Davis|
|School Location:||United States -- California|
|Source:||DAI-A 71/03, Dissertation Abstracts International|
|Subjects:||Economics, Health sciences, Public health|
|Keywords:||BMI, Compulsory schooling law, Education, Health outcomes, Obesity, Socioeconomic status|
Copyright in each Dissertation and Thesis is retained by the author. All Rights Reserved
The supplemental file or files you are about to download were provided to ProQuest by the author as part of a
dissertation or thesis. The supplemental files are provided "AS IS" without warranty. ProQuest is not responsible for the
content, format or impact on the supplemental file(s) on our system. in some cases, the file type may be unknown or
may be a .exe file. We recommend caution as you open such files.
Copyright of the original materials contained in the supplemental file is retained by the author and your access to the
supplemental files is subject to the ProQuest Terms and Conditions of use.
Depending on the size of the file(s) you are downloading, the system may take some time to download them. Please be