Three Essays on the Economics of Obesity

UNCG Author/Contributor (non-UNCG co-authors, if there are any, appear on document)
Christian Anthony Gregory (Creator)
The University of North Carolina at Greensboro (UNCG )
Web Site:
Christopher Ruhm

Abstract: The literature examining the relationship between obesity and wages has fairly consistently found that BMI has a negative impact on earnings for women, and less (if any) consequences for men. In the first study in my dissertation, co-authored with Christopher J. Ruhm, we relax the assumption--largely unquestioned in this research--that the conditional mean of wages is linear or piecewise linear in body mass index (BMI). Using data from the 1986 and 1999-2005 Panel Study of Income Dynamics, we estimate semi-parametric wage models that allow earnings to vary with BMI in a highly flexible manner. For women, the results show that earnings peak at levels far below the conventional threshold of obesity or even overweight. For men, our main estimates suggest a reasonably flat BMI-wage profile that peaks early in the ``overweight'' category. The findings for females (and the IV estimates for males) suggest that it is not obesity but rather some other factor -- such as physical attractiveness -- that may be producing the observed relationship between BMI and wages. In the second essay of this dissertation, I examine the effect of obesity--and body mass more generally--on wages across the age distribution, using conventional parametric and more flexible semiparametric approaches. My parametric results suggest that the literature may overstate the effect of BMI and obesity on wages for women and almost certainly understates any negative association for men. For women, my results show that the negative effects of BMI and obesity are concentrated among women between 25 and 35 years old. While women in this age group experience an average 0.5 to 0.7 percent decrease in wages for each point increase in body mass (roughly 7.5 pounds), women over 40 will suffer a 0.25 percent decrease in wages for each extra point of body mass, and may not experience any wage penalty at all. Similarly, women who are 31-35 years old experience a 7.7 percent decrease in wages for being obese, while women over 40 experience only a 3.9 percent decrease. More flexible models largely confirm these results. For men, my parametric results indicate that, for those who are in their 20's or early 30's, BMI has no effect on wages or is associated with a small increase; this is consonant with the rest of the literature. However, for men over 35, the effect of extra body mass is clearly negative: an extra BMI point brings with it a 0.3 percent decrease in wages for men 36-40 years old, and for men over 40, an extra BMI point is associated with a 0.5 percent decrease in wages. More flexible semiparametric models suggest that the negative association of BMI and wages may be as much as three times more than these estimates in some ranges of BMI. For instance, these models suggest that men 36-40 years old who have a BMI between 27 and 37 experience a 0.9 - 1.2 percent decrease in wages for each extra BMI point, as opposed to a .3 percent decrease predicted by the linear model. Finally, in the third essay, I examine whether it has gotten easier to be obese over the last 25 years. Improvements in treatments for co-morbidities of obesity--high cholesterol, diabetes, sleep apnea and heart disease--over the last 25 years have made obesity less burdensome. I use data from the NHIS to examine whether such improvements have been borne out in obese persons' self-reports of health. My results suggest that between 1982 and 1996 obese women enjoyed significant gains in health relative to their normal weight counterparts. However, these gains do not appear to be due to improvement in treatment for co-morbidities of obesity; rather, income and especially education explain a large share of these health trends. For men, there seems to be little in the way of trends during these years of the survey. Results from the later years of the NHIS survey (1997-2006) suggest very little in the way of trends in self-reported health for obese men or women, but they suggest very large and significant improvements in health for obese women with coronary heart disease and obese male diabetics. All of these results should be interpreted with caution, as evidence of reporting anomalies in health appear to be present.

Additional Information

Language: English
Date: 2010
Health, Obesity, Self-reported health, Semiparametric, Wages
Obesity $x Economic aspects $z United States

Email this document to