Understanding the effects of obesity on male and female labor outcomes Open Access
Downloadable ContentDownload PDF
When it comes to understanding the obesity effect on male and female labor outcomes existing studies largely focused on earnings. This study provides a broader analysis of the obesity impact on labor outcomes, including labor force participation, employment status, occupation type, hours worked, full-time employment, consecutive employment, and hourly earnings. This dissertation has three objectives: First, estimate the effect of being overweight, moderately obese and severely obese on male and female labor outcomes. Second, determine to what extent weight related labor inequalities are the result of individual, household, and local labor characteristics and whether discrimination plays a role. Finally, discuss a role for government interventions to reduce weight-related labor inequalities. This study used sixteen years of National Longitudinal Survey of Youth 1979 data. Both pooled OLS and pooled logit regression in combination with individual FE methods were used to measure gaps in labor outcomes. In addition, to address bias arising from endogenous relationship between obesity and labor outcomes I estimate the weight effect using instrumental variables approach with county obesity prevalence as instrument for obesity. Finally, I apply Oaxaca-Blinder decomposition method to examine the factors that contribute to the observed gap in hourly earnings. For both men and women the negative effects of being overweight, moderately and severely obese increases by weight category, but the labor market inequalities for men and women are distinct. In the first set of results that restricts analysis to employed people, female obese salaried and self-employed workers on average `lose' about $1,000 in annual earnings. For salaried and self-employed male workers, the effect of weight is not significant. These lost earnings do not take into account effects of obesity on individual choices and employment prospects captured by the other labor outcomes. For women, weight does not significantly affect those labor outcomes. But, weight does significantly affect male labor outcomes. Overweight men experience lower labor force participation and are less likely to be employed in white-collar occupation but have higher earnings than normal weight men. Moderately obese men are less likely employed in white-collar occupation and are more likely underemployed than normal weight men while severely obese men are less likely to participate in the labor force and experience underemployment than normal weight men. To get a sense of magnitude of these effects I find that these negative effects are even larger than the size of reported inequalities between white and African-American male employees. This evidence of labor outcome gaps should not directly be interpreted as a need for government intervention. The Oaxaca-Blinder decomposition results show that most earnings inequalities are explained by characteristics other than obesity and do not provide evidence of weight-based discrimination. Other grounds for government intervention may exist. Employers are burdened by growing health insurance premiums that pose a fiscal externality. Obese individuals use more health care services but its costs are shared among all employees and employers. As a result, obese employees do not feel the full impact of their behavior through health care costs and obesity prevalence may be suboptimal.