Project:
The Effects of Female Labor Force Participation on Adult and Childhood Obesity

Year: 2009

Investigator: Mitnik, Oscar A., Pedro Gomis-Porqueras, Adrian Peralta-Alvay, and Maximilian D. Schmeiser

Institution: University of Miami

Project Contact:
Oscar Mitnik
University of Miami
Department of Economics
P.O. Box 248126
Coral Gables, FL 33124-6550
Phone: 305-284-1626
E-mail: omitnik@miami.edu

Obesity is a critical issue faced by adults and children in the United States and in many other developed and developing countries. For the first time in history, the number of overweight individuals around the world rivals the number of those underweight. In 2003, the World Health Organization reported that more than 1 billion adults were overweight and at least 300 million of them clinically obese. Understanding the causes behind recent increases in obesity rates is fundamental for devising policies aimed at controlling and eventually lowering obesity rates.

This study assessed whether labor force participation of single mothers is associated with obesity of both mothers and their children. To understand and quantify the causal effects of changes in female labor force participation on obesity, the study exploited a “natural experiment” generated by the 1987 and 1994 expansions of the Earned Income Tax Credit (EITC).

The contribution of using changes in the EITC to study obesity is that they provide a credible exogenous variation in labor force participation. The study exploited two important features of the EITC. First, the expansion in the value of the credit starting in 1987 introduced an important incentive for women with children to increase their labor force participation. Therefore, comparisons of single women with and without children, before and after the 1987 expansion, were one of the sources of identification. Second, the study exploited the differential effects of the EITC for families with one child versus more than one, which became relevant after the 1994 EITC expansion. Thus, the second strategy for identification compared families with more than one child versus families with just one. These policy changes provided a credible empirical strategy with which to study the effects of labor force participation on obesity. Specifically, by comparing between groups and across time, we were able to control for other confounding factors that might be related to changes in both labor force participation and obesity.

To implement the empirical strategy, we used two different datasets with different characteristics, which allowed us to perform complementary analyses. The first dataset is the National Health Interview Survey (NHIS), a nationally representative annual sample of U.S. households. The second data source is the National Longitudinal Survey of Youth 1979 (NLSY79), which contains information on adults, and the NLSY79 Children and Young Adults (NLSY79CYA) for the information on their children. Estimating the adult models by using both repeated cross sections and longitudinal data made the study more robust as it can exploit different sources of identification with each dataset.

Using NHIS data, we replicated the findings of the EITC literature regarding the positive effects of employment for single mothers. These results are robust with regard to different identification strategies, either by comparing women with children to women with no children or by comparing women with two or more children to women with one child. In addition, we estimated differences- in-differences regressions and exploited differences in maximum EITC benefits. Using NLSY79 data, we also replicated most of those employment results, but for some groups, these are very noisily estimated. In general, we found that increased labor force participation modestly affected the adults’ body mass indexes (BMI), overweight rates, and obesity rates. In addition, based on NLSY79 data, we found evidence of an increase in the children’s overweight and obesity rates, particular for children of low-education mothers. However, we are reluctant to attribute these increases to changes in the labor force participation status of their mothers because the adults’ regression coefficients are very imprecisely estimated.

We interpret the employment effects and the results of the robustness checks as justifying our identification strategy. We also interpret the variety of obesity-related results as weak evidence of a causal link between female labor force participation and increased adult and childhood obesity. The results apply to a particular population, single women with children with relatively poor labor prospects, which make them potentially affected by the EITC policy changes. Future research could address whether our conclusions would apply to married women and women with higher earnings potential.

Direct inquiries about this study to the Project Contact listed above.