Search Results

Almost 1 out of 3 white women in the U.S. are obese and data on white women born in 1960-1964 from the National Longitudinal Survey of Youth 1979 (NLSY79) show that, at age 30, obese women have completed almost 1 fewer school grade, are less likely to participate in the labor market, and if they work they have wages that are lower by 17%; they are more likely never to have been married and, if they are married, their spouses have incomes that are lower by 27%.

I interpret these facts taking into account that (a) body mass may affect labor- and marriage-market opportunities; (b) behavioral factors, potentially influenced by schooling attainment and family income, play a key role in the accumulation of body weight; and (c) women may be heterogeneous in terms of their propensity to gain weight on the one hand and labor- and marriage-market endowments on the other hand.

To this end, I specify and estimate a dynamic model in which (a) wage and spousal income offers, as well as the arrival probability of marriage offers, depend on body mass; and (b) from the time they leave school, women make decisions about their labor market participation, marital status, and body mass to maximize lifetime expected utility. Their utility function represents preferences over consumption, leisure, marital status, and body mass, where preferences over body mass capture the psychic costs of keeping the balance between energy intake and energy expenditure. Such preferences depend on recent fertility history, consumption, and schooling attainment; reflecting the role played by family background factors in the accumulation of body weight, they also depend on unobserved permanent characteristics, as do wage and spousal income offers. The model is estimated by the method of simulated maximum likelihood using a pa nel dataset on the body mass, labor market, marriag e market, and fertility histories of white women born in 1960-1964 from the NLSY79.

My results can be summarized as follows. In line with results on schooling attainment and labor market outcomes, I find that unobserved permanent characteristics play quite an important role also in the determination of body mass outcomes, explaining 16% of the cross-sectional variation in obesity at age 30. Furthermore, these characteristics are important in accounting for the observed relationship between body mass on the one hand and wages and spousal incomes on the other hand: the type of women who are most likely to become obese receive wage and spousal income offers that are lower by respectively 53% and 29%.

In itself, obesity has negative consequences in the marriage market, but not in the labor market: the odds ratio of receiving a marriage offer declines by 15% for obese women, who also receive spousal income offers that are lower by 7%. Without such marriage market incentives, the prevalence rate of obesity at age 30 would be higher by 21%.

More than 10% of women in my sample are already overweight when they leave school. In the concluding counterfactual experiment I address the following question: how effective in preventing adult obesity would be policies aimed at eliminating excess weight at the time of school leaving? I find that the prevalence rate of obesity would be persistently lower, even though initially overweight women are intrinsically much more inclined to gain weight than the average.