The Returns to Increasing Body Weight

Two major health-related policy objectives are to increase the birthweights of children in low-birthweight populations and to reduce weight or body mass among most adults. There have been a number of previous studies on these topics. However, the literatures concerned with the consequences of weight gain at birth or in adulthood generally do not provide a clear interpretation of the observed linkages between physical characteristics, adult human capital investments and adult earnings. A problem with prior studies of the consequences of birthweight variation, for example, is that the policy-relevant effects of increasing the nutrition received by a fetus and possible genetic influences on fetal development are not distinguished. Moreover, generally researchers are inattentive to estimation problems that arise from endowment heterogeneity in assessing the effects of anthropometrics on adult human capital. Three recent exceptions to this generalization use data on relatives, but their resolutions of these estimation problems are incomplete. For example, the sibling-based studies ignore the responsiveness of resource allocations in the family to endowment differences.  This paper obtains improved estimates of the impacts of anthropometrics at birth and in adulthood on critical outcomes over the life cycle within a context in which individuals are born with health and earnings endowments that may be correlated with each other and across generations and optimally invest in health and earnings capacity. A simple three-period optimizing model is constructed that determines investments in human capital and physical characteristics and that incorporates endowment heterogeneity. The model (i) illustrates the difficulties in identifying the effects of at-birth and adult physical characteristics in the labor market and (ii) shows how both the distinct roles of endowments and of early health investments can be identified using data on identical twins. Estimates using use new survey data on female twins collected by the authors from a sample from the Minnesota Twins Registry provide a number of clear results. (1) Increasing fetal growth, for given endowments, has a significant positive effect on subsequent schooling attainment. The effect of increasing birthweight on schooling, moreover, is underestimated by 50% if there is no control for genetic endowments as is the case for cross-sectional estimates. This suggests that either nutrient consumption is negatively correlated with genetic endowments or that health and ability endowments are negatively correlated. (2) Intrauterine nutrient consumption does not have effects that persist to affect significantly adult BMI - increasing birthweight is not a cause of adult obesity. (3) The genetic birthweight endowment also is the component of birthweight that plays the dominant role in the intergenerational correlation of birthweights. (4) In contrast, intrauterine nutrient consumption plays the dominant role in determining adult height, consistent with the literature that makes use of height statistics to gage childhood nutritional investments over time and across countries. (5) The significant inverse association between adult BMI and wages found in cross-sectional estimates solely reflects a correlation between unmeasured earnings endowments and BMI, and disappears with control for endowments common to monozygotic (MZ) twins. (6) The significant positive association between adult height and wages found in cross-sectional estimates is increased substantially with control for endowments. There is thus evidence that augmenting birthweight in low-birthweight populations, which evidently increases both adult schooling attainment and height, has real labor-market payoffs, while the returns to controlling adult body mass in high-income settings may be illusory, at least in terms of labor market consequences for adults of prime labor-force age.

Download Paper

Paper Number
01-052
Year
2001