We use data on monozygotic twins to obtain improved estimates of the effect of intrauterine nutrient intake on adult health and earnings and thus to evaluate the efficacy of programs aimed at increasing birthweight. We use the results to evaluate the bias in cross-sectional estimates and to assess the proposition that health conditions play a major role in determining the world distribution of income. We show that there is considerable variation in the incidence of low birthweight across countries, and our estimates suggest that there are real payoffs to increasing body weight at birth. Increasing birthweight increases adult schooling attainment and adult height for babies at most levels of birthweight, but has no effect on adult body mass. The effect of increasing birthweight on schooling, moreover, is underestimated by 50x% if there is no control for genetic and family background endowments as in cross-sectional estimates. We also find evidence that augmenting birthweight among lower-birthweight babies, but not among higher-birthweight babies, has significant labor market payoffs. However, shifting the distribution of birthweights in developing countries to that in the United States would reduce world earnings inequality by less than 1%, far less than indicated by the cross-country correlation between per-worker GDP and birthweight.

New survey-based data on siblings are used to assess potential roles of bequests in redistributing income among siblings as implied by prominent models. The data are not focused on the upper tail of the wealth distribution and include both own and sib reports on own and sib's bequests, enabling use of a measurement model. Results indicate that bequests are received by almost two-thirds of children, average bequests are significant fractions of annual earnings, and there are significant differences between sibs' schooling and earnings. However, there are not significant sib differences in bequests once measurement error is incorporated into the analysis.

Nonexperimental data are used to evaluate impacts of a Bolivian preschool program on cognitive, psychosocial, and anthropometric outcomes. Impacts are shown to be highly dependent on age and exposure duration. To minimize the effect of distributional assumptions, program impacts are estimated as nonparametric functions of age and duration. A generalized matching estimator is developed and used to control for nonrandom selectivity into the program and into exposure durations. Comparisons with three groups—children in the feeder area not in the program, children in the program for ≤ 1 month, and children living in similar areas without the program—indicate that estimates are robust for significant positive effects of the program on cognitive and psychosocial outcomes with ≥ 7 months' exposure, although the age patterns of effects differ slightly by comparison group.