03 We show how comparisons between the within-twin correlations of human capital outcomes across identical and nonidentical twins can be used to identify the variability in the individual-specific component of endowments and the responsiveness of schooling to individual-specific endowments in the family and in the marriage market even when schooling is measured with error. Estimates from two twins samples indicate that 27 (42) percent of the variance in log earnings (obesity) is due to variability in individual-specific endowments, allocations of schooling reinforce specific endowments, and individual-specific earnings endowments of men and their wives' schooling are negatively associated.

Unequal access to financing for education may be an important source of educational differences. We develop a model relating sib schooling and earnings similarities to sibship size with and without equal access and estimate it for the education of veterans, for whom the GI Bill assured equal access, and for their children, who had no such government assistance. We find an inverse relationship between sibship size and sib schooling and earnings similarities for the children, but not for the veterans; we conclude that, in the absence of equal access policies, unequal access is an important source of educational differences.

This paper develops a general preference model for analyzing parental allocations of resources among their progeny. The implications fro this model for the distribution of educational resources and earnings potentials among siblings are examined. A particular version of the preference model is estimated using data on the education and earnings of adult male twins. The estimates imply that parents care about offsprings' earnings inequality and provide more (less) resources to the less (more) able than is consistent with an investment model.

Parents can transmit income and wealth to their offspring directly through bequests and gifts of financial assets. Parents can augment a child's skills and traits through the provision of, for example, education, diet and doctors. Biological parents also supply the genetic endowments which provide skills useful in the labor market. Although practically a virgin territory, it is important to estimate the effects of genetic endowments and common (mostly family) environment on earnings and on other indices of individual attainment in order to gauge the effectiveness of programs designed to improve equality of opportunity. Also if genetics and common environment are not controlled for, biased estimates of regression coefficients of observable variables, such as years of education, often are obtained.

We modify R. A. Fisher's model by incorporating measures of the environment that may be correlated across kin groups. We estimate a model of schooling attainment using data on eight kin groups and find a large contribution of genetic endowments to the variance in schooling though certain aspects of the environment matter.

The income elasticity of calories generally is substantially smaller than the income elasticity of food expenditure. One reason may be an increasing concern for food variety as incomes increase. Food variety can be linked with two characteristics of food indifference curves: (1) curvature and (2) location of the curves relative to the axes. Estimates suggest increasing taste for variety as food budgets increase. Therefore, such taste for variety apparently underlies in part the low income elasticities for calorie demand, which in turn cast doubt on the World Bank assertion that the nutrient intakes of poor populations will improve rapidly with income.

Unequal access to financing for education may be an important source of educational differences. We develop a model relating sib schooling and earnings similarities to sibship size with and without equal access and estimate it for the education of veterans, for whom the GI Bill assured equal access, and for their children, who had no such government assistance. We find an inverse relationship between sibship size and sib schooling and earnings similarities for the children, but not for the veterans; we conclude that, in the absence of equal access policies, unequal access is an important source of educational differences.

This paper develops a general preference model for analyzing parental allocations of resources among their progeny. The implications fro this model for the distribution of educational resources and earnings potentials among siblings are examined. A particular version of the preference model is estimated using data on the education and earnings of adult male twins. The estimates imply that parents care about offsprings' earnings inequality and provide more (less) resources to the less (more) able than is consistent with an investment model.