This project will exploit the """"""""natural experiment"""""""" of twin births based on a newly-collected sample of approximately 7,000 male and female identical and non-identical twins born in Minnesota. Models of intrafamily allocation, assortative mating, and intergenerational interactions will be applied to the new data, which contain contemporaneous and retrospective information about the twins, their spouses and their children and parents and permit for the first time separate analyses of female twins. The models incorporate (i) differences among individuals in innate or endowed labor market skills and health that may be intergenerationally correlated, (ii) assortative mating by observed and unmeasured traits, (iii) multiple dimensions of schooling, (iv) multiple dimensions of child endowments, (v) adult child-parental interactions, (vi) measurement error, and (vii) dependence of investments in the human resources of any one child in a family on its own endowments and on the endowments of other children in the family. The project will yield estimates of (i) the effects of endowments on school investments, on work experience, on marital sorting, on family size and child human capital, on personal interactions between adults and their parents, and on parental-child financial transfers and (ii) the effects of the amount and quality of schooling and health-related behaviors on own health and own fertility, on the characteristics of marital mates, on attention paid to one's parents and transfers received from and given to one's parents, and on children's health that are better than any available in the literature. The applications will yield insights into the importance of heterogeneity in endowments in biasing estimates of schooling relationships, assortative mating and intergenerational interactions generally in determining child outcomes. The estimates will also permit comparisons of the impact of male and female schooling on a range of outcomes, including earnings, work histories and the health and education for sons versus daughters. The results of the project will provide important information on both substantive and methodological issues concerning household behavior, the determinants and consequences of schooling, and intergenerational family linkages that are important for understanding the intergenerational transmission of equality and for evaluating the consequences of schooling policy.
