The effect of child-care costs on the probability that married women with children will participate in the labor market is examined. Child-care costs are estimated.using a generalized Tobit specification corrected for selection. Estimates of a structural probit model of labor-force participation provide evidence to support the prediction that increased child-care costs lower the probability of participation. It is also shown that the lower rate of labor-force participation among mothers of preschoolers is entirely the result of the higher child-care costs faced by these women and endogeneity of the number of young children in the participation equation.

This article develops analytic links between cohort composition and human capital accumulation across generations. By focusing on cohort composition rather than cohort size, it offers new links between demographic change and economic outcomes. The model shows that changes in the educational attainment of parents and changes in relative fertility rates between educational classes affect human capital accumulation of the next generation. The main prediction of the model, of a negative behavioral feedback, is shown to be consistent with data on the probability of attending college for men born between 1927 and 1962.