In the NLSY79, 69% of women in the lowest cognitive ability quartile have a first birth before age 22, compared with 22% in the highest. This gradient persists within education groups and within contraceptive method, suggesting that neither education sorting nor method selection fully accounts for it. I estimate a life-cycle model in which cognitive ability shifts the effectiveness of fertility control, beyond standard opportunity-cost channels. Restricting effectiveness to be equal across ability groups causes the model to underpredict the ability–fertility gradient by a factor of five. Equalizing effectiveness to high-ability levels reduces births before age 22 by 50% and raises college attendance by 15%. A cost-reduction policy generates welfare gains of 10% of lifetime consumption for the lowest ability quartile but near-zero gains for the highest, suggesting that improving effectiveness matters more than reducing cost for disadvantaged women.

I study how parental altruism shapes investment in college. Altruistic parents support children when resources are low, financing college and insuring income risk but—because they cannot commit to withholding support—creating a Samaritan’s dilemma. College disciplines it where cash cannot, by permanently raising the child’s income. In matched parent–child data, parents of college children consume about 8% more (roughly $2,500 a year) yet transfer about a fifth less often—the signature of releasing the precautionary resources held to stand ready to help. I quantify the mechanism in a continuous-time dynastic model estimated by simulated method of moments. The inability to commit distorts the college decision unevenly: it raises attendance for low-ability children (by 13–18 percentage points), for whom anticipated support makes college affordable, but lowers it for the medium- and high-ability (by up to 27 points), who can lean on the parent’s lifetime safety net instead of investing to become self-sufficient—a small net decline. The same transfers that distort the college margin also insure children, so severing the relationship to remove the dilemma—through a one-time transfer—lowers welfare for both parent and child, about 12% for the dynasty; only full commitment, which removes the dilemma while preserving insurance, raises dynastic welfare (+15%), and it does so by reallocating from child to parent rather than as a Pareto improvement. Free college operates through the same relationship: it reallocates resources to the child (about +5%) rather than financing many new entrants, leaving the dynasty roughly neutral.

We use the Penn Wharton Budget Model’s microsimulation of U.S. demographics projections to construct estimates of the U.S. federal fiscal and generational imbalances. The federal government’s fiscal imbalance (FI) calculated under current fiscal laws and purchases policies over the next 75 years equals $93.8 trillion, which is 7.0 percent of the present value of projected GDP (PVGDP) over that time horizon. Calculated in perpetuity, FI equals $202.9 trillion, which is 8.2 percent of PVGDP, also calculated in perpetuity. The FI/PVGDP ratio in perpetuity would be 9.4 percent under extension of provisions that are scheduled to expire under the Tax Cuts and Jobs Act of 2017.

We analyze the general equilibrium effects of human capital misallocation in Chile. First, we utilize tax and educational records to estimate the proportion of educationally mismatched individuals (high-ability individuals with low educational attainment). Second, we estimate the labor market returns on ability, education, and human capital investment. Finally, we construct and calibrate a dynastic overlapping generations model with both private and public human capital investment to decompose the causes of educational mismatch and the general equilibrium effects of changes in sorting.

The U.S. teen birth rate fell by roughly three-quarters between its 1991 peak and the early 2020s, one of the most dramatic demographic shifts of the period. This paper asks which economic mechanisms drove the decline and how they interacted. I estimate a dynamic life-cycle model of schooling, work, marriage, and fertility separately on two cohorts of women—the NLSY79 (teens in the late 1970s and early 1980s) and the NLSY97 (teens around 2000)—that together span the bulk of the decline. Comparing the estimated model across cohorts, I decompose the change in early childbearing into the contributions of distinct economic primitives, including the returns to schooling and work, the cost and effectiveness of fertility control, schooling costs, the marriage market, and preferences for children. By isolating the primitives behind early childbearing, the analysis contributes to explaining the causes of the broader decline in fertility in the United States and elsewhere.