These patterns raise an important question: Do college majors themselves shape later-life outcomes, or do they simply reflect differences among the students who choose them? Students who study Engineering, for example, may differ from Teaching students long before college begins. Chile’s centralized admission system helps us address this problem. Students submit rank-ordered lists of majors, and admission depends heavily on cutoff scores and seat availability. As a result, students with very similar backgrounds and preferences can end up in different fields of study for reasons that are partly outside of their control. Combining this feature of the admission system with detailed information on students’ preferences, degree completion, labor market earnings, and fertility allow us to estimate how different majors causally shape adulthood outcomes. 

We find significant differences across majors. Graduates from Medicine, Engineering, and Business experience the highest earnings payoffs. The differences are large enough that they explain an important share of inequality among college graduates. But the differences across majors are not only monetary. Fields of study also differ sharply in their relationship with fertility. Some majors are associated with earlier and higher fertility, while others are associated with delayed childbearing and lower fertility rates in the early thirties. For example, graduates from Teaching and Health tend to have children earlier than graduates from Law. These patterns are consistent with additional evidence we document: fields associated with higher fertility tend to lead to careers with greater work-family compatibility, including more stable employment, smaller child penalties, and more progressive gender norms. 

With gender-specific causal estimates of how majors affect earnings and fertility timing, we can examine whether men and women place different weight on these tradeoffs when choosing college majors. Using students’ reported preferences together with our estimates of the earnings and fertility returns to different majors, we show that both men and women respond to expected earnings. However, women place greater weight on how different careers may affect fertility, suggesting that career-family considerations play a more important role in major choice for women than for men. This distinction helps explain persistent gender differences in major choice. Some careers offer higher salaries but involve greater penalties for career interruptions or less predictable schedules. Others provide more stability and flexibility but lower earnings. Students appear to understand these differences and incorporate them into their educational decisions long before entering the labor market. As a consequence, women remain underrepresented in high-paying fields while they are overrepresented in fields associated with greater career stability and higher fertility. Our results suggest that these differences in preferences explain roughly 30 percent of the gender earnings gap among college graduates. 

What can policy do? Using our estimates and the algorithmic structure of Chile’s centralized college admissions system, we evaluate how alternative admission policies could affect representation and inequality. One option is to simply increase the number of seats in Engineering and other high-return fields. Another is to reserve part of those seats for women. Our results suggest that broad expansions alone improve overall access but do little to reduce gender inequality because men and women apply to these programs at different rates. In some cases, these expansions may even widen earnings gaps. By contrast, targeted admissions policies for women can substantially reduce representation gaps and have the potential to reduce earnings inequality with small impacts on aggregate earnings or fertility