Choosing what subject to study at university can have substantial long-term effects on a graduate’s life. Everything from wages to prospective romantic partners is shaped by the discipline a person chooses for their ‘major’. But this decision is itself affected by access to (and the structure of) financial aid. At the margin, students with non-repayable tuition grants are more likely to choose fields that are associated with high earnings and employment opportunities. They also have a greater chance of choosing a STEM subject as their major.
When prospective university students choose a subject for their ‘major’, they are making one of the most significant investment decisions of their lives. Research shows that earnings differences between the highest- and lowest-paid degrees are comparable to the wage gap between graduates and those who never attended university.
Besides the pure impact on future earnings, students’ choices also influence expected job stability, occupational opportunities, and even the pool of romantic partners (Eika et al., 2019). Different fields of study also vary in the share of students who drop out or fail to graduate on time. Since attending university is not without cost—there is both a direct cost of tuition fees and living expenses, but also an indirect cost of forgone earnings while being enrolled—choosing the right field is also a risky decision.
In a recent paper, we ask the question of how financial aid policies affect these trade-offs. We focus on two scenarios: one in which the government issues student loans that need to be repaid after graduation and one in which students receive the same amount in the form of a ‘no-strings-attached’ grant.
There are several theories about how students might respond to each case. If they do not have to worry about repaying loans, will they choose fields with lower earnings? After all, students have different preferences for each field, and some might only have chosen the high-paying major to make sure that loans can be repaid. Or would they be less concerned about the risk of dropping out and, as a result, take their chances in a more challenging major? If programmes that are harder to finish are also the ones that result in higher-earning careers, can we still make a prediction about how financial aid affects the expected earnings of students?
High tuition, high stakes
We study these questions in the context of Chile’s higher education system. Until recently, Chile charged substantial tuition fees, often equivalent to about half the annual median family income for a year of study. The need for financial support was widespread and the government provided such assistance by subsidising student loans and through direct grants. What makes Chile an ideal setting for our study is that access to either type of financial aid is determined fully by a combination of family income requirements and performance on a standardised test.
The overwhelming majority of Chilean high school graduates who want to pursue a degree participate in a multiple-choice test. This is then standardised across all test takers. Students scoring above a specific threshold have access to grants covering a large portion of their tuition fees. Students below the cut-off must instead pay for their studies, typically taking out loans to do so.
Our research design relies on contrasting the college major choices of students close to the grant eligibility cut-off. The idea is simple. Test results range from 150 to 850 points, and in most years the grant eligibility cut-off is 550. If we compare the choices of students with 750 points to those of students with 350 points, this is likely to be an ‘apples-to-oranges’ comparison, given that one group performed drastically better than the other. But the closer we get to the grant cut-off level, the more likely it becomes that the two groups are comparable. The only difference between those who scored 550 points and those who scored 549 points is that one group had a luckier exam day and, consequently, now has access to grants.
Replacing grants with loans increases STEM enrollment
Using detailed administrative records provided by Chile’s Ministry of Education, we link each test taker in the years 2012–2014 to information about their university enrollment. This includes the institution as well as the chosen college major. When comparing students with test scores close to the grant eligibility cut-off, we find that access to grants shifts the allocation of students across different academic disciplines.
Figure 1 displays the differences in enrollment rates by subject between students who marginally became eligible for grants and those who had to rely on student loans to finance their studies. Enrollment in majors related to Science, Technology, Engineering and Mathematics (STEM) increased the most. While 25% of students below the cut-off enrolled in STEM subjects, this share grew by 2.9 percentage points among the students with access to grants. Enrollment rates also increased in the social sciences, but fell for the humanities.
Figure 1: The difference in college major choices at the grant eligibility cut-off (%)
Source: Authors’ calculations
Grants holders choose better-paid and riskier majors
In Chile, the subjects that lead to the best paid jobs include STEM, the social sciences, and health (Hastings at al., 2013). At first glance, our analysis does not suggest that students use the extra money received through grants to enroll in fields with lower earnings potential.
We confirm this by characterising students’ choices based on information from past graduates in the same programmes. Overall, students with grants are more likely to choose fields that are associated with high earnings and employment probabilities, as well as steeper wage growth in the first five years after graduation. They are also less likely to select majors where the risk of taking longer than the formal study duration to graduate is low.
Using a theoretical model, we then try to assess the relative importance of various college major characteristics. The results indicate that it is mainly concerns about dropout rates that keep students from enrolling in financially rewarding programmes. But having access to grants substantially changes this trade-off. Students who rely on loans are willing to forgo 6.4% of total earnings growth over the first five years after graduation to choose fields with a 1% lower dropout rate. Those with grants are only willing to give up 3.6%, highlighting their increased tolerance to risk.
Given our emphasis on study-related uncertainty, a natural concern is that students entering more challenging fields because of grants might not manage to graduate. But in our data we find no evidence for this. One potential explanation suggested by previous research is that grants might alleviate financial stress and allow students to focus on their studies instead of working in part-time student jobs (Broton et al., 2016).
Implications for policy
Our findings suggest that financial aid is likely to shape the composition of college majors. While discussions on optimal higher education financing often focus on the ‘extensive margin’ of college attendance—whether students enrol in college at all—we emphasise that such policies can have (potentially unintended) consequences which are often overlooked. In cases where these effects are explicitly considered, the discussion shifts from whether financial aid should be provided to how it might alter students’ decisions in ways that could be of concern to policymakers.
Specifically, policymakers may worry that lower tuition could deter students from choosing high-earning degrees, increase dropout rates, or prolong the time to degree completion (concerns supported by previous research). But we highlight that if grants serve to mitigate the risks associated with the college investment decision, students’ choices may instead be well aligned with policymakers’ objectives.
Higher education is a risky investment, and risk-averse students may avoid programmes that are a better fit for their abilities and long-term earnings potential simply because they fear the cost of failure. By reducing this risk, financial aid can encourage students to pursue degrees that both match their interests and have a high monetary return, rather than settling for safer but potentially suboptimal choices. Tuition grants can ‘de-risk’ higher education, with potential long-run benefits for graduates as well as the wider labour market.
Authors: Adriano De Falco and Yannick Reichlin