This paper has now been published in the Quarterly Journal of Economics.
The choice of where to go to college and what major to study is one of the most complex and consequential decisions that an individual can make[i]. Yet despite its importance for future earnings, employment, and life trajectories, we know little about how the preferences and beliefs that drive this decision are formed and if they can be changed. Recent evidence indicates that family background and other social factors are important in shaping college and major choices, suggesting that relatives and social networks could signiﬁcantly inﬂuence them. However, causally identifying these social interaction eﬀects has been challenging.
In a new working paper (PDF) published with the Industrial Relations section at Princeton University, we investigate how college applications and enrollment decisions are influenced by one of the most important social peers a person has when growing up: older siblings. Using data on the universe of applicants from centralized school systems in three countries, we provide causal evidence that social interactions greatly influence where students go to school. Specifically, we find that younger siblings are signiﬁcantly more likely to apply and enroll in the same college and major as their older siblings.
These findings suggest that college access programs, like affirmative action, not only affect the higher education trajectories of their direct beneficiaries, but also the choices made by other members of their families.
Our research setting: Centralized college admissions systems in Chile, Croatia, and Sweden
Centralized admissions systems in Chile, Croatia, and Sweden provided a valuable setting for our research. In each of these systems, students applying to college are allocated to schools and majors only depending on their stated preferences and previous academic performance. These selection systems give rise to sharp admission cutoﬀs in all oversubscribed majors, where some students received their preferred placement based on random luck.
We look at students near the cut-off that receive their preferred placement and study whether and how their enrollment in a speciﬁc institution or major affects their younger sibling’s probability of applying and enrolling in the same program and institution.
College and major choice is often a family affair
We find that when an older sibling slightly above of the cutoff is able to enroll in his preferred major, the likelihood of his younger sibling applying to that same program increases between 2.5 and 4 percentage points. Further, students are between 9.5 and 15.5 percentage points more likely to apply to the college where their sibling is enrolled and between 4.5 and 9 percentage points more likely to enroll there.
The eﬀects that we document are stronger when students resemble their older siblings in terms of gender and academic performance in high school, and when older siblings enroll and are successful in majors that on average have higher scoring peers, lower dropout rates and higher earnings from graduates. Despite the diﬀerences that exist between Chile, Croatia and Sweden, we ﬁnd similar spillover magnitudes in all three countries.
We explore several hypotheses for why younger siblings are more likely to apply and enroll in the same college and major as their older siblings. We find little evidence that the decision is based in practicality (i.e. it’s easier to commute together), nor do we find evidence that it’s based in siblings wanting to attend school together.
Though we can’t perfectly distinguish the exact mechanisms behind our results, we provide suggestive evidence consistent with the idea that the effect is driven by older siblings being an important source of information about a school’s quality and the type of experience young siblings can expect.
How information and college access programs can improve student outcomes
Identifying these family spillovers has important policy implications. First, the findings could help to explain inequality in education uptake and trajectories across families and socioeconomic groups. Relatedly, the effects show how policies that change the pool of students admitted to speciﬁc programs and institutions, such as aﬃrmative action, can have an indirect multiplier eﬀect on members of the social network of their beneﬁciaries.
Finally, if the reason students are making education decisions based on their older siblings’ choices is because of incomplete information about other options, then it’s possible to improve the match of students and educational programs by providing that information.
Our research provides many reasons to be optimistic that policymakers can improve student’s higher education outcomes. With the right technology, tools, and platforms, higher education institutions and governments can help students forge the academic path that’s best for them.
[i] (Altonji et al., 2012; Oreopoulos and Petronijevic, 2013)