Picking a college is a complex process. One little-studied component is the role of students’ siblings. A new paper demonstrates what one might expect: The choices of younger siblings generally follow the patterns set by older ones. But there’s more to the story. I spoke with one of the authors, Joshua S. Goodman, about the paper and his research team’s planned follow-up work to tease out the causality in this relationship. The conversation with Mr. Goodman, an associate professor at Harvard University’s John F. Kennedy School of Government, has been edited for clarity and length.
Q. Why look at siblings as a factor in college choice?
A. There’s been a lot of work in economics of education about “peer effects,” the notion that the students who are in a classroom with you, or your college roommate, or other kids in the community affect various educational outcomes. There has been some work about how siblings affect each other, but it’s actually harder work to do, in part because unlike college roommates and classmates in school — where sometimes there’s some randomness to whom you get placed with — with siblings that’s not true.
The College Board’s data, because it has people’s last names and home addresses, allows us to identify people who are siblings, and then we connected that to all of these college outcomes.
We’re going to be writing a follow-up paper where we look at what happens when the older sibling is just barely admitted to college based on SAT scores, and studying how that affects younger siblings. So there’s going to be a paper where we’re trying to get at something causal. This paper was just to look at what the correlations suggest.
Q. How can you separate out the older siblings’ college choices from other things the siblings share by being in the same family?
A. We are able to control separately for each sibling’s SAT scores and high-school GPA, so we can control for the fact that siblings themselves can be different in terms of their academic skills. But you’re absolutely right: In the end it could be that, even after controlling for all of that, that the reason siblings’ choices resemble each other is because of something going on in the family. That’s why we’re trying to do a follow-up paper to see if we can pin that down a little more carefully.
Q. Describe the relationship you find in this paper.
A. Controlling for both the younger sibling’s and older sibling’s academic skills, younger siblings make very similar college choices to older siblings. If an older sibling enrolls in a four-year college, the younger sibling is something like 16 percentage points more likely to do so. If the older sibling enrolls in a highly competitive college, the younger sibling is something like 19 percentage points more likely to.
When we looked at siblings’ genders, and how far apart they are in age, and how far apart their academic skills are from each other, what we find is that siblings who look more like each other tend to follow each other’s college patterns more.
Q. Is that what you expected?
A. I was certainly expecting to see that younger siblings tend to do what their older siblings do. I guess I’m not surprised that if they’re of the same gender, they’re more likely to do similar things, but I didn’t have strong feelings about that one going in.
I went to Harvard as an undergrad, and my younger sister went to Yale, and it’s not clear if that was because she wanted to get away from me or not. She actually fits the pattern in this paper because she’s five years younger than I am, and a different gender, and a lot smarter than I am, and she went to a different college.
Q. But a different college that is, whatever people at Harvard and Yale might want to think, still pretty similar.
A. That’s right, in that sense she fits all the patterns. But she didn’t go to the exact same institution. One fact that we document, which was very striking to me, is that one-fifth of younger siblings go to the same college as their older siblings.
Q. Tell me more about the follow-up paper.
A. The challenge is, can we find some sort of randomness in the college choice that an older sibling makes that we can use as a part of an experiment on the younger sibling?
We have found a handful of colleges in the U.S. that appear to use SAT cutoffs as a part of their admissions process. Imagine two older siblings, each of whom has very similar SAT scores, but one scored 10 points less than the other one. One scored 990, and the other scored 1000. And both of these older siblings are applying to a college that uses 1000 as a cutoff. So one of these older siblings gets in, and the other doesn’t. Even though they’re very, very similar older siblings. And then we watch what happens to the younger siblings of those two kids.
If the older siblings are very similar, then, on average, the younger siblings will also be quite similar. The only difference is that one of them has an older sibling who squeaked into the college he wanted to get into, and the other one had an older sibling who was rejected. We can trace out the effect of that on the younger siblings.
Q. Walk me through the implications of this research for people trying to help students make good college choices.
A. Imagine you had a finite number of guidance counselors trying to help students go to college, and you were trying to identify which kids are either not going to go to college when they should, or they’re going to go to a college that is not a good-enough quality given their own skills. Part of what our work is pointing out is if you knew something about what their older sibling had done, you’d be able to make a better guess about what the younger sibling will do. Because these patterns run in families, you might use older siblings to help warn you about what might happen with younger siblings and catch them earlier.
If it’s true that when we do the causal work the older sibling’s college choices really do affect their younger siblings, I think it may teach us a little bit about Michelle Obama saying that if Princeton hadn’t recruited her brother for basketball, she never would have even thought to apply to Princeton. It would be really nice if we could get some more rigorous evidence that that pattern is more widespread; that for kids, particularly in families that don’t have great information or long histories of college enrollment, once you start that information flowing with one kid, it continues to flow to the other kids.
Beckie Supiano writes about college affordability, the job market for new graduates, and professional schools, among other things. Follow her on Twitter @becksup, or drop her a line at beckie.supiano@chronicle.com.
