Sending Students a Letter About Their Debt Might Not Be Enough to Change Their Borrowing

The Indiana University system saw a much-heralded drop in borrowing after sending students a letter about their debt. The system was doing other things to educate borrowers, but the letter is what caught the attention of policy makers and others.

A paper based on evidence from an experiment at the University of Missouri at Columbia, however, suggests that a letter alone may not be enough to change borrowing patterns. The Chronicle spoke with the researcher behind that experiment, Rajeev Darolia, an assistant professor of public affairs and education at Missouri and a visiting scholar at the Federal Reserve Bank of Philadelphia. The following conversation has been edited and condensed.

Q. You tested a debt letter like the one that made a big splash in the Indiana University system. Was the University of Missouri planning to send one anyway? How did you get involved?

A. The university was interested in doing something to help students understand their debt and make good student-loan decisions. After the Indiana University system initiative got so much press, both in the media but also among the financial-aid administrator community, this was one that we thought would make sense to try out.

Q. How did you conduct the experiment?

A. We took all student-loan borrowers, undergraduates only, at the university and selected those who would be returning the next year. We randomly assigned half of them to receive what we’re calling a loan notice, or the debt letter, as it’s been called elsewhere. They received this letter at two points in time. The letter includes information about their borrowing to date, their expected monthly payments, typical borrowing of their peers, and where they could get more data, how they could contact the financial-aid officer. We sent that last year, and then this year we observed how much they borrowed, their financial-aid receipt, how many credits they took, whether they dropped out. Then we compared the behavior of those who received the letter to those that did not.

Q. What did you find?

A. Something I would say was not surprising to me, though is a little bit counter to what’s been out in the public space based on some of those other studies. We found that the letter, in isolation, didn’t have an overall systematic effect on borrowing. We did find some behavioral changes for certain student subgroups, so it’s not to say that nobody changed their borrowing because of the letter, but over all we didn’t see a large-scale change. That’s somewhat different than the Indiana setting that found pretty large effects.

One other key thing that we found that didn’t manifest itself in differences in borrowing, but it’s still a really important outcome, is that students who got the letter were more likely to then seek more information. Those that received the letter actually came to the financial-aid office at a much higher rate than those who did not.

Q. Why weren’t you surprised your findings were so different from those in Indiana?

A. In the Indiana setting, there was a lot of attention given to this letter and the power of information — and I’ll say that I really believe in the power of information. What was less covered was the whole suite of other great programs that they had to encourage students to make active borrowing choices. What we did is only implement this low-touch informational letter, and not this whole suite of other programs. One of the main takeaways is that information in isolation, without other supports, may not be sufficient to really drive behavior. This is not a criticism at all of what they did in the Indiana setting; they did a lot of really great things, but I think they did a lot of really expensive things that took a lot of coordination, which was not possible in the setting that we had here.

Q. If information alone doesn’t change borrowing behavior, what would really make a difference?

A. I don’t want to make it seem as though we implemented the perfect debt-informational letter. There are certainly some design elements that could be improved. But even that takes some time and some resources to figure out. One of the big things is the availability of systematic support for students, not just to get information in a letter, but also the ability to sit down with somebody and really go through in detail about the implications of their decisions, and also to have that be delivered in numerous different ways. That’s really important, but it’s also very resource-intensive.

Q. You write in the paper that even without changing borrowing behavior, the letter might have some value in making students more informed. That’s interesting because we sort of elide the idea of people being informed and people making good decisions, but it seems like those could be two distinct things. Could you could flesh that out a bit?

A. A good outcome for students is not necessarily that students borrow less. For some students, borrowing less may be good for them; for others, borrowing more may actually make sense to allow them to complete. And so the outcome we found, that students were more likely to seek more information, I view that as a very positive thing in that if they’re making active and informed decisions about financial aid, even if it means they’re borrowing the same amount, or some students may be borrowing more. What we’re really concerned about is their burdens later in life. As long as they’re making active and informed decisions, then hopefully this will also lead to fewer issues down the road as well.

Q. You’re doing follow-up work interviewing the students. What do you hope to learn from that?

A. The experiment allows us to provide some level of information on what we can track within our data. But there’s a really important other piece of this: how students are dealing with the options that they have. A student may decide to borrow the same amount, but work less or more in response to the information. They could also decide to borrow less from the university, but they may be going out to a private lender, they may be going to their parents to ask for more money. These are things that we can’t track in administrative data. It will be helpful to understand more nuance about how students use or don’t use this information.

Q. There’s a big push to help students make better borrowing decisions, and it doesn’t seem as if anyone has that completely figured out. What can research contribute?

A. There’s a lot that can be done about ways we can reach students, whether that’s delivery mechanisms, through texting, or online, or other creative ways, but also how do we present that information to make it really salient to the student so they can take action.

There’s this whole big question about what is the proper debt load for students. There’s people saying that we should have no debt, and there are some saying that we should have manageable debt. But we don’t actually know what manageable means. Understanding what is manageable debt for students is very important.

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.