There’s been a lot of new data and analysis of student loans and colleges in the past week, including a new Brookings paper and the launch of the College Scorecard by the Department of Education. And with so much data coming out, it’s becoming more important that we keep our questions open-ended.
The Brookings Report, “A crisis in student loans?” (CSL) by the Treasury Department’s Adam Looney and Stanford’s Constantine Yannelis, is an important piece of research. It puts together a lot of evidence from other sources with a new, detailed data source. But there’s one criticism I want to explore: By focusing on a student loan “crisis” only in terms of defaults, we limit what problems we can see, with potentially serious consequences for an entire generation.
The case CSL makes is that there is no general student loan default crisis. Instead there is a serious, though limited, problem concentrated in for-profit schools and, to a lesser extent, community colleges. This was made far worse by the Great Recession, as many people went back to school, often falling into programs they weren’t prepared for.
I agree with this assessment. Student loans defaults from for-profit schools are a genuine problem, and the media often fails to recognize this. As Astra Taylor notes, when student debt activists in the wake of Occupy put forward people defrauded by the for-profit industry, the media prefers to talk about poetry majors with outrageous debt balances instead.
We should understand much of this is by design. For-profits were allowed to expand rapidly in the 2000s, and they’ve done a remarkable job in maximizing their profits. The researchers note that two-year community colleges have seen their borrowing increase $2,000 between 2000 and 2011. Yet, from the Delta Cost Project, the yearly net tuition increase for students attending a community college is up around $950 a year between 2000 and 2010. If the public policy is to shift costs from states to individual students, we shouldn’t be surprised when it goes perfectly to plan.
So there’s a crisis in which poor communities have very large debt balances relative to income, forcing such distress that the results are large default rates. But there’s another issue, and that’s the life effects of student debt. And default rates won’t catch this.
Distress Beyond Defaults
CSL and the coverage focuses largely on defaults. Yet the paper also mentions, but does not address, a wide body of research which finds that student loans could “impair students’ abilities to finance first homes and to live independently of their families, or could constrain their occupational choices, reducing rates of homeownership and marriage, or entrepreneurial risk taking.” That’s a lot of impacts! And it’s these channels that researchers have been investigating anf finding results in recent years.
The term crisis is loaded, and lord knows I’m guilty of abusing it as well, but this use overlooks two important points. A crisis can exist outside the narrow problem of defaults: Why aren’t students who don’t default, but are still forced to overhaul their lives in major, historically unprecedented ways, considered to be in crisis? It’s certainly a matter of public policy.
Yet if by “crisis” we mean a system that is going to collapse without major interventions (e.g. the financial crisis), there’s no crisis here. The economy and student loans can continue to chug along with large defaults among small groups. Our system has gotten by just fine by punishing the poor relentlessly, and if the idea is that high defaults will trigger a crisis moment, I don’t see it.
If the worry is about the government collecting its debts, remember that defaults are only “defaults” in a system in which bankruptcy isn’t allowed, as it isn’t for student loans. Many have made the comparison between student loans and subprime mortgages, yet the losses on defaulted student loans are near 0 percent. The debt stays there and the government goes to great lengths to collect it eventually. Compare that with financial losses on mortgage defaults, which are well over 50 percent.
On the other hand, distress doesn’t have to lead to defaults. Debtors who simply extend out the payment period, reducing the monthly payment, are considered fine. People with some resources and sophistication have many ways of avoiding default, including forbearance, deferral, or entering into an income-based repayment plan in which payments can be low enough to actually increase the balance of their loan.
CSL includes a rate of repayment graph that shows what a transformation is happening:
You can see the time to repayment expanding significantly in recent years. Something is changing in the lives of our young people, and it’s broader than the distress of defaults.
The College Scorecard has a data webpage, and the data documentation notes that default rates are “susceptible to gaming behavior that may push students toward forbearance and deferments, meaning they stay out of default but don’t make progress on repaying their loans and may continue to accrue interest. The repayment rate resolves many of those issues.” The repayment rate is simply how many people are making progress in paying down the principal of their loan, and I’d argue it’s a strong sign of another level of distress.
The College Scorecard data shows repayments by parental FAFSA income. Here’s the density of repayments by low- and high-income households for selective private, nonprofit schools (selective defined as having an admission rate less than the median private nonprofit one):
This is as elite as it gets, and even here there’s a clear divide between high- and low-income families and the distress their kids experience. (Low is under $30,000, high is above $75,000; this is generalizable to other types of schools.) If young graduates aren’t paying down their loans, that directly impacts their ability to build wealth, get loans, and start businesses, and has indirect impact on how they choose to get married or have kids. This goes double for kids from poor families.
Preliminary playing around with some College Scorecard data tells me there’s a there there. Schools that have higher repayment rates for high income families have higher ones for low income families. Yet regressing on all schools, whenever a school has a repayment rate go up 10 percent for high-income families a low-income family’s repayment rate only goes up 5.8. This looks directly related to student debt levels itself. Higher student debt correlates with a lower repayment rate, even when you control for incomes and school selectivity. This needs to be incorporated within the larger discussion, even if it means leaving the use of the term “crisis” behind.