Brookings Institution researchers Beth Akers and Matt Chingos set the internet in a tizzy today with some “counterintuitive” research on student debt, with the takeaway for some being that student debt is not, in fact, the burden that the media (and policymakers) would have you believe. I put counterintuitive in quotation marks because some of what Akers and Chingos reveal is not all that controversial to those who have been following the issue of student debt for some time. And yet, there are some pretty big caveats to their findings that should make anyone wary of concluding that student debt isn’t a major—and growing—concern for American families.
Let’s start with what the Brookings report gets right. First, the large increase in total student debt (which now sits well over $1 trillion) can be partially explained by the fact that more students are attending both undergraduate and graduate school. Graduate school debt in particular takes up a disproportionate chunk of total student debt, and is often the source of media stories about six-figure student loan balances. What the Brookings report also infers is that college (or, higher levels of education in general) is worth it despite increases in debt, because earnings for those with debt have generally risen. This is primarily because six-figure debts are absolutely the exception and not the norm.
But it would be a rather blatant mistake to take from this that we don’t face a real student debt problem in this country.
As David Leonhardt rightly notes in his recap of the Brookings study, one of the major issues with student debt is those who do accumulate debt but to not graduate. Almost by definition, these students will have amassed less debt (because they’ve attended fewer years of school), but the degreeless-in-debt are precisely the ones who are most likely to fall into trouble with their loans, becoming delinquent or defaulting.
This points to one of the major problems with the Brookings study. It only includes those actually repaying their student debt. In fact, half of all federal loans (including Direct as well as older Federal Family Education Loan Program loans) are not currently in repayment – they’re either in deferment, default, forbearance, or held by recent graduates or those still in school (and thus not required to pay). By not including those in deferment, or in default, it takes away a pretty big slice of students who are struggling to pay off loans. Even if balances in deferment or default are lower, again—almost by definition—these are borrowers whose loan burdens represent a financial hardship (and may have a much higher debt-to-income ratio than the Brookings study suggests), excluded from the study. We can’t begin to assess how burdensome these loans are going to be over the long run. (UPDATE: Matt Chingos reached out to me on Twitter, noting that the overall debt figures in the study include more than just those in repayment, but the monthly-payment to income ratios do not).
But perhaps the largest problem with the Brookings study is the dataset itself. Akers and Chingos rely on the 2010 Survey of Consumer Finances (and previous versions), and households within the survey between the ages of 20-40. There is every reason to think—from the rise in average debt balances, student loan defaults (and delinquencies), and college costs (both graduate and undergraduate) spurred in part by massive state disinvestment in higher education—that the situation is more dire in 2014 than 2010. Since then, college costs at public schools (which educate 3 in 4 students) have risen faster than private schools, average debt levels have increased, and graduates’ incomes have generally remained flat.
This isn’t the fault of Akers and Chingos; the 2013 version of the SCF won’t be released for another year. But even if there weren’t a lot of differences, one of the main problems with the use of the SCF is that it doesn’t do a very good job of isolating whether the circumstances of someone who left college in 2010 (at, say, age 24) is going to be different than someone in the same dataset who left college in 1995. The survey simply isn’t big enough to get that granular, nor does it have any value in assessing how big a problem rising balances are going to be in the future. But that’s the point – we seem to have entered a new phase in student debt, one where almost ¾ of graduates borrow and nearly all low-income graduates and students of color borrow—that is fundamentally different than what we saw a generation ago. Taking a look at debt in 2010 is somewhat helpful, but it’s also somewhat backward-looking.
A few other points: Data from the Federal Reserve seem to indicate higher balances than the Brookings study. Whereas Brookings claims that only 7% of borrowers have over $50,000 in debt, the Fed claims that number is hovering around 13%. In addition, this 2013 piece from the Fed—again using a much larger (and updated) dataset, shows that something happened around the time of the recession in which households with student debt saw homeownership and auto purchases decline faster than those without debt. Student debt also seems to be putting a damper on credit scores.
Finally, it will be interesting to see how the continued increase in income-based loan repayment plays into analyses like these. In IBR, monthly payments seem lower—which skews the total overall burden that student loans (particularly those with relatively high interest rates) will cause. To get to that overall burden, my colleague Robbie Hiltonsmith analyzed the same dataset in At What Cost? that Akers and Chingos use, coming to the conclusion that student debt indeed has a very large negative impact on lifetime wealth accumulation.
For some time now, the media has been hyping stories of extreme student loan balances to tell a story about the impact of student debt on most students. That has been an irresponsible move that could lead to bad policy. But just because most students don’t acquire $50,000 or $100,000 in debt does not mean that the student debt crisis is manufactured. From every piece of evidence we have, we know that the percent of borrowers is increasing, average loan balances are increasing, default and delinquency are increasing, and the share of those who do not finish school but take on debt is increasing. The Brookings study gets a few things right, but those who are using this research to downplay the obvious impact that student debt is having on both a household and macroeconomic level may be in for a rude awakening.