- Who's struggling to pay back college loans? Hint: it might not be who you think.
- @AndrewPKelly takes to @forbes to discuss what kinds of students are struggling to pay back loans & policy implications
- .@AndrewPKelly talks about new Brookings study on student debt on @forbes
The country is in the midst of a heated debate over student debt, but we don’t know nearly enough about how debt actually affects borrowers.
The conventional wisdom—perpetuated by the media—casts the relationship between debt and financial hardship as a linear one. Call it Newton’s Third Law of Student Debt: every additional dollar of debt leads to an equal and opposite reaction in the well being of borrowers. Every dollar of student debt is tragic, the logic goes, and the extent of the tragedy is proportional to the amount of money owed. You can see Newton’s Third Law in panicked New York Times stories about students with large debt loads and in the steadfast belief that more student loan debt must be bad news for the economy. Finally, belief in the Third Law pushes policymakers to focus on debt loads as the key problem to be solved.
An alternate take rejects the linear relationship and instead asks how the same debt load might affect different types of borrowers. Let’s call this the Theory of Heterogeneous Effects: the effect of each additional dollar of debt will depend on borrower characteristics. Did the borrower finish college? Is he from a family that is willing to let him live at home after college? Here, the relationship between an additional dollar of debt and financial hardship is anything but linear.
Instead of identifying large debt loads as the problem, this second theory puts us on the lookout for negative effects across the debt distribution; a borrower with $10,000 in student debt could be in much worse shape than one with $60,000. It also points to policies that minimize the probability students fall into particular high-risk categories—college drop-outs, for instance—and to then help those that do.
Growing evidence suggests that the Third Law of Student Debt just doesn’t fit the data. Last week, the blogosphere went into an absolute tizzy over a new Brookings Institution study of student debt. Unfortunately for the student debt alarmists, the findings were, well, not all that alarming.
Using two decades of data from the Survey of Consumer Finance (SCF), Beth Akers and Matt Chingos found that while student debt loads have increased markedly since the 1990s, borrowers were paying about the same percentage of their income toward monthly loan payments in 2010 as they were in the early 1990s. Just seven percent of borrowers had debt loads greater than $50,000. Akers and Chingos argue that we can attribute a significant portion of the increase in debt to those Americans who are best able to pay it off: those who have gone to graduate school.
Other evidence supports the theory of heterogeneous effects. In a recent AEI white paper (part of our series on reinventing financial aid), the same Beth Akers (using the same SCF data) found that the probability of experiencing financial hardship was largely uncorrelated with the amount of student debt owed. Instead, the incidence of financial hardship was highest among individuals with low levels of debt (less than $5,000) and those with some college and no degree. High-debt borrowers were often the least financially troubled (because they’d gotten more education!).
These data jibe with other evidence on who is struggling to repay their loans. Last year, the Consumer Financial Protection Bureau reported data on the average debt load for borrowers in various repayment categories: actively repaying, forbearance, deferment, and in default. The category with the highest average balance had to be those who defaulted, right? Wrong. Borrowers who defaulted on their loans had the lowest average balance across all of the repayment categories ($14,500 vs. $26,800 among those in forbearance).
One last data point on this: community colleges have the lowest tuition of any colleges in the country. Yet the percentage of community college students that default on their loans (the three-year cohort default rate) is among the highest across all sectors (21 percent in the 2010 cohort). True, only about 20 percent of community college students borrow, and average balances are low; 70 percent of community college borrowers take on less than $6,000. Then why the high default rates? Because borrowers at all levels of debt can struggle to pay back their loans. Those who do not finish their degree have it particularly rough. According to Akers and Chingos, among households with some college but no BA, borrowing rates increased from 11 to 41 percent between 1989 and 2010.
What does heterogeneity mean for policy? First, policymakers should stop worrying about total debt and focus their attention on students’ ability to repay. Existing income-based repayment plans try to do exactly this, but they are far too generous to graduate students that probably don’t need the help. And their forgiveness provisions are unnecessary to ensure borrowers are protected.
However, a focus on helping borrowers after the fact ignores the front-end problem: student loan programs encourage attendance at any program, at any college, and at any price. That means we subsidize a lot of failure. According to my analysis of the most recent federal data, about 37 percent of loan disbursements in the Stafford and Parent PLUS program (loans for undergraduates) in 2012-2013 went to colleges with six-year graduation rates under 40 percent in 2012. That’s a lot of loans to people whose chances of finishing a degree are worse than flipping a coin.
What we need are policies that push students toward more effective and affordable options on the front end: better consumer information, income-share agreements, and risk-sharing that give colleges skin in the game. Without such policies, we’ll continue to see low-debt borrowers who struggle to repay their loans while high-debt, high-income borrowers—graduate students in particular—get all the attention. It’s time to acknowledge what the data say and put our priorities in the opposite order.