Sort by

The Federal Reserve Just Released a Boatload of Information on Student Debt. Here’s What It’s Telling Us.

Mark Huelsman

The New York Fed has blessed us this week with a whole lot of new data points about student debt, in a series of three blog posts and the release of their quarterly Household Debt and Credit Report. Some of their findings validate some of our preconceived notions about student debt—it’s growing tremendously in the aggregate; it’s interacting negatively with the auto and housing markets; lots of folks have trouble repaying at some point—but there are also a few points worth hashing out further.

A Lot More People Are Defaulting on Student Loans than Official Federal Statistics Indicate

Every summer, the Department of Education releases their official 3-Year Cohort Default Dates (CDRs) that measure, by school, the percentage of student borrowers who default on a student loan (in other words, go 9 months without making a payment) within three years of leaving school. If a college’s CDR is above 30% for three consecutive years, the Department threatens to turn off the faucet of federal financial aid (loans, Pell Grants, etc.) that are the lifeblood of most college revenue streams.[i] It’s about the only accountability tool in the Department’s toolbox, one with pretty lame thresholds at that.[ii] But a three-year window provides us the best numbers we have on how many borrowers are defaulting. Last year, the Department reported that around 14% of borrowers defaulted on their loans within three years.

The Fed though, using a fairly cheeky methodology, created its own “cohorts” and looked at the percentage of borrowers who default over five, seven, and nine years. They found that the “official” default rate may be underestimating the problem by half.[iii] For those who left school in 2005, about 13 percent defaulted within three years. But nine years out, a full 25 percent of borrowers had defaulted:


It would be nice to have data on this by race, income, gender and college attended—or at least type of college or degree attempted or attained. The Department of Education is the only entity that could compel such information, but colleges are loathe to go along for fear of sanctions.[iv] But this Fed data shows that a lot more students are eventually defaulting on their loans than we might think from official statistics.

The Expansion of Income-Based Repayment May Not Be Getting to the Right People

In their quarterly report on household debt and credit, the NY Fed also updated information on delinquency rates. Unlike the default data, this is a snapshot of how many loans are delinquent by 90 days or more at any given time. As you might expect, people fell behind on pretty much every type of debt payment during the recession, and started paying off loans more regularly as the economy improved a bit (credit card delinquencies, for example, are at their lowest level in over a decade). Not so with student loans, where delinquency rates have shot up to 11.3%.

This is what happens when you treat student debt differently than other types of consumer debt and make it nearly impossible to discharge in bankruptcy. It’s also the inevitable result of rapidly increasing college costs, which has led us to a new world where most students now borrow to fund an education.[v]


One explanation could just be a lag—delinquencies are high now because enough borrowers deferred their payments (or received a forbearance), or were in a grace period while attending graduate school, and we’ll see delinquencies fall in a couple years’ time. Indeed, if you look at the right side of the student loan trend line, it looks similar to what we saw with other debts during the recession. More likely though, as people have basically swapped out mortgage and auto debt for credit card debt, and as student debt has become far more pervasive, we’re just now seeing the consequences.

This brings up a big issue. The Obama Administration and Congress have repeatedly attempted to solve the delinquency issue (among others) by touting income-based repayment (IBR) plans (such as Pay As You Earn) for student borrowers. There have been off-and-on pushes to enroll more people in these repayment plans that ensure a monthly payment will never take up an “unmanageable” portion of a borrower’s income and forgive remaining student loan balances after 20 or 25 years. And uptake in these plans has increased—nominally a good sign that people are taking advantage of the benefits offered to them.

But if one of the explicit target populations of income-based repayment is student borrowers who risk missing monthly payments, it’s clear that these plans aren’t reaching them enough to at least slow down the rise in delinquencies. After all, your monthly payment under Pay As You Earn could be zero if your income is low enough (though, crucially, your overall balance will increase).

This lends credence to the theory that a disproportionate amount of enrollees in IBR plans are those with graduate school loans. Why does this matter? Because graduate students are less likely to default on loans (due to higher earnings). In fact, as the Fed’s data shows, the default problem is most pronounced among borrowers who leave school with low balances – in fact, over one-third of student borrowers who took out fewer than $5,000 defaulted within five years, while fewer than one-in-five of those with $100,000 in debt ended up defaulting.


This sounds counterintuitive, until you consider two things: First, most of the folks taking on fewer than $5,000 in debt almost definitely didn’t complete a degree. While the average college graduate takes on just south of $30,000 for the credential, those who have only a semester or a year or two under their belt can’t accrue that much debt. But, with debt and no degree, these students face long odds of being able to pay off even a few thousand in debt without facing difficulty.

Second, graduate school is wildly expensive, but brings with it an increase in wages that prevents repayment trouble in the immediate term.[vi] Unlike undergraduate students, graduate students can borrow unlimited federal money up to the cost of attendance (which can easily exceed $70,000 a year at for law or medical school students). Basically, while paying off six figures of debt is no easy task, it’s just a fact that people with graduate degrees are less likely to find themselves unable to pay off a loan for a 3, 6, or 9 months.

One elegant way to fix this would be for the Department of Education to compel loan servicers to identify all student borrowers without a degree, and automatically enroll them in an income-based repayment plan. They can almost definitely do the identifying (though they may or may not need Congress to act to automatically enroll students), meaning there could be a major public relations effort tomorrow to find students who we know are most likely to fall into trouble, and enroll them in a plan that will almost definitely lower their likelihood of becoming delinquent.

High Debt Balances Are Still Posing a Problem

As I just mentioned, the worst consequences of student debt—default—is primarily a problem for those with lower balances (likely due to completion). But that doesn’t mean that high-balance borrowers aren’t facing hardship en masse.


Due to the sheer amount they took on in debt, people with high balances are more likely to have trouble reducing their total loan burden—basically, the payments they are making just aren’t keeping up with their interest. And while these student borrowers default at lower rates, they actually experience delinquency at higher rates than the average borrower.

The growing balances could be an interaction with income-based repayment (which would prevent default). If that’s the case, a lot of this would be on track to be forgiven after 20 or 25 years. But the higher rate of delinquency among these borrowers is something worth watching; obviously they’re having some trouble making consistent payments, just not enough trouble to default.

Different Tools for Different Jobs

As the Fed has shown for a while, there’s now a negative relationship between holding student debt and owning a home, at least among young people.

There’s also a negative relationship between student debt and credit scores. For low-balance borrowers, the biggest risk seems to be defaulting on a loan. For high-balance borrowers, it’s either becoming delinquent, having your balance rise, and maybe delaying home purchases or other big-ticket items and putting a kink in the economic recovery. Essentially, the “trouble” that student debt is posing in the economy is simply different for different populations.

What that means is that we need different tools to deal with each problem. If we want to reduce default and delinquency among those who don’t complete a degree—and I would argue that we do—we need to find a way to get these borrowers into IBR, or forgive their debt (particularly if they went to a school that provides almost zero quality in the labor market), and increase grant aid so they don’t have to borrow in the first place. If we want to remedy the fact that people with high balances, be they graduate degree holders or otherwise—are more or less unable to repay their obligations because their balances are growing (and thus paralyzing their financial decisions), we need to use a different tool—capping the amount of interest that can accrue on a loan, for example, or providing incremental forgiveness rather than a one-time shot in the arm after 20 years.

Student loan debt brings about different issues for different populations. It’s time we not think of it as a monolithic problem.

[i] Schools can lose eligibility to participate in the federal loan program if, in any given year, their CDR is 40% or higher.

[ii] For those playing the home game, only about 10-15 colleges out of over 7,000 each year fall under threat of sanction, and a good half are beauty schools.

[iii] On one level, colleges have a legitimate gripe about holding them accountable for a student who defaults on his or her loan many years after leaving school. On the other hand, evidence is so strong that some colleges manipulate the official three-year rate to avoid sanction that more data, and a longer window, would undoubtedly be better.

[iv] The Department has sneakily published a similar measure since 2007, found here. It releases information by type of school, but not information on individual institutions, or by other student characteristics.

[v] While they’ve been around forever, student loans didn’t become the primary financing mechanism for college until the 1990s.

[vi] To be sure, the cost of graduate school is a big deal, particularly if borrowers are taking 20 years after graduate school to pay off loans. It could have real consequences on savings, homeownership, and the like. But from a monthly payment perspective, people with graduate degrees are less likely to find themselves in immediate danger of default than those who took on undergraduate debt.