Do Income-Based Payment Plans Really Ruin Repayment Rates?

Flickr/401(K) 2013

Two weeks ago the U.S. Department of Education held meetings to discuss a proposal that would hold career programs accountable for not providing students with gainful employment in a recognized occupation. Those conversations dealt in part with a new accountability metric known as the student loan portfolio repayment rate. This measure gets at a similar concept – “the federal student loan repayment rate”– that was included in the last iteration of the gainful employment rule in 2011, but is calculated in a different way. Instead of measuring the percentage of loans held by borrowers that are being repaid, this new measure looks at whether the cumulative outstanding principal balance of all loans taken out by borrowers in a program is reduced from the start to the end of the award year. Though this measure needs some improvements, it’s a simpler and easier to understand way of measuring the concept.

Negotiators brought up a few concerns about this proposal when they discussed it on Monday, but one in particular stood out: Programs could fail the standard because Income-Based Repayment (IBR) allows borrowers to make extremely low payments that would not cover the amount of interest that accrues. Their loan balances would therefore increase over the course of the year. This is a common concern that came up the last time the Department put forth a gainful employment rule. That 2011 version solved this issue by counting up to 3 percent of loan balances as successfully being repaid even if they were in an income-based payment program, as well as counting all loans in Public Service Loan Forgiveness (PSLF) as a success.

But here’s the thing—these concerns are significantly overstated. They stem from a misunderstanding of how IBR actually works in practice. Citing these plans as a reason to not adopt a repayment rate measure as part of gainful employment rules, or to justify certain exemptions or carve-outs for loans in PSLF or IBR would be a mistake, one that could allow struggling programs to escape accountability.

Borrowers who are likely to make below-interest payments on an income-based plan are doing so not because of debt levels but because they just aren’t earning very much money

Individuals who express concerns about income-based plans assume that the plans are synonymous with growing, rather than shrinking loan balances. But they are not. It’s completely possible to be on an income-based plan where your overall payment is less than what it would be on a standard 10-year plan but still more than the amount of accruing interest. In those cases, the loan balance still shrinks, albeit very slowly.

True, those borrowers who use an income-based plan and make below interest payments are likely doing so either because they have extremely high debt balances or very low earnings. Yet the data show us that the latter is far more common than former in programs covered under the gainful employment rules–the median loan balance of these programs that have at least some debt on average is about $8,000 according to data on nearly 8,000 programs the Department. Therefore, borrowers who are likely to make below-interest payments on an income-based plan are doing so because they just aren’t earning very much money, a much more common issue with gainful employment programs. In fact, that’s the explicit concern – low earnings compared to debt levels– that the Department is trying to address with this regulatory work.

Here’s a more thorough explanation of why concerns about the interaction between income-based repayment plans and gainful employment are baseless. First, understand how the formula for programs like Income Based Repayment and Pay As You Earn (PAYE) operate. The PAYE formula takes a borrower’s income and makes a series of deductions for fringe benefits, student loan interest, and a few other things. That remaining amount is known as Adjusted Gross Income (AGI), which New America estimates is about 90 percent of a borrower’s actual income for people earning less than $68,000 a year. The formula then takes the AGI and deducts from it 150 percent of the poverty level for the number of people in the borrower’s family. For example, if a borrower has a household of one, then $17,235 gets subtracted from AGI; for two people the deduction is $21,255. A borrower’s payment under PAYE is 10 percent of the amount remaining after all those deductions and adjustments for family size.[1]

In more layman’s terms, this formula means that if your student loan payments would be more than 10 percent of the subset of your income that’s not going to necessities, then you will not pay quite as much a month as you would under the standard plan.

But there’s a big gap between income that results in “not paying quite as much” and “not paying enough to offset interest growth.” The table below demonstrates the extent of this gap by showing income level at which a single individual would see some reduction in her payments and the level at which her payments would not be less than the amount of accumulating interest. Anyone between these two income thresholds benefits from being in PAYE in the form of a lower payment than what they would owe under the standard 10-year plan, but they still pay off all the accruing interest. For these borrowers, their balances aren’t growing, they are just shrinking slower than they otherwise would.

Income Below Which a Borrower's Annual Payment is Less than...

Loan Balance | The 10-Year Standard | PlanAccrued Interest
$10,000                   32,5612                             3,439
$15,000                   39,2662                             5,583
$20,000                   45,9722                             7,728
$25,000                   52,6772                             9,872
$30,000                   59,3823                             2,017
$35,000                   66,0883                             4,161
$40,000                   72,7933                             6,306

So what about those borrowers in these income-based plans that actually do have low enough incomes where their payments aren’t enough to prevent their balances from growing? As we can see from the table, those people are pretty low income. For someone who owes $10,000, their payments will be above the amount that interest accrues unless their income is less than about $23,400, a not particularly impressive earnings level. It’s true that people with higher debt balances must make more reasonable incomes to avoid having their balances grow, but remember that many of the debt issues at career programs is not that they owe tens of thousands of dollars—three-quarters of gainful employment programs with at least some average debt owe less than $14,500. Rather, students in these programs may owe smaller amounts of debt they cannot repay.

One Repaying Borrower Has A Disproportionate Effect

It’s important to remember that the loan portfolio repayment rate is based upon the total performance of all loans. Pooling all the loans together allows for a borrower making successful payments to more than balance out several people making absolutely no payments, depending on the debt levels. To understand why this matters, consider the table below, which shows how much a borrower pays on the 10-year standard plan in a year versus how much interest would accrue if someone makes no payments. As you can see, the size of payments on the standard plan is always a little over three times greater than the amount of interest that accrues in a year.

Loan Amount | Standard Plan | Accruing Interest When No Payments Occur
5,000                     603                                  193
10,000                1,207                                  386
15,000                1,810                                  579
20,000                2,414                                  772

The greater size of payments under the standard plan compared to interest accrual for non-paying borrowers allows for one successful repayment to more than balance out several unsuccessful ones. Consider a program where there are four borrowers each owing $10,000. If one is on the 10-year standard plan, then she would make $1,207 in payments. If the other three make absolutely no payments whatsoever then their loan balances would each grow by $386 that year, or $1,158. Since that amount is less than the $1,207 paid by the borrower on the standard plan, the program still passes. The one successful outcome–the borrower making standard payments–has compensated for three completely unsuccessful outcomes.

In fact, the differences in both debt and repayment behavior between graduates and dropouts may allow programs to pass with even fewer successful outcomes. Students who graduate are likely to have more debt than dropouts because they were enrolled for longer. And students who complete a program are more likely to avoid default. In those cases, a more indebted borrower that graduated can make up for even more successful outcomes because they will make larger payments on higher debt levels. For example, a borrower with $15,000 in debt making payments on the standard plan can make up for five borrowers with $10,000 each in loan debt. And this process works the other way as well–a borrower making standard payments on a $10,000 debt can still make up for two borrowers making no payments on $15,000 of debt each.

To put it more succinctly, a handful of borrowers making the same standard payments as the vast majority of individuals with federal loan debt would allow for programs to have well over a majority of students making absolutely no payments through IBR or PAYE. It’s open to debate whether a repayment rate test that allows that to be the case is tough enough, but it’s highly unlikely to cause failures thanks to these income-based plans.

 Keep it Simple

A big benefit of the loan portfolio repayment rate is it tests a simple question–do a program’s borrowers owe more at the end of the year than they did at the start? But as with many regulatory or legislative proposals, there’s a tendency toward complexity through a series of tweaks, exceptions, and other alterations that respond to each and every critique. But based on what we know about how much students in these programs borrow and how income-based repayment works–a program whose borrowers aren’t keeping up with interest on their loans is almost surely one where the vast majority of the cohort has low salaries. It’s a sign of multiple types of negative results that should be measured and taken into account, not which student loan repayment plan borrowers are using. So as work on the gainful employment rule continues, resist the urge to tweak and fiddle with income-based plans and the like. Simplicity is better here.

[1] For example, let’s say someone is in a household of one and earns $30,000 a year. Their AGI is likely somewhere around $27,000 and $17,235 gets subtracted from this amount. Of that remaining $9,765 then, 10 percent can go toward student loan payments–$976.5 a year or $81 a month.