From Each According to His Earnings

Until about five years ago, few people said much about America’s federal student loan system. There was no talk of a “crisis,” and discussions about change were  pretty much limited to tinkering around the edges. 

Today, however, with rapidly rising default rates on student loans and numerous tales of the hardship that repayments cause recent graduates, we are hearing a lot about the need to dramatically change the system. I recently examined one writer’s argument in favor of eliminating the pressure on college graduates by making public higher education free. Here, I’m going to look at a far more reasonable reform, namely an income-contingent repayment system.

Last spring, Oregon made headlines with a bill, subsequently signed by Governor John Kitzhaber, to establish a pilot program for students in that state’s public colleges and universities. Under the proposed system, students would not pay any tuition while in school, but instead would repay the government a percentage of their income after graduation for some number of years. (The percentage that has been discussed is 3 percent and the number of years 24, but nothing has been carved in stone, since the bill only instructs Oregon’s Higher Education Coordinating Commission to design a program, which would then have to be approved through the political process.)

The Oregon proposal has been greeted with enthusiasm by some, but others question its feasibility. George Mason University economics professor Tyler Cowen, for example, thinks that it faces an adverse selection problem. He wrote on his Marginal Revolution blog, “At the margin, I would expect this to attract people who don’t have a vivid mental picture of the distant future. Furthermore, the terms of the program discriminate against those who expect high earnings or for that matter those who expect to finish.”

Thus, if Oregonians expect this plan to be self-sustaining, it probably won’t be. 

More recently, a fully-developed proposal has been released by the Brookings Institution. In their paper Loans for Educational Opportunity, authors Susan Dynarski and Daniel Kreisman argue that our present student loan system “turns reasonable levels of debt into crippling payment burdens that can prevent young workers from attaining financial independence and stability.”

Here is the essence of the Dynarski/Kreisman plan.

All existing federal student loan programs would be consolidated into a system called Loans for Educational Opportunity (LEO). Instead of the current fixed monthly payment obligations that borrowers must make (unless they have opted into an income-sensitive plan; only about 12 percent have and the authors note that it is difficult to do so), they would pay a percentage that rises or falls with their income. Dynarski and Kreisman would impose this sliding scale: 3 percent on the first $10,000 of earnings, 7 percent on the next $15,000, and 10 percent on earnings above $25,000.

To see how LEO would work, the authors provide several examples. Let’s consider these three hypothetical students who vary enormously in their post-college success.

Cathy graduates with student loans of $25,000 and soon finds a good job with annual income of $50,000, and earns steady raises up to $75,000. Under the sliding percentage formula, she would pay her loan off in seven years and save about $200 in interest cost compared with the current 10-year repayment system.

Avery also graduates with loans of $25,000, but starts off at an income of $23,000 annually, which slowly rises to $35,000 by the time she’s 35. Unfortunately, she suffers a period of reduced earnings for a year—only $20,000.  Avery would pay $100 per month based on her initial salary, only $82 per month during that period of reduced income; and $196 per month at her highest salary. Avery’s loan would be repaid in 13 years, and while she’d pay more in interest than under the current system, she avoids the severe distress of having to make what would have been her standard payment of $241 during that time of low earnings.

Dana takes out $10,000 in loans, but never graduates. At age 25 she finally lands a job that pays merely $6,000 per year.  By age 35, she is earning only $10,000 annually. Under LEO, she would pay only 3 percent of her earnings over a period of 25 years (the maximum number of years under LEO), amounting to $8,527. At that point, the remaining balance of her loan, $8,561, would be forgiven.

Dynarski and Kreisman point out that Dana’s low but steady payments are preferable, both for her and for the taxpayers, than if is she were to go into default early on because she could not afford the fixed payments. Under LEO, she would pay back more of her debt in total and the government would avoid the expenses of trying to collect on the bad loan.

There are other aspects to the LEO plan, including the very sensible recommendation that the government stop lending money to the parents of students, but let’s get to the crucial point: Is this a good change or not?

The authors make a strong argument that their plan is superior to the status quo because it better matches the payments with students’ ability to pay, thereby alleviating “financial distress and damaged credit records.” They also contend that their system would not cost taxpayers any more than the current system and might even cost less because the government would often collect at least 3 percent from borrowers who find themselves in financial distress rather collecting nothing when they fall into default. The government would also save some money by reducing loan servicing costs.

In short, LEO would make student loans easier to manage and less fearful of a prospect than they now are.  That seems appealing.

Our big problem, however, is not that some students find the present system stressful and onerous. Rather, our big problem is that far too many students borrow for college who should not do so. LEO would exacerbate it.

The higher education bubble has been inflated with enrollments of large numbers of young people who often derive little or no human capital benefit from their courses. After graduation (or dropping out), they’re caught between the rock of low earnings and the hard place of loan repayments. LEO would make it easier and more enticing for students with weak academic backgrounds to go to college on borrowed money. After all, if they don’t earn much, they won’t have to pay much and after 25 years their debt will be forgiven—that is, covered by the taxpayers.

Think about Avery and Dana, the hypothetical students discussed above. Avery might have gained some knowledge and skill in college, but her rather low earnings until age 35 would suggest that her college years were not much of an investment. Instead of borrowing $25,000, she might have been better off going to a community college without any borrowing, or looking into non-college skill certification programs.

As for Dana, those comments apply a fortiori. 

Marginal students like Avery and Dana, contemplating frightening scenarios under the current student loan system, might think twice about borrowing for college. Under LEO, however, they’d more apt to think that borrowing looks safe, so why not enroll and take out student loans?

There may something to be said for making the repayment of student loans contingent on post-college income, but for the good students who probably gain from college—those like Cathy—that is a small gain. For the much larger number of marginal students like Avery and Dana, making repayments easier would incline them toward what is apt to be a mistaken path after high school, namely borrowing for college.

The major flaw in our federal student lending system is that it does nothing to winnow out students who are unlikely to gain much from college. Until we fix that problem, I think it would be a mistake to change the system to make borrowing easier for everyone.