The Case for Income-Share Repayment of Student Loans

Student loans for higher education strike me as both sensible and crazy.

The sensible part is that on average, a college degree raises income levels by more than the cost of college. The crazy part is that the United States is a country where someone from the ages of 18-20 is not legally allowed to order a beer, but is allowed to accumulate tens of thousands of dollars of debt. Moreover, basing an individual  decision on average outcomes is a risky business: as a statistics professor of mine used to say, "On average, the Great Lakes don't freeze." Some students will complete college and have careers that easily allow them to repay college loans. At the other end of the spectrum, some student will take out loans and not complete college, or will complete college but end up on a lower-paid career path so that repaying the student loans is very hard. 

One way to think about income-contingent college loans is that they limit the risk of being unable to repay because of poor career outcomes, but they also can impose higher costs on those with good career outcomes. 

The basic idea is that when you borrow for your college loan, you promise to repay, say, 10% of your monthly income for the next 30 years. You never need to pay more than 10%: indeed, there are often provisions that if your income is especially low, you don't need to repay at all during that period. In addition, any remaining loans after 30 years are wiped out. So you know that your future payments will be a limited share of your income, for a limited time. 

On the other side, if you have a strong income-earning career, paying 10% per month may cover what you borrowed within a few years. However, in this case the rule might be that you need to keep paying until you have repaid twice the principal of the original loan. The underlying idea here is that in exchange for the flexibility to avoid repaying your loan, depending on your your career turns out, you need to promise to repay extra if your career turns out well. But even so, there is a cap on the total amount you need to repay. 

Proposals along these lines are not new. For example, back in 1993 the Journal of Economic Perspectives (where I work as Managing Editor) published an article on the topic by Alan B. Krueger and William G. Bowen ("Policy Watch: Income-Contingent College Loans," 7:3, pp. 193-201). As they point out, the idea can be traced back to a 1955 essay by Milton Friedman titled "The Role of Government in Education"(in Economics and the Public Interest, edited by Robert A. Sol). Friedman wrote 65 years ago: 

This underinvestment in human capital presumably reflects an imperfection in the capital market: investment in human beings cannot be financed on the same terms or with the same ease as investment in physical capital. It is easy to see why there would be such a difference. ... A loan to finance the training of an individual who has no security to offer other than his future earnings is therefore a much less attractive proposition than a loan to finance, say, the erection of a building: the security is less, and the cost of subsequent collection of interest and principal is very much greater.

A further complication is introduced by the inappropriateness of fixed money loans to finance investment in training. Such an investment necessarily involves much risk. The average expected return may be high, but there is wide variation about the average. Death or physical incapacity is one obvious source of variation but is probably much less important than differences in ability, energy, and good fortune. The result is that if fixed money loans were made, and were secured only by expected future earnings, a considerable fraction would never be repaid. ... The device adopted to meet the corresponding problem for other risky investments is equity investment plus limited liability on the part of shareholders. The counterpart for education would be to "buy" a share in an individual's earning prospects: to advance him the funds needed to finance his training on condition that he agree to pay the lender a specified fraction of his future earnings. In this way, a lender would get back more than his initial investment from relatively successful individuals, which would compensate for the failure to recoup his original investment from the unsuccessful. ...

One reason why such contracts have not become common, despite their potential profitability to both lenders and borrowers, is presumably the high costs of administering them, given the freedom of individuals to move from one place to another, the need for getting accurate income statements, and the long period over which the contracts would run. These costs would presumably be particularly high for investment on a small scale with a resultant wide geographical spread of the individuals financed in this way. Such costs may well be the primary reason why this type of investment has never developed under private auspices. But I have never been able to persuade myself that a major role has not also been played by the cumulative effect of such factors as the novelty of the idea, the reluctance to think of investment in human beings as strictly comparable to investment in physical assets, the resultant likelihood of irrational public condemnation of such contracts, even if voluntarily entered into, and legal and conventional limitation on the kind of investments that may be made by the financial intermediaries that would be best suited to engage in such investments, namely, life insurance companies. The potential gains, particularly to early entrants, are so great that it would be worth incurring extremely heavy administrative costs. ...

Individuals should bear the costs of investment in themselves and receive the rewards, and they should not be prevented by market imperfections from making the investment when they are willing to bear the costs. One way to do this is to have government engage in equity investment in human beings of the kind described above. A governmental body could offer to finance or help finance the training of any individual who could meet minimum quality standards by making available not more than a limited sum per year for not more than a specified number of years, provided it was spent on securing training at a recognized institution. The individual would agree in return to pay to the government in each future year x per cent of his earnings in excess of y dollars for each $1,000 that he gets in this way. This payment could easily be combined with payment of income tax and so involve a minimum of additional administrative expense. The base sum, $y, should be set equal to estimated average --or perhaps modal--earnings without the specialized training; the fraction of earnings paid, x, should be calculated so as to make the whole project self-financing. In this way the individuals who received the training would in effect bear the whole cost.

But there are various signs that the idea of income-contingent loans is gaining some momentum. For example, back in 1998 the United Kingdom underwent a seismic shift in higher education. It shifted away from a model where tuition was government-paid, but free to students, and toward a model where universities would charge tuition. But at the same time, it set up a program of income-contingent loans. Here a quick overview from a report by Jason D. Delisle and Preston Cooper ("International Higher Education Rankings: Why No Country's Higher Education System Can Be the Best," American Enterprise Institute, August 2019), which I posted about last summer. They write: 

In England, where the vast majority of the country’s population is concentrated, universities charge undergraduate students tuition of up to $11,856, making English universities some of the most expensive in the world. ... To enable students to afford these high fees, the government offers student loans that fully cover tuition. Ninety-five percent of eligible students borrow. Repayment is income contingent; new students pay back 9 percent of their income above a threshold for up to 30 years, after which remaining balances are forgiven. Despite the lengthy term, the program is heavily subsidized: The government estimates that just 45 percent of borrowers who take out loans after 2016 will repay them in full ... England’s high-resource, high-tuition model is relatively new. Until 1998, English universities were tuition-free, with the government directly appropriating the vast majority of higher education funding.

Another sign of the attractiveness of income-contingent loans is that at some institutions--Purdue University is a leading example--have started providing such loans. Tim Sablik tells the story in "Education without Loans: Some schools are offering to buy a share of students' future income in exchange for funding their education" (Econ Focus, Federal Reserve Bank of Richmond, First Quarter 2020). Rather than referring to this arrangement as an "income-contingent loan," Sablik's article refers to it as an "income share agreement" or ISA: 

ISAs provide students with funding to cover their education expenses in exchange for a portion of their income once they start working. Under a typical contract, recipients pledge to pay a fixed percentage of their incomes for a set period of time up to an agreed cap. For example, a student who has $10,000 of his or her tuition covered through an ISA might agree to repay 5 percent of his or her monthly income for the next 120 months (10 years), up to a maximum of $20,000. ISAs typically also have a minimum income threshold before payments kick in; if the recipient earns less than the minimum, he or she pays nothing. This means that ISAs offer students more downside protection than a traditional loan.

A company called Vemo "reports that it works with more than 75 schools and training programs to offer ISAs." Many schools limit such programs to those who have already made progress toward completing certain courses of study, and just need a financial boost to make it over the finish line. 

The United States is seeing a shift toward income-contingent loans as well. Here's a comment from a e Congressional Budget Office report on "Income-Driven Repayment Plans for Student Loans: Budgetary Costs and Policy Options" (February 2020), which I discussed in a post earlier this year:

Between 1965 and 2010, most federal student loans were issued by private lending institutions and guaranteed by the government, and most student loan borrowers made fixed monthly payments over a set period—typically 10 years. Since 2010, however, all federal student loans have been issued directly by the federal government, and borrowers have begun repaying a large and growing fraction of those loans through income-driven repayment plans.

Under the most popular income-driven plans, borrowers’ payments are 10 or 15 percent of their discretionary income, which is typically defined as income above 150 percent of the federal poverty guideline. Furthermore, most plans cap monthly payments at the amount a borrower would have paid under a 10-year fixed-payment plan. ... Borrowers who have not paid off their loans by the end of the repayment period—typically 20 or 25 years—have the outstanding balance forgiven. (Qualifying borrowers may receive forgiveness in as little as 10 years under the Public Service Loan Forgiveness, or PSLF, program.)

For me, the idea of income-contingent loans is a useful way of striking a balance between financial access to higher education and protecting students from being stuck with large and lifelong debts. (There are even legal provisions for garnishing Social Security benefits to repay student loans. The idea that this step is either necessary or possible seems demented.)  I also think that for many undergraduate students, telling them that they are promising to pay a certain percentage of income over the next 2-3 decades would put such loans in a more honest and open context. 

But as usual, I'm all about acknowledging the tradeoffs, which arise for both economic and political reasons. Let's make the plausible assumption that students have some ability to know in advance whether they are going to be able to repay a conventional student loan within a few years. Those who are more likely to repay will take out conventional student loans rather than income-contingent loans--so they can avoid the risk of repaying more than they borrowed. Those who are less likely to repay will take out income-contingent loans, so they have greater protection against difficulties in meeting a conventional repayment plan.  This adverse selection dynamic suggests the underpayers are likely to outnumber the overpayers. 

In addition, politicians aren't great at setting up loan repayment plans. When setting the share of income to be repaid, or the length of time, political forces will tend to choose numbers that are unrealistically low in an actuarial sense. Politicians are also continually tempted to come up with lists of exceptions where repayment doesn't need to be made: for certain careers, in certain locations, during certain economic condition, and so on and so on. 

Putting these forces together, it seems likely that a substantial share of income-contingent loans will not be repaid in full. A rough estimate based on the UK experience and on the US experience since 2010 is that about half of income-contingent loans, at least under the current rules, will not be repay the full principal and interest. I'm OK with some government subsidy of higher education, at both the state and federal level. But an honest plan for income-contingent loans may need to have higher repayment rates, for longer periods, than politicians and borrowers would prefer.