Sections

Research

Fixing student loans—the right way

Wellesley College

The amount of outstanding student loan debt is at $1.3 trillion, having surpassed the outstanding debt associated with credit cards, auto loans, or home equity loans. According to some estimates, there are 8 million borrowers currently in default of their student loans. Meanwhile, the popular media is filled with stories about graduates struggling to pay down the high levels of their student debt, contributing to the perception that there is a student loans “crisis.” It is largely in this context that a number of recent proposals have suggested making public colleges free and having the federal government refinance student loans at much lower rates. The next president should address the problems of the existing student loan program but can do so without resorting to costly and regressive policies.

A better solution is to replace the current array of repayment options with a simple income-contingent repayment system for all borrowers. Income-contingent loans ensure that payments are capped at a certain fraction of income (e.g. 10 percent) and allow borrowers who earn below some threshold to postpone their payments. After a certain period (e.g. 25 years), all debts are forgiven. These features help insure borrowers against the downside risk of their investment in higher education. While going to college is an extremely worthwhile investment on average, it is also inherently risky. So we need ways to help borrowers when this investment takes time to bear fruit or does not pan out at all.

The amount of outstanding student loan debt is at $1.3 trillion, having surpassed the outstanding debt associated with credit cards, auto loans, or home equity loans.

In contrast to the perception that emerges in the press, most defaults come from people with relatively low levels of student debt, as documented in a recent Brookings paper. Borrowers in default are much more likely to have attended for-profit colleges and community colleges than selective colleges, and many have dropped out before earning their degrees. Such borrowers simply don’t earn enough to pay back their student loans, even if those loans are relatively small. The adoption of a broad-based income contingent loan program would prevent defaults and reduce financial hardship for these struggling borrowers. And—if implemented directly through the tax system—such a program would substantially lower administrative costs.  It’s a win-win.

Income contingent loans are already available to borrowers in the United States. In fact, income-contingent repayment (ICR) was first introduced under the Clinton administration in 1994. The Obama administration has further expanded these opportunities for borrowers. An income-based repayment (IBR) plan was introduced for borrowers with financial hardship in 2009. Another income-based plan, Pay As You Earn (PAYE), was introduced in 2012 and is now available to all borrowers. Take-up of such income-contingent loans is steadily increasing but, as documented in a recent Government Accountability Office study, the share of borrowers on income-based repayment plans is still below 15 percent. Because none of these programs are implemented directly through the tax system, it is also quite costly to administer them.

…the share of borrowers on income-based repayment plans is still below 15 percent.

Replacing the current array of repayment options with a simple income-contingent repayment system is relatively straightforward and has been done by other nations. Australia introduced such a system in the late 1980s, and New Zealand and Great Britain have also successfully adopted income-contingent student loan programs. Meanwhile, many other nations are also currently experimenting with income-contingent loans. A key to success has been the use of the existing income tax system to administer the programs. There is no reason we cannot replicate this system here. Indeed, a recent Hamilton Project proposal describes how such a system could work in the U.S. After leaving college, borrowers would repay their student loans directly through their paycheck, just as with social security contributions. The interest rate on these loans could be set so that the program is revenue neutral. But the level of monthly payments would depend only on income, not on the interest rate which would affect only the duration of payments. And any student loans that were not paid off after 25 years would be forgiven.

To be sure, there are some issues with income-contingent loans. As with any type of insurance, there is potential for moral hazard and adverse selection. Since some students could pay back smaller amounts of their loans, they may borrow more or be less deliberate about their choice of college or major, compared to the current system. These are real concerns but it is possible to mitigate them. For example, in Australia’s student loan program, the fraction of income paid increases with the total amount borrowed. Moreover, the experience of other countries suggests that adverse selection is not a major concern in such broad-based systems. Another potential issue with income-contingent loans is that borrowers who take longer to pay off their loans may end up paying more interest than they would otherwise. This could be addressed, with origination fees that impose lower effective interest rates on struggling borrowers (as is done in Australia) or by capping the amount of interest that can accrue on student loans.

Adopting income-contingent loans for all students would eliminate the need for loan forgiveness programs, like the Public Service Loan Forgiveness Program, which are difficult to target effectively and are extremely costly. It would require us to re-formulate the existing accountability measures that are defined in terms of default rates (by using repayment rates instead). However, it would help shift attention away from regressive proposals to make public colleges free or having the federal government refinance student loans at lower rates; the benefits from such policies would go disproportionately to advantaged students who take out large loans to attend more selective colleges or attend graduate school.  Instead, we will be able to focus on tackling other important policy issues in higher education—such as the persistent low quality in some segments of the higher education sector.

Author