Is Student Debt Too High?

Some members of the new Congress are claiming that the debt of students to finance their college education is too high, and that more generous federally funded student grants should be available. Reforms of the college loan program are desirable, but when placed in a proper perspective, college students generally receive an excellent deal on their student loans.

Over 60 per cent of students who finished in 2003-04 college or graduate studies with a Certificate or a Degree had taken out a loan. This percent was highest at 70 to 80 per cent for students who received a professional degree, was also high for students who attended for-profit colleges, while the percent was lowest for students who graduated from two-year public institutions. This difference by type of college is partly explained by the fact that the fraction taking loans is much larger for students from families with low incomes since poorer students are more likely to go to for-profit colleges.

Even after adjusting for inflation, the average student loan increased by about 50% in the decade prior to 2004. The average size of the loan for those with loans was about $15,000 for graduates in 2003-04 with Bachelor's Degrees, it was much lower naturally for those who received certificates or degrees after two years of college, and was substantially higher for those with Masters and other post-graduate degrees. Perhaps surprisingly, the average loan did not vary much between for-profit, other private, and public colleges.

Although the debt of graduating students is not a minor burden, it is not usually a major one either, if the size of loans is related to benefits from college as well as to financial and other costs. Costs measured by tuition did increase at a rapid rate since 1980. According to calculations by Pablo Pena at the University of Chicago, tuition at private non-profit four-year colleges rose 140 per cent in real terms from 1980 to 2005, which means an annual rate of increase of over 3.5 per cent. Public schools charge a lot less but they too had rather rapid increases in tuition. Students who are from poor families have the most trouble paying for college, and obviously that burden gets heavier when tuition is higher. This helps explain the increase over time in both the fraction of students who take out loans, and the size of the typical loan.

On the other side of the ledger, higher tuition over time was related to sharply higher financial benefits from a college education. The typical college graduate earned per hour about 50 per cent more than the typical high school graduate in 1980, and the gap is now about 95 per cent. Earnings of graduates with a professional degree or other post-graduate education grew even faster over time than did earnings of college graduates.

The net benefits from graduating from college are determined by the higher earnings college graduates would receive over their lifetime compared to what they would receive if they started working after high school, minus tuition and any other costs of a college education. The data I have just given show that the increase in the earning advantage from a college education during the past couple of decades was far greater than the increase in tuition, so that average rates of return on a college education--a measure of the net benefit--increased greatly. In addition, various non-monetary benefits of a college education also grew over time. Probably the two most important of these benefits are that higher education increases health through the improvements it induces in lifestyles and medical care, and higher education also improves investments in the learning and behavior of one's children.

How big a burden is the average loan for college graduates who take loans, which is about $15,000 to $20,000? The net present value of the earnings of typical graduates of four-year colleges over their lifetimes after discounting future earnings and subtracting out tuition and other costs has been shown to be over $300,000 more than what high school graduates earn. Even a $20,000 student loan debt is small relative to such a large benefit. Put differently, if the only way to go to college would be to borrow $20,000 under a student loan program at the prevailing 7 per cent interest rate on these loans, the returns from college to a typical graduate would be big enough to allow the borrower to pay off the loan and have a lot left over.

Of course, such loans would be a much greater burden for students who only received two years of college, perhaps because they dropped out of a four-year program, or because they received an Associate Degree. Such students do not earn nearly as much as graduates of four-year colleges. However, the burden of a fixed amount borrowed is not the right comparison since as I indicated, graduates of two year programs borrow much less than do graduates of four-year programs. In reality, the burden of what graduates of two year programs typically borrow is not much greater compared to their discounted earnings than the rather minor burden of the actual loans taken by graduates of four-year colleges.

Within any category of graduates, earnings vary considerably by type of job-- teachers and clergymen earn a lot less than investment bankers--and by degree of success within jobs. Fixed interest loans are not the best way to borrow when loans are used for risky activities. Returns on higher education are rather risky, even after adjusting for how they co-vary with returns on assets. Businesses often borrow with the equivalent of equity to finance start-ups and other risky activities, where the equity pays off well if the venture is successful, and pays little if the venture fails.

This suggests that student loans should not have fixed interest rates that require a fixed amount to be repaid per $1.000 borrowed, but rather should have the equivalent of an "equity" repayment system. That would mean that persons who earn very little repay little, while those who earn a lot repay a lot (per $1,000 borrowed). Requiring individuals who are repaying student loans to submit their income tax statements each year, so that lenders could document what the borrowers earned, could enforce such an income-contingent repayment system.

The United States already has a small student loan program that allows repayments to be conditional on the incomes of borrowers. But a system with both fixed interest loans and income-contingent loans has a "moral hazard" problem. Students who expect to go into well-paying jobs would tend to borrow at fixed interest rates since that would be cheaper to them than repayments that rise with higher earning. A possible reform of the federal program that would reduce this moral hazard would be to shift entirely to an income-contingent system, where persons with student loans who earn little would repay relatively little, and those who earn a lot would repay much more.

A full income-contingent loan program would not be without its own problems since it would attract students who expect to go into low- paying occupations, and repel students who expect to higher earnings. In addition, such a program would "tax" high earnings that would further discourage effort by high earners, and further encourage them to try to hide incomes. But it might work better than either the present largely fixed interest system, or a dual system that allowed both fixed interest loans and income-contingent loans, with the choice among these systems determined by students.