Don’t make colleges pay for student-loan defaults

Incentives matter. This is a fundamental tenet of economics: People respond to their incentives. If something in a market seems to be going wrong, it’s because the incentives have gotten screwed up.

Looking at the market for education, it’s hard not to think that there’s something wrong with the incentives. Tuition keeps going up and so does debt. The percentage of people who are not paying off that debt - either because they are in default, deferment, or an income-based repayment program - is staggering. Naturally, a lot of folks would like to get the government in there to start tweaking those incentives until the market stops being so crazy.

One issue involves the incentives that schools have to ensure that their graduates get value out of their degrees. At the moment, a school can enroll you in practically any program, and the government will lend you money for tuition and living expenses, whether or not that degree is likely to produce the means to repay the loan. Since schools are often in a better position to know the economic value of their degrees than naive potential students, that twists the incentives. Eventually, the student will pay, either with money or trashed credit. If the loan defaults, taxpayers will pay too. The school has the most information about the transaction and yet it has the least at stake. No wonder we have such high tuition, so many dubious degree programs and such a troubling rate of default.

So why not make the schools care? That’s the idea behind “risk sharing,” a reform plan that is hot in Washington. The details come in various flavors, but the core is the same: When graduates default, their schools would be on the hook. That’s already true to a small degree, but a school’s default rate has to be egregious before the government will take action. Ideally, risk-sharing would remove those thresholds for government action. Every time a student defaults, the school would pay some fraction of the lost money.

It sounds like an economist’s dream: simple, elegant and even just. Unfortunately, when you start digging into the details, it starts looking less elegant and more complicated.

It’s tempting to think that the reason so many students default must be that colleges are charging people too much money for worthless degrees. And in the case of some for-profit schools that operated like borderline scams, that’s a fair assessment. But community colleges also have extremely high default rates, and it’s not because they’re charging outrageous tuition. In fact, it doesn’t look like tuition has much to do with the problem; the smallest loans have the highest default rates.

This suggests that the biggest problem is not cost, but income. A small loan often indicates a student who dropped out before acquiring enough earning power to pay it back. And dropping out of college, in turn, is often a sign of low income: Kids from poorer families are more likely to quit college than kids who grew up on the top of the income distribution.

Penalizing schools with high dropout rates would undoubtedly deter schools from preying on naive students with no college graduates in their family background. But it would probably also discourage schools from admitting those sorts of students, because they are the ones who are more likely to default.

In 1990, New York State began releasing “report cards” on cardiac surgeons with data on the mortality rates of their patients. The idea was simple. Greater transparency would allow patients to make better informed choices about their doctors, and would give surgeons the incentive to perform better. And in some sense, that’s what it did. Surgeons started operating on healthier patients who were more likely to survive the procedure, while refusing to treat sick patients who might have benefited but were more likely to die on the table. A 2003 study published in the Journal of Political Economy concluded that the report cards improved the matching of patients with hospitals, increased the quantity of cardiac surgery and changed its incidence from sicker patients toward healthier patients.

“Overall this led to higher costs and a deterioration of outcomes, especially among more ill patients,” the study said. “We therefore conclude that the report cards were welfare-reducing.”

How did this intervention go wrong? In short, because measuring health outcomes is really, really tricky. A lot depends on the skill of the medical team. But a lot also depends on the patient - on genetics, behavior and environmental factors over which the medical team has no control. So when the state started measuring outcomes, surgeons started trying to control their patients more. Since they couldn’t fix people’s genes or make smokers quit, they exerted that control by excluding sicker patients from their operating rooms.

You can try to prevent this sort of thing by adjusting for patient risk and other variables. But complicated metrics are often easier to game, because as you keep adding variables, the interactions between all them become harder and harder to understand.

Education has similar problems, for similar reasons. You can take the greatest teachers in the world and put them in a state-of-the-art classroom, but if a kid doesn’t show up for class, or hand in work, or if she lacks the preparation needed to handle advanced material, that kid is not going to graduate. And it is hard for an outsider, using the sort of objective measures that our legal system demands, to distinguish between the cases where the school failed, and the cases where the student did.

That doesn’t necessarily mean we shouldn’t make schools keep some skin in the student-loan game. But if we do, we should be clear about the risk that some of the current players will be benched.

Megan McArdle is a Bloomberg View columnist. She wrote for the Daily Beast, Newsweek, the Atlantic and the Economist and founded the blog Asymmetrical Information.