Bryan Caplan  

Why Not a Free Market in Educational Loans?

Bruce Bartlett's Mistake... On Human Evil: Implications fo...
Suppose investments in education are every bit as fantastic as we're supposed to believe: Ability bias and signaling are myths, so the entire observed education premium is causal and socially valuable.  Even so, it's hard to see why government should subsidize education.  Why can't students simply fund their ever-so-valuable investment in human capital with unsubsidized educational loans?

Non-economists' favorite argument is something like: "The interest rates would be so high that few people would borrow."  At least on the surface, though, this objection clashes with the "education is a fantastic investment" premise.  If education really has enormous benefits, people should be happy to pay high interest rates to acquire it.  Furthermore, if education yields such reliable returns, lenders should be confident of repayment, and therefore happily lend at a low rate.

At this point, many economists will leap to the non-economists' defense.  Free-market educational loans would have high interest rates despite the fantasticness of the investment.  Why?  Because of imperfect information.

Now things get really interesting.  Imperfect information, you say?  Which kind?  Symmetric or asymmetric?

Case 1: Symmetric Imperfect Information

You might say, "No one really knows if an educational investment will pay off."  If so, we've got symmetric imperfect information.  Contrary to much loose talk, this is not a "market failure."  If an investment turns out to be worthless 5% of the time, the efficient response is to take this bad eventuality into account.  Maybe the investment will still be worth it.  Maybe it won't.  But it's stupid for government to subsidize loans so borrowers and lenders act as if this 5% downside didn't exist.

Case 2: Asymmetric Imperfect Information

You might say, "Borrowers know better than lenders if an educational investment will pay off."  If so, we've got asymmetric imperfect information.  This can be a market failure.  But it depends.  If desire to borrow and default probability are positively correlated, you can get the standard market-for-lemons "unraveling" outcome.  But is this really likely in the market for student loans?  It seems like the people most willing to borrow will be the students with unusually promising post-graduation career prospects.  So even with hidden information, the market could still work fine.

If desire to borrow and default probability do happen to be3 positively correlated, simple market responses remain.  Can borrowers offer collateral?  Down payments?  Guarantees?  If so, the market can still work very well despite the information asymmetry.  To take an extreme case, suppose that educational lenders had as much latitude to recover bad debts as the IRS.  Do you really think they'd still be reluctant to lend students money?  Or consider this keyhole solution: For a small handling charge, the IRS directly collects student loan payments when you pay your income tax, and remits payment to your lender.  Unless you flee the country, you're on the hook for whatever you borrow - and the asymmetric information problem vanishes.

Before you take extreme measures to overcome asymmetric information, of course, you might want to double check that the problem is genuine.  Would borrowers really have a big information edge over lenders?  In a free market, lenders could - and probably would - verify students' test scores, grades, school, intended major, and so on.  Given all this information, it's far from obvious that borrowers do have superior information.  Yes, students know lots of details about their lives, but lenders have the power of actuarial science behind them.

Overall, then, neither symmetric nor asymmetric imperfect information provide compelling arguments against a free market in educational loans.  But maybe this just reflects the narrowness of neoclassical economic reasoning.  When people say "imperfect information" they often mean "irrationality."  Perhaps the problem isn't that interest rates are too high, but that people are too myopic to see that even high-interest educational loans are, all things considered, a great deal.

While I'm sympathetic to this argument, it's a double-edged sword.  Yes, irrationality might lead borrowers to spurn good loans.  But as we've seen in recent years, irrationality can just as easily lead lenders to make bad loans.  In fact, lenders' recklessness, not borrowers' paranoia, had turned out to be the more serious psychological pitfall.  Why then are we so sure that we need heavy government subsidies to make educational lenders even more reckless than they'd be on their own dime?

COMMENTS (18 to date)
marksjo1 writes:

Isn't affordability before the investment pays off the issue? Buying shares in company x may be a great investment, but I still might not be able to afford to take advantage of it in the near-term--it may pay off but not fast enough to make it wise for me to borrow the money to invest--the payments may be easy in ten years but prohibitive in five. While the government won't step in to help me make an investment in company x, they may step in with a subsidy to help me make in investment in my education, if it thinks inequality in access to higher ed is a problem.

Bostonian writes:

In a free market for educational loans, students who are less likely to graduate would be charged higher interest rates. Student with low SAT scores are less likely to graduate. Low income students, blacks, and Hispanics have lower SAT scores than high income or white students. Therefore in a free market for loans, the "underprivileged" would be charged more for loans on average, and some would be priced out of the market. Liberals cannot accept this, since they think of education as an equalizer, rather than the multiplier (of innate intelligence) that it is.

Rohan writes:

I believe you're over-thinking the process. I would wager the logic goes something like:

1. If government subsidizes X, X will be cheaper.
2. Borrowing money for a college education should be cheaper.
3. Therefore, the government should subsidize borrowing money for a college education.

(1) is suspect, but I think that many people believe it is true.

"Why can't students simply fund their ever-so-valuable investment in human capital with unsubsidized educational loans?"

The answer is that they could, but if the government was involved, it would be cheaper and thus they would be more likely to do so. You don't need to go so far as to argue that the loans wouldn't exist at all.

MikeDC writes:

Two things.
1. The biggest beneficiary of subsidies are generally producers of the subsidized good.

2. The stock econ 101 answer for subsidizing education is that it's a public good. This may well be complete BS, but the idea is that education is subsidized for the same reason that every other public good is subsidized; we'll miss out on the benefit we get from others being educated if they skip out on that education.

another bob writes:

the most likely lenders would be the universities themselves. they would balance loan terms with student profiles and sell collateralized student obligations into an opaque market with implicit government guarantees. but, that is irrelevant.

democracy elects those politicians who promise the highest benefits and the lowest taxes. loan guarantees promise free money now (while I'm running for office) and loan defaults and other higher costs later (after I've been elected and served my term and made my money through graft and moved on and it's some other politicians problem.) and so, loan guarantees for everyone for any purpose - farm loans, house loans, student loans, small business loans, car loans, hamburger-today-pay-you-on-tuesday loans.

Peter H writes:

You can't take out the government intervention in the market. The bankruptcy code is a -major- factor in this model you are neglecting. Send the IRS all you want, if the debt is bankruptable they'll be stopped short. If the debt is non-bankruptable, the lender can accomplish all that the IRS can, with a bit more effort.

The question can't be "will there be government intervention," it will be "what is the nature of the government intervention?" If the nature of the intervention is to treat student loans like all other unsecured debts, then the loan market will collapse except for borrowers who can either qualify on their own merits (i.e. are 5+ years established into steady high-paying careers and have substantial net worths) or can get a co-signer meeting those standards. The government will be intervening through the court system one way or another, whether just to enforce or to deal with bankruptcy. Literally no government intervention would kill the market because obtaining repayment would depend on guys with no necks breaking kneecaps.

Realistically, for a 22 yr old just graduated (or dropped out of) college, the NPV of bankrupting her loans versus the cost of the credit score annihilation is a no brainer. It would only make sense to refrain bankrupting if she had substantial assets to protect, in which case, she probably didn't need a student loan in the first place.

Alan Shields writes:

Milton Friedman (I think this is referenced in Free to Choose) wanted legal support to allow students to sell a percentage of their future income. Students could compare competing offers in a market.

Yes, it sounds severe to sell off 4% of your gross income over 30 years, but there's nothing to say it would be so bad. Further, a percentage of your income is a much better deal than enormous debts you have to pay regardless of whether or not your education was worthwhile.

Students would receive further information about the value of various degrees in the percentages that colleges would demand. An engineering degree may be very cheap while an English degree may simply be unavailable.

Mordatar writes:

I am curious why you didn't mention positive externalities.
This paper, which is the first one I found and didn't analyze well, suggests they are significant

andy writes:

I heard an interesting thing about Australia (I'm definitely NOT sure if this is correct..but..) - the Asian governments subsidize their student education in Australia. This (likely) raises the education fees in australia thus making it more expensive for australians to get you cannot stop other governments subsidize their people, you should start subsidizing yours to equalize thing a bit.

Bostonian writes:


State universities in the U.S. are admitting increasing numbers of full-pay foreign and out-of-state students, and they say that doing so allows them to continue to enroll in-state students at relatively low prices. Foreign and out-of-state students may be taking the places of some in-state students, but they are supposedly subsidizing the ones who do matriculate.

Sara writes:

I have always thought that the argument for government intervention in student loans is that otherwise everyone should default on them. After all, even doctors experience years of low income after medical school in which their income to debt ratio would qualify them to declare bankruptcy. Indeed, most people graduate at a young enough age that it would make sense to default on all loans and suffer through ten years of bad credit. After all, who wouldn't put off buying a house for 10 years when they are 22 years old and could repudiate $30K, $50K, $100K in debt? Theoretically, student loans could be non-dischargable in bankruptcy, but otherwise functionally "free-market," but once they are non-dischargable and the IRS gets involved in making sure they are paid, the government can't help itself but insert itself into every other aspect to make sure the loans they are enforcing are "fair." Afterall, which politician wants to be caught collecting a loan a with a 15% APR for Bank of America from some kid who just wants to volunteer at the local peace and justice charity?

The entire discussion of the student loan industry presumes the current institutional structure of the education industry. This structure imposes enormous costs. Direct costs include operating budgets. Indirect costs include opportunity costs to students of the time that they spend in school and the opportunity cost to society in lost innovation in educational technology (broadly understood to include curriculum and matching of students to curriculum, as well as the methods of delivery).

Becker (Human Capital) defines "school" as an institution whose principal product is education. Whether those institutions we call "schools" qualify is an empirical queston.

It does not take 12 years at $12,000 per pupil-year to teach a normal child to read and compute. Many post-secondary schools charge tuition close to the mean annual income for a full-time employed adult. Curricula at most US universities include a core that most students could acquire through independent study.

The ponderous and obsolete structure of the education industry magnifies the cost of "education" to the point that students need loans and lenders need protection from default.

Peter H writes:

"Theoretically, student loans could be non-dischargable in bankruptcy, but otherwise functionally "free-market," but once they are non-dischargable and the IRS gets involved in making sure they are paid, the government can't help itself but insert itself into every other aspect to make sure the loans they are enforcing are "fair." Afterall, which politician wants to be caught collecting a loan a with a 15% APR for Bank of America from some kid who just wants to volunteer at the local peace and justice charity?

This is not theoretical. As of the 2005 bankruptcy law change, this is the current law of the United States (save direct IRS enforcement). Private student loans are not dischargable in bankruptcy. Kids ARE paying BofA 15% interest when they just want to volunteer at the local peace and justice charity. And politicians don't care.

egd writes:
I have always thought that the argument for government intervention in student loans is that otherwise everyone should default on them.
And that rationale holds so long as Universities are providing education, not credentials. But if you accept the signaling model then you can attach a penalty for defaulting on student loans.

If the option is to repay the loan and keep your diploma or discharge it and and lose your education credentials, bankruptcy becomes a more difficult choice. I suspect most students wouldn't be willing to give up their educational credentials - even if they only have partial credit and haven't graduated - in order to eliminate debt.

Matt C writes:

So the interesting argument is that the loan market will mostly collapse if ed loans became dischargable in bankruptcy.

Would that be so awful? It's plausible that tuition prices would then fall to a level where an ordinary person could work their way through school (used to be common, yanno). I admit I haven't thought it through, but it doesn't sound so bad to me.

I like egd's idea. I think that would do the trick, as long as employers played ball.

Peter H writes:

egd, Matt C:

It's an interesting thought, and one I've seen floated before, but I'm not sure it'd be as effective as you think. First, enforcement of that is difficult to impossible. How do you stop someone from truthfully saying that they attended a school? Have the school supply false records saying they didn't? Or true records saying they attended but defaulted their loans? Most potential employers don't check school records, especially for people not right out of school.

It might make sense for professional licensure, where you could effectively bar someone from say practicing law, but even with the rise of licensure lately, 2/3 of jobs don't require it.

Additionally, it only covers half the story. It would help with the propensity of borrowers to repay, but it doesn't improve the collateral of the loan. When you take a mortgage, the rate is lower than a credit card because you're more likely to repay (under threat of losing your house) and more importantly, because the bank gains your house to repay the loan if you default. Negating the degree has the destructive side of that, but there's no upside for the lender, it's just a weapon to be used in the collections process. You can have all the collections weapons in the world, but if the borrower has no income and no assets, it does not matter.

Also, Matt C, why on earth would employers play ball? The incentive to defect is high (cheap high-quality labour) and the incentive to adhere is as far as I can see nonexistent.

My view is that the student loan market is wildly distorted by the non-bankruptability provision, and will likely collapse under its own weight, but not before ruining many many people's lives.

theodorerud writes:

In reading this post I kept trying to interpret the word "fantastic" with its original meaning---a fantastic investment would be the same as an imaginary investment. I may have been reading this blog for too long.

Nathan Vomocil writes:

The reason the market fails for loans is that there was and currently is 0% risk involved with student loans. It is nigh unto impossible to get them discharged, even for a legitimate hardship. So lenders do not do their due diligence on loans, and push students loans onto people who should not be getting them. For profit schools and strapped state colleges push up tuition because demand is so high now that everyone and their dog can get a student loan. So you begin a vicious cycle of ever increasing tuition and increased need to finance. The only way that a market solution should be allowed is if lenders are allowed to make bad loans and borrowers are allowed to default. No risk means there should be no real reward, but that is not currently the case.

Comments for this entry have been closed
Return to top