Bryan Caplan  

Student Loans and Bankruptcy

Robert Conquest RIP... Dennis Hastert and Colorado po...
One of the main things I learned while writing The Case Against Education: Contrary to almost everyone, student loans are virtually the least objectionable government subsidy for college.  Why?  As you know, I firmly subscribe to the view that - from a social point of view - Americans spend far too long in school.  Student loans are benign because they do almost nothing to increase schooling.

The key facts I learned:

1. While the federal government originates over a hundred billion dollars worth of loans each year, most of the loan is not a subsidy.  The highest-quality estimate of the average 2010-2020 subsidy is 12% of the face value of the loan.  Not 12-percentage points per year, mind you, but 12% of the face value total.  On an annual basis, that's 4% of all government subsidies for higher education.  In 2011, that comes to merely $13 billion a year. 

2. Contrary to the credit market imperfections story, loan subsidies have about as much effect per dollar on college attendance as tuition cuts.  The reason, on reflection, is pretty obvious: "Giving" people money for tuition isn't very motivating if they know they'll eventually have to repay what they received.  Loan subsidies sweeten the deal, but knowing you'll effectively have to repay 88% of what you borrow is only slightly more enticing than knowing you'll have to repay 100% of what you borrow. 

But isn't it awful that student loans survive bankruptcy?  It's definitely awful for unemployed college graduates - and even worse for dropouts.  For society, however, it's for the best.  If borrowers could escape student debt via bankruptcy, taxpayers would make up the difference.  This would probably substantially raise the implicit loan subsidy.  And if I'm right about the social value of education, that's bad.

Still, as long as policy-makers ignore all the complaining about social loans and bankruptcy, the complaining serves a useful social function.  I strongly suspect that few 18-year-olds fully realize how hard it is to weasel out of their student loans.  The more often they hear pundits moan, "You can't escape student loans with bankruptcy!  How terrible!" the scarier borrowing gets - and the less likely kids are to borrow or attend. 

It's great that education reformers are talking about student loans and bankruptcy.  Our goal, however, should not be to change bankruptcy law, but to publicize it.

COMMENTS (15 to date)
Paul writes:

I have always assumed that student loans are not dischargable in bankruptcy to be because many (most) recent graduates are technically bankrupt. When I graduated (in UK in an era when we got outright grants, so this is a hypothetical) I had a small overdraft, no job, and no assets. If I had a dischargable loan then it would be the perfect time to not get a job, go bankrupt, get rid of my loan. Then got get a well paying job (actually I went and did a PhD, then moved to silicon valley, so it was a few more years before I got that well-paying job).

rp writes:

Please enlighten me here.

You say: "If borrowers could escape student debt via bankruptcy, taxpayers would make up the difference."

Do you mean that the amount of new loans would stay the same/increase? With the defaults eaten by taxpayers?

My thinking was that new student loans would plummet because the loans are riskier, leading to less schooling, which is what I thought you wanted.

Hazel Meade writes:

IMO the key problem with the student loan system is that neither loan eligibility nor interest rates are tied to income (and hence repayment) prospects after graduation.

A purely private student loan system would have restrictions on loan eligibility set by private banks. You would not be able to get money to study a subject with little or no employment prospects, or if you did, the interest rates would be much higher. This would discourage students from studying subjects with poor employment propests, unless they could pay for it themselves. And it would push students into studying subjects that give them skills that are actually in demand.

For instance, instead of studying (say) psychology, because it sounds like fun, a student might decide to major in computer science and study psychology as a minor, in order to qualify for the loans and get a good interest rate.

Vivian Darkbloom writes:

""Giving" people money for tuition isn't very motivating if they know they'll eventually have to repay what they received. Loan subsidies sweeten the deal, but knowing you'll effectively have to repay 88% of what you borrow is only slightly more enticing than knowing you'll have to repay 100% of what you borrow."

This (and the related discussion) is wrong or misleading in several respects.

First, "student loans" federally granted, guaranteed, or privately granted, are not given solely for "tuition". Logically, most students would first borrow from the federal programs and then resort to other sources of loans. While money is fungible, it is also likely that a greater percentage of the money borrowed from private sources than from federal sources goes towards living and other expenses and not towards tuition.

Second, student loans *are* dischargeable in bankruptcy if "undue hardship" is shown. While that has historically been a high hurdle to overcome, the standard seems to be softening. Once upon a time one had to be "disabled" to get social security disability. Standards for "undue hardship" are likely going the same route. And, if bankruptcy courts don't provide the relief, many students today have the not unreasonable expectation that the federal government will step in and provide "relief". As long as (former) students don't acquire any attachable assets, they are effectively judgement proof. The immediate gratification for many "students" of getting these loans to meet living expenses and postponing reality seems today to far outweigh any hypothetical possibility that they will have to repay those loans in the future, in part or in whole.

Third, "knowing that you will have to effectively repay 88 percent of what you borrow" is an incorrect interpretation of the CBO study, even with the use of the term "effectively". The CBO concluded that under "fair value" terms, the federal government was effectively granting a 12 percent subsidy. That has nothing at all, however, to do with the "amount borrowed". The amount borrowed is the amount borrowed, full stop.

Finally, Paul, the first commenter got it right---were it not for the higher burden placed on discharging student loans in bankruptcy, one could expect nearly all students with substantial loans to immediately file for bankruptcy on graduation. (A different, more lax standard would also have the perverse effect of encouraging more borrowing so that precisely that could be done). This is a very practical rather than theoretical consideration. And, as far as theory and practice are concerned, the bankruptcy law, not unreasonably, may be considering the education obtained from those loans to be an "asset" that is not on the personal balance sheet, as such, but should be taken into account, even if indirectly, in fashioning the bankruptcy policy and determining who should ultimately bear the burden of that financing. Were it not for those rules, kids who don't go to college and choose to work instead would be subsidising the more studied slackers.

JLV writes:

Note that the choice facing a student with too much student loan debt right now is basically, "default and face penalties, do IBR, or get a job in the public/nonprofit sector and do PSLF". For a variety of reasons, the last option is best, since it essentially offers a sort of delayed bankruptcy without the penalties.

Seems to me that if we just replaced IBR/PSLF, etc with changes in bankruptcy law there would on net be less distortions.

Daublin writes:

I'm not sure, I believe Bryan is anticipating the comments. If the feds created students loans and then saw everyone going bankrupt to get out of them, then it would start covering the losses to keep the student loans market alive.

FWIW, I feel very good about non-dischargeable student loans existing. For the people that it helps, it provides a life-changing opportunity to start their career on a much better path than anything their parents ever had. It's a humongous good for the people it benefits, and for everyone else, caveat emptor.

Moreover, I don't feel good about direct subsidies for education. A loan is a much better structure: people who really have something to gain from a better education can easily pay off the loans. For people who major in _____ Studies, I really don't think the public should be funding their party years.

IVV writes:

"If borrowers could escape student debt via bankruptcy, taxpayers would make up the difference."

This assumes that the universities would necessarily be made whole. I agree with you that student loans aren't the problem. Skyrocketing tuition is. In the old world, where only maybe 30% of high school graduates went to college but 70% could find gainful employment, the average new worker could start at zero. Now, where 65% got to college but only 50% find gainful employment, the average new worker begins at -$50K. Tuition (and the resulting student loans to cover it) becomes the method by which the new worker is indentured instead of acting as a free agent.

That's the real problem.

Matt Moore writes:

I appreciate that you are probably done writing the book, but I was reading the other day and noticed something that might be of interest to you - you probably already know about.

In the UK, Tony Blair's government set a target of getting 50% of school leavers to go to university, in order to "raise productivity and prepare people for the jobs of tomorrow". They succeeded in meeting that target, but the actual result was that exactly the same profile of people did all same professional jobs as before, plus there was massive inflation in the entry requirement of previous non-graduate trades.

Also, there was a big rise in student debt and a lot of disillusioned young people who were pretty disappointed that the unemployment office was looking at their archaeology degree and sending them to stack shelves.

What about the argument that student loans raise the price of college? To the extent that both demand and supply are inelastic, that effect can be very significant.

(I suppose Bryan could say "great, let's raise the price so fewer kids go to college, and let's focus on telling them not to go in the first place," but at the same time his view is that college is already overpriced, no? Shouldn't it be cheaper, according to that view?)

Or, for that matter, what of the fact that the subsidy is large and going to an unproductive sector (in Bryan's view)? Surely there are other, less objectionable subsidies than that... Agricultural subsidies are small, and at least they go to a productive sector...

Mike Hammock writes:

Jacob, the question is not "do student loans raise the price of college," it is "do government student loans raise the price of college, compared to private student loans?" I think Bryan is saying the answer is "not by much".

I do have a question for Bryan, though. Are government student loan standards similar to private student loan standards? That is, is it easier to get a government student loan? If so, that is also a sort of subsidy, right? Do current estimates of government loan subsidies take this into account? I couldn't get the CBO study link to open.

ColoComment writes:

Be they public or private student loans, the college or university should have at least a substantial piece of the debt, and should have to provide annual disclosure of paying v. defaulting borrowers for the benefit of those shopping for an education.
We have the same problem here as with, for example, the Ex-Im Bank and the U.S. health care system, in that there is a third-party payor (lender or guarantor) situated between the provider of the goods/services and the consumer of same. If the university bore the burden of default, that would motivate it to ensure to the greatest degree possible that its consumers would be prepared to earn sufficient to satisfy the debt.
As it is, since the education consumer has neither income nor collateral to secure his debt, its non-dischargeability in bankruptcy is the only "security" available to the lender. Otherwise, the price of student lending would be/should be far, far higher, yes?
I may not have the precise language to explain my thoughts, but is any of that inaccurate?

Hazel Meade writes:

@ColoComment: most colleges teach many subjects with widely varying employment rates, so the aggregate numbers for paying vs. defaulting borrowers are not very informative. A high repayment rate in a school with a large engineering department may mask poor repayment rates for students in social sciences.

What students are more likely to need is information about repayment rate by major. Unsurprisingly, this is the sort of information that would be encapsulated in price information in the form of interest rates in a private market.
Because the government makes loans available to everyone regardless of their field of study, on the same terms, and at the same interest rates, the student lacks price signals telling them what they ought to study.

AlexR writes:

Income-sharing agreements, in which the lender receives a percentage of post-college earnings, look to be on the rise. See (ungated):

Such arrangements offer an insurance feature to the borrower: repayments are low if post-college earnings turn out to be low.

Critics think it's an "icky" idea because some borrowers (those with unexpectedly high post-college earnings) will pay more, and on average the cost of the loan could be higher than for a student loan. In other words, critics object to the insurance because a premium is paid for it.

ColoComment writes:

Hazel: yup, good points. And you laid it out far better than I. Thx!

Nathan Smith writes:

Hmm. Caplan says that he thinks people should spend less time in school, but doesn't object strongly to student loans.

But surely one of the big reasons it's bad for people to spend so much time in school is that they end up in so much debt.

The only reason I can see why student debt could be regarded as a small cost of widespread over-education, is if the monetary costs of studying, which show up as student debt, comprise a small part of the opportunity cost.

Monetary costs, please note, might include lost earnings. Thus, if a person has $50,000 of school debt, because their education cost $100,000 and they spent $20,000 of savings and got $30,000 of subsidies, it makes little sense to say that the real cost is the foregone earnings and not the debt. If they had been earning, they wouldn't have needed to go into debt.

Education is very costly. If we're getting a lot more education than is socially optimal, a lot of those costs are falling on students and showing up as student debt. It would be odd to see the costs that fall on (mostly poor) students as greatly less objectionable than those that fall on (wealthier in a weighted average sense) taxpayers. Caplan seems to be mistaken about the ramifications of his own argument.

Comments for this entry have been closed
Return to top