Zingales on Education Equity

Luigi Zingales has a good op-ed on education in today’s NYTimes:

… scholars like me…work in the least competitive and most subsidized industry of all: higher education.

We criticize predatory loans by mortgage brokers, when student loans can be just as abusive. To avoid the next credit bubble and debt crisis, we need to eliminate government subsidies and link tuition financing to the incomes of college graduates…Just as subsidies for homeownership have increased the price of houses, so have education subsidies contributed to the soaring price of college.

…These subsidies also distort the credit market. Since the government guarantees student loans, lenders have no incentive to lend wisely. All the burden of making the right decision falls on the borrowers. Unfortunately, 18-year-olds aren’t particularly good at judging the profitability of an investment…

Last but not least, these subsidized loans keep afloat colleges that do not add much value for their students, preventing people from accumulating useful skills.

Instead of subsidies Zingales, drawing a page from Milton Friedman, proposes income-contingent loans.

Investors could finance students’ education with equity rather than debt. In exchange for their capital, the investors would receive a fraction of a student’s future income — or, even better, a fraction of the increase in her income that derives from college attendance. (This increase can be easily calculated as the difference between the actual income and the average income of high school graduates in the same area.)

As I wrote about earlier, Bill Clinton received a loan like this from Yale’s law school and later created a national program but it didn’t get very far (although Obama wants to expand the program). Australia, however, implemented an income contingent loan program in 1989. Australian students don’t pay anything for university when they attend but once their income reaches a certain threshold they are charged through the income tax system.  Many other countries are experimenting with income contingent loans.

Lumni is a private organization, started by economist Miguel Palacios (here is his book and Cato paper on human capital contracts), that is funding loans like this right now.

One point that Zingales doesn’t examine is adverse selection – an income-contingent loan will appeal most to people who want careers with low-income prospects, say in the non-profit sector. (Redistribution of this type was one of the reasons for the Yale law school program.) Thus, the program works best when incomes differ due to luck. My guess is that the adverse-selection problem can be handled if education venture capitalists are left free to price.


I didn't really like the whole thing about the IRS acting as a collection agent for private lenders.

They already collect on federal student loans, which are about 90 percent of the market. Back payments can be taken right out of your tax refund.

I'm not a fan of that status quo either.

The US student loan system has had an income-contingent option since 1993.

Right, that's the program that Clinton created. I have rewritten slightly and added a link to make this clear. Thx.

Two thoughts on the NYT piece:
1. When talking about the responsiveness of higher education and professors to the market, we probably should not lump of all higher education together. Fields like science that often depend on industry funding or a competitive grant process are more accountable to external constituencies than fields that do not, even if grants are also government subsidies. We should probably also distinguish between the research and educational roles the university plays when debating how we want to fund them.

2. I wonder which kinds of students the market would be willing to invest in under a percentage of your income model. Another twist on the adverse selection point: how would it deal, for example, with facts like women being more likely to take time off work and thus having lower incomes?

Incentives will matter.

Also, let me say again that if any other sector was saddling teenagers with six-figure non-dischargeable debt for a product of questionable value, there would be talk of jail time.

Indeed, in 2009 Congress passed the CARD act to restrict credit card marketing to students because of a perceived need to protect these young adults from the likes of Visa. But student loan debt is still all you can eat.

Higher lifetime earnings of $400,000 on average is of questionable value? How many other assets can you purchase that reliably produce that return? Maybe a house during the bubble.

If it were that sure a thing, there'd be no lobbying for publicly subsidized education.

The adverse selection effect can be solved by allowing the initial price of an income slice to vary. For example, 2% of a future engineer's income might buy much more up front tuition money than 2% of a puppet studies student's income.

What if the loan corpus was tied to the current average wages made by, say, the last 5 years of that university+major combination? i.e. making the available loans tied to the track record of a program in leading to well paying jobs?

That should also incentivize colleges in getting their programs more employer friendly too?

There are variables known to the borrower but not the lender that will influence future earnings. The engineering major who is planning on joining the peace corps will take the loan, the engineering major who is planning on going into management consulting will not.

> or, even better, a fraction of the increase in her income that derives from college attendance. (This increase can be easily calculated as the difference between the actual income and the average income of high school graduates in the same area.)

Wow, what an astonishingly terrible idea. Your student loan payments based on where you live?

It's not unreasonable.

$30k in New York = poverty.
$30k in Tuscaloosa = party time.

Others should not subsidize New Yorkers because they choose to govern themselves poorly. The absurd cost of living in New York stems mostly from bad policy.

If those doing the lending could discriminate the amount they funded based on the student's field of study and their performance, income-contingent loans would be an excellent idea. If they cannot, it's a far less nice idea, since the high-paying borrowers would subsidize the low-paying borrowers.

Anyone with the ambition and smarts to make good money after college would not be dumb enough to sell equity in their future earnings. What a terrible idea.

It's harder to predict personal wages accurately than the average wages of a relatively homogeneous cohort. Pretty analogous to insurance.

Those students could try to convince a bank to make them a conventional loan instead, then.

This feels very similar to the abusive credit card offers that are shoveled at college students. Smart people know to avoid them. Less smart people are dupe'd into debt serfdom.

I wonder how many supporters of this type of financing would actually consider it for themselves instead of for others...

The culture of narcissism strikes again...

There was a short-lived attempt to set up a for-profit company that worked this way in Germany. It was a nice idea, they even offered quite extensive career coaching for free as a way to maximize their returns. However, the whole company collapsed when its borrowing costs increased during the financial crisis. It would have been interesting to see if the model had worked out in the long-run. Unfortunately, I can't remember it's name.

I'd be curious to learn more if you can remember the name.

There are a lot of problems with these sorts of plans that never seem to get addressed in the proposals. First, how big an income slice are we talking here, and for how long? I'm guessing something like 2-5% for life for a typical bachelor's degree, right ? How big a work disincentive effect will that create? Second, how do you handle non-cash compensation? If these sort of plans became commonplace, I would expect to see a lot of employers offering jobs with lots of perks (free on-site meals and the like), rich benefit packages, and 8 or 10 weeks of paid vacation every year.

It just seems like a really bad idea. Investing in the equity of a business makes sense because businesses exist primarily to make money. People have lots of interests besides making money, and a loan payback that scales with their pay just encourages them to prioritize those other interests over the ones that generate income.

Agreed. A lot of perverse incentives against work in this case. The borrower might structure the employment contract to reduce the # of hours worked. That's what I would do!

Even if you build it in stealthily into the existing tax system, there is an added drag against doing work above a certain marginal #.

Sure there will be a distortion; but will the distortion be large enough to outweigh the advantages of a income-linked loan?

Just because local taxes are mostly related to property-value didn't result in people moving into tiny dwellings. So also, there is a fairly large utility in earning most of your wages in cash rather than kind and I doubt any substitution would be so large to nullify the purpose of an income-contingent loan.

The advantages of income-linked loans seem pretty scant, so outweighing them with disadvantages doesn't seem like too high a hurdle to clear. At what income level do people imagine the break-even point between payments under the current system and payments under the new system will be? Presumably it's going to be lower than the median income for college graduates (i.e., the median graduate will be paying a higher rate than he would have under a fixed-repayment scheme, to account for the additional risk), so we would expect more students to be made worse off than to be made better off. And what, if anything, do you do about stay-at-home spouses? How do you handle the adverse selection problem that others (including you, I see) have glossed over? (Rebuttal to your previous attempt to dismiss the issue: I don't have to predict my wages very well. If I think I'll be able to earn something like the median or better, I know I'll be better off with a fixed repayment. Furthermore, the more people who make that calculation, the lower the bar for being better off with a fixed rate will be. People's natural tendency to overconfidence about their career prospects will also work against the scheme by biasing people toward fixed-repayment loans.)

To your last point, note that most of us already get a substantial fraction of our compensation in noncash form (health insurance, life insurance, paid vacation, and holidays, just to name a few). At the margin, this kind of loan (if it becomes commonplace) will push the trend further in that direction. To the extent that it pushes down cash salaries that just exacerbates the problem of higher base rates that I've been alluding to.

Really, the more you look at this proposal the worse it looks. I'm baffled by the enthusiasm it generates in some quarters. Honestly, would you have taken such an option had it been offered when you were going to school? Looking at how your life turned out, would it have been a good move? I'm hard-pressed to think of anyone I know who would have been better off under the income-contingent scheme.

John Silber, former president of BU, and one-time candidate for Governor of Massachusetts proposed something like this in his book in 1989. As I remember it, the take would be somewhere around 20% of the difference between the person's earnings and the average earnings of a high-school graduate with no college, and the total lifetime take would be capped at around 150% of the money loaned. With a cap, some of the work disincentive disappears, as paying the loan off early has larger benefits down the road.

Zingales makes practically no effort to differentiate public universities from the actual culprit: for-profit universities.

From businessweek -

Students at for-profit colleges carry the biggest loans in U.S. higher education. Bachelor's degree recipients at for-profits have median debt of $31,190 compared with $17,040 at private, nonprofit institutions and $7,960 at public colleges, according to Washington-based nonprofit Education

Leaving the actual numbers aside (the average debt for public universities is closer to $25,000), the point is, a student at a public universities have much lighter debt loads. There are a few students who seem to be dragging the average up. Still, a university educated student (at a decent school, not Phoenix) can expect an extra $400,000 in future earnings on that debt.

Students of for-profit universities receive 1/4th of the aid and make up 1/2 of the defaults. You can see the default rate of students attending these schools here:


Not surprisingly, the "libertarians" and "deficit hawks" like Paul Ryan refuse to curb grants and loans that end up in the pockets of the owners (because, not surprisingly, libertarians never cared about "free markets," just advocating on behalf of businesses).

I'm not sure what your complaint is. Don't "libertarians" and "deficit hawks" either want to end the subsidies entirely, for both for-profit and not-for-profit alike, or want to apply the same standards to not-for-profits as are applied to for-profits, which would, on your theory of the world, protect taxpayers from being fleeced by both for-profit and not-for-profit institutions. The fact that problem loans are concentrated in a particular sector doesn't mean that a solution to the problem needs to be focused on only that sector. After all, half of the defaults are in the not-for-profit sector.

"The fact that problem loans are concentrated in a particular sector doesn’t mean that a solution to the problem needs to be focused on only that sector."

Jesus. Is this a serious statement? There is a major disservice done by lumping together two segments that are that are producing substantially different consequences (nonprofit education and for-profit education).

You are right, they should not be lumped together.

Universities serve mostly semi-elite to elite 18-24 year olds who can move into campus housing. For-profits are everyone else.

So far you've told us that people with higher debts default more often. Thanks.

Now go internalize the taxpayer cost subsidizing those elite 18-24 year olds and get back to us.

Seriously though, are you really just for high efficiency in government expenditures? Or are you just against for-profit education? If so, are you just against it because you think we are for it?

I think that the argument is that for-profit schools have even worse incentives than nonprofits.

Yeah, it's in the name, "for-profit," I get it. Beyond that, what's the argument?

My argument is simply this. Traditional schools have been working on their craft for centuries. For-profits have made up amazing ground, while breaking a lot of eggs, in a decade.

You don't want to give them money? I never asked for it.

A question I never see addressed in these schemes is the effect on marginal taxes rates. If people must pay a percentage of their future income, this raises their effective marginal tax rate, potentially by a non-trivial amount (the effect is the same whether the payment goes to the government or to private investors). Would the percentage be large enough to create the same kind of disincentive effects that higher marginal tax rates are supposed to create?

The highly skewed distribution would also create interesting incentives for investors. Paying for Mark Zuckerberg's college makes you much more than paying for the colleges of thousands of computer programmers who only reach normal levels of success. What information would investors have about entering college students? Would investors be allowed to/encouraged to provide access to other resources (career counseling, networking) that might improve future earnings in the same way that VC investors help start-ups? Would Wall Street firms end up creating complex models that attempt to predict the future success of high school seniors? (And would the creators of those models make a lot of money, thereby repaying their own backers?)

Its worth noting that to the extent that existing universities operate on endowment income, much of their funding already comes from (voluntary) taxes on the earnings of alumni. Princeton can offer low tuition to low- and middle-class students because so many of its alumni give a substantial portion of their earnings back as donations.

Finally, I can't remember if this was discussed here, but University of California students proposed a similar model: http://www.insidehighered.com/news/2012/02/02/uc-system-weighs-shift-tuition-payments-after-graduation . I'm not sure if anything has happened in response, though supposedly the UC administration was going to run the numbers on the proposal.

I fear what the end-game of this is. Colleges perfectly pricing tuition to capture the entire education surplus, followed by the loan companies perfectly pricing loans to make sure college education is only barely worth it.

We need price pressure on college costs before anything else.

Seems the universities themselves would be the best folks to do this.

Offer a full ride to candidates based on grades/SAT/Major then collect a percentage of income after graduation for say 10 years.

They then can sort for the candidates that are the best investment and offer only the classes that produce a rate of return.

Someone was floating this as a proposal in the UC system.

Here you go:


This proposal is one I've used for years in my class in Securities Regulation. I use it to get students to think personally about what is and what is not a security. Unfortunately, this would be a security, and therefore bring with it enormous regulation, with the burden especially falling on the issuer of the security, that is, the students. Under current law, it seems a non-starter.

Depending on how this is structured, one consequence of this would be universities making sure they admit only students who both can earn and want to earn high incomes, same with the organizations making the loans. If you want to get a university degree just to get an education and then work in the non-profit sector, too bad.

But if you want to get a university degree just to get an education and then work in the non-profit sector, too bad with the current system. In both cases college education has been reshaped so its main use is to get the credentials some think are needed for a high-income career. With the current system, students who want to try to use the system differently get loaded with debt they have trouble paying. It might be better for universities not to admit them in the first place.

Now if the government underwrites the income-contingent loans or the income contingent tuition, so the universities get their tuition and the lenders their interest regardless, the incentives will work in the opposite direction. Now the government is effectively subsidizng the student who just wants an education and will work at low income jobs afterwards. If you want to make money, it might be better to avoid going to college. This is not necessarily a bad outcome, the point is that this idea can have completely opposite results depending on the details of how it the program is operated.

this is like the worst idea I have ever heard.

So it's up there with financing college by going into peoples' homes at night, cutting out their organs, and selling the meat to zoos?

"This increase [in actual income derived from college attendence] can be easily calculated as the difference between the actual income and the average income of high school graduates in the same area"

That sentence is very nearly proof that one college education, at least, was wasted. Sadly, I suspect Zingales does earn more than the average high school graduate.

Or, instead of indentured servitude, they could simply allow student loans to be written off.

Of course, that would mean private lenders throwing millions at naive [financially illiterate] and vulnerable borrowers would lose money.

You beat me to the punch re indentured
servitude. There is a point at which income capture schemes would run afoul of the 13th Amendment.

The rational thing to do upon graduating is declare bankruptcy. Would you lend in that environment?

Lots of debt is dischargeable in bankruptcy but people still get loans because they can bargain for pricing and contractual provisions that mitigate the risk of default. If you're not creditworthy, tough toenails. In an actual market for education financing, C students and Gender Studies majors would find credit the pool of loanable funds very small. Colleges would have to get their snouts out of the rent-seeking trough, admit less students and eliminate subsidized departments, as they should. Subpar students would have to explore alternatives to a four-year degree like, God forbid, vocational training. Philosophy and the arts would have to be funded by wealthy families; one time we did that, we had this thing called the Renaissance.

How much is non-dischargeability a factor anyway? You can't squeeze blood from a stone, which is why we have to use the Fed and the GSE's to make an artificial secondary market for all the crap-ass debt anyway. I thought we learned how all this ends up back in 2008.

The system in Australia works quite well. All universities get funded by the government. The government then recovers half of that money from students' taxes at a few percent of their income until they have paid out their half of their tuition.

The government and the universities effectively decide how many doctors, lawyers, engineers, etc there should be. All admissions are decided purely on high school exam results, which are based on standard tests.

I know according to some theories it shouldn't work, but it does seem to work fairly well.

The system has a nice safety valve too: if the government miscalculates the number of engineers, we just import them.

One thing about the Australian system.... it is heavily subsidised by overseas students.

The amount locals pay is relatively small, and so the income contingency works more or less.

But as the overseas income stream dries up, the Australian funding model is in trouble.

Partly I think because of the voodoo around the word "university". We used to have a variety of post secondary education of which research universities were only one part. Now the landscape is badly distorted, funds for all universities are tied to undergraduate enrolment, faculty salaries and employment are judged on research output for everyone....

Yes the HECS thing works now, but will it continue to?

The problem with income-contingent loans is that the most highly successful people will resent them the most. And they will be precisely the sort of people who are most capable of finding a way to get out of their contract, one way or the other, by hook or by crook.

Something very similar was routine in the pre-piracy recording industry. Record labels would sign a whole bunch of groups, most of whom who go nowhere, but a few of whom would become stars. Income from the latter would cover the losses from the former. That was the theory.

In practice, when a recording artist became a star, they would angrily demand to renegotiate their original contract. Their success, in hindsight, was predestined, and the bloodsucking corporate suits simply exploited them when they were young and naive.

The recording artists did things like releasing Metal Machine Music. The billionaire hedge fund manager will simply emigrate to Switzerland or Singapore and tell the loan company to go eff itself. The resulting losses might make the whole scheme fall apart.

Creative ways will be found to disguise income. Maybe instead of hiring individual superstar employees, companies would subcontract to small companies that just happen to be family firms with extremely redistributive profit-sharing programs.

Australia does have income contingient loans, but it doesn't affect incentives for universities or investors (other than the government). Universities get paid on a fixed schedule, if the student never pays, then the taxpayer takes the hit.

Wouldn't a lot of incentive problems be eliminated if we turned this into a vendor financing system, where the colleges themselves gave out the loans?

They probably have all kinds of information that is useful for predicting who can/will repay. And such a system would help nudge them towards making sure education was relevant.

I think he misses the real issue. Normally whether an expenditure should be made and how it should be financed are two separate questions. Normally the former is determined first then the latter. The issue is not how to finance college tuition but to fix college tuition in the first place, by expanding alternatives to the "4 year degree or nothing" model. And by getting colleges' delivery of education out of the 19th century and into the 21st.

I also meant to say he errs in assuming that one's income is a benefit that is attributable to one's college education. That is quite a fallacy. I went to a four year liberal arts college and, while it was a very good one and I had much enjoyment there, I have never been able to identify any connection between my income and my learning in college. Those years merely served as a test to confirm that I was as good academically as my high school transcript and SAT's had implied, even in more select competition. I think it is accurate to say that the only economic beneficiary of my college education was the next school that admitted me, in that the college record gave them more information to sort applicants with greater efficiency than if it had been forced to admit people directly out of high school.

The problem is not debt-financing versus equity-financing. The problem is that the government is guaranteeing student loans and manipulating interest rates. The result: jacked up tuition prices, students overwhelmed by the money they owe, and zero accountability for universities that aren't producing sufficient outcomes. How do we solve the real problem?

Yale Law still has a loan repayment system that pays off the loans of students who do not take high paying jobs.


Comments for this post are closed