Hacker News new | past | comments | ask | show | jobs | submit login
Risk Based Student Loans by Michael Simkovic (ssrn.com)
31 points by protomyth on May 6, 2012 | hide | past | favorite | 30 comments

> ...there may be no reliable price signal about the long-term financial risks inherent in different courses of study.

So credit availability and interest rate is going to be directly related to what other people ahead of you studied in school and how well they settled the debt. Also:

> Educational institutions may have incentives to funnel students into areas that do not maximize students’ future incomes or employment prospects.

I guess they do. Unfortunately, the instituions are not even a factor. As far I can tell, the proposed solution is to deprive the student of the opportunity to get involved in that "low-value" field. Then, indirectly, we punish the institution by reducing revenues. Interesting.

It's an interesting read.

I didn't read past the abstract, but it seems that fair risk assessment would have to include not only the potential income from the intended course of study but also likelihood of completing the course and overall credit history of the borrower (actually his/her parents, as most students won't have a credit history yet).

Easy credit for tuition has no doubt contributed to increases in tuition that are way out of line compared to overall inflation. This works to the detriment of everyone pursuing higher education, as you (or more likely your parents) are either spending a lot more than the education is really worth, or you're ending up deep in debt as you start your professional life.

As far I can tell, the proposed solution is to deprive the student of the opportunity to get involved in that "low-value" field. Then, indirectly, we punish the institution by reducing revenues.

Institutions need some price pressure. Tuition costs are out of control. Even state institutions are regularly raising tuition every year by the maximum allowed by their legislatures, and tacking on various other "fees" to raise the overall costs even more.

Disagree, if a student wants to pursue computer science he/she should have a lower interest rate than a philosophy major. There is a much higher likelihood that the CS major will pay back the loan soon.

There is a bit of a naivete among college students, that having any degree from "X prestigious university" is enough for the big boys to respect them and give them a high paying job. As a consequence I know many graduates who end up working as waiters etc. Perhaps had this tradeoff been made more evident from the beginning, they would have made different choices.

The student isn't deprived of their opportunity to get involved in low-value fields, but that low value is more apparent, which should help their decision process.

And more high value graduates, with higher incomes and a better ability to donate back to the university would probably increase revenues in the long run.

I've heard an even crazier idea.

Have colleges charge a percentage of a student's future income. This aligns economic incentives for the student and the college.

What if I just want to take a few foreign language classes? Does a percentage of my income as payment make sense?

Or I'm self employed and want to get another degree at night just because I can. Does a percentage of my income as payment make sense?

What if I studied art history, but the side project I was working on after classes was acquired by Google for $30 million and now they're paying me $200k/year as part of the deal. Does a percentage of my income as payment make sense?

You're arguing with a straw man. Nothing about that plan precludes a fixed fee option in situations where a future percentage doesn't make sense.

> Nothing about that plan precludes a fixed fee option in situations where a future percentage doesn't make sense.

That doesn't really answer his questions though.

Professionally, I work as a software developer and a farmer. If I had taken the percentage of future income option while receiving my CS degree thinking that's all I'd ever do, would the school be entitled to the profits from my farm too?

I didn't even consider farming until I was 26, so it's definitely in the realm of possibility.

I'm massively curious about how you got into farming and how you do that alongside software development.

I grew up on a farm, so I was not without connections, but it was actually on the suggestion of my father that I got started down that path.

I knew pretty early on that programming was my thing. That was where I put all my focus, getting my first part-time dev job in high school. Farming wasn't even a consideration in my mind. It was only after working full-time in the software game for a few of years, dad approached me and suggested that I take some of the money I had earned, find some land to rent, and buy some crop inputs with it. I agreed that it sounded like a good plan, and he helped me put it in motion.

My farm is strictly cash crop, so the work is highly seasonal. Right around this time in the spring and later in the fall there is a lot of work to do, but the rest of the year is fairly light on the time required of me. The second thing that made it work is that I was already telecommuting and my hours were flexible. There have been days where I've been in the field all day and working on code all night. Additionally, it's not completely unheard of to find me writing code from the seat of the tractor. Mobile internet access has been a lifesaver.

I hadn't really considered this to be an interesting topic to the HN crowd, but maybe I should write up a more detailed blog post about it?

>but maybe I should write up a more detailed blog post about it?

Yes you should write more about this. It's fascinating.

And to answer your previous question about whether the college would get a percentage of your farm income, I would have to say yes.

The option of limiting their stake to a particular profession would introduce too many loopholes.

For instance I could get a math degree, and sign something giving them a share of all future earnings as a mathematician--then work as a programmer instead. That being said I'm sure there could be some sort of maximum.

Student loans already work somewhat similarly. There is an option to pay off loans based on a percentage of your total income, so you are in effect paying x percent of your income for y number of years.

Do you mean that, for example, studying law would be more expensive because there is a high probability of future income? Whereas, say, astrophysics would be very cheap because there's a good chance that you'll stay in academia?

uhhh no, opposite. the university would have a greater incentive to foot the bill for your law degree because lawyers will probably make more money, therefore giving them more money.

too extreme. I like the concept: less students would be funneled into liberal arts majors that don't have a high return if the schools realized that they could get a percentage of every bill gates they spit out.

some cs and engineering programs are shut down because they are the most expensive to run. this would reverse that trend.

seems kind of like VC applied to schools. but then the schools would have to be completely free: we are giving away a percentage of our company, we share in returns with our investors, but we sure as hell aren't paying them.

I wonder if the institutions would drop the price on credits in "low-value" fields and move away from one $ per credit rate?

This is a great idea.

Place all the information in the hands of students, including the estimated present value of anticipated future earnings, as well as statistical distributions.

Estimating default risk, completion of degree, and so forth should be child's play for analysts who are familiar with mortgages and credit risk. It would be great to apply similar risk comparisons to the higher education bubble.

The irony is that the rising freshmen who are actually numerate are probably already entering high average return fields (CS, engineering, sciences, business) and those who are innumerate (and thus can't interpret those data) are probably entering low average return fields (humanities, sociology, gender studies).

The discussion of formulas for pricing risk and forecasting methods is a great (if unintended) example of the problem with student loans: any government subsidy will distort price signals in unintended ways. Even if government tries to price in risk (which is a good idea), the signals will only be as good as the regression analysis, will still be subject to political pressure, and will still concentrate rather than distribute risk. A poor forecast of the future demand for philosophers, or the decision to train more teachers for the public good will still result in human capital misallocation, even if it's not as bad as the current system.

Compare this to private loans, which use the price system and mechanisms like futures markets to evaluate and distribute risk. Instead of one formula to price risk, many formulas compete and the information they contain is aggregated and summarized by relative prices. Risk is distributed among many lenders instead of concentrated in the hands of government.

The worst possible equilibrium seems to be the system in place now: price signals are distorted by subsidies that cover most of the loan market, while private lenders serve those whose educations are so expensive that they aren't covered by already-generous government subsidies. Private lenders encourage these students to take out risky loans, while strict rules on default shift the risk to borrowers.

There are all kinds of alternatives to student loans, but they will not be competitive while federal subsidies continue to distort the student borrowing market. One of my favorite ideas is human capital contracts (here's a good paper: http://www.cato.org/pubs/pas/pa462.pdf), which would work like equity instead of debt. Students could sell "shares" of their future earnings to investors in exchange for the start-up capital for an education. There are lots of potential pitfalls with this model, but I think it aligns the incentives of lender and borrower (or investor and future earner) in a much more positive way: investors would have strong incentives to help students succeed that do not exist under the current debt-based model. Unfortunately, big subsidies are a big barrier to entry: there's no reason to offer equity when one can easily obtain a subsidized loan.

I chose to attend a large state university on a generous scholarship and graduate without debt rather than attend one of the more prestigious schools where I was accepted and take on student loans. So far, I think it was the right choice. But if I had the necessary capital, I'd start a Y combinator for students: find students like me who would otherwise stay in-state and finance "prestige degrees" at an Ivy or other top university in exchange for a share of future income. The model would operate very much like a start-up incubator: providing mentorship and advice and placing students in well-paying jobs after graduation would have a direct effect on the profitability of the investment. It's just a matter of figuring out how to pick (or make) students who will succeed.

The potential of buying shares in someone's future worries me, because it begins to look like indentured servitude. High school seniors are not financially or legally savvy, and a large oversight system would need to exist to insure that they aren't signing away their futures to predatory lenders.

If the system is like the current loan system, where student debt can't be discharged even in bankruptcy, students run the risk of being settled with a tithe that will prevent them from making ends meet if they can't find a good job.

How would it be worse than the existing situation? As you said, students already can't bankrupt out of student loans.

There appears (I think that is the key word, I'm not actually sure if its true) that paying off debt is more flexible than fulfilling your 'future shares' option.

What about this?

If you make less than x$/year you keep it all. On income above that, you pay y percent (but not more than z $ total). After w years, the contract expires.

How does such a system factor that STEM fields are more difficult and have a higher drop out rate? What is half a Chemistry degree worth compared to an anthropology major and chemistry minor, for example?

Statistics, dear boy. Sure, it's complex, but this is loan companies' bread and butter.

Ah, but changing the rules will change people's behavior, changing the statistics. We will have far more people failing out of STEM fields, with nothing to show for two years of study. Remember that a third of people who start in the sciences end up with a degree in the social sciences and humanities.

Do you really want our CS courses to be filled with people who have no chance of succeeding in the major? People who cannot pass college algebra wasting the time of the class? Who does that benefit?

The problem isn't statistics. The problem is not having an outlet for those drummed out of the sciences.

Closing off (or at least making one think twice about) the option of "do a degree that's not actually any use to them" for such people is a good thing. That they have other bad choices available to them does not diminish this.

So I don't suppose I could persuade anyone that education is a public good?


The professional victims will cry racism and destroy any such scheme. (IQ is a strong predictor of future income, is mostly genetically determined, and some human groups drew the genetic short straw.)

Not sure whether you are being serious or not, but beyond a certain IQ (I think around 120) there is no correlation between IQ and income (or success in general). This has been studied over and over.

You are misinformed; the correlation shrinks a great deal (diminishing returns) and becomes dominated by other factors like personality (Extraversion and Conscientiousness especially), but it's still positive. Look at the Terman study's lifetime incomes.

>some human groups drew the genetic short straw.)

Let me guess which groups: the ones you don't belong to?

I drew a number of genetic short straws, but apparently not one limiting IQ too badly.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact