Student loan debt: it’s still a good thing

Is education a right or an investment? The debate continues here


Firstly, I apologize for not getting involved in the comments last time around. Blog posting involves a responsibility to engage in the resultant debate, but jumping into the pond would have been much too haphazard, particularly given that I wasn’t expecting that volume of response and got to it late. So I’m going to try and address some of the points I feel aren’t being fully understood while trying to cover a lot of ground with respect to how economists typically think about educational issues. There are responses to some of the previous comments at the end of this post.
The decision to attend any postsecondary institution revolves around a basic choice: preferences over potential lifetime income streams. The idea, for most, is that by sacrificing a bunch of weekends and the chance to earn a salary for a few years to read textbooks and write exams, we can emerge from university with expanded labour market opportunities. There are very, very few people who can be credibly told that regardless of whether they attend university or not, they’ll never earn more than minimum wage – and who still choose university. Similarly, there are not many people who’d pay the full costs of university simply for the pleasure of it.

Since different people have different abilities and different feelings toward being stuck in lecture halls for years on end, not everyone makes the same decision to attend university or not. It’s entirely possible the story could end there. The oldest university still operating today was founded as a private charitable religious endeavor. Yet in most countries throughout the world, government has decided to intervene and distort the incentives facing those people making the choice whether to invest in PSE.

Why? Most economists (if they had no idea what university education consisted of) would first ask whether PSE is a public good, i.e. something that can only be realistically provided by government. The leading example is national defense – it’s very hard to imagine all Canadians teaming up to collect the money and fund a military force without some sort of a central body to oversee things. Public goods are characterized by being (a) unexcludable and (b) nonrivalrous. If I’m scared of the Americans invading and I buy a military to defend the country, I can’t decide to not provide military protection to my neighbour; I cannot exclude him. It is nonrival because we can both consume it simultaneously, unlike a hamburger.

However, education is certainly excludable – the professor can close the door – and at least somewhat rival. Lecture seats cannot be filled by an infinite number of people. Instructors do not have infinite time. Lab space in the hard sciences. The straightforward case is therefore very weak for direct public provision of educational services, so it’s incumbent on us to start being cautious about whether the government has a legitimate role.

Let me throw out some reasons from economics textbooks why governments might get involved. Democratic values rely upon an educated populace. Citizens could be boundedly rational or misinformed about the impact of education. Private markets could be flawed and unwilling to extend loans to those wanting to borrow against their future high income to attend university – this is the consumption smoothing argument I advanced the other day. People who earn more will remit more taxes. (For the record, each year of PSE confers about a 5% annual wage premia and an 8% total return, though that’s only one estimate, there remains much uncertainty about the true numbers; there are some tricky statistical issues.) Providing a home for academics to work might result in favourable conditions for other enterprises – this happens a lot in biosciences. Etc.

Since most of us can probably agree these are good things, not as many people will attend university as is socially optimal without government intervention. For example, suppose the benefits to me of getting a degree are $150,000, with an additional $25,000 in benefits to society, for some combination of the reasons listed above. (Just because these numbers are in dollars doesn’t mean they only include monetary benefits. If the student receives pleasure from learning, it’s included in the benefits. This way, I don’t have to introduce utils. Any cost-benefit study will monetize nonmonetary benefits in this way.) If the degree costs me $160,000 (including things like the income I choose to forego by working less), then I won’t attend. But if the government gave me $15,000 to attend university, then I would be better off by $5,000, and society would be better off by $10,000. Actually, more like $8,000 – raising $1.00 in taxes costs about $1.20 in wealth – at least, that’s the figure I used in cost-benefit class – in what is called the marginal excess burden of taxation. But the important point is that there’s scope for win-win here.

The problem arises in that the government is not all-knowing. If I value the degree at $200,000, then the $15,000 from government is free money in my pocket and a waste of resources in generating the funds and transferring them to me. Bigger problems arise as the subsidy increases, say with the same costs and societal benefits as before, but with the government subsidizing the cost down to $50,000. Then someone who only values the education at $60,000 could enroll, be personally better off by $10,000, but cost everyone else collectively $65,000 (or $87,000, counting that 20%)! But clearly it’s impossible to get everyone (anyone?) to accurately state their valuation of the degree before they even start classes. It’s equally impossible to accurately measure the benefits to society from one more person having letters after their name.

This brings us to student loans. Since large subsidies have the potential to generate these losses, which economists would term “deadweight losses,” other methods of reaping the ‘positive externalities’ of education might be worthwhile. Credit markets are imperfect, as illustrated over the last year. If the majority of the benefits of education accrue to the individual, rather than society, then student loans become a very effective tool for achieving the win-win described above: if I value my degree at $300,000, it costs $150,000, but I cannot raise the money as an 18-year-old, the government can step in with a loan to cover the discrepancy, enhance my welfare by the $150,000 and grant society the $25,000 in externalities. Even better, by placing the choice of whether to accept the loan or not on the student, the government implicitly learns the worth they place on their education, thereby significantly reducing the chance of spending a lot of money on someone who doesn’t value the service.

Intentionally or not, the combination of subsidies and student loans both serve different purposes: to tip people over the edge and rake in the social benefits, and to fix credit markets, respectively. Whether you think subsidies should be larger or smaller depends on your assessment of the magnitude of the externalities, but the dominance of student loans over grants/subsidies/etc. makes much more sense from an efficiency criterion. Both, yes, but loans should be first. Given that many Canadians have willingly – willingly! – shouldered billions in student debt, I cannot say that the government offering this choice has proved anything but a benefit to most (obviously, university provides no future earnings guarantees, some do lose out on this lottery, see my response to Joey below). If student debt was such a horrible thing, people would avoid it accordingly. Conclusion, as before: student loan debt is a good thing.

I’m not saying students are better off because they have to pay tuition. But if we decided to implement free tuition, that would basically be making a large transfer of funds from society to (a) students and (b) an incinerator – unless you think that someone else attending university is worth hundreds of thousands of dollars to the rest of society, above and beyond what the individual earns from their degree. I don’t think the numbers add up that high and cannot find any references that would support such a magnitude of externalities.

Beyond this framework, most points raised about how much the individual should be expected to pay are not economics but religion. “Social justice”, “equality of opportunity”, and so forth are termed normative statements. These arguments are ultimately subjective judgments about how an individual believes society should operate, and as such are removed from the microscope of scientific analysis, which is only equipped to discuss how society does operate. Some people may be willing to sacrifice large amounts of resources to ensure that tuition is free to all comers, which is fine in the context of one’s moral viewpoint. All economics can do is compute the likely effects. Now, I’ll try to address some of the specific points raised in the comments.

Joey: I agree in that nobody – or at least very few people – want(s) to default. Maybe, maybe there are one or two people who attend university and plan a personal bankruptcy the week they get out the door. Student loan debt is difficult to erase from the books, harms the credit rating and so on. I agree. There is a fine line between being too harsh on those who didn’t realize dreams through university – particularly since the government subsidies prodded them into taking the risk in the first place – and being too generous and opening the door to large-scale losses. Without the numbers, it’s difficult to judge. As Bob Whitney points out two comments down, there’s no guarantees here, but there’s no coercion, either. University is a risk that people take willingly.

Dale: I’m not presenting a single argument that cannot be found in an introductory microeconomics textbook anywhere in the country. I have probably moderated my points here a little bit relative to last time, but I don’t think anything fundamental has changed. It is ideological only in the sense that evolution is ideological: what I’ve said above enjoys virtually as strong a consensus among economists as evolution does among biologists, except none of us are sure of the numbers and thus have accordingly different responses. As I said, an individual may have moral judgments as to what policies should be pursued – economists like to maximize net benefits – but if one does not accept that social goal, then we enter a philosophical sphere. With respect to your first citation, it is misleading. Here is chapter 7 of the first reference, for example. Every nonmonetary benefit they quote accrues to the individual, not society, so free tuition is not justified no matter how large the nonmonetary benefit is. In fact, if the benefits were only monetary and nonmonetary – but both reaped by the individual – there would be zero case for government. The only way to justify free tuition is the belief that benefits accruing to the rest of society, not the individual, are larger than the costs of education, which I find very difficult to argue and cannot find any evidence to support. I seriously doubt you’ve read any part of either book.

Jeff: I do believe that people who are paying $10,000 to sneak across from Mexico to the U.S. illegally are doing it because it’s a good deal. For them. It may not be good in our eyes – anyone reading this article probably has better options than cramped factory hours at sub-minimum wage, but they wouldn’t do it if they had better options. Or do they just like to torture themselves? Similarly, sweatshops are a dream in some places. Certainly, the career and lifestyle that many Canadians desire are probably only obtainable from university education, barring exceptional cases. But we can’t promise everyone a good job just because they want it. I would still like an explanation for why people shoulder student loan debt if it’s against their best interests. Are they simply stupid? No. They’re taking a calculated risk to improve their lives. They expect that they will be better off with the loan and the degree than without either. On average, thanks to the student loan, they become better off. The government does not force anyone into student loans. As I said previously, how can “no university” be better than the choice between “no university” and “university plus loan”?

jessica: I agree entirely, you’re right, that’s a problem. Assuming that all parents are willing to fork out for their kids education is inaccurate. The education is an investment for the individual, not their parents, and should be calibrated as such. The intent of such legislation is to prevent rich families from using the student loan money – which comes at low, low interest rates – to buy a summer home or play the stock market with, but that doesn’t mean it’s blameless. Once anyone can vote, they should be free to conduct any financial arrangement without the status of their parents being factored in.

Chris: In my eyes, there is a difference. A graduated income tax has nothing to do with education. It is purely an issue of income equality. I think a progressive income tax is a good thing. But it’s got nothing to do with education. Why is paying back a student debt after graduation worse than paying an equivalent amount of higher taxes after graduation? The only difference I see is the former reflects how the individual gets most of the benefits of education and takes responsibility for that, while the tax and transfer invokes all sorts of bad incentives that I’ve talked about at length.

Josh: I cannot find “poverty” in my article, so I can’t respond there. But unequivocally, raising taxes reduces the incentive to work. Repudiating that is equivalent to a book-burning of every economics text on the planet. Like I said, education isn’t free, someone has to pay for it. Again, consider the extreme case: 100% tax rate. Why would anyone bother working? Number of professors = number of universities = 0.

patrick: ‘Default’ does not mean that the student still owes money. I am unsure where you got that impression, but you have it completely wrong. To quote, page three, last paragraph: “default (loans that are deemed uncollectible and lost)”.

Finally, I think it’s also important to note that anyone reading this, by virtue of visiting the oncampus site, is considerably more tied up in academic life than most. Whether a professor or a debt-ridden student, priors on these issues, mine included, are probably biased from the national mean, which is why I think it’s necessary to be objective, rather than advance policies that ‘would be nice’ or ‘sound good’.

Addendum: Fiscal policy, since it’s a hot topic. Whether you believe fiscal policy is effective or not is not the point. The idea is that the government can spend today in order to raise the aggregate demand for goods and services in the economy, which requires employment to produce those goods, etc. From first principles, direct spending is more effective than tax cuts: the idea is to get more money into the economy, so spending a dollar certainly does more than handing someone a dollar (say through tax cuts) and letting them decide how much to save and spend. Conversely, forgiving student loan debt is starting off by dedicating all the money to savings, so you have to count on the second-order effect of the individual to spend out of the payments they would have otherwise made on their debt, so it gets the least money moving of all per dollar of government spending. I won’t make claims about consensus here, because right not the profession doesn’t have any real consensus about whether fiscal policy is a sound idea; though probably the majority are in favour, there is certainly not a hint of agreement on what the proper spending targets could be.

Anyway, there. Way too much text, but I wanted to be as clear as possible. I’ll respond to comments in the morning and at least a few times after that, between bouts of tackling a problem set.


Student loan debt: it’s still a good thing

  1. Andrew, you wrote:

    “5. Government transfers have bad incentives. When the government raises tax rates, it reduces the incentive for people to go out and earn more money – if the tax rate was 100 percent, how much would you work? How much would you thus pay in taxes to fund the educational system? Higher taxes make us collectively poorer as a whole, so if you want to use taxes to fund public projects, you should be sure it is worth it.”

    This remains nonsensical and, yes, is a right-wing canard, not any kind of axiom of economic science (not as applies to the real world anyway). Raising the GST back to 7% is a perfect example of a tax that will have little effect on the “incentive to work”, but then raising the top marginal income tax rate a bit (or establishing a new, higher bracket) will do little either.

    If it were the case that, generally speaking, people could choose to earn more whenever they want, but that each additional dollar earned was taxed at a higher rate, then there might be something to this. However, that not being the case – tax brackets are few and widely spaced – I challenge you to provide empirical justification for this claim. Would you, yourself, refuse the 100k job to keep the 20k job even though you’d end up paying a higher rate on the portion falling into higher tax brackets? It’s still more money.

    Back to the issue at hand, you’ve spent many words by sidestepping the most significant issue – the size, not existence of student loans. While I suppose one could argue that students choose “willingly” to take on loans, they often have little choice – it’s not as if it is in any way rational or reasonable to expect young people to decide NOT to improve their skills and education. In my parents’ time, it was generally possible to save money in the summer to pay for tuition during the year – but few jobs now cover the cost of tuition much less living expenses (which are certainly not optional if you don’t happen to live nearby to a university).

    Beyond that, if indeed large debts are “bad” for homeowners and other people with actual equity, why should they be considered reasonable for students without such assets? That they enjoy the prospect of better careers is not in dispute, but especially during a downturn, the pickings will be slim. What I call for is reasonable tuition fees that can be reasonably expected to be paid by students working summers and perhaps during the year, or else loans that are not as onerous as they currently are.

    (And for all the talk about public goods re: PSE, hypothetical examples are of little value – one could start questioning the “value” of libraries and public schools, and it would still miss the macroeconomic forest for individual level trees.)

  2. You’ve hit on something with the GST, but I don’t think it’s quite right. You can find lots of economists who would support higher GST to lower income taxes, but that’s besides this point. The idea that ‘just kinda bad’ is okay because it isn’t ‘a whole lot of bad’ still concedes that it’s ‘bad’. How much does a “little” incentive to work cost us? On the back of my envelope, an extra four minutes every day per adult is about $10 billion. It’s not nonsensical at all: society is collectively poorer off paying 41% tax than 40% tax from bad incentives, excess burden, etc, unless you attribute the government superhuman powers in identifying and explicitly targeting public goods. This is something I am unwilling to do.

    Here are a few studies to illustrate my point. Lowering the effective tax rate on single mothers substantially increased labour force participation and incomes. Or maybe how tax reforms decreased deadweight loss. And here’s another paper, and a blogspot, if you don’t want to read papers. Basically, for everyone except males working full-time (unless they’re self-employed or well-off), they – wait for it – respond to the incentives of tax rates.

    When you tell me to choose between the 100k job and the 20k job, it’s misleading. Rather, how much overtime will you work when it’s taxed at 50% and when it’s taxed at 35%? That number might be different for lots of people.

    “it’s not as if it is in any way rational or reasonable to expect young people to decide NOT to improve their skills and education”. Except there is no free lunch, as we love to say. It’s entirely rational. If someone will be better off going to trade school than university, or maybe no PSE at all, that’s what they should do – because they’ll be better off. There are costs and benefits to education. Since the individual gets the vast majority of the benefits from the education, it’s only reasonable they should be willing to pay the majority of the costs – which is more than can be earned during summer jobs.

    I didn’t say large debts are bad. Take mortgages. Suppose nobody was allowed to buy a house until they could pay for it upfront in cash? Would that be a better world?

    With respect to your last point, a rational person would question the value of those things as well. Public schools are much cheaper than universities, though, and the social benefits many, many times larger. It’s pretty much unequivocal they are a good investment.

  3. It makes little sense to use public schools as a model for PSE, unless you actually made PSE fit the public school model:

    1) You made PSE compulsory.
    2) You provided the resources necessary so that every student could live at home and obtain a high-quality university education.

    If that’s not going to happen, then you shouldn’t claim that what works in one system should be applied to the other.

  4. Since different people have different abilities and different feelings toward being stuck in lecture halls for years on end, not everyone makes the same decision to attend university or not. It’s entirely possible the story could end there.

  5. Yeah. For those of us who went to university, it’s sometimes easy to forget that many of the people with whom we went to high school wouldn’t have gone to university at gunpoint.

  6. Pingback: Student loan debt: it's still a good thing – Macleans.ca | Easy Loan

Sign in to comment.