« Even more on the ineffectiveness of minimum wages as an anti-poverty measure | Main | Saturday Night Fever: Roger Farmer, multiple natural rates, search theory, and share prices »


Feed You can follow this conversation by subscribing to the comment feed for this post.

I think the only other concern would be the lower-middle class in this case. I also knew many students from middle class or upper-middle class families whose parents would not help with the cost of their education. I don't like the idea of basing financial aid on parental income/wealth for this reason, especially since many students are adults. I guess there's no good way to go about this without screwing a good portion who are ostensibly too wealthy to qualify for assistance but not wealthy/fortunate enough to get much help from their parents.

But those cases can't be typical, or else kids from upper-income households wouldn't be twice as likely to go to university as those from lower-income households.

Stephen Gordon, is that the case for Canada, or Quebec?

We had this debate in English Canada in the early 1990s. Governments ended up raising tuition and CUTTING student aid at the same time.

England, on the other hand, did simultaneously raise tuition and student aid, but I'm not sure how successful the policy has been. Interestingly, it was implemented by Labour and opposed by the Conservatives.


OK, demand for a bachelor's degree is inelastic (but -0.03? That low?), the universities would get more and the government would pay less.

But what about the impact on the labour market? A 9.6% decrease in the universities output of graduates would increase demand of skilled personnel asking for higher wages. Good for now, but... demographic projections indicate unusually high level of retirements in the next five years. Have these costs factored in?

I think tyronen is right: the important point is whether governments will actually move the savings from higher tuition to student aid for those who need it. An extreme case is the changes in some of the elite universities in the US. For example, Princeton University now charges roughly $33,000 (US) for tuition, but students with family incomes less than something like $200,000 go free (and that means no loans at all, and applies even to students coming from outside the US). Princeton used to provide less aid to middle class students, and load them down with more loans, but changed its policies to increase participation from this group, and leave them debt-free so they could pursue careers in the less lucrative sectors like non-profits and so forth. Of course, Princeton has its own independent funding which the government can't reduce on a whim (i.e. an endowment of about $1M/student), and I think that's a crucial point. (By the way, this leads to weird distortions: it's far cheaper for a middle class Ontario student to attend Princeton than the University of Toronto.)

I see the problem as political: governments feel limited political danger from underfunding student aid. For quite a while now, the University of Toronto has put more emphasis on soliciting its alumni in order to spend less time with a tin cup at Queen's Park with little to show for it. One would hope to see a virtuous circle develop such that the higher the proportion of the population that gets a university degree, the greater the political pressure to fund student aid--or am I just dreaming here? Economists have a role here by explaining the causal factor that education plays in economic growth and wealth.

I don't think tuition is the full story as some people have pointed out. Part of the issue is that Canada over-relies on public universities. In the US, the Ivy league and other successful private schools have managed to obtain huge endowments and were able to put money into areas that public universities didn't have mandates for. Princeton, Stanford, etc weren't always in the position to give free tuition. They also act as venture capitalists and patent holders; companies like Sun, Google, and Yahoo all had close ties to Stanford. Sun itself stood for Stanford University Network when they were doing early research on networking and computers back in the day. You don't see that happen often at public universities in either Canada or the US.

While I don't think we necessarily need to privatize our universities (technically most of them are non-profits and many do have their own endowments), they need to start being more innovative in how they be successful themselves with the students/alumni they have. Waterloo has been quite successful with it's legendary co-op program, for example. People want to go to Harvard or MIT because they know students there are successful, but if we just try to increase enrollment for sake of enrollment, we're not benefiting the students or society.

Further to Andrew's comment: Do we have the cause and effect backwards? My, admittedly anecdotal, observation is that kids whose families value education go to university. That usually means rich white kids whose parent also went to university, and most immigrants, especially Asians. Both will probably pay just about anything to get their kids into university, knowing it's the only possible way for them to have anything resembling a middle class existence. Poor, rural, or 'blue collar' folks tend to be very ... shall we say suspicious ... of those pointy headed academics in their ivory towers who clearly work too little and are paid too much. After all, real men don't think. The only cases I know of where the parent refused to pay for university were peers from the backwater in SE QC where I grew-up whose families were long time rural poor. They simply didn't see the point. "Go work at the mill" was the attitude - of course, the mill is now gone and they have to scrap a living as best they can doing seasonal odd jobs - or they just give-up and go on welfare.

Furthermore, it's really, really difficult for the rural poor to get out of the trap if they don't have family support to go to university. They have to move away from home to go to school, and that is really expensive and more than a little intimidating. Student loans just don't cover the costs, so they have to work at least part time, and probably full time on top of going to school full time (to maintain their full loan). Compare that to the rich kid who lives at home and pays no rent, doesn't work, drives his/her car to school, has mommy or the housekeeper make their meals and do their laundry, etc ... So not only does the poor kid start with a handicap (no family support), we further punish him/her by insisting they work harder than the rich kid to get the same education. It's perverse.

All the above just to say that raising tuition and substantially increasing transfers to the poor is probably a good idea.

Christopher: The Ivy leagues are notorious for refusing to touch their endowments and NOT providing funding to students who merit funding.

Patrick: That's not what I hear.

It seems our economics discussions almost always seem to "degenerate" into social policies (I guess that's what economics is supposed to serve), but if the poor choose not to go to school for whatever reason then there's not much we can do about it short of forcing mandatory schooling past post-secondary education. I think that would be a bad idea. Not everybody is academic or learns best in a structured learning environment. If we wanted to help the working class, we'd remove a lot of the barriers into the trades where the shortages are getting worse. The apprenticeship limits in many provinces are killing entry.


"degenerate" into social policies

I don't see it that way. People matter.

"... but if the poor choose not to go to school for whatever reason ..."

Oh man.

Reasons like they can't afford it and the barriers to entry are higher for them then for the rich? You may be OK with dooming the poor to perpetual poverty, and wasting human capital, but it doesn't sit right with me.

Look, there's no doubt that some people make poor choices. Forgive the lousy sport metaphor, but when we KNOW the playing field is tilted against the weaker team it strikes me as being stupid and wrong to want to further tilt it against them AND take pleasure in our good fortune to not be on their team.

"If we wanted to help the working class, we'd remove a lot of the barriers into the trades where the shortages are getting worse. The apprenticeship limits in many provinces are killing entry."

If that is the case, then by all means it should be fixed.

"Reasons like they can't afford it and the barriers to entry are higher for them then for the rich? You may be OK with dooming the poor to perpetual poverty, and wasting human capital, but it doesn't sit right with me. y"

OK, please don't accuse me of that. That's not what I said. I do believe strongly in the market, but it' not because I believe it screws people. I very firmly believe that it's mostly barriers that are preventing people from getting ahead, not lack of government support for them. My point was that the private institutions were doing a fine job of helping any poor people that are capable of going to their schools. If anything, it benefits the school via reputation more than it benefits the students.

Back to the topic of tuition, Stephen's points on funding sources caught my eye when I read the post again. Where do we want it to come from if we want secure funding? If we leave it at tuition, the school's funding would probably be more secure so long as students wanted to go to the school. If we left it coming from the government, they'd be at the mercy of the party of the day. This could be further compounded if the government cut funding, but left the price ceiling in effect for political reasons.

As somebody else said, enrollment is pretty inelastic with respect to tuition levels. In actuality people tend to go back to school during bad economic times to upgrade skills. I think higher tuition would be better for the schools, which would in turn be better for the students.

Also, tuition of $1600/year is $130/month. It's essentially free now, only the wealthier are being subsidized as well.


I'm not arguing against raising tuition, so long as it's accompanied by a increased transfers to the poor and lower middle class students. Especially in QC. Tuition there is way too low.

"it's mostly barriers that are preventing people from getting ahead "

Like being poor? Seriously, if money isn't the biggest barrier - especially to the rural poor - then what is? I was rural middle class and it was big stretch for my family to get me through McGill in the early '90's because of living expenses, not tuition. And back in those days rent on a rat hole apartment - especially if you had flat mates - was very cheap compared to today. It would have been impossible (and clearly was) for lower income students. Today, I'm sure many don't even consider it.

This might be of interest:


As long as we do consider the lower-middle class. My concern is that the poor will do ok, but the lower middle class will be left out in the cold.

I think this article from Policy Options provides a good introduction to the policy questions surrounding access to higher education. Then again I co-authored it, so take what I say with a pretty big grain of salt :-)


Two things are clear, I think:

(1) Debate about tuition - about the amount - is inadequate. The literature on access to education tells us that it's much more complex than a simple discussion of price (and keep in mind that much of the literature in Canada is written by economists). In fact the literature seems to tell us that price is fairly low down the list of factors that affect educational outcomes. This is not to say that money (or income) doesn't matter, only that it needs to be examined with a sufficient amount of care.

(2) A discussion of access to higher education that ignores the K-12 system is similarly insufficient. In the article, we argue that policymakers need to consider a "life-course" approach; raising or lowering fees and waiting for something to happen will probably do more to exasperate activists than change enrolment behaviour.

[Edited to add the click-through link - SG]

Thanks for the link. And you're quite right - once of the interesting facts I blogged about over here is that most kids who go to university make the decision before they're 14. Once you've decided that university is part of your future, you take school more seriously. And of course, if you've decided that it isn't, then you're unlikely to make the effort to get the marks necessary in case you change your mind.

JB and Stephen,

I agree exactly on your points. Ability, desire, and preconceptions about higher education matter more than it's price. If somebody grows up in a poor learning environment without access to a good k-12 or home environment, post-secondary education is irrelevant.

People often forget that the poor aren't poor because they don't have money, but have limited access to obtain money or opportunity. If enrollment of the poor at the most prestigious universities in the world is falling, despite price being all but eliminated, then there's something else going on. If people get laid off at a poor rural mill, what's keeping them from moving to Alberta where there's high paying unskilled work?

I think the problem with arguing in favour of raising fees and offsetting the effects with increased financial aid to those most heavily affected is that many of us simply don't trust that the financial aid part will actually happen. My suspicion is that the fees would go up and the financial aid would be put off. I suspect this because it is what usually happens. In the 1990's we were told that health care cuts would be offset by more home care, but the home care on offer didn't increase for another decade. At around the same time, the Americans were told that cuts to welfare would be offset by more job training, but most states didn't fulfill that promise, either. These sorts of cuts in benefits or increases in fees are usually sold to the public with some offset included, but it's usually just a bait and switch.

Christopher Hylarides's description of high tech is inaccurate.

BSD stands for Berkeley System Distribution and there are (at least) 3 versions still alive and independent. Earlier there were WATFOR and WATFIVE (origin determination left as an exercise for the student). There is a reason RIM has its headquarters in Waterloo. There are a bunch of things associated with the U of T that are invisible because of the size of the Toronto economy (Palo Alto is rich, but it's not a big city, neither is anyplace else in the valley).

Jim, I'm a UNIX administrator so I know what BSD is. :-) The difference there is that 1) Berkley is a public university 2) they made BSD open source (a wonderful thing, I might add), but BSD cost them a lot especially in fighting the lawsuit with AT&T.

I also didn't say there weren't spin off effects around public universities, but that a lot of the private universities were doing a better job of directly benefiting it. Also, San Jose and San Francisco aren't far from stanford/Palo Alto (within an hours drive). They're both rather big and benefit rather well indirectly.

JB: Thanks. I paste with emphasis added:

In terms of the desire to work rather than study, it should be noted that a booming resource-based labour market has also had an impact on post-secondary planning, luring potential students away from higher education and toward well-paying manual labour. A study of high school graduates in Alberta, Saskatchewan, Manitoba and New Brunswick found that the proportion of youth choosing to work rather than study immediately after high school increased from east to west, in step with the east-west increase in the youth employment rate.


To quibble with one detail, many of those high-paying jobs in the resource sector available to high school graduates and drop-outs are semi-skilled jobs. Some but not all of those skill set acquisition costs are sunk. Heavy-equipment and machine operators can earn as much or more than a typical university graduate.

One thing I find about the original post's analysis is the classic case of assuming a can-opener - in this case, assuming the existence of an error-free, costless means by which to distinguish between the students who are not going to school because of the increased tuition and a) those who will go anyway but who will qualify for the 'payments to make up the difference anyway, and b) those who would not go (compared to some base case) but for whom the benefits somehow make it worthwhile, and (perhaps most important!): c) those who would go (again, in the arbitrary 'base case') but for whatever reason cannot qualify for the benefits (or are missed by the system, or don't know about them, whatever).

Any analysis that attempts to quantify this and make policy recommendations is missing the point by a wide, wide berth. Whatever system is used to identify those worthy of support has costs of its own (deduct from savings), suffers from leakage (some getting benefits that don't 'need' them), and misses some false negatives that should get support but don't.

There are other areas where this is the case: means testing sounds good but is not cost and friction free.

Sorry, but I've heard this before: "It's too hard to figure out who are poor, so it's easier to just keep giving free money to rich kids."

Means testing is easy and cheap - governments already have all the information they need, and the IT necessary to interpret it.

Stephen: Do you know the current of student loan defaults? I believe it is rather high.

That is something that ideally should change.

I did not argue that we should just continue 'giving free money to rich kids,' and don't see why you've attributed this claim or argument to me. Seems like the economist equivalent of Godwin's Law.

It's disingenuous to claim that it is costless, and it's certainly not obvious that it is 'easy and cheap' - which implies close to 100% accurate and so cheap in the context of the savings/other benefits that it swamps other concerns. I don't see any evidence for this claim - do we know that means testing is that accurate particularly for minors that become adults during the period discussed? And as with other transaction costs, the costs and error rates do not need to be very large to provide quite different results in terms of efficiency.

My point is, with friction and accuracy costs included, the 'optimal point' of subsidisation may be quite different than what is implied here. The approach also doesn't seem to incorporate much in the way of externalities of education.

Another issue is that (like many other social programs) we don't and likely can't know how accurate our data is, and may not find out for years. And the relative impact of these types of errors may be large - if means-testing (primarily tax) data is wrong by a year, this may not be a big issue for a 55-year old with a steady job (gets worked out next year), but for an 18-year old considering whether or not university is a feasible option, error-driven signals on future affordability may discourage beginning university at all.

While this doesn't mean automatic support for 'giving free money to rich kids', any formulation that incorporates these issues is quite likely to find a different 'second best' (compared to assuming an omnipotent means-testing can-opener). I don't know where that point comes out and didn't say (hey, it may mean much higher tuition but much more credible support for subsidies, forgivable loan-grants like in Australia, I don't know), but the issue can't be avoided and hand-waved away by throwing 'free money to rich kids' calumny at those who point out there's no can-opener.

It's worth noting, getting back to Vierstraete's paper, that her data source (the SLID) excludes many of the key factors others (like Frenette) have identified as crucial in understanding the income gap in university participation. I'm not sure to what extent David Johnson's MESA paper includes factors like grades, academic engagement in high school, parental support and expectations, etc., though he does use the same data set (YITS) as Frenette. I wonder if Vierstraete's paper might have led to different results if she had some of those variables.

The comments to this entry are closed.

Search this site

  • Google

Blog powered by Typepad