Some facts about tuition and access to post-secondary education

The issue of tuition fees and access to post-secondary education is one that keeps popping up, and it's one that is often a source of frustration to me. One problem is that the basic facts are not well-known. And many of the factoids that have made it into popular circulation are either taken out of context, or simply wrong.

So here are some facts, almost entirely lifted from the excellent survey The Price of Knowledge 2004: Access and Student Finance in Canada (393-page pdf), by Alex Usher and Sean Junor.

I'm going to avoid editorial comment here, and I'll update it as new data become available. If I've missed something, or if you know of more recent data, please let me know in the comments.

Here's a summary:

1. People decide whether or not they want to go to university well before they reach the end of high school.
2. University students are more likely to come from higher-income families.
3. Both tuition fees and participation rates have been increasing.
4. Notwithstanding the increase in tuition fees, the percentage of those who cite financial barriers as a reason for not going to university has decreased.
5. The financial benefits from going to university greatly exceed the financial costs.
6. The benefits of going to university are underestimated, particularly among youths from low-income households.
7. Academic costs account for about 25% of the total costs of attending university.
8. Tuition and other fees have been rising faster than inflation, except in Quebec.
9. Student assistance has remained roughly constant over time.
10. There's been a shift away from need-based to universal  expenditures in student assistance.
11. Financial barriers explain only a small fraction of the difference between the participation rates of people from low- and high-income families.
12. The trend in real spending per student is positive; but it's a declining share of GDP.
13. Student debt loads have been increasing.
14. The incidence of student debt has also increased.

Details below the fold.

1) People decide whether or not they want to go to university well before they reach the end of high school:

This surprised me, but it really shouldn't have. Once you've made the decision to go to university, you make the effort to do well at school.

2) University students are more likely to come from higher-income families:

A university student is twice as likely to come from a household in the top income quartile than from the lowest quartile. Interestingly, this correlation is largely driven by the education level of the parents: once you condition on parents' education (which is of course highly correlated with income), income per se has very little extra explanatory power. This is consistent with point 1): the key isn't having rich parents, it's having parents who have already set the example of going to university.

3) Both tuition fees and participation rates have been increasing:

4) Notwithstanding the increase in tuition fees, the percentage of those who cite financial barriers as a reason for not going to university has decreased:

5) The financial benefits from going to university greatly exceed the financial costs:

Of course, the benefits associated with a lifetime of higher earnings should be discounted. My own back-of-the envelope estimate of the net present value of the education premium is about $450,000. If they face a (say) 45% marginal tax rate, that brings it down to around $250,000. But that doesn't change the fundamental conclusion: the financial benefits exceed the financial costs by a wide margin.

6) The benefits of going to university are underestimated, particularly among youths from low-income households:

7) Academic costs account for about 25% of the total costs of attending university:

8) Tuition and other fees have been rising faster than inflation, except in Quebec:

9) Student assistance has remained roughly constant over time:

10) There's been a shift away from need-based to universal  expenditures in student assistance:

11) Financial barriers explain only a small fraction of the difference between the participation rates of people from low- and high-income families.

In a recent StatsCan study, Marc Frenette finds that only 12% of the gap can be attributed to financial constraints. But there are limits to how far we want to push this overall finding:

Despite the weak evidence on credit constraints, there are two important caveats to keep in mind. First, even if credit constraints do not matter a lot for the population of youth as whole, they may matter for certain groups of students in some instances. For example, Ontario students from middle-class backgrounds saw a large decline in their probability of pursuing a professional degree following the large and sudden tuition fee deregulation in these programs in Ontario universities (Frenette, 2005b). Another example is students who grew up living out of commuting distance from a university. The additional cost of attending a university away from the parental home is greater than $5,000 (Barr-Telford et al., 2003), which appears to reduce enrolment among students from lower-income families who must move away to attend (Frenette, 2004). Second, even if credit constraints could be 'ruled out,' it is important to note that this would be conditional on the existing financial aid system. Removing that system may (or may not) introduce credit constraints.

12) The trend in real spending per student is positive; but it's a declining share of GDP:

There's presumably a certain amount of demographics driving the share of GDP number. After the baby boom went through the system, the share would have decreased; the recent increase is the baby boom 'echo'.

13) Student debt loads have been increasing.

The average debt load of a university student who takes out student loans has more than doubled in the past 15 years, and is now about $24,000. The proportion of those with debts greater than $15,000 has almost doubled between 2003-2006.

14) The incidence of student debt has also increased. The proportion of students who graduated debt-free has gone from 55% to 40% since 1990.

Debt loads and the incidence of debt seem to have leveled off recently, but the authors of the 2006 edition of The Price of Knowledge (21-page pdf) warn that the previous trend may resume:

On the surface, the stabilization of university graduate debt is reassuring, considering how quickly debt levels grew prior to 2000. A deeper look reveals reasons for concern. First, it appears that the proportion of university students graduating with debt is rising. Second, recent changes to the Canada Student Loans Program (and complementary changes by a number of provincial loans programs) will make more middle-income Canadians eligible for loans, and provide existing borrowers with high levels of financial need with larger loans. This raises the prospect of further increases in both the proportion of students graduating with debt and the amounts owed by borrowers.

Comments

[tom s.]

As someone who is just about to be directly affected in my bank account by university tuition, this is fascinating. Thanks.

I wondered two things: in some cases it seems there is a confusion between the prospective student and the prospective student's parents. For example, (4) "the percentage of those who cite financial barriers as a reason for not going to university has decreased" - you may get a different opinion from my son and from me if you ask about financial barriers to him going to university.

Second, looking at the (average) benefit of a university education seems to miss a big point (which I have not seen taken up elsewhere). I wonder what the standard deviation is on that number? A big average benefit with a significant prospect of a negative benefit would have a different impact on many people's decision making than a guaranteed big average benefit.

[Stephen Gordon]

I'm pretty sure they ask the students themselves. I agree that the issue of just how much parents are willing - as opposed to able - to contribute is often a thorny one. But it's probably hard to get much more than anecdotal evidence on that.

And there certainly is a lot of variation in benefits, which should affect the willingness to take on debt. But apparently, the link between debt and post-graduate income isn't as strong as might be expected. I'll see if I can track something down.

Thanks!

[Christine]

They ask the students, not the parents. One of the problems, though is that the question is vague, and can be interpreted as meaning (a) university costs too much; or (b) I don't have the up-front funds to go, but would if I could get access to a cheap enough loan. These two are very different things from a policymaker's perspective.

There's a recent AER paper on the risks associated with university education, and some research suggesting that it is a quite risky investment, so university students tend to be relatively risk-taking compared to the non-student population. This wasn't exactly my feeling when I was going through, though: I considered not going to university very risky, and thought uni was the safe option. I suspect perceptions of the risk of uni depend on your family background, and on your peer group. This makes sense, because actually measuring the return to education is very difficult - measuring the risk around that return in any clearcut way is extremely nasty.

Stephen also mentions the differences in estimated costs of attending university depending on family background. If true, this would systematically cause students from low income/education families to attend uni less.

Averages are particularly meaningless in the area of student financial aid. There was a huge jump in the maximum amounts students could borrow under the CSLP in 2004 and 1994, which caused big increases in borrowing among those near the maxima, but didn't have much effect on anyone else. This still showed up in the averages, but doesn't look nearly as dramatic as it actually was for the affected students.

[true dough]

It might be interesting to look at the variation in government spending on education across Canada and the number of tax-paying citizens in each region. Under the current system, Atlantic Canada seems to have too many universities chasing too few tax dollars from its small population to prop up the number of universities that it does. As a consequence, some students in this region pay higher fees than anywhere else in Canada.

[Christine]

Not sure that's true, actually: NS still spends quite a lot per capita on universities, but not so much per student. What that means is that NS, eg, is doing is oversupplying places relative to the rest of Canada, but undersubsidising them, meaning out of province students pay more. And they do have programs to provide a bit of extra assistance to students from their own province (tax credits in NB, eg). If you net it all out, it'll probably look something like students FROM the region end up paying a bit more on average than those in other provinces, but students from other regions studying in the maritimes probably bear most of the cost. But this is a reasonable response if other provinces are trying to free ride on NS universities - have low tuition fees in their own provinces, but few places, forcing students out of province.

Optimal industry structure is another matter, and something I know less about (do universities have significant increasing returns to scale? I don't really know).

Basically - it's more complicated than it looks, and I don't think anyone's really going to be able to figure it all out. Though The Price of Knowledge does an excellent job.

[music]

What do you mean ?