Entre los días 19 y 23 de enero de 2015, Alex Usher publica en su blog académico un pequeño curso elemental, en Chile sería avanzado, sobre economía de la sala de clase (universitaria) según lo llama él. Sin duda, un aporte directo para el debate que comenzará a la vuelta de vacaciones, cuando el gobierno envíe al Congres Ncional su propuesta de reforma del modelo de financiamiento universitario.
Let’s start by stating what should be – but too often isn’t – the obvious: universities are paid to teach. They are paid specific amounts to do specific pieces of research through granting councils and other kinds of research funding arrangements, but the core operating budget – made up of government grants and tuition fees – relates nearly entirely to teaching. This is not in any way to suggest that teaching is all professors should do. It is, however, to say that their funding depends on teaching. Want a bigger budget? Teach more students.
This link is more obvious in some provinces than others. In places like Ontario and Quebec, which have funding formulae, the link is clear: each student is worth a particular amount of money based on their field and level of study. In others, like Alberta and British Columbia, where government funding comes as a block, it’s not quite as clear, but the principle is basically the same.
So the issue within the institution is how to get the necessary amount of teaching done. One way to work out how much teaching is needed is this little formula:
X = aϒ/(b+c)
Where “X” is the total number of credit hours a professor must teach each year (a credit hour here meaning a student student sitting in one course for one term – a class with 40 students is 40 credit hours), “ϒ” is average compensation per professor, “a” is the overhead required to support each professor, “b” is the government grant per student credit hour, and “c” is the tuition revenue per credit hour.
Now, let’s plug in a few numbers here. Average professorial compensation, including benefits, is approaching $150,000 in Canada. Faculty salaries and benefits are about 44% of total operating budgets, meaning that for every dollar spent on faculty compensation, another $1.27 is spent on other things. For argument’s sake, let’s say the average income from government is about $6,000 per student (or $600 per credit hour) and average tuition income, including that for international students, is about $8,500 per student (or $850 per credit hour). These last two figures will vary by field and level of study, and by province, but those numbers are about right for undergraduate Arts in Ontario.
So, what does our equation look like?
X = 2.27*150,000/($600+$850) = 235.
In this simplified world where all students are undergraduate Arts students, at current faculty salary rates and university cost structure, professors on average have to teach 235 credit hours in order to cover their salaries. If you’re teaching 3/2, that means 5 classes of 47 students each; if you’re teaching 2/2 that means 4 classes of 59 students apiece.
Now, I know what you’re going to say: there’s not actually that many profs teaching that many students. And that’s true mainly because I’m low-balling the per-student income figure. Add in graduate students and the per-student income rises because of more government subsidy. Choose another discipline (Engineering, say), and income rises for the same reason. But at universities like, say, Wilfrid Laurier, Saint Mary’s, or Lethbridge, which are big on Arts, Science, and Business, and low on professional programs, this is pretty much the equation they are looking at.
Assume the following starting parameters:
Where a credit hour = 1 student in 1 class for 1 semester.
Here’s the most obvious way it works. Let’s say the government decides to increase funding by 10%, from $600 to $660 (which would be huge – a far larger move than is conceivable, except say in Newfoundland at the height of the oil boom). Assuming no other changes – that is, average compensation and overhead remain constant – the 10% increase would mean:
X= 2.27($150,000)/($600+$850) = 235
X= 2.27($150,000)/($660+$850) = 225
In other words, a ten percent increase in funding and a freeze on expenditures would reduce teaching loads by about 4%. Assuming a professor is teaching 2/2, that’s a decrease of 2.5 students per class. Why so small? Because in this scenario (which is pretty close to the current situation in Ontario and Nova Scotia), government funding is only about 40% of operating income. The size of the funding increase necessary to generate a significant effect on teaching loads and class sizes is enormous.
And of course that’s assuming no changes in other costs. What happens if we assume a more realistic scenario, one in which average salaries rise 3%, and overhead rises at the same rate?
X= 2.27($154,500)/($660+$850) = 232
In other words, as far as class size is concerned, normal (for Canada anyway) salary increases will eat up about 70% of a 10% increase in government funding. Or, to put it another way, one would normally expect a 10% increase in government funding to reduce class sizes by a shade over 1%.
OK, let’s now take it from the other direction – how big an income boost would it take to reduce class sizes by 10%? Well, assuming that salary and other costs are rising by 3%, the entire right side of the equation (b+c) would need to rise by 14.5%. That would require an increase in government funding of 35%, or an increase in revenues from students of 25% (which could either be achieved through tuition increases, or a really big shift from domestic to international enrolments), or some mix of the two; for instance, a 10% increase in government funds and a 17% increase in student funds.
That’s more than sobering. That’s into “I really need a drink” territory. And what makes it worse is that even if you could pull off that kind of revenue increase, ongoing 3% increases in salary and overhead would eat up the entire increase in just three years.
Now, don’t take these exact numbers as gospel. This example works in a couple of low-cost programs (Arts, Business, etc.) in Ontario and Nova Scotia (which, to be fair, represent half the country’s student body), but most programs in most provinces are working off a higher denominator than this, and for them it would be less grim than I’m making out here. Go ahead and play with the formula with data from your own institution and see what happens – it’s revealing.
Nevertheless, the basic problem is the same everywhere. As long as costs are increasing, you either have to get used to some pretty heroic revenue assumptions (likely involving significant tuition increases) or you have to get used to the idea of ever-higher teaching loads.
So what are the options on cost-cutting? Tune in tomorrow.
Classroom Economics (Part 3)
Back to our equation: X = aϒ/(b+c), where “X” is the total number of credit hours a professor must teach each year (a credit hour here meaning one student sitting in one course for one term), “ϒ” is average compensation per professor, “a” is the overhead required to support each professor, “b” is the government grant per student credit hour, and “c” is the tuition revenue per credit hour.
I noted in Part 1 of this series that most profs don’t actually teach the 235 credit hours our formula implied. Partly that’s because teaching loads aren’t distributed equally. Imagine a department of ten people, which would need to teach 2350 credit hours in order to cover its costs. If just two people teach the big intro courses and take on 500 credit hours apiece, the other 8 will be teaching a much more manageable 169 credit hours (5 classes of under 35 students for those teaching 3/2).
Now, while I’m talking about class size, you’ll notice that this concept isn’t actually a factor in our equation – only the total number of credit hours required to be taught. You can divide ‘em up how you want. Want to teach 5 courses a year? Great. Average class size will be 47. Want to teach four courses? No sweat, just take 59 students per class instead. It’s up to you.
When you hear professors complain about increased class sizes, this is partly what’s going on. As universities have reduced professors’ teaching loads (to support research, natch) without reducing the number of students, the average number of students per class has risen. That has nothing to do with underfunding or perfidious administrators; it’s just straight arithmetic.
But there is a way to get around this. Let’s say a university lowers its normal teaching load from 3/2 to 2/2, as many Canadian institutions have done in the last two decades. As I note above, there is no necessary financial cost to this: just offer fewer, larger courses. Problem is, no university that has gone down this path has actually reduced its course offerings by the necessary 20% to make this work. Somehow, they’re still offering those courses.
That “somehow” is sessional lecturers, or adjuncts if you prefer. They’ll teach a course for roughly a third of what a full-time prof will. So their net effect on our equation is to lower the average price of academic labour. Watch what happens when we reduce teaching loads from 3/2 to 2/2, and give that increment of classes over to adjuncts.
(.8*150,000) + (.2*50,000) = $130,000
X= 2.27($150,000)/($600+$850) = 235
X= 2.27(130,000)/($600+$850) = 195
The alert among you will probably note that the fixed cost nature of “a” means that it would likely rise somewhat as ϒ falls, so this is probably overstating the fall in teaching loads a bit. But still, this result is pretty awesome. If you reduce your faculty teaching load, and hand over the difference to lower-paid sessionals, not only do you get more research, but the average teaching load also falls significantly. Everyone wins! Well, maybe not the sessionals, but you get what I mean.
This underlines something pretty serious: the financial problems we have lay much more on the left side of the equation than on the right side. However much you think professors deserve to be paid, there’s an iron triangle of institutional income, salaries, and credit hours that cannot be escaped. If you can’t increase tuition, and more government money isn’t forthcoming, then you either have to accept higher teaching loads or lower average salaries. And if wage rollbacks among full-time staff isn’t in the cards, then average costs are going to be reduced through increased casualization. Period.
Or almost, anyway. To date we’ve focused just on ϒ – but what about “a”? Can’t we make that coefficient smaller somehow?
Good question. More tomorrow.
Classroom Economics (Part 4)
Yesterday we looked at ways to get the teaching budget down. Today, we’re going to look at the other half of the cost equation: all that overhead. And we’re going to look at it by asking the question: how big a cut in overhead would it take to equal the effect of replacing 20% of your credit hours with sessionals (which, as we saw yesterday, reduces overall teaching loads by 17%)?
Recall the equation: X = aϒ/(b+c), where “X” is the total number of credit hours a professor must teach each year (a credit hour here meaning one student sitting in one course for one term), “ϒ” is average compensation per professor, “a” is the overhead required to support each professor, “b” is the government grant per student credit hour, and “c” is the tuition revenue per credit hour. Given that equation, the answer to our question is simple: you need to drop overhead by 17%. But how might one go about achieving a cut that size?
On average across Canada, universities spend about $16,300 per FTE student on things other than academic staff compensation (yes, really). Over half of that – 54% or so – goes to non-academic staff compensation: the professional staff, the cleaners, the lab techs, the janitors, etc. They’re all in there. No other single item comes close. The table below shows the full breakdown. Most of those categories are pretty self-explanatory, except perhaps for “other” operational expenditures (which is mostly long-term space rental and property taxes, with a few miscellanies thrown in for good measure).
Now imagine you want to achieve your 17% reduction without firing anyone, or trying to get them to give back salary – what are your options? Well, to start with, it’s important to acknowledge there’s a bunch of things in here that are difficult to touch. Scholarships, for instance. And not paying interest isn’t too smart.
So that leaves only 35.8% of the whole non-academic budget. Squeezing 17% out of that would be pretty horrific; it would require cuts of as close to 50% as makes no odds. What do you think our universities would look like with half the library acquisition budget gone? Half the travel and communications budget gone? Half the budget for light and heat gone? It’s simply not an option.
All of this, of course, means that balancing budgets this way leaves you with very few options other than reducing labour costs. Say you had a way to reduce your non-academic staff costs by 10% – either by wage rollbacks or layoffs, or some combination of the two: you’d still have to find a way to squeeze 20% out of the rest of the non-academic budget to make the math work. And that would be tough.
Bottom line: there is no easy salvation here. Any serious reduction in costs on this side will require some bloodletting in terms of staff. That’s never easy to stomach.
My wrap-up on all this tomorrow.
Classroom Economics (The End)
So we spent Monday looking at the economic basics of classroom and teaching loads, and Tuesday looking at how difficult it is to improve the situation by increases in tuition or government grants. Wednesday we saw that reducing average academic compensation (presumably via increasing the proportion of credits taught by adjuncts) can be quite effective in reducing teaching loads, while on Thursday we saw how trying to achieve a similar effect through attacking costs other than academic compensation would require enormously painful – and probably unrealistic – cuts.
What can we conclude from all this?
There is no silver bullet here. You can’t solve everything on the revenue side because governments: i) aren’t going to fork over the stonking huge amounts of money required to change things; ii) aren’t going to permit large tuition increases; and, iii) at some point are going to put limits on the extent to which universities can escape domestic fiscal problems by becoming finishing schools for the Asian middle class. At the same time, you can’t solve everything by decreasing average academic wages because: i) tenure; ii) unions; and, iii) casualization can’t go on indefinitely. Finally, you can’t solve everything by cutting “fat” on the non-academic side because the size of the bloodletting would simply be too big.
So, realistically, the solution to keeping teaching loads (and hence class sizes) manageable is to work at the margins on all three, at once. The income one is probably the easiest: even if government does not have more money, it could (as I argued back here) allow tuition to rise without students being unduly affected if it simply reformed student aid to make it more efficient and transparent.
On non-academic costs, vigilance is key. Costs need to be kept in check. There is a need to continually become more efficient – which probably means looking more seriously at outsourcing certain functions. Bits of IT come to mind, as do bookshops.
On academic salaries, there’s no big secret about what needs to be done. Every time wages increase, universities either have to get more income, or increase the number of sessionals, or raise teaching loads. That’s simple arithmetic. To the extent an institution can keep enrolments up and get a little bit more money per student, on average, the situation can stay relatively stable indefinitely (though it isn’t going to get any better).
Where this gets tricky is where student numbers – and hence income – start to fall. We didn’t explore that this week because our equation – X = aϒ/(b+c) – assumes that there is budget balance. But when enrolment drops, expenditure has to drop in the medium term because the lack of students means you can’t release the pressure by increasing teaching loads.
So when you see the number of applicants to an institution drop by, say, 20% (as first-choice applications have now done at Windsor) over two years, you start to worry. Without the option to increase loads, expenditures have to fall, and as we’ve seen, the least disruptive way to do that is to increase sessionals. But since tenure exists and you can’t force out a professor and replace them with a sessional, that’s a marginal solution at best. Academic compensation will have to fall: either through wage freezes, pension changes, or a reduction in the number of academic positions. Either that or the institution will close.
There’s no sinister conspiracy here, no evil administrative plots. It’s just math. More people should pay attention to it.