HESA

Higher Education Strategy Associates

Category Archives: tax credits

May 21

Rationalizing Regressive Subsidies

A couple of weeks ago, I wrote a blog analysing the distributional effects of tuition reductions vs. targeted grants, and concluded that the latter was far more progressive in their impact than the former.  In response, Carleton professor Nick Falvo wrote a piece on OCUFA’s Academic Matters website saying that I was “wrong about tuition”.  Because some of his arguments are interesting – to his credit, he didn’t reach for the appalling argument that a regressive distribution of benefits is OK because the rich pay more taxes – I thought I would take some space here to respond.

Although Falvo claims to be demonstrating that my thesis about regressivness is wrong, at no point does he actually address the distribution issue.  Rather, he essentially concedes this point, and then make a series of arguments about why tuition reduction is preferable to targeted grants despite their regressiveness.

Falvo makes five separate arguments about the superiority of a free-tuition arrangement over a tuition-plus-grant arrangement.  The first is that free tuition is more “efficient” than grants because the administration costs are lower. But this is silly.  In fact, SFA administration costs in Canada run about five cents on the dollar.  Why you’d spend billions of dollars on one type of subsidy, just to save a few tens of millions by getting rid of the few hundred public servants who administer the existing programs, is a bit beyond me.

The second argument  is, essentially, that grants don’t work because sometimes tuition rises faster than grants.  But the more efficient solution to this – were this indeed a problem – would of course be to spend more on grants, not decrease tuition.

His third and fourth arguments are mutually contradictory.  One is that targeted subsidies create disincentives to work (the “welfare wall” argument); the other is that targeted grants are too complicated to understand, and that free tuition is more effective because it is easier to understand than a fees-plus-aid strategy.  The first implies that families have quite a good understanding about how subsidies work, and adjust their behaviour accordingly; the second implies that they don’t.  My view is that the second theory is more likely to be the correct one.  Sticker prices are simpler to understand than net prices.  The question really is whether this actually matters.  How much damage does poor communication actually do to access?  Is it sufficiently bad that we should spend an extra couple of billion on it?  For that to be the case, one would need to prove not simply that some people are deterred by financial barriers (undoubtedly true), but that they are deterred in large numbers because of their misunderstanding of extant financial incentives.  On the basis of existing evidence, I’d guess that’s not the case.

Falvo’s final point is that free tuition is a more politically saleable proposition than grants – because more people will benefit, it is easier to create and maintain voting coalitions in favour of it.  Even if that’s true, it is an appalling argument.  Stephen Harper certainly sold the Universal Child Care Benefit much easier than Paul Martin sold the (targeted) expansion of daycare spaces, but that doesn’t make it good public policy.

The argument that the only way, politically, to get a dollar to the youth from the poorest quartile is to give three dollars to youth from the richest quartile is an awfully convenient one… if you’re from the top quartile.  I simply don’t believe we have to settle for a system where the only way to get money to the needy is to buy-off the rich.  And I remain completely baffled why people who claim to be progressive actively promote such an idea.

May 09

Who’s Progressive?

To the extent that finances act as a barrier to higher education, they are an obstacle to those without resources – that is, those who tend to come from lower-income backgrounds.  It is, therefore, simply common sense that if you want to relieve financial barriers, you concentrate resources among those with the fewest means.

Except, it doesn’t seem to be common sense among many of those who consider themselves “progressive” in Canada.  “Progressives”, for reasons that are almost incomprehensible, prefer solutions that give far more money to students from high-SES backgrounds.  Why?  Good question.

The best way to focus aid is to use grants based on income (or, second best, on assessed need).  By using income-targeting, you can get money to exactly who you want.  Say you have $100M that you want to put towards affordability.  Want to give all of it to students in the lowest income quartile?  You can do that.  Split it between the bottom two income quartiles? You can do that, too.  Or maybe spread it out a little more thinly so that it cuts out gradually – say, 60% to bottom income quartile, 30% second quartile, 10% third quartile?  You can do that, too.  Grants make many different types of investments possible.

Figure 1: Some Possible Distributions of Grants Across Income Quartiles

image001

 

 

 

 

 

 

 

 

 

 

 

 

But some people despise this idea.  Some people say things like, “lower tuition is the best form of student aid”.  Implicitly, because people from richer families are likelier to attend post-secondary education, the distribution of $100 million, if delivered in the form of a tuition cut, looks like this:

Figure 2: Distribution of Benefits of a Tuition Reduction, by Income Quartile

image002

 

 

 

 

 

 

 

 

 

 

 

 

That doesn’t actually look so good compared to a grant, does it?  In fact, it’s even worse than it seems.  That’s because a $100 million reduction in tuition ends up affecting other types of aid as well.  For every dollar of tuition reduced, students and their families lose 21-28 cents (depending on the province) in tax credits.  And, to the extent that anyone has provincial need-based grants (as opposed to the mainly income-based federal ones), a dollar less in tuition means a dollar less in need, which in many cases means a dollar less in grant.  Thus, for high-need students (which is not quite the same thing as low-income students, though there is some overlap), a dollar less in tuition can mean $1.25 less in non-repayable aid.  That is to say, they are worse off after the tuition reduction than they were before.  But the rich kids who had no need of student aid to begin with?  They’re better off by $0.75.

All told then, if you spend $100 million to reduce tuition, the spread of benefits looks something like this:

Figure 3: Distribution of Benefits of a $100 Million Tuition Reduction, by Income Quartile, in Millions

image003

 

 

 

 

 

 

 

 

 

 

 

 

Of the $100 million in reduced tuition, $42 million gets recouped by one level of government or another, either through reduced tuition tax credits or reduced grants.  Of the remainder, only 13% goes to the poorest quartile, and only 38% goes to the bottom half.

So, ask yourself: who’s progressive?  The folks who want to give 50, 60, or 100% of their money to kids from the bottom income quartile?  Or the folks who want to give almost three times as much to the top quartile as to the bottom?

April 30

Good and Bad Arguments Against Education Tax Credits

One of the things that has become clear to me in much of the commentary about the Net Zero Tuition material last week was that a surprising number of people really don’t understand how tax credits work, or what their distributive impact is.  Worth a review, then.

Bad Argument: Poor students don’t benefit from tax credits.  It is quite true that students whose income is not high enough to be taxable cannot use the credits themselves in the current tax year – indeed no credits get used that way.  But they can pass them on to parents or spouses who are supporting them, and who presumably find the tax relief of great benefit.  Nearly 40% of tax credits get distributed in this way.  Or, if their parents or spouses have no taxable income (rare), or if they just don’t want to give them up, they can carry them forward until such time as they have taxable income.  Fortuitously, this is usually about the same time their student loans start coming due.

Better Argument: Tax credits would be better if they were refundable.  No one would be better off in the end, but this way, at least, one could get rid of the carry-forward provision, and those students who currently have to wait to get their money could get it faster.

Bad Argument: The rich benefit more from tax credits than the poor.  This is a tricky one, because it has a different answer if you’re making this claim at the individual level, or on aggregate.

It is certainly true that some tax expenditures are worth more to the rich than the poor.  Tax deductions, for instance, reduce taxable income, which obviously is worth more if you’re in a higher tax bracket.  But our whole system shifted from deductions to credits twenty-five years ago.  And tax credits – by definition – are worth the same amount to everyone, regardless of their income.  The only case where this is not true is if someone has no taxable income – but that’s irrelevant for education tax credits, because of the carry-forward provision.

Where this argument is true is with respect to aggregate spending.  On aggregate, upper income families do receive more money from tax credits, because youth from upper-income families are more likely to attend PSE.  For most people, that’s a good reason to dislike them.  What’s hysterically funny, though, is that at least some of the people who use this argument simultaneously argue for greater tuition subsidies – which have exactly the same distributional consequences.  Charitably, these people could be described as “confused” (less charitable descriptions include: “cynical”, “ridiculous”, “dumber than a bag of hammers”).

Good Argument: Money spent on tax credits would be better spent on up-front, need-based student aid.   There are too many people receiving it who really have no need of it.  Spread that money – that big chunk of over $2 billion/year – less thinly, focus on those who need it most, and our system would be much more effective and equitable.

April 28

Mobility Responsibilities

Saying that we should remove barriers to student mobility sounds like a motherhood issue.  But scratch a little deeper, and you’ll see that, in fact, Canadians are pretty equivocal on the concept.

For starters, while everyone loves inbound mobility (come here!  It’s a great place!), there’s a pretty deep streak of protectionism in Canadian provincial governments on the issue of outbound mobility.  The sentiment of “let’s keep our kids at home” runs deep in many parts of the country.  It wasn’t until the advent of the Millennium Scholarship Foundation that all student aid programs became portable in most provinces (though Quebec maintains a policy of not funding students to go out of the province, unless the program is not offered in the province, “and it is in the interest of the Quebec collectivity”.  Yes, really).

But whose responsibility is mobility in the first place?  In Europe, with respect to tuition and student aid, it’s the receiving country who bears the cost – no matter where they’re from, students pay whatever the locals do, and have access to whatever student aid program the locals provide.  In Canada, our default assumption is that host provinces are supposed charge equal tuition to all Canadians regardless of their place of origin, but the responsibility for mobility on student aid lies with the sending province.  One can move from Nova Scotia to Manitoba and pay Manitoba rates of tuition, but one still has to rely on Nova Scotia Student Aid.

But there are exceptions.  The most clear-cut is Quebec’s insistence on charging tuition fees to out-of-province students.  Less clear-cut (but still clearly discriminatory), are the cases of Nova Scotia and Ontario.  The former provides tuition rebates to Nova Scotia students, but not to other Canadians.  Ontario has a variety of subsidies that are only available to Ontario students attending Ontario institutions (the tuition tax rebate is one, as are the many provincially-mandated, institutionally-managed access funds – funded through a “tax” on tuition paid by Ontarians and non-Ontarians alike).  These are all anti-mobility measures: they effectively create a two-tier tuition policy within a province, and (in Ontario’s case at least) provide extra subsidies to students who chose not to leave.

Interestingly, while Quebec’s two-tier tuition system is usually portrayed as a piece of xenophobia and rank insolence to other Canadians who, through equalization, are partially picking up the tab for Quebec’s lower tuition, Nova Scotia and Ontario are given a total pass, despite their policies having almost identical effects.

What that tells us is that Canadians don’t mind mobility barriers as long as they are dressed-up as “affordability enhancements”. Ultimately, such measures are self-defeating; as with trade barriers, eventually everyone is left worse off.  But there are enough small-minded politicians out there to ensure that these kinds of tactics always have a potential audience.  As budgets get tighter over the next couple of years, there’s a good chance we’ll see more examples of discriminatory tuition fees, loans, and grants being made non-portable across provincial borders.  I hope that’s not the case, but history doesn’t give me huge grounds for optimism.

April 25

Nova Scotia Ditches a Bad Subsidy

About a decade ago, a really bad policy idea started making its way across the country’s “have-not” provinces.  I can’t remember if it started in Saskatchewan or New Brunswick, but within a couple of years it had spread to Manitoba and Nova Scotia, as well.  The details (and generosity) of this policy varied somewhat, but the gist of it was this: “let’s pay our graduates not to leave the province by refunding a portion of their tuition, via tax reductions, once they graduate”.  Sometimes this was dressed up as a “talent attraction policy”, in the sense that it would benefit people coming in from outside the province; in the main, however, it was understood that this money was mainly about keeping “our” kids at home.

Now this was a dumb idea from almost any policy perspective you can imagine, but the two most obvious ones are:

1)      Effectiveness.  Most youth, even in the most economically depressed provinces, tend to stay where they are: in the provinces where these programs were introduced, the “stay” rate ranged from about 75-85%.  So even if you bring the stay-rate up by 10 percentage points, that still means that for every student you successfully keep, eight others will get to keep cash for doing exactly what they were going to do anyway.

2)      Horizontal equity.  If you have a couple of tens of millions in cash that you want to devote to youth – and lord knows there isn’t much of that around – why in the name of all that’s holy would you hand that money over to the group of youth who are the most employable, and have the best prospects?  Especially if you’re not actually changing their behaviour, all this does is reduce the cost of education and make it easier for tomorrow’s upper-middle class to start accumulating assets.

Anyways, though it wasn’t much noticed outside the province, Nova Scotia dropped the tax rebate, largely because it was ineffective – young people continued to leave the province.  While it drained a lot of money, it simply wasn’t big enough to change many people’s minds about leaving.  And this makes sense: if the reason someone moves from Halifax to Toronto is a $10K/year difference in pay, a $2k tax rebate isn’t going to change their mind.

Of course, it would have been a lot better if the Nova Scotia Government had actually put that money back into some other youth-serving purpose – the community college, say, or student assistance (a category in which Nova Scotia remains among the most miserly in the country).  But with the province hemorrhaging money, it’s not exactly a surprise that this money is going straight to deficit-reduction, no matter how unfortunate that might be.

Interesting trend, though: first Quebec and now Nova Scotia have started dialling back on tax credits – with no apparent backlash.  Hopefully, this is the start of a trend that allows us to restore some sanity to the way we subsidise higher education.

November 22

Higher Education Tax Credits

Last week, CD Howe released a very good paper (available here) written by my colleague, Christine Neill, on the subject of tax credits in higher education.  It’s an important piece, because it not only puts in one place a number of key factual aspects of tax credits (what they cover, how much they’re worth), but it also places them in the context of research on behavioural economics.  Given what we know about behavioural economics, she asks, what should we expect these huge – largely hidden – subsidies to achieve in terms of improving access? (The answer: not much.)

The three key bits for me are: 1) she has a nice, useful breakdown of how the credits are used (current students: 31%; transferred to parents/spouses: 38%; carried forward: 31%), which to my knowledge hasn’t been done before; 2) she has a great finding, showing that 42% of the total value of tax credits go to families in the top income decile; and, 3) she makes a reasonable case that the tax credits would be improved if they were refundable rather than non-refundable.

Christine and I agree on pretty much everything about this subject, so my comments are mostly niggles.  One thing missing is any discussion of the role of federalism in the development of tax credits.  The reason the feds have often preferred credits to student aid as a means to fund education is, simply, that they can do it without faffing around with the provinces.  Throw a couple hundred million at student aid, and its tough to avoid displacing existing provincial aid; throw it in tax credits, and it all gets to its intended recipients.  That’s not a negligible factor in their rise.

A second niggle is that, while I agree that refundable credits are preferable to non-refundable, I’m not sure how it would significantly improve many of the issues Neill raises with respect to behavioural economics. If the problem is that students don’t get the money right away, the fact that they’d get it 9 months after enrolment, instead of 21 or 33 months, doesn’t seem like a huge gain to me.

I’d actually go a bit further: In addition to refundability, focus on transferability.  Currently, transfers are limited to parents, grandparents, and spouses.  But why not include educational institutions on that list?  That way, institutions could prepare T2202As at the time students actually enrol; the student could sign it back to the institution at the time of fee payment, and the institution could provide the student with an instant rebate, and then claim the money back from the government itself.  Then we’d really be cooking with gas.

Food for thought, anyway.

October 01

How the Zero-Tuition Crew Could Learn to Love Tax Credits

So, let’s say you’re among those who clings to the idea that tuition isn’t just a massive give-away to upper-income families.  Let’s say you really, really believe that tuition – sticker-price tuition, none of these “net price calculations”, thank you very much – affects access.  How would you go about gathering evidence for your point of view?

Ideally, of course, there would be some data showing that, as fees went up, participation went down.  Problem is, the data doesn’t show this.  To wit:

Tuition + Ancillary Fees (in $2013) and Participation Rates, Canada, 1993-94 to 2012-13

 

 

 

 

 

 

 

 

 

 

 

 

Well, OK then.  Maybe if you’re desperate to prove a point, you might relax your scruples about counting grants against tuition.  Maybe it’s all the extra student aid pouring into the system which has made a difference?

Tuition + Ancillary Fees (in $2013), Net Fees (i.e. Minus Grants) and Participation Rates, Canada, 1993-94 to 2012-13

 

 

 

 

 

 

 

 

 

 

 

 

Well, no.  Damn!  Now what?

Well, here’s one possibility.  Check this out:

Real Tuition + Ancillary Fees, Minus Grants and Tax Credits, as a Percentage of Average After-Tax Family Income

 

 

 

 

 

 

 

 

 

 

 

 

If you add grants and tax credits, and adjust for inflation, and then adjust for median family income, you get to the point where you realize that, in fact,  fees were more or less constant in terms of affordability over the last decade or so.  So voila!  That damnable rise in tuition fees that so inconveniently accompanied the increase in participation rates?  Turns out it didn’t happen – at least if you properly count subsidies.  In fact, one could now convincingly argue that the lack of a relationship between fees and participation is a huge hoax, and that the rise in participation was actually fuelled by the massive infusion of tax credits and grants over the past fifteen years, which effectively netted out the impact of tuition.

There’s a slight problem with all this, of course.  And it’s that the zero-tuition crowd has a long track record of claiming that subsidies have no effect whatsoever, because sticker price is the only thing that matters.  No one thinks tax credits have any effect on participation, and the more hardcore of the zero-tuition crowd don’t believe grants matter either (True story: a long-time student leader informed me the other day that the whole concept of net tuition was bogus, and that grants should never be included in discussions about affordability, because “grants don’t reduce tuition, they just help you pay for it”.  Yes, really).

So the zero-tuition crowd has three choices here:

  • They can keep their beliefs about subsidies, but accept that participation has risen along with tuition, and admit that, perhaps, their views on tuition are wrong.
  • They can renounce everything they’ve ever said about subsidies, run with the idea that affordability has not been deteriorating, and arguing that this is what has fuelled the rise in participation.
  • Ignore all empirical evidence, and continue their evidence-free approach to the whole question.

No prizes for guessing which is likeliest.

September 06

Grants and Net Prices

Yesterday, we saw how tax credits lowered net prices by refunding students (or their families) roughly one out of every three dollars spent on tuition.  But that’s not the whole story, because there are a lot of university students who also get some form of non-repayable assistance (i.e. grants); for them, tuition is even lower.

Let’s start with Quebec, where net tuition after tax expenditures is a mere $1,555.  Data from the latest Aide Financiere aux Etudes annual reportadjusted for known changes in student aid expenditures, suggests that somewhere in the neighbourhood of 50-55,000 university students are receiving grants, which, on average, are worth $6,380 apiece.  Meaning that net tuition for grant recipients in Quebec is in fact negative $4,825.

In Ontario, net tuition after tax credits is $5,680.  Everyone with a family income under $160,000 is eligible for the Ontario Tuition Grant, which is (effectively) worth $1,730.  So that means that, in fact, for a considerable majority of the full-time undergraduate population, net tuition last year was is $3,950, which is lower than it’s been at any time since 1998-99.

Figure 1 – Net Real Tuition in Ontario, After Tax Credits and Tuition Rebate, 1995-96 to 2012-3

 

 

 

 

 

 

 

 

 

 

 

 

Here’s where the analysis gets tricky.  In the CSLP zone, many people receive more than one grant, mainly because of the overlap between federal and provincial aid.  But while we know the average size of each grant, there’s no method of working out how many of the 320,000 recipients of federal grants (who receive on average 1.18 federal grants each – you can get more than one) also receive one of the 250-300,000 provincial grants.

However, based on a little bit of policy analysis – and some phoning around to friends in provincial governments – I reckon that between half and 2/3 of all provincial grant recipients are getting federal aid, as well.  That would give us a ballpark of about around 430,000 total grant recipients, of which roughly two-thirds are in universities.  With roughly $1.2 billion being given out in the CSLP provinces, that suggests that the average grant recipient there receives about $2,800.

Taking that data and merging in the Quebec numbers gives us the picture we see in Figure 2: 

Figure 2 – Actual Net Costs, Canada, 2012-3

 

 

 

 

 

 

 

 

 

 

 

 

Across Canada, the sticker price of tuition and fees last year was $6,331.  As we saw yesterday, that falls to just over $4,300 when you take tax credits into account.  And that’s the real net cost for about two-thirds of the full-time student body.  But for the other third, the third that gets grants, real net tuition averages just over $1,000 – and it would appear that for a substantial proportion of these students, the actual cost is negative.

So, when the Statscan tuition numbers come out, just remember: no one actually pays the amounts Statscan reports.  Most students pay about 66% of the sticker price, and the neediest third (proportions may vary by province) pay about 17% of the sticker price.

September 04

The Impact of Tax Credits

One of the many ways that Canada stands out as unusual in its financing of higher education is the degree to which its subsidies to students and families runs not through loans or grants but through tax relief.  Well over $2 billion/year goes out to students that way; for full-time university students in Canada last year, tax credits on average amounted to $2,200, or almost a third of the sticker price.

But given how central tax credits are to our system, what’s incredibly puzzling is that no one seems to actually understand how they work.

These are what you call “universal subsidies” – everybody gets them, regardless of need.  Right-wingers should (and often do) like this because it’s a straight voucher-like mechanism.  Left-wingers should (and often do) dislike this because, sans need assessment, they are much more likely to end up in the hands of wealthier families than poorer families.

But while it’s true that, when it comes to tax credits, the political right usually lines-up in favour, and the left usually lines up against, the fact of the matter is that the distributional impact of these grants is absolutely no different from a tuition subsidy or rebate – as the NDP have implemented both in Manitoba and Nova Scotia.  There is not one iota of difference.  And yet, left and right line-up completely differently.  What is brilliant/heresy if done through the tax system becomes a waste of money/the epitome of progressiveness if done through tuition subsidies.

(Shorter version: most partisans are stupid.)

The other objection to tax credits is that, “they don’t deliver aid when students need it”.  And while that’s a cogent critique, I’m not sure it’s as powerful as some people think.  Thirty-five percent of credits get transferred to parents, and I’d guess their need for them come tax time is probably as acute as their need for them in September.  For the 45% which get claimed by students in the year they are issued (most of whom keep the credit because they need it to offset earnings), they actually see the benefit every single time they get a paycheque, via lower tax withholdings.  Pretty useful, no?  It’s really only the 20% that carry the credit forward who might really have a serious complaint, and even they probably find the bigger tax rebates handy for repaying student loans after graduation.

The point here is not that tax credits are ideal; their goofy distributional consequences alone are enough to put them outside the pale.  But they do help reduce costs – a lot.  Tax credits mean that every time tuition goes up by a dollar, governments effectively pay for $.33 of it themselves.  We should pay more attention to them when thinking about the real costs of education.

November 29

Tax Deductibility

Here’s a question which I think may get a little more attention over the next few years, largely because the OECD has made it one of the principal recommendations under its new Skills Strategy. Should tuition be tax deductible?

We’ve been here before, actually. Back in 1959, Diefenbaker was looking for something to counter the “national scholarships” idea being touted by Lester Pearson which was (a) less costly and (b) less irritating to Quebec. Eventually, he glommed onto a pitch of making tuition tax deductibility made by then-Tory Youth President Ted Rogers (yes, him). Needing to square it with Duplessis, he then sent an emissary – Laval law-student and then-Tory Youth Vice-President Brian Mulroney (yes, him) – to negotiate the details. And thus federal tax assistance for students was born.

One peculiarity of the system was that fact that the tuition amount was made transferable – it wasn’t just students who could claim it, but Mom and Dad as well. This introduced a serious element of regressiveness to the policy, because deductions are worth more to rich families than poor ones. This problem was solved when then-Finance Minister Michael Wilson converted a number of deductions into tax credits in his 1991 tax reform. This eliminated most of the policy’s regressiveness; the remaining bits were taken care of in 1997 when students were allowed to carry-forward any unused portions of the credit.

By and large, our current policy on tax assistance for tuition is the right one: there should be some kind of tax offset for investments in human capital, and for traditional-aged students it makes sense that the sum be a flat one rather than one which grows with parental income. But the case looks slightly different from the perspective of older learners paying to upgrade their skills. For them, a very good case can be made that their human capital investment should be treated like any other type of investment – i.e., fully tax deductible. It would significantly lower the cost of education for working people and require a lot less bureaucracy to administer than some sort of new direct loan or grant program.

So, a proposal. Let’s make tuition deductible if it is claimed by the person who actually obtains it. This would provide a benefit to older workers taking shorter-term courses without giving any new money to current students, most of whom would see no change in status because they’re in the bottom tax bracket anyway. But to avoid regressiveness, any transferred amounts would remain credits, as they currently are. And we can pay for it by clipping the education and textbook tax credits, which make less than no sense anyway.

Agree? Disagree? Let us know what you think.