One huge misconception about student debt is that it is mostly a function tuition fees, or the cost of living. But that’s only partly true.

In fact, borrowing is a function of assessed need (that is, assessed costs minus assessed resources). And, in turn, assessed need is subject to maximums. Governments—at least, governments outside Quebec—don’t simply hand out whatever amount students need. Instead, they put maximums on total aid, which for reasons that defy easy explanation are expressed in weekly terms. So, for instance, a student might have assessed need of $20,000, but might only be eligible for $350/week x 34 weeks = $11,900.

Got that? It’s about cost, resources, and maxima. Now, very broadly, there are a number of ways to increase or lower overall borrowing simply through administrative fiat, without actual student costs even moving a single cent.

For instance, in terms of assessed costs, governments can:

• Increase or lower the limits for allowable costs for non-tuition items: (e.g. expenses for tools or laptops).
• Increase or decrease allowances for living costs or childcare (unlike in skills retraining programs, student loan programs use a system of allowances rather than actual expenditures).

In terms of assessed resources, governments can:

• Change the definitions of dependent and independent students (the former are assessed a “deemed” parental contribution, the latter are not); an example of this is the Ontario government raising the number of years from high school graduation a student must be to become independent from four to six in 2019.
• Change the thresholds and rates at which parental income is “taxed” to create a deemed contribution.
• Change the thresholds and rates at which student income is “taxed” to create a deemed contribution.
• Change the threshold at which other types of income (scholarships, savings) are taxed to create a deemed contribution.

And finally, of course, it can simply raise or lower effective loan limits, either by changing the borrowing limit directly, or changing the amount of grant available to meet a greater or lesser portion of need (e.g. Ontario raised grants enormously to create its targeted free tuition policy in 2016, before seeing them lowered again in 2019 when the Ford government slashed both student aid and grant at the same time) thus leaving loans to take up either a bigger or smaller portion of the aid package. And again, all of this can change debt levels without actual student costs changing a cent.

Now let’s look at what’s been going on in student aid for the last decade or so. There hasn’t been a lot of movement on the cost side, but there have been significant changes in assessed resources. At the federal level, Budget 2016 significantly relaxed both student and spousal contributions, both of which should have had the effect of increasing borrowing (provinces tend to get carried along by changes like this at the federal level, so to all intents and purposes these expansions affected provincial loans as well). This of course was applauded by student groups because the immediate effect here was to make students better off in the short-term by reducing income claw-backs on student loans. I don’t think anyone really thought about the trade-off for the long-term.

More important, though, is the very large increase in loan limits that was announced in the 2023 Budget. Prior to this, since 2006, the federal weekly loan limit was $210/week. Because of the way the Canada Student Financial Assistance Program works (most of the time) on a 60-40 split between the feds and the provinces, that meant provinces were ponying up $140/week and hence students could access $350/week. Then last year the feds raised their maximum to $300/week and most if not all provinces roughly followed suit which means the weekly limit just jumped from $350/week to $500/week. That’s a 42% increase in loan maxima and—theoretically—the potential for a 42% increase in borrowing and hence debt.

Have there been any countervailing forces that might reduce borrowing? On the grants side yes, first with Ontario injecting tons of cash via the targeted free tuition policy of 2016-2019 and then, when that got axed, the feds doubling the Canada Study Grant for the duration of COVID. But then the feds pulled back on that policy last year, which leaves the system as a whole slightly richer than it was in 2015 but not inordinately so. Perhaps more important was another, less-noticed 2019 change in Ontario, which declared that for purposes of provincial student aid, students could not become “independent” of their parents until 6 years after finishing secondary school. This matters because a fairly significant proportion of student borrowing happens in “fifth year”: students whose parents are too wealthy to permit borrowing in the first four years suddenly become eligible to borrow (essentially) the maximum amount allowed once they become “independent” and this exempt from parental income-testing. This is less generous to students in the sense that they are eligible for less money in the short-term, but boy is it effective at reducing total student debt.

These measures restraining debt are important, but they are feeble in the face of a 42% increase in borrowing limits. It will take a while for higher student debt to really register in the statistics: basically, it takes four years for a change like that to really cycle through the system; this year’s graduates, for instance have three years borrowing under the old system and only one under the new system, and the effects will be pretty negligible. But my strong guess would be that by the end of the decade, barring some massive cash infusion of grants, we’re probably going to see rises in average student debt on the order of 20-25%, meaning the national average student debt for a bachelor’s degree (among those who borrow) would top $35,000. And that’s without factoring in much in the way of tuition or other cost increases.

Increasing the generosity of loan programs to make students better-off in the short-term might be good politics and good policy (though not necessarily both). But in the long-term there are consequences. Always.