Many of my Canadian readers will likely have read a piece that has been circulating on the internet from Kim Siever, a self-described leftist internet journalist from Lethbridge.  The headline says it all: UCP Government to Cut Post-Secondary Spending by $1.5 Billion; That Number Rises to $3.5 Billion if you Factor in Inflation and Population Growth. 

You know how I am always on about Economic Impact Analyses always being forms of competitive counting? Methodologically speaking, this is worse.

Ok, so here’s what Siever found.  By looking at the Advanced Education budget (he seems to be looking at the table on page 105),  he notes the difference between the ministerial budget for the years 2020-21 through 2023-24 (though that last year is probably not within the UCP’s current mandate, which will end in the first half of 2023) and 2019-2020, and comes up with figures of $345 million, $408 million, $392 million and $359 million.  Therefore, says Siever, higher education can expect a cut of $1.159 billion for the next three years, or – if you include last year – $1.504 billion.

Analysts do not describe cuts this way.  A cut is the difference between spending in Year T⁰, in this case 2019-20, and compare it to spending in year T¹, in this case 2023-24.  Under this method the cut from 2019-20 to 2023-24 is $359 million.  This $1.504 billion number is not so much a cut as it is a change in lost aggregate total income over time.  Total income over a combined 3 (or 4) years will be $1.159 (or $1.504) billion, compared to what it would have been if some idealized baseline had been maintained.  The 3-year cut of $1.159 billion turns into a 4-year cut $1.504 billion cut not because income is being reduced (income is actually projected to increase by $31 million between 2022-23 and 2023-24), but because we are simply adding another year to the total in which income is lower than it was in 2019-20.  Add another few years to the projection and we could get this up to $2 billion PDQ.

You know who else does this kind of thing?  The federal government, when it wants to describe the size of its investments in something.  Like, say, research.  Back in 2018, in response to the Naylor report, the feds decided to increase federal expenditure on research by 14%.  This was described as THE HUGEST INCREASE IN RESEARCH FUNDING, LIKE, EVER mainly because when the feds totted up spending, they did so over the space of five or six years.  In fact, the 2018 boost to spending was nothing like the increases of the late Chretien years in percentage terms: it’s just that if you add up increases over a more years, you can get to a bigger absolute number.  This is exactly what Siever is doing with the math here.

Now, to make the cut seem even bigger, Siever adds in inflation.   This is totally legit in the sense that inflation matters – I always try to show changes in real dollars.  But because Sievers is working on a much larger multi-year aggregate base, his “cuts” jump from $1.5 billion to $2.6 billion, or from about 27% of annual expenditures to 44% of annual expenditures.  In a more normal T⁰ to T¹ change, that increase goes from $359 million to $786 million, or from about 6.5% to 13.3%.   

Then – this is the fun bit – Siever adds population growth to the mix.  Sievers notes that Alberta’s population grew by 1.6% between 2019 and 2020.  Adding that figure to the baseline – in other words, what would income have been if money had risen in line with population growth? – then he gets to add about another billion dollars to the total, which brings things up to $3.5 billion. 

This is – how can I put it? – utter nonsense.  Not only are there very few government programs which cover both inflation growth and population growth, even in good years, but general population growth is a really bad baseline.  More sensible would be the 18-24 age group, which is still the primary age group served by universities and colleges.  If you used that number for 2019-20 (the same year as Siever uses), you get an increase not of 1.6% but rather about .15%, which would reduce the overall “loss” by about a billion dollars.  If you used the change from 2019 to 2021 – a year in which the 18-24 population fell by over a percentage point, you’d find in fact a reduction in “cuts” and the overall figure would drop from $3.5 billion to something closer to $2 billion.

So, that’s two lessons so far: first, when numbers seem too big to be true, it’s probably because people are finding ways to count things in a non-intuitive way. And second, pay attention to how people are choosing to index numbers.  These can have big effects, particularly if someone is choosing to aggregate figures across years rather than comparing one year to a baseline.

But, there’s a third lesson here, which is: pay attention to how governments use data.  The budget figures Sievers is using do not portray government expenditures at all, but rather aggregate institutional expenditures.  That’s because in Alberta, institutions are under the reporting umbrella of the government, as they are in British Columbia and – IIRC – Newfoundland, but nowhere else in the country.  These numbers therefore cover money projected to be received not just from government but also from student fees, ancillary operations, tri-councils, etc.  And you can bet that those numbers are huge undercounts as far as future revenues go because all institutions are working hard to offset recent cuts with new revenue streams.  In fact, most of the expenditure drop in 2020-21 that Siever assumes is due to government cuts is actually a drop in revenue from sales, rentals and services due to COVID.

More accurate for these purposes is the data found in the annual estimates, which only portrays what government transfers to institutions.  That shows a drop of $136 million from 2019-20 to 2020-21, and a further drop of $141 million to 2021-22.  The problem of course is that the estimates don’t allow you to project forward a year or two and so you can’t get those nice, juicy ten-figure numbers.  

One could try to use the Ministry Business Plan for projected future expenditures, but super-annoyingly, this document uses *yet another* method of accounting which does not reconcile the other two figures and does not specifically lay out what the government expects to spend.  However, if you read between the lines, it seems to suggest a cut of a similar magnitude to the previous two years ($140 million) this year and then levelling off thereafter.  Using that estimate you get the following picture.

	Cut	Difference from Baseline	Siever “Cuts”
2019-20	–	–	–
2020-21	$136M	$136M	$136M
2021-22	$141M	$277M	$443M
2022-23	@$140M	@$417M	$857M
2023-24	Unclear	@$417M	$1.268B

This isn’t as exact as the incorrect numbers Siever quotes, because of the uncertainty of the post 21-22 numbers, but they actually get you to more or less the same place, if you’re determined to get to an Alberta-government-doing-ten-figure-bad-thing result.  So maybe there is a fourth lesson: every once in awhile, incorrect numbers lead you to the right spot, a bit like a stopped clock telling the correct time twice a day.

In any case, yes, Alberta is cutting funding to post-secondary, as one would expect given the reduction in the province’s revenue sources.  And it is good to measure them and hold the government to account.  But please, not this way.