You all know I complain a lot about data in Canada.  So today, I thought I’d assemble a wish list: a set of priorities for developing a better system of higher education data, along with some thoughts about how these measures could be implemented as part of a larger, overall accountability agenda

Now, I am going to focus on the need for new data but there is a lot that could be done to make better use of existing data.  Just last week, someone asked me about data on completion rates and times-to-completion data for graduate students, and I had to answer that although such data was theoretically available through the Post-Secondary Information System (PSIS), no one at Statscan or ISED or ESDC – or anyone else for that matter (the Canadian Association of Graduate Studies, maybe?) has used this data to look at this issue.   And I’m not going to talk about the timeliness of data publication, either: everyone knows that Statscan trails most of the world’s developing countries (let alone developed ones) in terms of publishing enrolment numbers.  I might do something on all that next year.  I’m just going to leave all that kind of stuff aside and focus on the real holes in the system.

Nationally, the single most important thing we can do is to resurrect the Youth in Transition Survey (YITS).  This great survey followed two cohorts of young people (one starting at age 15 and the other at 18-20) every two years from 2000 to 2012.  The survey’s strength was that it followed both people who went to post-secondary and those who didn’t.  One could learn a lot about pathways into and through post-secondary education (including credit transfer between institutions), and get more pinpoint data on access rates by ethnicity and income as well as barrier to access and success. The fact that our best data on the high-school-to-PSE transfer is now close to twenty years old is a complete disgrace.  Statscan and ESDC should bring this survey back. 

Another piece of work, which Statscan should have done years ago, concerns community college tuition fees.  We have had data on university tuition for about 50 years now.  But we have literally no idea about tuition fees in our community college sector.  Why?  Basically, because Statscan simply doesn’t prioritize community colleges. I mean, to be fair, coming up with useful program averages is tougher in the community college sector than universities because they are on average so much smaller, but I’m sure there are some reasonable ways to construct something parallel.

I don’t know if it should be done by Statscan, or by institutions themselves acting collectively, but we need better data on students, their socio-economic origins, how they earn and spend money, etc.  For that we need a good comprehensive survey of post-secondary students.  This is different from YITS because it looks at the whole student body rather than just one or two cohorts passing through.  This could be achieved through something like what used to be called the Post-Secondary Education Participation Survey (which Statscan ran a couple of times in the oughts and then abandoned), or through something like the Canadian Undergraduate Survey Consortium (if they still did surveys of the entire student body rather than just first year/middle years/senior year) but with broader participation, plus an equivalent coalition on the college side. 

At present, and for most of the past twenty years, the national database of student aid disbursements consists of us filing Access to Information requests to individual provincial student aid programs every spring.  This is patently ridiculous.  I’m happy to do it, but one day I’m not going to be around and y’all are going to start asking yourselves…say, I wonder how much student aid is given out in Canada?  And no one will know.  Because provincial governments, with notable exceptions in Alberta, Saskatchewan and Quebec, instinctively is to hide loan data wherever possible because, who knows, some of it might make them look bad.  When the feds get around to making the “temporary” COVID-era doubling of student grants permanent in the next budget, they should require provinces to publish their student aid outlays as a condition, and we’d get this one out of the way.

Something similar could be done with data on student debt.  Weirdly, for a country that loves talking about the evils of student debt, nearly all the “data” we have is self-reported.  Why?  Because the only national student aid database is the one run by the Canada Student Financial Aid Program, and it contains no data on provincial programs.  Each province has a complete set of data on provincial and federal student aid, but since the cardinal rule of running a provincial student aid program is “never give anyone data because it might make us look bad.” Aid data is never published, never exploited by linking to other data sources (like for instance Statscan’s Post-Secondary Student Information System or the T1FF tax database), which would allow us to look at things who takes on student debt, how debt does (or doesn’t) affect borrowers’ subsequent life course via such things as family formation.  Again – all the feds must do is tie the distribution of new dollars to provincial co-operation in data sharing. 

Then there is the stuff where institutions – mostly universities – need improvement.  The first big area is publication of staff numbers.  Institutions do not publish data on the number of non-permanent instructors they hire.  Provincial governments should make them do so.  Ditto publication of the number of full-time and part-time staff hired to cover support and administrative positions.  Collectively, the country spends something like 1% of GDP on people in these two broad areas, and we know next to nothing about them. 

And the last area concerns the outcomes of graduate studies.  Everyone knows that the number of students being brought into doctoral studies in the expectation that they have a realistic shot at an academic job is ridiculously high.  Departments should be required to post i) 6-year completion rates and ii) the percentage of PhD doctorates from the last five years who are in full-time, permanent academic work.  Indeed, it would be a trivial thing for the tri-councils to make their grants to researchers conditional on institutional compliance in publishing these outcomes.  Forget to publish last year?  Oops, sorry, your faculty members aren’t getting those research grants they though they would get.

None of this will happen, because this is Canada and the no-data-for-you Soup Nazi approach to educational statistics is pretty well ingrained across the system.  But it’s what should happen.