Sorry for the late blog appearance: I’ve been bouncing around Alberta and British Columbia for work and play this week (#ICETECA, baby) and it’s tough to write on these terrible little short-haul flights.  Anyways, today I want to talk about a paper which came out a few weeks ago called Overqualification among 2012 and 2013 bachelor’s graduates, by Statistics Canada’s Diane Galarneau. 

“Overqualification” is a fraught topic to define and measure.  This study uses a snapshot of bachelor’s graduates a couple of years after graduation, which some people might think a restricted way of talking about “overqualification”.   One snapshot, 24 months after graduation, is not really the same as long-term overqualification.  It is, arguably, a measure of friction in the labour market and how well job-matching is going. 

(This next paragraph is for the methodology nerds and can be safely skipped if you do not count yourself in this blessed tribe). 

The paper has nifty methodology because it shows how new methods of data linkage are freeing StatsCan from the need to use the National Graduates Survey (NGS) which, let’s face it, is looking increasingly dodgy as a survey due to low response rates.  Instead, this paper takes data on graduates from the Post-Secondary Student Information System (PSIS) from the years 2012-2013, and links to data from the 2016 census to check on their occupation in 2015 (i.e. 2-3 years after graduation).  This is excellent because it gives us a sample of about 130,000 bachelor’s graduates, which allows for a reasonably comprehensive look at some minority populations. Also – no new survey required!  The study defines individuals as “overqualified” if the individual has a bachelor’s degree but is employed in an occupation where the occupational skill factor is considered to be “high school” or less.  So, one is considered “overqualified” if one has a degree and is in a low-skill occupation roughly three years later.  Graduates who have returned to school full-time are excluded from the sample.

(Welcome back)

Anyways, here are the paper’s key findings:

• Men with bachelor’s degrees are slightly more likely to be overqualified than women (17.7% vs. 15.2%), adjusted for various other factors including field of study.  This is a big difference from 20 years ago, when women’s rate of overqualification was higher by a roughly similar magnitude.
• On the whole, visible minorities – particularly Black Canadians, were more likely to be overqualified than others, again adjusted for other factors like gender and field of study.  There’s a bit of nuance here because the ethnic categories were divided up (South Asian, Chinese, Black, other VM) and how data was reported by immigration status, and not all sub-groups have the same pattern, but largely this is the case.  The magnitude of the effect seems to be about two percentage-points (i.e., similar to the male-female gender gap) for non-Black visible minorities and about four times that for Black Canadians.
• The biggest gaps are in field of study.  The lowest levels of overqualification are in health and engineering (which is pretty much as you’d expect).  They are a bit higher in math and computer science, and from there on out the numbers start taking big jumps.  business and science students have overqualification rates that are 11-15 percentage points higher than Engineering students (i.e. 2.5-3x more likely to be overqualified), and for the social sciences and humanities the gap with engineering is anywhere from 17-24 percentage points (i.e. 3-4x more likely).
• There also appears to be a significant difference in provincial rates of overqualification, with Ontario having the highest rates (Manitoba close behind), and the then oil-boom provinces of Newfoundland, Saskatchewan and Alberta having the least.
• Most of the “overqualification” effect disappears if students complete a master’s degree.

There’s a second reason why “overqualification” is a fraught topic.  Now that it has been measured, what does it mean?

Just think about your reaction to each of the bullet points above.  When you look at the statistics on race, chances are, you thought to yourself “at least to some degree, this represents structural racism” (based on other research in various fields, I’m pretty sure this is actually case, btw).  But when you looked at the gender on data, did you think to yourself that “this represents structural sexism against men”?  I suppose that’s possible, but probably there’s something else going on there.  Not sure what, exactly, but worth an examination.

What to make of different rates by province?  Clearly, here we are not talking about biases in hiring, but is it so simple as to say some provinces are overproducing university graduates and others not?  Ontario, which has much higher participation rates than other provinces, also has the highest rate of overqualification, so perhaps so.  But then again, Saskatchewan and Manitoba have relatively similar attainment rates and – according to this paper – quite different overqualification rates so perhaps there is more to this story as well.

And what to make of different rates by field of study?  Is it simply that we have too many graduates in sciences, social sciences and humanities?  That is one interpretation.  Another would be that we are not in an “education” economy but rather a “skills” economy, and outside the regulated professions universities are not very good at teaching the skills necessary to be employable.  The regulated industries have a lot of work based in practicums or other forms of work-integrated learning and we simply haven’t worked out how to replicate that in other parts of the academy.

And there is a final issue here: maybe overqualification is a good thing.  Part of the reason we subsidize higher education is to bring down the cost of specialized labour so industries can compete.  So overqualification might in one sense be seen as the system working the way it should.  Also, overqualification might be seen as a form of technological up-skilling.  Over time, many occupations that were seen as semi-skilled became more skill-intensive.  Presumably, to some degree, that is because people with higher levels of skill entered the profession and began using their knowledge to make these positions more useful and productive, to the point where employers began to re-think the actual qualifications for this job.  Under this view, overqualification is a necessary part of the transition to a higher-skilled, more credentialled economy.  I am not sure I would take either of those positions, myself, but they are perspectives which need to be considered.

Anyways: it’s a good study with lots of food for thought: just don’t jump to conclusions about causes and remedies.