Usher: Viewpoint diversity [at universities]

Sound critique of the methodology used and resulting conclusions:

Last week, the MacDonald-Laurier Institute released a truly bad paper on “viewpoint diversity” at Canadian Universities.  How bad was it, you ask?  Really bad.  Icelandic rotting shark bad.  Crystal Pepsi bad.  Final Season of Game of Thrones bad. 

The basic thrust of the paper, co-written by Christopher Dummitt and Zachary Patterson, is that

• The Canadian professoriate is well to the left of the Canadian public
• Within the academy, those who describe themselves as being on the right are much more likely to say they “self-censor” or find work a “hostile environment”
• This is an attack on academic freedom
• There should therefore, in the name of academic freedom, be a significant government bureaucracy devoted to ensuring that right-wingers are hired more often and feel more at home within the academy.

Dummitt and Peterson are not, of course, the first to note that the academy is somewhat left-leaning.  Back in 2008, MR Nakhaie and Barry D. Adam, both then at the University of Windsor, published a study in the Canadian Journal of Sociology showing that university professors were about three times as likely to have voted NDP in the 2000 general election than the general population (the NDP got about 8.5% of the vote in that election), about as likely to have voted Liberal, and less likely to have voted Bloc, Conservative, or Reform.   Being at a more prestigious institution made a professor less likely to support the NDP, as did being a professor in business or in the natural sciences. 

(This effect of discipline on faculty political beliefs is not a Canadian phenomenon but a global one.  Here is a summary of US research on the issue, and an old but still interesting article from Australia which touches on some of the same issues). 

Anyways, this new study starts out with a survey of professors.  The sample they ended up with was ludicrously biased: 30% from the humanities, 47% from the Social Sciences and 23% from what they call “STEM” (where are health professions?  I am going to assume they are in STEM).  In fact, humanities professors are 13% of the overall faculty, social science profs 23%, and the rest of the professoriate 64%.  Despite having read the Nakhaie/Adam paper, which explains exactly how to get the data that would allow a re-weighting of the data (you can buy it from Stastcan, or you can look up table 3.15 in the CAUT Almanac, which is a couple of years out of date but hardly incorrect) , the authors claim that “relatively little information was available for the population of professors in Canada so no weights were developed”.  In other words, either through incompetence or deliberate feigning of ignorance, the authors created a sample which overrepresented the known-to-be most leftist bits of the academy by 2 times and underrepresented the known-to-be less leftists bits of the academy by a similar factor, and just blithely carried on as if nothing were amiss.

Then – this is the good one if you are familiar with conventions of Canadian political science – they divided respondents into “left-wing” and “right-wing” partially by asking them to self-locate on a four-point likert scale which left no space to self-identify as a centrist and partly by asking them about their views on various issues or how they self-described on a simple left-right scale.  If they voted Green, Bloc, NDP or Liberal they were “left-wing” and if they voted Conservative or People’s party they were “right-wing”.  Both methods came up with a similar division between “left” and “right” among professors (roughly 88-12, though again that’s a completely unadjusted figure).  Now, generally speaking, no one in Canadian political science forces such a left-right choice, because the Liberals really aren’t particularly left wing.  That’s why there is nearly always room for a “centre” option.  Certainly, Nakhaie and Adam did so – why didn’t Dummitt and Peterson? 

Anyways, having vastly exaggerated the degree of polarization and the pinkness of professorial views, and on this basis declared a “political monoculture”, the authors then go on to note that the embittered right-wing professors appear to have different feelings about the workplace than do the rest of their colleagues.  They are three times more likely, for example, to say their departmental climate is “hostile”, for instance, or to say that they “fear negative repercussions” if their political views – specifically, on social justice, gender, and Equity Diversity and Inclusion – were to become known. They are twice as likely to say they have “refrained from airing views or avoided pursuing or publishing research” (which is a hell of a conflation of things if what you’re interested in examining is academic freedom).  On the basis of this, plus a couple of other questions that conflate things like job loss with “missed professional opportunities” or that pose ludicrous hypothetical questions about prioritizing social justice versus academic freedom, they declare a “serious crisis” which has “disturbing implications for the ability of universities to continue to act as bastions of open inquiry and rational thought in modern Canada” which requires things like legislation on academic freedom, and a bunch of things which would effectively ban universities from anti-racism initiatives.

Look, this is a bad study, full stop.  The methodology and question design are so obviously terrible that it seems hard to avoid the conclusion that its main purpose was to confirm the authors’ biases, and clearly whatever editorial/peer review process the Macdonald Laurier Institute uses to oversee these publications needs major work.  But if a result is significant enough, even a bad methodology can find it: might this be such a case?

Maybe.  Part of the problem is that this paper spends a lot of effort conflating “viewpoint diversity” with “party identification diversity”, which is absurd.  I mean, there are countries which allocate academic places based on party identity, but I doubt that those are places where many Canadian academics would want to teach.  Further, on the specific issues where people apparently feel they have a need to “not share their opinions” on issues concerning race and gender, there are in addition to a censorious left a lot of bad faith right-wing concern trolls too, which kind of tempers my ability to share the authors concern that this is a necessarily “bad thing”.  And finally, this idea that the notion of being an academic means you should be able to say whatever you want without possibility of facing criticism or social ostracism – which I think is implicitly what the authors are suggesting – is a rather significant widening of the concept of academic freedom that wouldn’t find universal acceptance.

I think the most you can say about these issues really is first that viewpoint diversity should be a concern of every department, but that to reduce it to “party identification” diversification or some notion of both-sidesism (anti-vaxxers in virology departments, anyone?) should be seen for the grotesquerie that it is.  Second, yes, society (not just universities) is more polarized around issues like gender and race and finding acceptable and constructive common language in which to talk about these concepts is difficult, but, my dudes, banging on about why someone who happens to have a teaching position is absolved from the hard work of finding that language because of some abstract notion of academic freedom is not helpful. 

And in any event, you could make such points without the necessity of publishing a methodologically omnishambles of a report like this one.  Just sayin.’

Source: Viewpoint diversity