Conservatism on Canadian campuses

In the US, liberal bias in academia has long been a subject of concern, especially to those on the right of the political spectrum. Now here in Canada, pundits and politicians are increasingly bothered by a perceived lack of openness to conservative views on campus (see, for example, here, here or here). Yet Canada is not the US. Is there good reason to believe that pro-liberal or anti-conservative bias on campus is a serious issue here? 

One way of answering this question is to consider the political leanings of Canadian university professors. In Canada people do not affiliate with a political party as part of their voter registration, so the only data available on professorial political affiliation comes from surveys. The only published survey of professorial attitudes that I am aware of was carried out almost twenty years ago, in 2000. Nakhaie and Adam sent out a mail-in questionnaire to 10,000 Canadian faculty members. Thirty-four percent of those surveyed responded. The political affiliations of the respondents are shown below the fold.

Professors were less likely to belong to one of the  parties on the right than the typical Canadian - see the "professor to population ratios" in the last rows in the table below. Professors were also more likely than a typical Canadian to support the NDP. However, a plurality of Canadian university professors responding to Nakhaie and Adam's survey self-identified with the centrist Liberal party - as did a plurality of the Canadian population as a whole, at that point in time. Professors were not much more likely to be Liberals than the typical Canadian.

Source for Table 2: Nakhaie and Adam 

Some might explain the relative lack of Reform or PC supporters in Nakhaie and Adam's data by claiming that there is discrimination against conservatives on campus; that academics with right-leaning views are less likely to be hired. One way of testing this theory is with polling data. Are the views of university professors systematically different from those of other highly educated workers? If so, perhaps the possibility of discrimination needs to be investigated further. However if university professors' political views are similar to those of other highly educated individuals, it is harder to argue that people are selected into the academy on the basis of their ideology.

To compare the views of Nakhaie and Adam's professors with other similarly educated Canadians I used November, 2000, Gallup polling data. This poll asked 585 decided voters their federal voting intentions. Here is the relationship between education and federal voting intentions in the Gallup data (the tall blue columns on the left represent likely Liberal voters):

It is not surprising that Nakhaie and Adam did not find a lot of support for the conservative Reform party among university professors; apart from CEGEP-educated voters, who are largely located in Quebec, university educated Canadians were those least likely to support Reform. The majority of university-educated Canadians - in November, 2000, at least - supported the Liberal party. Hence it is not surprising that a plurality of the university professors Nakhaie and Adam surveyed reported Liberal leanings.

I am reluctant to put too much weight on a single monthly poll by a single polling company. However here are some data that I crunched from a September, 2016 Forum Research Poll of Ontario voters. It finds a similar pattern. Although Patrick Brown's Progressive Conservatives led Kathleen Wynne's Liberals among all educational groups, people who had undertaken post-graduate studies were the most pro-Liberal and least pro-conservative of any demographic group - see below. (Somewhat confusingly, in this graph, the tall blue column on the left of each group represents the Progressive Conservative Party):

 

I also did some tabulations with an Angus Reid poll. This considered 2011 Federal party choices, and again finds that more highly educated Canadians are more likely to lean Liberal and less likely to lean Conservative:

These surveys do not provide conclusive evidence of an education-political affiliation link. They are based on a series of relatively small polls (about 1000 respondents) conducted via telephone. Non-response is a concern, as are confounding factors, such as the interaction between age and education on the one hand (older people tend to have lower levels of education) and age and political affiliation on the other (older people tend to be more conservative). Also none of the figures and tables above do proper tests for statistical significance, as one could do with, say, a multinomial analysis [Bleg: is it worth putting in the effort to do some multinomial analysis? how would one aggregate polls to get a decent-sized data set?].

The evidence presented above is consistent with the idea that part of the reason that conservatives are underrepresented among the professoriate is that conservatives are underrepresented among the group of people who could potentially be university professors: highly educated Canadians. 

Yet this begs the question: why are highly educated Canadians more likely to support the Liberal party and less likely to support parties towards the right of the political spectrum?

In the US, where the positive correlation between education and liberalism is well documented, much ink has been spilled on the subject. There are good reasons not to import this debate into Canada.

First, although Canadians and Americans have similar overall levels of education (see here), Canada is exceptional among OECD countries in having a large "college" - as distinct from "university" - sector (see here). As the Angus Reid and Gallup polls discussed above show, is pretty difficult to find any evidence of pro-Liberal bias among college graduates. If college instructors have left-leaning orientations similar to those of university professors, the relative conservatism of college graduates would be evidence against a straightforward  "liberal teachers pass liberal values to their students" hypothesis. It would, however, be consistent with the idea that the content of the university curriculum inclines people to become more liberal.

Yet U of T is not Harvard; UBC is not Oxford or Cambridge. Some evidence suggests that the effect of university education on political attitudes stems from students' interactions with other students, rather than what they are taught by professors. For example, Kyle Dodson's research summarized here and partially ungated here suggests that it's engagement in student activities, rather than interactions with faculty members, which creates campus radicals. Faculty interaction has, if anything, a moderating effect on students' views. Dodson's study suggests that any systemic differences in the on-campus experiences of Canadian and American students might be expected to affect the values they absorb while at university. 

There are also differences between the demographics of university-educated Canadians and Americans. Canadian university graduates are disproportionately likely to be immigrants - a reflection of Canada's immigration system which gives points for education level. Visible minority Canadians are much more likely than other Canadians to have a university degree - this is particularly true of visible minority men. These demographics provide a straightforward explanation of why many people with university degrees would be reluctant to embrace any party perceived to be anti-immigration or nativist - and despite Conservative repudiations of racism e.g. here, it is not obvious that the party has succeeded in creating an image of itself as being pro-immigration.

Indeed, this raises a more general point: it is foolhardy to try to explain the education/political affiliation linkages without looking at the substantive content of political parties' platforms. For example, parties that promise austerity and reduced government spending are unlikely to appeal to doctors and nurses, teachers and social workers, and all of the other highly educated public sector workers. It is entirely possible that the relationships between education and political affiliation in Canada are driven, in whole or in part, by self-interest.

Self-interest, in the largely unionized Canadian university context, might explain why the typical university professor is more likely to support the NDP than the typical highly educated Canadian, given the NDP's historically pro-union stance.

Another possibility often raised in the US debate is that smart conservatives seek out better paid jobs in the private sector. It is not obvious that this explanation would apply to Canada, where professorial pay is relatively generous - though, at the same time, it is also worth pointing out that many undergraduates are taught by contract instructors earning a fraction of what their tenure- or tenure-track colleagues make. It is not obvious why someone would want to work as a contract instructor, rather than an accountant, unless they valued the non-monetary aspects of the job.

The non-monetary aspects of professorial life - working all morning in one's pyjamas, working all night marking papers - are not valued equally by everyone. Hence personality characteristics - what one enjoys and values in life and in work - are another potential driver of the correlation between being left-leaning and being a university professor. Below is a table from Carney et al that provides a sense of the literature on the personality/politics correlation. The notation "Slovenly, ambiguous, indifferent (C-)" in the Liberal/Left-wing column means that people who are low (-)  in conscientiousness (C) are more likely to be Liberal. The "Reliable, trustworthy, faithful, loyal (C+, A+)" in the conservative/right-wing column means that people who are high (+) in the selected conscientiousness (C) and agreeableness (A) traits are more likely to be conservative. 

If creativity and openness are essential prerequisites to having a successful and happy career in academia, then any overrepresentation of liberalism in academia could simply be a function of the type of personality needed to be successful on the job.

This has been a long post. The underlying motivation for it is to better understand the political context under which Canadian universities are operating. My gut feeling is that over the next five to ten years universities will find themselves under increasing pressure to show that they are delivering value to Canadians. I'm trying to figure out where that pressure will come from, and what form it will take. Suggestions/thoughts/ideas for regressions to run would be welcome.

Note: I have not been able to find any information on the attitudes of Canadian academics other than Nakhaie and Adam's work. Mohsen Javdani has carried out a survey of Canadian academic economists' views, but the results are not yet available. 

Comments

[Nick Rowe]

Good to see this data. And a fairish interpretation of it.

My own view is it's a mixture of preferences and discrimination (I am very uncertain about what proportion of each).

I think the simplest and most plausible preference explanation is this: lefties are generally against "profits", and so are more likely to want to work in (what they see as) the "non-profit" sector. Which includes universities. Plus the whole idea of being a "Progressive" includes the idea that you proselytise the progressive cause, rather than leading a quiet life doing what you are responsible for doing and trying not to do too much harm.

But especially since the Damore firing, and the left's reaction to that firing, I'm not going to let anyone on the left get away with making any preference explanation for why the distribution of occupations is not totally uncorrelated with everything. They are going to eat gander sauce if they try. We can use that argument; they can't.

On discrimination: you have to be careful if arguing that educated people tend to be less conservative, so it's evidence against discrimination if profs equally tend to be less conservative. If the education system discriminates against conservatives/women, it is unsurprising that educated people would tend to be less conservative/women.

As usual, I can't remember the source of that recent US study which found some lefty profs quite openly saying they would discriminate against hiring those with conservative views. And having been a conservative at university for the last 45 years, I totally get why they would do that, and not even see they are doing it, or see it as a problem. "Someone with those views just doesn't feel right for the sort of person we imagined we would hire for this position, and we know those views are wrong anyway, and someone who holds them can't be very smart or very well-informed, and we only want to hire the best".

Remember dear old Marvin, from Philosophy, who was a very open member of the Communist Party? I'm trying to imagine a world where Marvin had instead been a member of an equally extreme right wing party, that had an equally bad track-record. It would be a very different world. There is a massive double-standard when it comes to the teaching of Marx and advocacy of Marxian and Neo-Marxian views, which is very common, almost orthodox in some departments. Similarly, it's hard to imagine an Institute of Gender Studies built on an explicitly masculinist perspective, or even allowing such views to be taught (except to condemn them).

There seemed to be a brief moment of self-doubt and reflection on the left in the early 1990's, when communism collapsed. Maybe History was not on their side after all. It didn't last. But maybe it's beginning again, with recent events, and what we are observing in universities is a confused reaction to cracks in the Progressive edifice?

I self-censor on campus. And I have learned to develop a persona that lets me get away with a little more than others, so they laugh rather than hate. There's a word for it.

I used to worry about level playing fields more. But the left seem to reject the concept. It's all about power. OK, but two can play that game. Fortunately universities' power seems to be waning, and it's harder for them to have a monopoly on the wider conversation. But it's too early to say for sure.

Anyway, enough reflections from me (though you did tweet that you wrote it to trigger me!). Worthwhile post.

[Frances Woolley]

Nick, on much of this we're just going to have to agree to disagree. The point that you make that interests me is this one:

"universities' power seems to be waning"

What evidence do you have of this? And why might this be the case?

This particular blog post actually originated as a long digression from another post I was doing, where I was trying to learn about Canadians' attitudes towards universities. There is almost no data on the subject - I've ended up relying on the Forum poll that I quote in this post. I think, like you, that there's something in the air - but what it is, where it's coming from, and whether or not it matters - this I don't know.

[Nick Rowe]

Frances: yep!

I don't have any hard data. The web provides much lower barriers to entry for those expressing views that are not tolerated or barely tolerated at universities, and some of them do seem to be getting very large audiences (including young people who are likely to be students or recently students). And those views seem, after a lag, to be impossible for the "MSM" to ignore. And even negative publicity creates a Streisand effect, if anyone can search the web to find the original source. Jordan Peterson is one example. I rarely watch videos (except how to fix cars and Russian trucks in Siberia and follies on the Nurburgring etc.) but videos seem to get very large audiences of (presumably) younger people and having a bigger impact than the blogosphere once did. Then there's the rise of "populism" in politics, which seems to mean "views that are unpopular at universities". I can't help but say "or unpopular in The Cathedral (of which profs are the priests)". Memes don't just come from universities anymore.

A couple of days ago I saw a tweet about a survey of US university presidents' worries about the public perception of universities (will conservative donors stop supporting us?). There probably is no Canadian equivalent.

But, as usual, I am useless at finding data, or remembering stuff, and can only report my impressions. And there's always the danger of generalising from n=1=me. The financial crisis/recession must have done a lot of damage to faith in the educated elite like me, but I would be surprised if that's the whole story.

Yes, there's something in the air.

[Frances Woolley]

Nick, thanks so much for that suggestion, the study you're thinking of, and a summary of it, can be downloaded here: https://www.insidehighered.com/news/survey/survey-college-presidents-finds-worry-about-public-attitudes-confidence-finances. It's actually worth downloading the full report as some of the reporting of it doesn't quite convey the complete picture.

My own view - because I'm a die-hard materialist - is that ... nah, I'll save that for the next blog post...

[Bob Smith]

"universities' power seems to be waning"

"What evidence do you have of this? And why might this be the case?"

The one obvious example, in both the US and Canada, is the rise of policy work being done by non-university think tanks. It's more pronounced in the US, where think tanks are both more numerous and powerful - Obamacare has its roots in a policy proposal put out by the Heritage Foundation in the late 1980s, but you see it in Canada. And while there are think tanks that span the political spectrum, anecdotally in Canada, the ideological distribution of think tanks is probably a mirror image of that of Canadian academia - for every Canadian Center for Policy Alternatives, there's two or three Fraser Institutes, Atlantic Institute for Market Studies, or Montreal Economic Institute who are (generally) on the libertarian/conservative side of the political spectrum. I probably have a bit of a selection bias, since I tend to spent more time on the right of the spectrum, but I don't think I'm far off.

I do second Nick's comment about Marxists on campus. I certainly recall roaming the halls of the sociology department at Queens and professors and grad students unashamedly having hammer and sickles on their walls or doors. The irony the people who speak of "safe spaces" proudly brandishing the symbols of the 20th century most murderous ideology. It doesn't take much imagination to imagine the likely reaction if a grad student decided he wanted to put a swastika on his door. No wonder those department are increasingly ideologically homogeneous - how many conservative students would be remotely interested in studying with such people?

[Frances Woolley]

Bob: "the rise of policy work being done by non-university think tanks"

Interesting. I don't know if there has actually been much of an increase in think tank output in Canada - CD Howe, Conference Board, Canadian Tax Foundation have been putting out policy work for decades. Also think tank output often draws quite heavily from academic sources. I suspect one big difference between Canada and the US, especially when it comes to econ policy, is that there is far more academic work done on US than Cdn econ policy issues.

One thing that I've heard a couple of times is that policy makers don't have time to read through academic journal type articles, or the means to access them. Something that they can read in five minutes while they're waiting for a meeting to start, and access on their phone, is far more likely to be read and have influence. Which would contribute to declining university influence.

The economist (or is it the Marxist?) in me can't help observing that certain interests benefit materially from a waning of universities' influence.. Some of the research carried out in universities can be highly inconvenient, e.g. research on climate change, or the public health consequences of air pollution, or the extent of income inequality and income mobility.

[Bob Smith]

Sort of curious about this conclusion:

It is not surprising that Nakhaie and Adam did not find a lot of support for the conservative Reform party among university professors; apart from CEGEP-educated voters, who are largely located in Quebec, university educated Canadians were those least likely to support Reform. The majority of university-educated Canadians - in November, 2000, at least - supported the Liberal party. Hence it is not surprising that a plurality of the university professors Nakhaie and Adam surveyed reported Liberal leaning

In November 2000, ~22% of university educated Canadians voted for the Reform Party. Reform Party supporters made up 6% of the Professoriate. It may not be surprising that a plurality of university professors supported the Liberals, isn't it somewhat surprising that that Reform supporters were vastly underpresented in the professoriate relative to the population of university educated people? Now, 2000's a bit of an awkward datapoint, since the state of the right was sort of in flux. But clearly the ideological leanings of the highly educated population doesn't do much to explain the under-representation of conservative viewpoints in academia.

Also, I think you need to approach the "attitudinal" explanation for conservative under-representation with a grain of salt, because it's an answer that likely masks implicit biases. Are "creativity and openness" helpful attitudes for a career in academia. I suppose so, but I would have also thought that being "self-controlled", "conscientious", "orderly, organized", "reliable", tenacious" - apparently conservative traits - would also be useful (wouldn't those be useful traits for someone trying to tease data from complex datasets). And I would have thought that other, supposedly liberal traits - "slovenly", "indifferent", "impulsive", "unpredictable" - would be detrimental to such a career.

Imagine that explanation is another context. If a lawyer believes that you need to be aggressive and combative to succeed in law, he might hire lawyers, like him, who exhibit that traits. And if those traits are more common among men then woman, such that he mostly hires men, well, so be it. But is it true that one needs to be aggressive and combative to succeed in law? Are there other pathways to success that play on other strengths. Might a lawyer who exhibits collaborative traits - perhaps traits more common among woman, then men - also find a way to be successful? Because, if so, the starting assumption that you need to exhibit certain traits to be successful is a form of hidden bias and discrimination. To the extent that attitude is an explanation for underpresentation of conservative voices on campus, you need to ask the question as to whether those attitudes are really necessary for a career in academia, or whether those attitudes simply happen to be those shared by the people on the hiring committee.

[Bob Smith]

"Interesting. I don't know if there has actually been much of an increase in think tank output in Canada - CD Howe, Conference Board, Canadian Tax Foundation have been putting out policy work for decades."

Yeah, I suppose we need to define what we have in mind when we're thinking of "think tanks". I think the CTF is an example of one type - it supports and sponsors great work. But it doesn't really employ anyone itself (and those it does are mostly administrative people) - it relies on others (like me, or like tax policy academics) to produce its policy work. It's a think tank that complements universities, since without a steady pool of skilled people employed elsewhere it wouldn't be able to produce any work.

Compare and contrast that with, say, AIMS or the Fraser institute - which employ people full time to do research (as well as contracting with people do work for it). I think the latter type of think tank is more of a substitute for a university, rather than a compliment, since you can have a career working at it. And we've seen a growth of these sort over the past few decades (the political slant of Canadian academia is hardly a new development). Now, I don't want to overstate the distinction, the Fraser Institute leans on academics as well, but I do think that is a distinction that's relevant to the point at issue.

Beyond that, I think it is fair to say that the scope of work that Think Tanks look at has expanded enormously. The CD Howe institute is a great example. Yes, it has been putting out policy work for decades, but over it's life, the scope of areas it works on has expanded enormously. When it was first founded, its mandate covered economic relations with the US. By the 80s it had expanded into broader areas, including labour policy. Now, it's expanded into a host of other areas. Similarly, the Fraser Institute has seen it's focus expand into covering a range of policy issues from aboriginal policy, to health care to education. So it's not just that we're seeing more think tanks, but their footprint is expanding.

[Frances Woolley]

Bob: "isn't it somewhat surprising that that Reform supporters were vastly underpresented in the professoriate relative to the population of university educated people?"

Not considering the Reform's party's base. Older and rural Canadians are much less likely than others to have a university degree, and that was Reform's core base. Also Reform was very much anti the Laurentian consensus. I remember Preston Manning speaking at Carleton University. He said something like "try standing up in the middle of the Saskatchewan prairie and say 'this is meeting of two founding nations'". The line fell kind of flat.

Interestingly, the CD Howe business model is changing right now. They're moving away from the long, well-researched commentaries that made them famous, and more towards providing business clients with timely information, networking opportunities, and research that meshes with their interests. It must be lucrative - you can look up Bill Robson's salary on the CRA charity listings, and he's not doing too badly. Not sure if this change is particular to CD Howe or if something is happening in the think tank sector more generally.

[Tel]

"universities' power seems to be waning"

"What evidence do you have of this? And why might this be the case?"

This guy Bryan Caplan has been pushing his book "The Case Against Education" and he was claiming that close to 100% of the benefit to graduates of a university education is signalling.

I'm not in a position to prove him right or wrong, I think I learned something at university but not sure if I would have learned more by just studying books in the library on my own.

Let's presume that Caplan is correct (and he has collected some evidence, so he might be right) then the value of the universities hangs on the positive feedback effect, where each previous generation of signalers proudly hires other signalling graduates in order to reinforce their own belief that they did the right thing. If the general impression of university behaviour around the business community was to degrade, thanks to a cohort of exceptionally self-entitled and whinging graduates who are disinclined towards real work but very enthusiastic about social justice, this feedback would be broken very rapidly.

If it became easy to teach yourself whatever skills you like, the signalling would be weakened dramatically as high quality candidates just skip the cost and quickly and efficiently bring themselves up to speed.

Further, if it was fast and easy for business to accurately screen potential employees using automated means, this would be more useful than the indirect metric that signalling provides.

It's tempting to think that all of the above, and probably more are working against the universities right now. There will be some adaptation of course, but a lot of them appear to be adapting in the wrong direction as the extremists drive out the moderates and start to hunker down.

[Frances Woolley]

Tel, "If it became easy to teach yourself whatever skills you like, the signalling would be weakened dramatically" " if it was fast and easy for business to accurately screen potential employees using automated means, this would be more useful than the indirect metric that signalling provides"

These are excellent points. Related to your second point: software developers can participate in Github http://git-awards.com/about which provides rankings for coders. Someone with few education credentials and a high Github ranking might be attractive to an employer.

A friend who works in tech, and has been on sides of the interview process, tells me that the big tech firms typically have a three or four step interview stage. There'll be an initial "this person is not a psycho" interview with HR. Then there'll be a technical interview, where the person logs in to a computer remotely and is asked to solve a bunch of coding problems in a short period of time. Then there's something else. Finally, if all of these go well, there's the on-site. Even though it's not an automated screening, it is a way of assessing ability independent of credentials, and that's potentially deadly to universities.

My first ever WCI blog post was on signalling, BTW. There's definitely a signalling element to education, but Caplan overstates the case.

[Derek]

Bryan Caplan puts his highest estimate of signalling at 80% rather than 100%. He also discusses lower estimates where only sheep skin effects are assumed to be due to signalling. A key to his argument is that one characteristic students are signalling is conformity. This rules out many easier and less costly ways for students signal their type. Although there are other ways to signal intelligence and consciousness (other traits he assumes education signals) they tend to display nonconformity. BTW, in reference to the comment about reinforcing existing beliefs, it is not really necessary for employers to know why the educated are more productive, as long as they are more productive.
More related to the original topic, he finds no evidence that being educated by left wing professors makes students more left wing. He even reports evidence suggesting small affects in the opposite direction.
I highly recommend his book. One does not have to agree with all his points to find it intellectually stimulating.

[Jacques René Giguère]

Two things about Caplan.
He writes for the US. Was it Miles Corak (correction much appreciated ) who said that the difference between Canada and the US is that in Canada what matters is what you study and in the US is where you study? The signaling effect of Harvard vs South Mississipi àt Tupelo is much bigger than McGill vs Memorial (maybe because the objective "quality" is more equaly distributed between the Cdn universities).

And what if his thesis was just "Let's justify pulling out one more ladder in a fast ossyfying society in which I am a privileged member?

[Frances Woolley]

Jacques Rene - "in Canada what matters is what you study and in the US is where you study" - I hadn't heard that, but I like that quote very much.

I don't know of any studies that try to compare the degree of signalling in Canada and the US (I haven't looked) but that would be a really good topic for a blog post.

There are lots of reason to think that the objective quality is more equally distributed at Cdn universities because of the way government funding works, and the size of the big universities.

[Frances Woolley]

Derek - "A key to his argument is that one characteristic students are signalling is conformity."

I wonder what he means by conformity - and why conformity would have value in the job market. Yes employers don't want to hire sh*t disturbers, but employers don't want to hire people without any original ideas or thoughts of their own - at least to high skill/intellectual jobs.

[Derek]

Frances - Caplan goes into great detail on what he means by conformity. By giving a short response, I’m going to be guilty of ignoring many of his nuances. However, I particularly liked a quotation he has from Peter Wiles: “What employers need is intelligent conformism or great independence and originality within a narrow range”
In my experience, this is what university administrators want, faculty who can come up with original ideas to be published in good journals but don’t rock the boat internally. This, it agrees with my experiences.
I agree that any argument based so strongly on a single assumption is weaker than if it is not based so strongly on the assumption. However, this assumption agrees with my own confirmation bias.
BTW, I’m also not sure how many of our students get “high skill/intellectual jobs”. I run into many of my former students as taxi drivers, best buy sales people, etc. Some of my best former students manage to get jobs at the university (e.g. in the university finance department, which would pay good CUPE salaries but don’t relate much to what I taught them). Admittedly, I’m teaching at a lower quality institution. Still, even at better institutions, I’m doubtful that a high proportion of students do jobs that really relate to what university students learn.