April 06, 2005

Politics in the Universities

There has been a lot of play in the media recently about the so-called liberal tilt of university faculty. Let's see what the actual numbers are. As far as I can tell, the most comprehensive and authoritative data comes from HERI (Higher Education Research Institute) based in the UCLA Graduate School of Education and Information Studies, which has been studying trends in higher education for a long time.

HERI's 2001-2002 report on national norms for college teachers, finds that "34 percent of college and university faculty identify as "middle-of-the road" politically (down from 40 percent in 1989). Although the percentage of faculty identifying as "conservative" or "far right" (18 percent) has changed very little, the percentage identifying as either "liberal" or "far left" has grown from 42 percent to 48 percent", compared to a previous survey in 1989.

It turns out that women faculty are more liberal than men. The report finds that "54 percent of women, compared to only 44 percent of men, identify as politically "liberal" or "far left." In 2001, 21 percent of male professors and 14 percent of female professors defined their political views as either "conservative" or "far right.""

The report continues:

The latest survey involved 55,521 faculty and administrators at 416 colleges and universities nationwide. Of those, questionnaires from 32,840 full-time undergraduate teaching faculty at 358 institutions were used to compute the national norms. The numbers were adjusted statistically to represent the nation's total population of approximately 442,000 college and university faculty.

So those are the numbers. What are we to make of them? Is this imbalance in political leanings a sign of blatant political discrimination in the hiring of university faculty?

(At this point I have to reiterate my own belief that the terms 'liberal', 'conservative', 'Republican', 'Democrat' have ceased to have much meaning in terms of defining coherent political philosophies, but since this discussion and the data are framed in those terms, I have little choice but to use them for this post.)

That conclusion of hiring discrimination does not follow automatically. For one thing, the word 'liberal' in university circles does not have the same meaning it has outside. A 'liberal education' is what universities strive to provide for their students. It is used in contrast to 'vocational education'. To call someone a 'liberally educated person' is not to describe his or her political beliefs but to describe a person with breadth of knowledge and depth of understanding, as opposed to someone who has acquired a fairly specific set of knowledge and skills in order to perform a trade or profession. So the word 'liberal' has a fairly well-defined and valued meaning in universities, and one would expect people to want to identify with it.

Another point is that while it is true that universities have intense political struggles, they are based on parochial academic politics, and those divisions do not parallel national political splits. In academic departments the biggest battles over a new hire are likely to be based on field of study (in physics, it might be whether the department wants to grow the condensed matter field or the astrophysics field, or whether it should be a theoretician or an experimentalist) or rank (whether they want to hire a promising newcomer or an established star), and so forth. Similar battles occur in other departments.

These battles can be quite hard-fought, but leave little room for other considerations based on party affiliation and the like. Those are not considered important. The prestige of a physics department depends on the physics knowledge it produces, not on the ideological spectrum its faculty encompasses. No department is likely to hire an incompetent researcher to a rare and potentially lifetime appointment just on the basis of that person's party political affiliation.

But if national political considerations are not the cause of this difference in political leanings in universities, what could be the cause? I am not aware of any studies that have looked carefully at this causal question. But people have been willing to speculate.

Jennifer Lindholm, associate director of the Higher Education Research Institute's Cooperative Institutional Research Program and lead author of the faculty survey said: "The disproportionately greater shift we see toward liberal political views among women faculty may be attributable to their dissatisfaction with the Republican Party's current position on issues that often impact women's lives more directly such as abortion, welfare and equal rights."

Writing in the New York Times on April 5, columnist and Princeton economist Paul Krugman points out that registered Republicans are almost as rare in the hard sciences and in engineering (where clues as to ones political affiliation are hard to discern) as in the social sciences, suggesting that the reasons lie with more subtle causes..

Krugman postulates that "One answer is self-selection - the same sort of self-selection that leads Republicans to outnumber Democrats four to one in the military. The sort of person who prefers an academic career to the private sector is likely to be somewhat more liberal than average, even in engineering."

But the more serious charge that he levels is that the Republican party (and by association the conservative movement) are making themselves unappealing to academics by taking stands on issues that ignore evidence and that are anti-research. He pointed to a recent April Fools' Day issue spoof editorial by Scientific American entitled O.K., We Give Up in which the magazine "apologized for endorsing the theory of evolution just because it's "the unifying concept for all of biology and one of the greatest scientific ideas of all time," saying that "as editors, we had no business being persuaded by mountains of evidence." And it conceded that it had succumbed "to the easy mistake of thinking that scientists understand their fields better than, say, U.S. senators or best-selling novelists do.""

Krugman continues:

Scientific American may think that evolution is supported by mountains of evidence, but President Bush declares that "the jury is still out." Senator James Inhofe dismisses the vast body of research supporting the scientific consensus on climate change as a "gigantic hoax." And conservative pundits like George Will write approvingly about Michael Crichton's anti-environmentalist fantasies.
Think of the message this sends: today's Republican Party - increasingly dominated by people who believe truth should be determined by revelation, not research - doesn't respect science, or scholarship in general. It shouldn't be surprising that scholars have returned the favor by losing respect for the Republican Party.

Krugman argues that such an anti-research message is unappealing to any academic (whatever their political stripe), and so it should be no surprise that academics are distancing themselves from it. When Dennis Baxley, a state legislator from Florida who has introduced in that state a bill similar to Ohio's Senate Bill 24, cites professors who teach that evolution is a fact as a prime example of "academic totalitarianism", he should not be surprised that serious academics start giving him a wide berth.

As I said in an earlier post, universities are ultimately reality-based communities, which depend on evidence as an essential part of their knowledge structure. Academics in any field respect that scholars in other fields also use evidence in reaching their conclusions. They may not know that field in any detail but they tend to respect the way scholars go about reaching their conclusions and know that they can back it up with evidence if called upon to do so. The fact that their conclusions are evidence-based does not make them infallible, of course, just that they are grounded in reality.

Academics also suspect that the people who are upset about biology professors teaching that evolution is a fact are closely aligned with those who think that the Earth is only 6,000 years old and that Adam and Eve are historical figures. They suspect that the current attack on biology teaching is just the precursor to similar attacks on geology, physics, anthropology, archeology, and everything else that challenges a particular religious revelatory interpretation of the world.

Krugman argues that it should not be surprising that overtly linking such a world-view to a political movement should result in that movement losing ground in universities, even though it might be politically advantageous.

As I said, I don't know of any studies that have examined the causal reasons for this seeming ideological imbalance, but Krugman makes a point that is worth considering seriously.


Trackback URL for this entry is: Should college presidents take a stand on evolution?
Excerpt: In response to a previous post, Becky posted an interesting comment that I responded to briefly but which requires a...
Weblog: Mano Singham's Web Journal
Tracked: April 15, 2005 08:02 AM


I enjoyed your take on this, and I agree with the Krugman point that the rise of anti-intellectualism on the right is a cause of the liberalism in academia. (Perhaps it's a feedback loop?) Even serious fiscal conservatives are turning away from the Republican party nowadays, and I must admit I would have a problem with an academic who still voted Republican and was not actively fighting the anti-intellectualism in his/her own party.

Have you seen the SciAm perspectives write-up of a roundtable with university presidents? It turns out that no matter how liberal the faculty are, the administration is still conservative on many issues.

Posted by Becky on April 6, 2005 11:13 AM

Interesting comment, and I enjoyed the visit to your blog, which is full of interesting stuff. I particularly appreciated the link you gave to the Meg Urry article.

I do not quite agree with asking college Presidents to take a stand on the evolution question. I do not think it is their role to take such stands on this kind of specific issue.

Where they should take stands is in support of the basic mission of the universities which is to provide a place for scholars and students to seek, create and disseminate knowledge, free from coercion or political pressure.

So I am not sure that I would assign cowardice as the reason. This may have motivated many of the preesidents to duck the issue, but I think there is also a principled reason that can be given for their stand.

I may write in more detail about this later, because it is complex issue.

Thanks for the thought-provoking comment.


Posted by Mano Singham on April 6, 2005 11:57 AM

I think that survey, by relying on self-identification, may be misleading.

A person's self-conception as "liberal", "conservative" or "middle of the road" is affected by the people he sees around himself. If the hypothesis that colleges are significantly lefter-wing than the rest of the country is true, then one might be in the middle of one's peer group, but well left of national center.

For this reason, I think that party registration information is the better data set to use. Numbers I've seen based on this show college professors to be significantly further out of the mainstream than the self-identification numbers might lead one to believe.

Posted by Jason Colby on April 7, 2005 05:56 PM

Yes, all such self-identifications are suspect and you are right that the norms could be different.

I picked that HERI survey because of its size and comprehensiveness, and the fact that they have a lot of experience doing such things. I also have no reason to suspect that they have any ax to grind.

Party affiliations are more concrete but they have different problems, especially in America where the party structure is so loose as to be unrecognizable to anyone from a country where party identification means a lot more. What numbers did you get from the source you refer to? And what kind of sample did they use?

How do we go from party affiliation to political philosophy?
Do we make the identification Democrat=liberal, Republican=conservative, Independent= middle of the road? Is that a valid assumption? As I said many times, my feeling is that these labels are not very meaningful in terms of defining a coherent political philosophy.

Posted by Mano Singham on April 7, 2005 06:50 PM

Sure "Republican" and "Democrat" have broad meanings, and it's possible for a particular Democrat to be to the right of a particular Republican. But that doesn't make the labels useless. If we could plot histograms of the two groups on an ideological axis, while the edges of the curves would run into each other, the means would still be quite distinct.

A survey on actual concrete issues - "When should abortion be illegal?" "What should the top tax rate be?" "What is the proper attitude of the US towards the UN?" etc. - sounds like it would be more the kind of thing you would prefer, but I don't know that such data is available.

The latest party affiliation survey I've seen is from Dartmouth:

Yes, this study was done by a publication with an admitted agenda. It seems, though, that they merely accessed public information for registration data, and, unless you think they're flat-out lying, the numbers should be trustworthy.

They found: "of the 341 professors registered to vote in Hanover, NH, Lebanon, NH, and Norwich, VT, 225 (66 percent) are Democrats and eighteen (5 percent) are Republicans. Ninety-eight (29 percent) did not register a party. Put another way, there are 12.5 registered Democrats for every registered Republican."

The article also refers to a "Center for Responsive Politics" survey that found: "72 percent of the faculty at American universities and colleges describe themselves as liberal, while only 15 percent are conservative."

Those are two quite striking results. I find this kind of information strongly supportive of the thesis that college professors, as a group, are well to the left of the rest of the American populace.

PS - I'm a Case astro grad from, wow, 8 years ago now. I used to grade physics papers for Dr. Brown and once took a course that you helped teach. Hi!

Posted by Jason Colby on April 9, 2005 06:59 PM

Hi, Jason.

Your name sounded familiar but I just could not place you and I am glad that you identified yourself!

I think the HERI study already established that the number of people who self-identify as liberal is larger among college faculty than the general population. What we are talking about now is the extent of the difference.

None of these studies address the interesting questions of the cause of this difference and what, if any, policy implications this has.

The party affiliation study that you cite has a very small sample size and is from a very narrow region of the country, so that allows for two kinds of biases. (I am using "bias" in a statistical sense and not in the sense that the people who did it are not trustworthy!)

The reason that I like the HERI study is that it has the kind of sampling methods that can reduce this kind of statistical biases.

I do not know what kinds of sampling the CRP study involves. I did a quick check of their website and could not find any information on their survey results. I did not scour the website, though, and you may have better luck.

Posted by Mano Singham on April 11, 2005 10:42 AM