Dear my colleague, let’s think carefully about this latest bit of Horowitziana.

1. Anecdote is not the singular of data.

Observe: I say, old chap, our colleges are run by Republicans. How do I know this? Our dean, who has been dean for an unbelievable ten years — nobody does that! — is a former Reagan Treasury official and a George W. Bush-supporting Republican.

Now, you know this is true — our dean is a Republican. You also know it doesn’t prove anything, except maybe that it is certainly not true that Republicans are stigmatized always and everywhere. Which is to say that at least this story, unlike the ones in the column about not getting hired or getting shunned because you’re a Republican, meets the standard of falsifiability.

I could use this decanal decade as an opportunity to complain about the sinister GOP domination of the college of Letters and Science, and I could insinuate all kinds of nasty stuff — I could say, oh, “I doubt this dean has set out to purge the college of Democratic dissenters,” or “I believe that for the most part the biases liberal academics face from our dean are subtle, even unintentional.”

But I wouldn’t do that because I’m not a hack who’s careless of the reputations of others. Also because I know that under our dean, irrespective of his politics, our college has flourished.

2. Party Affiliation/voting ≠ ideology

Maranto cites three positions “practically verboten in much of academia”:

(i) welfare reform helped the poor
(ii) the United States was right to fight and win the Cold War
(iii) environmental regulations should be balanced against property rights

Couple things, here. (1) He has no proof of this. (2) Even if every academic voted Democrat, his claim probably still wouldn’t be true, as these are mainstream Democratic positions.

For myself, I’d say on (i) I’m agnostic; (ii) is incontestably true; (iii) is so innocuously phrased with its demand that these goods be “balanced” that only a flat-out anarchist could oppose it.

In fact, (ii) is arrant red-baiting. Nobody with any degree of seriousness opposes (ii). We had wild-eyed Nation lefty Eric Alterman here some weeks ago and when he talked the Cold War to graduate students, this was his opening and closing line. (Now, if Maranto had said, “right to fight and win in Vietnam,” it would be different. And not only because of the dubiety of the “and win” claim.)

Anyway, despite being close to Maranto on these issues, I’ve voted a lot less for Republicans in recent years than I used to. Why?

Well, partly because these are not the issues on which I vote. Nor are they the issues on which the electorate votes. What’s important to the electorate? In order: (1) War in Iraq, (2) Economy/Jobs, (3) Health care, (4) Immigration, (5) Environment. And I’m going to take a wild flying guess that “Environment” doesn’t mean an overwhelming concern that property rights are not being adequately balanced with environmental regulations, but much more a general fretting that Manhattan is going to be knee-deep in the Atlantic fisheries before the century is out unless we do something, y’know?

What’s Maranto doing here? He’s saying that because academics vote Democratic more than they used to, they must be wild-eyed anarcho-commie lefties who hate “balance” and want to cede Missouri to Stalin.

Suppose he’d listed among his defining positions, say, “global warming is a hoax,” or “evolution is a hoax.” Then he’d have something — those are positions that academics are incredibly unlikely to hold, but Republicans are much more likely to. Somehow, he doesn’t think to draw the line there, perhaps because that would make Republicans sound like an anti-intellectual fringe group.

Stipulate that for reasons like these, academics vote more Democratic than they used to. Why might that be?

3. Modern Republicans don’t like academics.

If academics have fled the Republican Party, it’s because the GOP changed, not academics. This is the anti-expert administration, the censor-scientists administration. Would you maybe have a little trouble supporting it if your job required you to be a dispassionate pursuer of truth? This administration drove away even its strongest academic supporters because its leaders refuse to listen to sober analysis.

Yes, this has long been true to some degree — if you speak in complete sentences, you must be a commie, if not in fact gay. That was, after all, the rap on Adlai Stevenson, who was about as conservative as a Democrat could be, yet somehow Nixon and McCarthy made him out a flaming lefty because he was articulate.

But it’s even truer today, when many Republicans are “straight-talking men of faith,” with utter contempt for reason. And yes, it’s a little worrying if you happen to think that expertise trumps gut in making high-level decisions.

4. These studies deserve strict scrutiny and the presumption that they are always-already bogus.

Maranto’s editorial sounds awfully like one that David Brooks wrote four years ago and which I helped knock down, four years ago (you can find this oldie pasted below the fold). The study on which Brooks based his argument was no good. The ones Maranto’s using are maybe more refined, but probably fishy too. For example, Rothman and Lichter, whom he cites, have in the past done not-quite-right stuff to prove anti-conservative bias. Klein and Stern got a severe critiquing by Zipp and Fenwick, (and also here) who pointed out that Klein and Stern have selection bias and also conflate party preference with ideology.

Zipp and Fenwick found that in fact, academia has been growing a bit less liberal in recent years, owing to an age-cohort shift from baby boomers to younger, more conservative folk.

Not only that, as far as indoctrination goes, Zipp and Fenwick found that conservatives were “more committed than liberals or moderate liberals to the importance of shaping students’ values.”

Still, there are more liberals in the humanities and social sciences than in business or engineering (though again, the social sciences have become less liberal over time). Zipp and Fenwick suggest one big reason why: women tend to be more liberal than men, and the humanities and social sciences have been famously more likely to hire women than the sciences. But there are other reasons, too.

5. What’s Liberal about the Liberal Arts?

Jeez-o-Pete, with all these people alleging discrimination in hiring and those like you, my colleague, who say you don’t know why there might be an imbalance, you’d think nobody had written a book with this title already. There is something intrinsically liberal about the liberal arts. It’s there in the name. It’s procedural liberalism, fostering continued debate, refusal to accept claims of authority, reluctance to reach an unassailable conclusion. These are not the straight-talking men-of-faith values.

Also, you know, if you’re a go-along-to-get-along type, if you don’t really like to argue, if you’d rather be employed than right, business is a better place for you than academia — and it pays better, man. Academia is a high-risk, low-reward career option. You would only do it if you really believed in those procedural liberal values.

And as far as that goes, despite what Maranto says, a think-tank job isn’t some poor second-best for most academics. A think-tank job — of which there are vastly more high-profile, well-paid ones for conservatives than for liberals — means getting paid to read and write, usually to live in a major metropolis, and not to teach. Now me, I’m crazy, I like teaching: but many folks don’t, and would leap at the opportunity to trade their university job in a small town for, say, The Manhattan Institute in, well, Manhattan.

Oh, man. I’ve been doing this kind of thing, as noted above, on and off, for four years. It’s getting tiresome. But we’ll have to keep doing it because, as Maranto indicates, there’s a lot of money behind these (to cite one study of this subject) “seriously misleading” claims.

Is there some short-hand term of art for what happens when rich foundations keep funding studies that use selection bias and play shell games with terms of analysis to make seriously misleading and publicly pernicious claims?


MSNBC.com’s “Altercation” for 9/16/2003:

David Brooks is all wet in the Atlantic.
Brooks discovered recently that academics tilt Democratic. No, beyond that: He discovered that academics “are drawn from a rather narrow segment of the population,” by which — he goes on to say — he means that thin swath of the populace who are not only not registered Republicans, but people who are not “pro-life, a member of the National Rifle Association, or an evangelical Christian.” And why this peculiar demographic tilt, you might ask? Thus Brooks: “It’s likely that hiring committees would subtly — even unconsciously — screen out any such people they encountered.”
Now, I know Brooks is the Republican everyone’s supposed to love because he’s urbane and witty and on PBS and NPR and in the NYT. But this is fishy, even pernicious stuff for a couple reasons worth going into in some depth.
First of all, academics surely do tilt Democratic. But, according to political strategists for both parties, so does the entire pool of people from which academics are, of professional necessity, chosen — the pool of people holding advanced degrees. As Karl Rove says, “As people do better, they start voting like Republicans — unless they have too much education and vote Democratic, which proves there can be too much of a good thing.” (Nicholas Lemann, “Bush’s Trillions,” The New Yorker, 2/19/2001; sadly TNY doesn’t keep this online.)
Second of all, Brooks is not really talking about academics per se, he’s talking about academics at “elite universities,” in which category he includes “Harvard, Stanford, Yale, Princeton, Caltech, MIT, Duke, Dartmouth, Cornell, Columbia, Chicago, or Brown” and also, later in the piece, “Penn State, Maryland, and the University of California at Santa Barbara.” Quarrel with the listing if you like; anyway each except Duke and Dartmouth is in a state that went for Gore in 2000. Are the elite universities cause or effect? Well, you may claim if you like that Princeton makes New Jersey vote Democratic, but party politics going back before Boss Jim Smith might have something to do with it too.
Third of all, to demonstrate the bias Brooks relies on a study published a year ago in The American Enterprise (I had to rely on another polite Republican, Eli Lehrer, who kindly told me its whereabouts; Brooks doesn’t say. If you’re interested, here’s the PDF version with numbers and graphs.) People better with numbers and fiercer than I have critiqued this study for various reasons. Let me point out two things that would be obvious even to the statistically disinclined.
Brooks says the study looks at “professors in the arts and sciences who had registered with a political party.” He goes on to note — without pausing to think how odd this is — that “fifty-seven professors at Brown were found on the voter-registration rolls.” If you take Brooks at face value, he’s suggesting that 90 percent of Brown’s faculty are not registered to vote. Unlikely, and in fact, not quite what the TAE study says. TAE researchers didn’t look for all registered professors: “we stuck mostly to major, uncontroversial, and socially significant fields of study.” Brooks’s exaggeration is not an inconsequential mistake — everyone makes mistakes — it’s an important misstatement that makes the TAE study sound more comprehensive than it is.
Which means the TAE study is both narrower and more credible than Brooks’s version of it, but it has problems of its own:
(a) the researchers’ choice of fields varies from one university to another, marring the comparability of data (why include Economics at UC Berkeley but not at UCLA?);
(b) the chosen fields don’t represent the faculties. Among the Crimson crowd in the TAE study 21 are listed as being from Political Science. Therefore this department alone accounts for 40 percent of Harvard’s registered professors found by TAE. Surely this doesn’t represent voting habits in the Yard. (There’s a subsidiary curiosity here — this means that 70 percent of the Harvard Government faculty aren’t registered to vote — which does seem most unlikely for a Government department.)
Similarly, 26 percent of the TAE group found for University of Texas at Austin come from Women’s Studies; again, probably not a representative sample of ‘horns profs. In fact, I couldn’t resist looking it up: Women’s Studies at UT Austin counts about 145 faculty affiliates at a university of about 2700 faculty, for about 5 percent of the total. Which now that I think about it is a huge Women’s Studies program — but still not a quarter of the academic population at UT.
OK, so what — if, as I said in the first point, university faculty surely do tilt Democratic, then what’s the big deal? Well, the study, and Brooks’ article, make it sound like the tilt is maybe 80 degrees to the left of upright — four to one, say, against conservatives. Such a severe slant can be taken to suggest foul play — Brooks says it urbanely, but he does say it, just as noted above: “It’s likely that hiring committees would subtly-even unconsciously-screen out any such people they encountered.” TAE is blunter, calling its findings actionable “discrimination by ideology” and worth a lawsuit.
The truth about politics in academia — as indeed about everything to do with academia — is almost certainly much less exciting. According to the Chronicle of Higher Education’s brand-new almanac (online only to paying subscribers), faculty at institutions of higher education report their own politics as follows: 42.3% liberal, 34.3% middle of the road, 17.7% conservative, .3% far right and 5.3% far left. This is 52% reporting as either conservative or middle of the road; that’s almost 10% more for the right and center than for the liberals. This is not a wildly lefty population.
There are problems in academia as in any profession. But looking to hire Democrats isn’t a major one. And ultimately this is where Brooks goes badly wrong. From personal experience I’d argue that academics aren’t looking for like-minded people when they go a-hiring. We’re looking for people we find it interesting to disagree with. That’s what intellection is about — worrying over ideas and what they mean — not about signing up as a member of someone’s team.
Notably, the TAE study doesn’t do anything with registered independents who vote Democrat sometimes and Republican others, depending on the tempora and the mores. I myself am proud to say that at the end of each term I invariably get at least one student evaluation accusing me of being a right-winger and another grousing that I’m a bleeding-heart liberal.
Hey, if I’ve confused them by exercising independent judgment, then good.

Advertisements