Selection Effects, Homogeneity, and Diversity in Academia

February 9, 2011 at 22:07 (Politics, Psychology)

I had been meaning for a while to suss out this issue in depth for a couple of reasons, but what really crystallized everything was this New York Times article covering Jonathan Haidt’s provocative presentation on political homogeneity in academic psychology at a recent conference in San Antonio. The journalist, John Tierney, nicely covered Haidt’s point, though also chose to focus on some interesting stuff, which I’ll get into momentarily. The comments (I only read the top-rated ones) are similarly enlightening and nicely illustrate the range of opinion on the topic.

The upshot of Haidt’s presentation was that an overwhelmingly vast majority of social psychologists (really, psychologists in general) are situated firmly in the left wing of the American / Canadian political spectrum. Judging from the quick straw poll he took, a small handful of conference-attending social psychologists identify as political moderates, and a vanishingly minuscule percentage were willing to out themselves as avowed conservatives. Haidt suggests some explanations for how such a statistically uncanny state of affairs might have emerged, and, quite importantly, suggests that it has and will continue to “hinder [social psychologists’] research and damage their credibility — and blind them to the hostile climate they’ve created for non-liberals.”

Now, before I get into a broader discussion of the significance of Haidt’s position for psychology and beyond, I’ve got to take specific issue with a particular claim in his talk – and besides, it will tie in with the other issues that I mentioned got me thinking about this. The relevant paragraph from Tierney’s piece is as follows:

Similarly, Larry Summers, then president of Harvard, was ostracized in 2005 for wondering publicly whether the preponderance of male professors in some top math and science departments might be due partly to the larger variance in I.Q. scores among men (meaning there are more men at the very high and very low ends). “This was not a permissible hypothesis,” Dr. Haidt said. “It blamed the victims rather than the powerful. The outrage ultimately led to his resignation. We psychologists should have been outraged by the outrage. We should have defended his right to think freely.”

Now, lest I just come off as grinding an axe, let me just say that I love Jonathan Haidt’s work and I would honestly count him amongst my personal heroes (see my Facebook for proof! even if I didn’t put him in the top 5, haha). That’s why it’s such a letdown for me to see him come so close to getting this issue right, only to get it wrong.

What he said that was spot-on: “[the hypothesis] blamed the victims rather than the powerful.”

Where he, alas, jumps the rails and goes careening off into Fail Country: “We psychologists should have been outraged by the outrage. We should have defended his right to think freely.”

Misinterpreting the Larry Summers Debacle, Part Umpteen

Look, I’m not here to deny that some of the public reaction crossed the line into knee-jerky calls for actual censorship or other anti-free-thinking what-have-you. It’s quite understandable to be outraged at attempts to shackle a researcher’s ability to think freely within the context of the academy, and let the truth lead us where it will, even into uncomfortable territory. (Whether outrage, as contrasted with restrained, reasonable discourse, is the most pragmatic response on either side in any case is a whole different discussion entirely, and one that I’m sure I’ll have later.)

But in implying that that’s all the flap was about – that it’s just reducible to a reactionary attack on academia’s cherished and crucial environment of free thought, which is unfortunately just the kind of “it-all-boils-down-to-this” move that Haidt pulled here – is Missing The Point Like Whoa. Not only is it just intellectually sloppy, it inadvertently reinforces a slew of really vile tropes that degrade the quality of discussion on this topic (at a minimum, The Feminists Are Out To Censor Us and The Women Are Overemotional). Several commenters on the NYT thread took Haidt to task for dredging up Summersgate in the way he did:

The outrage towards it was not caused by his free thinking. It was caused by the fact that he put forward a hypothesis justifying the unequal status quo without providing any serious argument for it. … So let me ask: what do you do when somebody puts forward some strong hypothesis (e.g., he was abducted by green people from a UFO), but all he does to justify it is to say “Oh, but I was!” – with no hard evidence, no nothing? What do you do if that person is a Harvard president talking to a group of very smart people, Harvard professors? I see no other legitimate reaction than outrage. He was ostracized not for his “free thinking”. He was ostracized because he did not do his homework.

Larry Summers was vilified because his saying that men were naturally better than women at math completely ignored the large body of social science and humanistic literature explaining why so few women made it to the top of math and science fields. It had nothing to do with liberal or conservative politics. Summers’s comments were indicative of his narrow conception of scholarly research. It’s no accident that he was replaced by an historian capable of thinking far more broadly than he.

I’m particularly inclined to talk about this issue right now for a few reasons (one of which is a troubling rash of “omg criticism is the same thing as censorship” foolishness in another corner of the internet; also, just today, my law professor was telling us about a student who complained to him that “university is about opening doors, but telling me I can’t call women ‘chicks’ is closing doors!” – at which I could only laugh and say, “no, all anyone has done is correctly label that door with a sign that says Assholes Only”), but the one I’ll share here is a recent email I wrote to some friends when the Larry Summers incident came up in the course of an email conversation:

… my view on his “intrinsic differences” debacle has matured over the course of several years’ mental cellaring.

My take on the issue used to begin and end with Pinker’s incisive quote: “Look, the truth cannot be offensive. Perhaps the hypothesis is wrong, but how would we ever find out whether it is wrong if it is ‘offensive’ even to consider it? People who storm out of a meeting at the mention of a hypothesis, or declare it taboo or offensive without providing arguments or evidence, don’t get the concept of a university or free inquiry.” 

Now, I still think that’s true. I think it’s worth knowing whether intrinsic differences contribute to the gender gap in science. The better we understand how it works, the better chance we stand of fixing it, which is almost certainly what Pinker (and Summers!) would love to see happen. But I now think it also sort of misses the point of the criticism.Once you hang around the culture of science for a while, you (at least hopefully) realize that the kinds of hypotheses you form and pursue, and the way they get framed, and the degree of importance assigned to answering the questions to which they pertain – none of these things just fall into your lap via some perfect, unbiased rationality that’s immune to the influence of inherently unjust sociocultural narratives.

Nor does the decision to pursue that kind of inquiry without challenging the culture from which it flows exist in some ramification-free vacuum: it perpetuates the same problematic attitudes that spawned the notion that THIS is the important question that is worth our immediate attention (e.g. “how can we be sure about whether men are/n’t intrinsically better at activity x than women?” rather than “how can we improve the practice and culture of science in such a way that any intrinsic difference become so trivial as to be irrelevant?”) and THIS is how you frame the hypothesis (“how can we isolate the immutable genetic effects that account for the presumed essential difference between the sexes” rather than “how can we tease apart the complex soup of genetic, environmental, and stochastic contributions to scientific aptitude in a way that helps us understand them with minimal interference from the intuitive but unscientific narratives we’re so inclined to read into the whole thing”), and that if you critique those background assumptions you must HATE SCIENCE OMG.

So that, in retrospect, in my opinion, was the problem with that.

Structural Selection Effects, How Do They Work?
Now, at this point, you might be wishing you could ask me: but Roland, didn’t Steven Pinker also make the very reasonable point that allegations about the damaging nature of, e.g., articulating a hypothesis laden with problematic assumptions about sex and gender, need to be backed up with data demonstrating concrete harm?
Well, you’d be right, although maybe just because you cheated and read the whole NYT article, during the course of which you would have come across the following:

Instead, the taboo against discussing sex differences was reinforced, so universities and the National Science Foundation went on spending tens of millions of dollars on research and programs based on the assumption that female scientists faced discrimination and various forms of unconscious bias. But that assumption has been repeatedly contradicted, most recently in a study published Monday in The Proceedings of the National Academy of Sciences by two Cornell psychologists, Stephen J. Ceci and Wendy M. Williams. After reviewing two decades of research, they report that a woman in academic science typically fares as well as, if not better than, a comparable man when it comes to being interviewed, hired, promoted, financed and published.

“Thus,” they conclude, “the ongoing focus on sex discrimination in reviewing, interviewing and hiring represents costly, misplaced effort. Society is engaged in the present in solving problems of the past.” Instead of presuming discrimination in science or expecting the sexes to show equal interest in every discipline, the Cornell researchers say, universities should make it easier for women in any field to combine scholarship with family responsibilities.

I’ll say at the outset that I have not gone and checked out Ceci and Williams’ study. I wish I had the time to carve up its methodology; if I were a betting man, I’d lay odds that it is riddled with issues. I say that not only because I have a prior stake in that turning out to be the case (since a lot of my views are premised on the notion that there is a problem here), but also because when faced with the apparently contradictory data points of “many female scientists encounter sexism in the workplace” and “study finds that female scientists don’t encounter sexism in the workplace,” I’m going to make an inference to the best explanation, i.e. the most likely one, and you shouldn’t need me to tell you that “whoops lol those stories of sexism were made up, silly wimminz” is not gonna fare too well as a candidate, especially compared to “well, the real issues flew under the study’s radar, which is unsurprising given the nature of institutional privilege.”

But in fact, we don’t actually have to play hand-wavy guessing games in the first place, because it’s entirely possible that a profession (particularly an academic one) is not only equitable to all social categories when it comes to interviewing, hiring, promoting, financing, and publishing, but in fact favors members of underrepresented social categories in these matters, and yet still exhibits a ridiculous degree of homogeneity.

Not only is such a thing possible, it exists. Exhibit A: academic philosophy.

Over the past few years, Wesley Buckwalter and Stephen Stich have been carrying out a brilliant program of research investigating the yawning, abysmal gender gap in philosophy departments and programs. (Stich even gave the prestigious Frege Lectures on this topic.) Their basic hypothesis goes as follows: philosophers rely extensively on intuitions as a source of evidence in puzzling out important thought experiments, and these intuitions are generally quite uniform (in at least quite a few important cases), with people who don’t share the consensus intuition being said to “have a tin ear” or to “just not get it.” But maybe it’s the case that at least a few philosophical intuitions vary systematically by gender, in which case, given that philosophy has been always-already male-dominated, a woman entering philosophy is exceedingly likely to encounter frequent exasperated “you just don’t get it”s. If that turns out to be the case, it’s not surprising that there’s a gender gap in philosophy, because we’ve created a selection effect that winnows out women who would otherwise want to do philosophy but are faced with so much discouragement that all but the tough-as-nails get exasperated and go elsewhere. (If your intuitive response here is to say “well they just oughta toughen up then,” I’ll curtly remind you that victim-blaming is not a very nice response to a manifestly unjust structural problem.)

Well, wonder of wonders, as best anyone can tell from Buckwalter and Stich’s research, the hypothesis is vindicated. Big time. Here, then, is a situation in which there’s no amount of slanted hiring / recruiting in the world that will close the gender gap, and even to whatever extent it helps matters, it’s dizzyingly inefficient. This particular example (and the Summers flap that prompted me to supply it) are about gender, but we can apply this thinking equally well to race, sexuality, able-bodiedness, religion, politics, what have you.

Post-Partisan Psychology: How to Get There from Here?

Wait … politics? A-ha. There’s the rub – now we come full circle. It seems pretty likely to me that social psychology, as a discipline, isn’t actively discriminating against conservatives any more than philosophy is actively discriminating against women. Instead, what we’re likely seeing in these cases – and maybe in pretty much every case of homogeneity in academia – is a combination of selection effects and some difficult-to-determine measure of “certain kinds of people are just not as interested in x or y field.” It’s particularly difficult to determine how much the latter factor contributes since there is no simple or ideologically neutral way to separate those two things out.

The important implication of what I’m suggesting is that, if it’s true, adding conservatives to the list of underrepresented groups eligible for “special funds to pay for travel to the annual meeting” at which Haidt delivered this talk will not work, for just the same reason that philosophy departments’ attempts to level the playing field for women haven’t worked. So it’s a tricky problem: neither completely created nor completely soluble by the action of consciously maintained biases. (Pretty familiar state of affairs, in that sense.) I have to say, in all honesty, I don’t know of a good way to tackle this problem. I hope someone figures it out.

You might be inclined to ask why this all matters; who cares if science is super-liberal? That’s not something I’m going to go into in this post, except to say that you can’t then turn around and be surprised when conservatives brush off scientific results as the fluff-laden discharge of so many pointy-headed left-wingers. Well, you can be surprised if you want, but I will roll my eyes and sigh in exasperation.

You also might be inclined to run the argument that the political homogeneity of psychology and science in general shouldn’t surprise us, because there’s something about what it is to be a conservative that’s fundamentally at odds with what it is to be a scientist. I guess that, to some extent, I have just postulated something like this myself: part of what it is to be a conservative is to be grated upon by, for instance, your psychology professor’s jokes about what a big dummy GW Bush is. (Can you imagine the furor that would issue forth if a professor made some other social category the butt of a joke? And then defended it, as so many of us are inclined to defend unnecessary jabs at Dubya’s intellect, with “but it’s TRUE!” Yeah. Not pretty.) But that’s still kind of different; I’m willing to believe that this tension can be fixed.

You might think it cannot be fixed, though. Certainly, many of the NYT commenters did. In its basest form (which I hope we can all agree to reject), the suggestion goes that “but what it is to be a scientist is to be intelligent, and what it is to be conservative is to be a moron!” Even if one doesn’t resort to outright insulting the other side of the aisle, there are still plenty of clever ways to substitute in your favorite intellectual virtue (one common one is “committed to knowing the truth rather than passing off what’s expedient as true”). It is my opinion that people advancing these claims have failed to stop and really consider them from a place of intellectual honesty. That, or, as some more astute NYT commenters noted, they have conveniently forgotten about William F Buckley Jr, Friedrich von Hayek, Thomas Sowell, even W.V.O. Quine (today I learned that he was a Reagan supporter!).

There’s a further irony in explaining away this issue by appeal to some intrinsic harmony between science and liberalism: if one takes seriously Jonathan Haidt’s own research on the moral psychology of political affiliation, then the argument that conservatism is essentially (rather than just incidentally) anti-science is dead in the water from the get-go. I won’t go into a ton of detail here, but the basic idea is that a key determinant of political affiliation is a person’s set of moral intuitions. Haidt identifies five basic universal categories of moral concern – harm/care, justice/reciprocity, loyalty/ingroup, authority, and purity – and has collected much data indicating that being a liberal is largely a matter of emphasizing the first two items in that list while minimizing the other three, whereas conservatives largely tend to value all five with about equal weight. (Whether we should conclude from this that conservatives are hallucinating moral concerns that cannot possibly be of actual moral relevance, or that liberals suffer from moral myopia, is a topic for another time.) If being conservative is a matter of having a particular set of moral intuitions, well, how is that supposed to make one a bad candidate for a career in science? Or rather, let’s put it this way: how would holding up loyalty, authority, and purity as realms of special moral importance make one any worse of a scientist than one who holds up harm and justice as the exclusive domains of moral concern? It seems that a) neither of those are intrinsically threatening to good scientific epistemology, even if they can incidentally degrade it in certain cases, and b) to the extent that either one is unfriendly to good scientific epistemology, I cannot see how they are not equally so.

I suppose this post is getting lengthy, and that’s as good a place to stop as any. Having covered so many issues, I’m sure I have opened up endless opportunities for objections, and we’ll see which claims I end up wanting to stand by.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: