Friday, December 25, 2015

Meg Urry on Science and Gender

When I joined the ceasefire in the Tim Hunt controversy, I stipulated that I would continue to defend Hunt against fresh instances in which his name is invoked as a symbol of sexism in science. It didn't take long before such an instance came up. Meg Urry recently published a comment at Nature that opens in a way that I hope we'll soon see the last of. "Gender equality in science made headlines repeatedly this year," she begins, and then goes on to mention Shrinivas Kulkarni, Geoff Marcy and, yes, Tim Hunt.

There are at least two problems with this framing.

The most important is that Hunt and Kulkarni clearly don't belong on a list with Geoff Marcy. They were accused of expressing casual sexism while discussing their work. Marcy was accused of sexually harassing his students inside and outside the lab. And while Hunt has now been completely exonerated of actually being a sexist, the case against Marcy seems to be holding up. (This, of course, is because the journalism in the Marcy case was competent, while the journalism in the Hunt case was not.) My hope for 2016 is that this thoughtless use of Tim Hunt's name to stand for "sexism in science" simply stops. Though it is true that he "made headlines", continuing to use him in this way amounts to propagating a falsehood.

The other problem is that it uses cases of supposed sexism in science to frame a discussion of gender inequality in science.** The conflation is these two issues is actually something that Tim Hunt has an opinion about, and he went on the record with it long ago, without causing any particular controversy. In the Lab Times in July of 2014, he said that he did not "think there is any discrimination" in science but had to "admit the inequalities in the outcomes, especially at the higher end, are quite staggering." He also asked "why women being under-represented in senior positions is such a big problem" and, on all these issues, admitted that he didn't know the answers. That opinion, in fact, is really the only corroborating piece of evidence that was ever offered to suggest that Tim Hunt is the sexist he was made out to be after his toast in Seoul. If Urry wanted to use him to frame a discussion of gender inequality, she could have engaged with these much more considered remarks.

Indeed, when Hunt started defending himself against being called a sexist, Deborah Blum accused him of "turning the issue from the main point - the status of women in science - to a focus on sympathy for himself". But here again, if Blum really had wanted to discuss the status of women in science with him, she could have simply engaged with his sense of the situation as stated the year before, rather than falsely accusing him of being a sexist, resulting in his marginalization from the conversation she'd like to have. In any case, in this post I would in fact like to turn attention away from Tim Hunt and onto the issue that Blum and Urry would like to discuss. Let's see how that goes.

While they should not, in my view, be simply conflated, gender inequality and sexism are, of course, related issues. Where there is gender inequality, it is reasonable to hypothesize that sexism is among the causes. But it is also reasonable to suppose that sexism is among the effects of gender inequality in a particular discipline. If, for whatever reason, a field is dominated by a single gender, one will imagine that the sense of humor and even the overall intellectual style in that discipline will be "gendered" accordingly. But this assumes something important, namely, that there are notable differences between the genders.

Urry clearly believes that there are. "Women can be more likely to apply to institutions that describe themselves as 'collegial' and 'student-oriented' than 'top-rated' and 'world-class'," she tells us. "Different ideas lead to scientific advances," she also says, "Sameness leads to stagnation." That is, the presence of women in science is tantamount to the presence of ideas that wouldn't otherwise be considered. Women presumably think differently.

I happen to share Urry's belief that we need women in science because their presence, as a gender, increases the variety of ideas that may be considered and approaches that may be taken. (I'm less comfortable with Emily Grossman's version of this argument, namely, that women should be valued for their emotionality and self-reflection.) Indeed, I suspect that there are some truths about our universe that a woman will discover sooner than a man. Perhaps there are even a few discoveries that are simply beyond the imagination of men to make (though, once made, these men of science would of course be able to understand them and replicate them.)

But it seems to me that Urry fails to consider the consequences of accepting such gender differences for our understanding of what Hunt calls the "staggering" inequality in outcomes. Here I'm of course going to get on the old libertarian (high?) hobby horse of insisting on the difference between equality of outcomes and equality of opportunities. Once we agree that women bring something distinct and different to science, we must consider the possibility that these differences also correspond with other unequally distributed aptitudes and attitudes.*

The to my mind best "best practice" that Urry proposes for fostering true equality is to hire and promote researchers in a rigorously blinded way. She reminds us that the pool of real talent for philharmonic orchestras was greatly increased when the unconscious bias of conductors was checked simply by letting musicians audition behind a curtain. We can see this as a victory for women, who now had the same chance of being hired based on their musical ability as their male competitors. But we can also see this as a victory for the arts, which were now enriched by perhaps a different kind of playing, which could now be noticed as competent rather than merely "feminine".

But if this all makes sense, then why are we so eager to blame sexism for the inequality of outcomes? We can show that girls do as well or better than boys in schools in most subjects. We can show that young women are accepted into university at higher rates than young men. We can show that they are as likely or more likely to graduate than their male counterparts. We can show that they populate the graduate schools in roughly equal numbers. And we can show that women are increasingly populating permanent and high-level academic positions in most disciplines. That "staggering" inequality at the top persists, but given the lack of inequality at the lower levels, it's hard to imagine that sexism is the mechanism that is responsible for it.

This can be true despite the existence of sexism. There may be sexists of the kind that Tim Hunt has wrongly been accused of being, but they just don't seem to be very effective at keeping women away from or out of science.

Sensitivity to gender differences is also what drives Urry's mild "affirmative action" proposals. "Recruiters should note that female applicants, being more selective in their attempts, are likely to be well suited to the position that they have applied for." I'm not sure how to read this, except that somehow the very fact that a woman has applied for a job should be seen as an indication that she's qualified, which seems odd when you've got her CV right there in front of you. It also contradicts, as far as I can tell, the very good idea I just noted, namely, to make recruitment "blind" to gender. And it suggests a reason for inequalities that sexism can't really be blamed for: women simply don't apply for jobs they are qualified for as often as men. Fixing that might equalize things; but it doesn't require us to hunt down sexists and humiliate them on Twitter.

I have to admit I'm a bit incredulous about the anecdote she tells about her "male colleague's" attitudes about affirmative action. (I'd like to get his side of the story, actually, so if he's reading this he's welcome to step forward in the comments.) Here's how she tells the story:

Recently, a colleague worried openly about young men who, in the face of added competition from women, might not land that coveted assistant-professor position. If a woman of equal ability were hired affirmatively in place of a man, he suggested, the unsuccessful male applicant should be compensated with $100,000. My jaw dropped. By that reasoning, shouldn't we compensate the thousands of women or other underrepresented scientists who were preferentially not hired over the past 50 years, despite being as talented as — or substantially more so than — the men who got the jobs?

It hard for me to believe that he made his appeal for compensation in the hypothetical situation where the applicants are equally qualified, though I suppose one could argue that the only fair thing to do in that case would be to flip a coin. The problem, however, is that no hiring decision ever ends up with two applicants whose only distinguishing characteristic is gender, so that if it happened to be two men or two women you'd really be in a coin-toss situation. What her colleague is saying is precisely that a woman who is discriminated against because of her gender, and therefore not hired despite being more talented, does, in fact, deserve compensation, and that a man who is not hired, simply because of his gender, should have similar recourse.

Again, I'd invoke the libertarian argument that in a competitive environment a department or lab does well to hire simply the most qualified applicant and offer a competitive salary to attract and retain that person. The important thing is to hire and promote according to ability, not gender. Given the obvious differences between the genders on so many points (including the specifics of the "female brain", I would add) I don't understand why an inequality of outcomes should today be seen as an indication of the effectiveness of sexism. As I see it, sexism is a failed ideology. We're giving it too much credit here.

Tim Hunt says he doesn't think sexism is what is keeping women out of the upper echelons of the scientific establishment. It may be that the particular combination of ambition and curiosity that is required for one's career to lead to such a post is differently distributed among the sexes. He doesn't actually say this, and I don't know if that's true either. But it's something that's worth talking about. It's not something we should be calling people out on simply for not being sure about. I, for one, really don't see how the idea can be shocking in a world where we have to be sensitive to the fact that women are unlikely to apply to a post they are qualified for simply because the lab is advertised as "world-class" rather than "collegial".

Let me conclude by noting something that Mary Collins pointed out about her husband. Tim Hunt, she says, doesn't himself think he has what it takes (if you will) for a top administrative post in the sciences. Marry Collins, by contrast, clearly does. In this light, why we have to begin conversations about gender inequality in science by remembering Hunt's misreported toast in Seoul is beyond me.

*In order to avoid needless controversy, I've added the words "and attitudes" here. I'm not trying to suggest that "an aptitude for science" is less prevalent among women in some simple sense. As I point out below, it will be some combination of aptitudes and attitudes, as well as curiosity and ambition, that might be distributed differently in the population of scientists by gender and therefore explain differences in outcomes as something other than an effect of discrimination.

**Update (05/01/16): On Twitter, Sarah Kendrew has emphasized to me that Urry never uses the word "sexism" in her piece. But she does use the words "discrimination" and "bias", and by leading with Hunt and Kulkarni she is, as I see it, framing her argument as one about needing to push back against the sexism that remains in the minds of individuals. It is that framing that has occasioned my criticism.

Like Hunt, I don't think subjective bias notably determines outcomes at the top of the scientific establishment. Like Hunt, I do think that the differences between the sexes cause "trouble" for men and women in the workplace, as they do in all workplaces. We've gotten past the stage where we need to call men out for thinking of women as, in a variety of ways, different from men. Correcting these residual subjective biases (in a word, sexism) by calling out individuals when they occasionally express them does more harm than good. In any case, when we talk about inequalities of outcomes we should certainly take the many possible confounders that stem from demonstrable differences—some of which are invoked by Urry—between the sexes into account and weighing them before we conclude that the inequalities are a significant effect of sexism.


Johan Fynbo said...

Thanks for this clear analysis and thanks for following this so thoroughly. Here is another recent example of a "misuse" of the Hunt case:

Johan Fynbo said...

It is interesting that the reference to the Impact of "Blind" Auditions on Female Musicians actually is not quite precise. In fact that study reached a more complex result (can be found here: They actually write : "The relative success of female candidates appears worse for blind than for non-blind auditions and this finding also hold for each round of the audition process." They interpret this as the adoption of screens makes more (less qualified) women apply. Hence, the outcome of that experiment is not so straight forward an argument for a bias against the selection of women in hiring situations.

Thomas said...

I've only looked at it quickly, but they do seem quite categorical that the screens thwarted a bias against hiring women. There's an interesting little detail though. On page 6, they quote Zubin Mehta saying, "I just don't think women should be in an orchestra." In the context of Goldin and Rouse's paper, it looks like he believes women aren't competent enough musicians. But it turns out that he's been sort of "Hunted", if you will, here.

In a 2010 paper, Amy Louise Phelps provides the context (a NYTimes interview from 1970): 'While he was conductor of the New York Philharmonic, Zubin Mehta was quoted as saying that, "I just don’t think women should be in an orchestra. They become men. Men treat them as equals; they even change their pants in front of them. I just think it’s terrible!" Nevertheless, Mehta prefaced this statement by qualifying that "They [women] are very good—they got their jobs in competition with men." To Mehta, if women were given equal treatment they would somehow lose their feminine qualities and that was a capital sin, even if they were just as qualified to work as their male counterparts.' (33)

This is still sexism, of course. But, interestingly, as Goldin and Rouse note in a footnote (p. 6, fn 12), it seems to have had little influence on the opportunities for women under his tenure. 45% of his new hires were women. The NYTimes interview is subtitled "Baton in Hand, Foot in Mouth", suggesting that they were spinning this a bit like the Tim Hunt toast. Now, many of us don't believe Hunt is a sexist, but in his case, too, it can in any case be demonstrated that his presence did not reduce opportunities for women.

While it's not definitive proof, this does suggest to me that structural innovations like auditioning behind a screen, blinded to gender, is a good thing. (As far as I can tell, Mehta's hiring would have been done using the blinded approach.) It also suggests that once such "screens" are instituted, the actual "sexism" (the views held about women) is rendered essentially irrelevant. Again, this suggests to me that going after micro-aggressive casual sexism may be pointless (accomplishing nothing) and, perhaps, harmful (resulting in false positives and the loss of harmlessly sexist but musically brilliant conductors, for example). This makes it a bit like PSA screening for prostate cancer.