“Some of these stories are closer to my own life than others are, but not one of them is as close as people seem to think.” Alice Murno, from the intro to Moons of Jupiter

"Talent hits a target no one else can hit; genius hits a target no one else can see." Arthur Schopenhauer

“Why does everything you know, and everything you’ve learned, confirm you in what you believed before? Whereas in my case, what I grew up with, and what I thought I believed, is chipped away a little and a little, a fragment then a piece and then a piece more. With every month that passes, the corners are knocked off the certainties of this world: and the next world too. Show me where it says, in the Bible, ‘Purgatory.’ Show me where it says ‘relics, monks, nuns.’ Show me where it says ‘Pope.’” –Thomas Cromwell imagines asking Thomas More—Wolf Hall by Hilary Mantel

My favorite posts to get started: The Self-Righteousness Instinct, Sabbath Says, Encounters, Inc., and What Makes "Wolf Hall" so Great?.

Tuesday, October 2, 2012

The STEM Fempire Strikes Back: Feminists’ Desperate Search for Smoking Guns

Sean Carroll, Your Disdain for Your
Own Readers is Showing
            Bad news: lots of research points to the inescapable conclusion that you, Dear Reader, whether you’re a man or a woman, are a sexist. You may be inclined to reject this label. You may even try to insist that you don’t in fact believe that one sex is inferior to the other. But it doesn’t matter, because the research suggests that what you claim to believe about the relative statuses of the genders doesn’t align with how quickly you attach positive or negative labels to pictures of women and men in a task called the Implicit Association Test. Your sexism is “subtle,” “implicit,” “unconscious.” If this charge irks you or if you feel it’s completely unfair, that probably means you’re even more of a sexist than we might have originally assumed. You can try to find fault with the research that demonstrates you’re a sexist, or offer alternative interpretations of the findings, but why would you do that unless you’re a sexist and trying to cover it up—unless, that is, you’re secretly harboring and seeking to rationalize hostile feelings toward women? Sexism is like original sin. It’s in us whether we like it or not, so we must work hard to avoid succumbing to it. We must abase ourselves before the altar of gender equality.
At least, this is what the feminists involved in the controversy over women’s underrepresentation in STEM fields—the STEM fems—would have us believe. Responding to the initial fifty-eight comments to his blog post “Scientists, Your Gender Bias is Showing,” in which he discusses a new study that found significant bias in ratings of competence and hireability depending on the sex of unseen applicants to a lab manager’s position, physicist Sean Carroll ungraciously—you might even say unbecomingly—writes, “At least the trolls have moved on from ‘there is no discrimination’ to ‘discrimination is rationally justified.’ Progress!”
By Carroll’s accounting, I am a troll (by mine, he’s a toady) because I happen to believe gender-based discrimination accounts for a very modest portion of career segregation and pay differentials in industrialized societies—and it may not account for any. And, this latest study notwithstanding, nearly all the available evidence suggests the underrepresentation of women in STEM fields is based on the fact that men and women, on average, prefer to pursue different types of careers. Indeed, the study Carroll so self-righteously trumpets, which didn’t track any actual hirings but only asked participants to treat application materials hypothetically, may have produced the findings it did because its one hundred and twenty-seven participants were well aware of these different preferences.
Dennis Junk
The underrepresentation of women in science, technology, engineering, and mathematics fields is taken by feminists as self-evident proof of discrimination. Since most people who work or teach in these areas understand that sexism is wrong—or at least recognize that it’s thought to be wrong by an influential if possibly misguided majority—not many of them openly admit to deliberately discriminating against women. Yet the underrepresentation continues, ergo the discrimination still exists. That’s why in the past decade there’s been so much discussion of unacknowledged or unconscious bias. Anyone who points out that there is another possible explanation—women and men are essentially (in the statistical sense) different—is accused of being a biological determinist, being a misogynist, having a reactionary political agenda, or some combination of the three.
Now, “essentially different” isn’t all that far from “naturally different,” which is of course part of the formula for sexism, since the belief that one sex is inferior assumes they are somehow inherently different. (I’m excluding genders besides male and female not as a statement but for simplicity’s sake.) But the idea that the sexes tend to be different need not imply either is inferior. Historically, women were considered less intelligent by most men (fewer records exist of what women thought of men), but most educated people today realize this isn’t the case. The important differences are in what men and women tend to find interesting and in what types of careers they tend to prefer (note the word “tend”).
So we have two rival theories. The STEM fems explain career segregation and pay gaps with the theory of latent sexism and rampant discrimination. My fellow trolls and I explain them with the theory that women disproportionately prefer careers focusing on people as opposed to objects and abstractions, while also prioritizing time with family over the achievement of higher rank and higher pay. The fems believe that gender roles, including those associated with career trajectories, are a bad thing, that they limit freedom, and that they are imposed on people, sometimes violently, by a patriarchal society. We preference theory folk, on the other hand, believe that gender begins with individuals, that it is expressed and enacted freely, and that the structure of advanced civilizations, including career segregation and a somewhat regular division of labor with regard to family roles, emerges from the choices and preferences of these individuals.
The best case that can be made for the feminist theory is historical. In the past, women were forbidden to work in certain careers. They were kept out of higher education. They were tethered with iron bonds to their children and their husbands’ homes. Men, meanwhile, had to live with the same type of rigid gender definitions, but at least they had some freedom to choose their careers, could count on their wives tending to the children, and enjoyed the highest position of authority in their families. So we can assume, the reasoning goes, that when we look at society today and find income inequality and segregation what we’re seeing is a holdover from this patriarchal system of the past. From this perspective, the idea that the different outcomes for each gender could possibly emerge from choices freely made is anathema because it seems similar to the rationalizations for the rigid roles of yore. Women naturally want to be mothers and homemakers? Anyone who dares make such a claim belongs in the 1950s, right?  
Though this take on history is a bit of a caricature (class differences were much more significant than gender ones), it has been easy, until recently, to take as self-evident the notion that gender roles erode in lockstep with the advance of civilization toward ever greater individual freedom for ever greater numbers.  Still, tying modern preference theory to policies of the past is nothing but evil rhetoric (almost as evil as accusations of unconscious thought crimes). No one wants to bring back educational and professional barriers to women. The question is whether in the absence of those barriers career segregation and differences in income between the genders will disappear altogether or if women will continue to disproportionally occupy certain professions and continue to spend more time on average with their families than men.
Catherine Hakim
Catherine Hakim, a former Senior Research Fellow at the London School of Economics, and the mind behind preference theory, posits that economic sex differences emerge from what she calls work-life preferences. She has devised three categories that can be used to describe individuals: work-centered people prioritize their careers, adaptive people try to strike some kind of balance between employment and family work, and home- or family-centered people prefer to give priority to private or family life after they get married. In all the western democracies that have been surveyed, most men but only a small minority of women fit into the work-centered category, while the number of women who are home-centered drastically exceeds the number of men. This same pattern emerges in the US even in samples restricted to elite math and science students. In 2001, David Lubinsky and his colleagues reported that in their surveys of high-achieving students 31% of females said that working part-time for some limited period in their careers was either important or extremely important, compared to only 9% of males. Nineteen percent of the females said the same about a permanent part-time career, compared to 9% for males.
Careers in science and math are notoriously demanding. You have to be a high achiever and a fierce competitor to even be considered for a position, so the fact that men disproportionately demonstrate work-centered priorities goes some way toward explaining the underrepresentation of women. Another major factor that researchers have identified is that women and men tend to be interested in different types of careers, with women preferring jobs that focus on people and men preferring those that focus on things. A 2009 meta-analysis carried out by Rong Su, James Rounds, and Patrick Ian Armstrong compiled data from over 500,000 surveys of vocational interests and found that gender differences on the Things-People dimension produce an effect size that is probably larger than any other in research on gender and personality. Once differences in work-life preferences and vocational interests are taken into consideration, there is probably very little left to explain.
Corine Moss-Racusin
Feminism is a social movement that has many admirable goals, most of which I share. But it is also an ideology that has a fitful relationship with science. Unfortunately, the growing body of evidence that gender segregation and pay gaps emerge from choices freely made by individuals based on preferences that fit reliable patterns in societies all over the world hasn’t done much to end the furor over discrimination. The study on that front that Sean Carroll insists is so damning, “Science Faculty’s Subtle Gender Biases Favor Male Students,” by Corinne A. Moss-Racusin, John F. Dovidio, Victoria L. Brescoll, Mark J. Graham, and Jo Handelsman, is the most substantial bit of actual evidence the STEM fems have been able to marshal in support of their cause in some time. Covering the study in her Scientific American blog, Ilana Yurkiewicz writes,
Whenever the subject of women in science comes up, there are people fiercely committed to the idea that sexism does not exist. They will point to everything and anything else to explain differences while becoming angry and condescending if you even suggest that discrimination could be a factor. But these people are wrong. This data shows they are wrong. And if you encounter them, you can now use this study to inform them they’re wrong. You can say that a study found that absolutely all other factors held equal, females are discriminated against in science. Sexism exists. It’s real. Certainly, you cannot and should not argue it’s everything. But no longer can you argue it’s nothing.
What this rigorous endorsement reveals is that prior to Moss-Racusin et al.’s study there was only weak evidence backing up the STEM fems conviction that sexism was rampant in science departments all over the country and the world. You can also see that Yurkiewicz takes this debate very personally. It’s really important to her that women who complain about discrimination be vindicated. I suppose that makes sense, but I wonder if she realizes that the point she’s so desperately trying to prove is intrinsically insulting to her male colleagues—to all male scientists. I also wonder if in any other scientific debate she would be so quick to declare the matter settled based on a single study that only sampled 127 individuals. 
The preference theorists have some really good reasons to be skeptical of the far-reaching implications many are claiming for the study. Most importantly, the authors’ conclusions contradict the findings of a much larger study that measured the key variables more directly. In 2010, the National Academies Press published the findings of a task force that was
asked by Congress to ‘conduct a study to assess gender differences in the careers of science, engineering, and mathematics (SEM) faculty, focusing on four-year institutions of higher education that award bachelor’s and graduate degrees. The study will build on the National Academies’ previous work and examine issues such as faculty hiring, promotion, tenure, and allocation of institutional resources including (but not limited to) laboratory space. (VII)
The report, Gender Differences at Critical Transitions in the Careers of Science, Engineering, and Mathematics Faculty, surprised nearly everyone because it revealed no evidence of gender-based discrimination. After reviewing records for 500 academic departments and conducting surveys with 1,800 faculty members (a larger sample than Moss-Racusin et al.’s study by more than an order of magnitude), the National Academies committee concluded,
For the most part, male and female faculty in science, engineering, and mathematics have enjoyed comparable opportunities within the university, and gender does not appear to have been a factor in a number of important career transitions and outcomes. (Bolded in original, 153)
But the two studies were by no means identical, so it’s important to compare the specific findings of one to the other.
            Moss-Racusin and her colleagues sent application materials to experienced members of science faculties at research-intensive institutions. Sixty-three of the packets showed the name John and listed the sex as male; 64 had the name Jennifer and sex female. The study authors gave the participants the cover story that they were going to use their answers to several items on a questionnaire about their responses to the applications in the development of a mentoring program to help undergraduate science students. The questions focused on the applicant’s competence, hireability, likeability, how likely the rater would be to mentor the applicant, and how much the rater would offer to pay the applicant. The participants rating applications from females tended to give them better scores for likeability, but lower ones for competence and hireability. The participants, whether male or female themselves, also showed less willingness to mentor females, and indicated they would offer females lower salaries. So there you have it: the participants didn’t dislike the female applicants—they weren’t hostile or “old-fashioned” sexists. But you can see how women forced to deal with this type of bias might be discouraged. To me, the lower salary offers are the most striking. But a difference in medians between $30,200 and $26,500 doesn't seem that big when you consider the overall spread was between $45,000 and $15,000, there was no attempt to control for differences in average salary between universities, and the sample size is really small.

            Moss-Racusin et al. also had the participants complete the Modern Sexism Scale, which was designed as an indirect measure of gender attitudes. On the supporting information page for the study, the authors describe the scale,
Items included: On average, people in our society treat husbands and wives equally; Discrimination against women is no longer a problem in the United States; and Over the past few years, the government and new media have been showing more concern about the treatment of women than is warranted by women’s actual experiences (α = 0.92). Items were averaged to form the gender attitudes scale, with higher numbers indicating more negative attitudes toward women.
Aside from the fact that it defines a lack of support for feminism as sexism (and the middle item, which bears directly on the third, is precisely the matter the study is attempting to treat empirically), this so-called sexism scale introduces the second of two possible confounds. The first is that the cover story may have encouraged many of the participants to answer even direct questions about their own responses as if they were answering questions about how they believed most other people in their position would answer them. And the second problem is that for obvious reasons it’s important that the participants not know the true purpose of the study, which the authors insist was “double-blind.” But we must wonder what conclusions the participants might have drawn about the researchers’ goals when they came across the “Modern Sexism Scale,” a really odd set of questions about the responders’ own views in a survey of their thoughts about an applicant.

           We also need to distinguish sexism—the belief that one sex is inferior—from biased behavior. Bias can be based on several factors besides sexism—but the feminists fail to acknowledge this. The authors of the study explain the (modest) difference in ratings for wholly imaginary applicants as the result of arbitrary, sexist stereotypes that have crept into people’s minds. (They of course ignore the sexist belief that men are less likeable—rightly so because the methods don't allow them to identify that belief.) The alternative explanation is that the bias is based on actual experiences with real people: the evaluators may have actually known more men who wanted lab management positions, more men who had successfully worked in that role, and/or more females who didn't work out in it. The conflating of sexism (or racism) with bias is akin to saying anyone who doesn't forget everything they’ve experienced with different types of people when making hiring decisions is guilty of perpetrating some injustice.
Jo Handelsman
            In a live chat hosted on Science’s webpage, one of the study authors, Jo Handelsman, writes, “We know from a lot of research that people apply more bias in decision making when they have less information, so I think this type of quick review is the most prone to ‘gut level’ decisions, which are colored by bias.” Implicit or gut-level reactions are notoriously sensitive to things like the way questions are framed, the order in which information is presented, and seemingly irrelevant or inconsequential cues. This sensitivity makes complex results from studies of implicit associations extremely difficult to interpret. Handelsman and her colleagues tried to control for extraneous factors by holding the conditions of their study constant for all participants, with the sole difference being the name and sex on the forms. But if I’m a scientist who’s agreed to assess an application in a hypothetical hiring situation for the purpose of helping to design a mentoring program, I would very likely be primed to provide information that I believe might give the students who are the beneficiaries of the research some useful guidance. I might, for instance, want to give female scientists a heads-up about some of the obstacles they might encounter—especially if in the course of the survey I’m reminded of the oppression of wives by husbands, discrimination in society at large, and the fact that some people are so callous as to not even want to hear about how bad women have it.
Another possibility is that the omnipresent and inescapable insistence of STEM fems that sexism is rampant is actually creating some of the bias the studies by STEM fems then turn around and measure. Since Moss-Racusin et al. report that high scores on the so-called Modern Sexism Scale correlated with lower ratings for females’ competence and hireability, we have to ask if the study participants might have been worried about women primed to make excuses for themselves, or if they might have been reluctant to hire someone with an ideologically inspired chip on her shoulder who would be ready to cry gender discrimination at the first whiff of rough treatment. Such alternative interpretations may seem like special pleading. But the discrepancy between the findings of this study and those of the National Academies committee, which, again, were based on a sample that was more than ten times larger and measured the variables directly, calls out for an explanation.
Perhaps the most troubling implication of the study is that women applicants to scientific positions will be less likely to make to the interview stage of the hiring process, so all the implicit stereotypes about women being less competent will never be overridden with more information. However, the National Academies committee found that in actuality, The percentage of women who were interviewed for tenure-track or tenured positions was higher than the percentage of women who applied” (157). Unless we assume males tend to be worse candidates for some reason—sexism against men?—this finding rules out the possibility that women are discriminated against for interviews. Are the women who make it to the interview stage thought to be less competent and hireable than their male counterparts? According to the committee report, “For all disciplines the percentage of tenure-track women who received the first job offer was greater than the percentage in the interview pool.” This finding suggests that for some reason women are thought to be better, not worse, candidates for academic positions. If there’s any discrimination, it’s against men.
It could still be argued that the Moss-Racusin et al. study suggests that the reason fewer women apply for positions in science and math fields is that they get less encouragement to do so because participants said they were less likely to mentor female applicants for a hypothetical position. But how do we square this finding with that of the National Academies finding that “Female tenure-track and tenured faculty reported that they were more likely to have mentors than male faculty. In the case of tenure-track faculty, 57 percent of women had mentors compared to 49 percent of men” (159). Well, even if women are more successful at finding mentors, it could still be argued that they would be discouraged by offers of lower starting salaries. But how would they know, unless they read the study, that they can expect lower offers? And is it even true that women in science positions are paid less than men. In its review of the records of 500 academic departments, the National Academies study determined that “Men and women seem to have been treated equally when they were hired. The overall size of start-up packages and the specific resources of reduced initial teaching load, travel funds, and summer salary did not differ between male and female faculty” (158).
Real world outcomes seem to be completely at odds with the implications of the new study, and at odds too with STEM fems insistence that discrimination accounts for a major portion of women’s underrepresentation in math and science careers. The National Academies study did however offer some strong support for preference theory. It turns out that women are more likely to turn down job offers, and the reason they cite is telling.
In 95 percent of the tenure-track and 100 percent of the tenured positions where a man was the first choice for a position, a man was ultimately hired. In contrast, in cases where a woman was the first choice, a woman was ultimately hired in only 70 percent of the tenure-track and 77 percent of the tenured positions. When faculty were asked what factors they considered when selecting their current position, the effect of gender was statistically significant for only one factor—“family-related reasons.”

Stephen Ceci
The Moss-Racusin et al. study was probably conceived of as a response to another article published in the same journal, The Proceedings of the National Academy of Science, in February of 2011. In “Understanding Current Causes of Women’s Underrepresentation in Science,” authors Stephen Ceci and Wendy Williams examine evidence from a vast array of research and write, “We find the evidence for recent sex discrimination–when it exists–is aberrant, of small magnitude, and is superseded by larger, more sophisticated analyses showing no bias, or occasionally, bias in favor of women” (1-2). That Moss-Racusin et al.’s study will likewise be superseded seems quite likely—in fact, it already has been superseded by the NAS study. Ceci and Williams' main conclusion from their review is a good summary of preference theory:
Wendy Williams
Despite frequent assertions that women’s current underrepresentation in math-intensive fields is caused by sex discrimination by grant agencies, journal reviewers, and search committees, the evidence shows women fare as well as men in hiring, funding, and publishing (given comparable resources). That women tend to occupy positions offering fewer resources is not due to women being bypassed in interviewing and hiring or being denied grants and journal publications because of their sex. It is due primarily to factors surrounding family formation and childrearing, gendered expectations, lifestyle choices, and career preferences—some originating before or during adolescence. (5)
Moss-Racusin et al.’s study should not be summarily dismissed—that’s not what I’m arguing. It is suggestive, and the proverbial further studies should be conducted. But let’s not claim it’s more important than it really is just because it produced the results the STEM fems were hoping for. And let’s quit acting like every study that produces evidence of gender discrimination is a victory for the good guys. Far too many people assume that feminism can only be good for women and good for science. But if discrimination really doesn’t play that big a role for women in science—which everyone should acknowledge the current weight of evidence suggests is the case—the infusion of gender politics has the potential to cause real harm. The standing accusation of sexism may not in the end lead to better treatment of women—it may lead to resentment. And the suggestion that every male scientist is the beneficiary of unfair hiring practices will as likely as not lead to angry defiance and increasing tension.

           To succeed in the most elite fields, you have to be cut-throat. It would be surprising if science and math careers turned out to be peopled with the nicest, most accommodating individuals. Will the young woman scientist who has a run-in with a jerk frame the encounter as just that—a run-in with an individual who happens to be a jerk—or will she see it as a manifestation of patriarchal oppression? It seems to me the latter response embodies the same type of prejudice the STEM fems claim to be trying to end.

Read Catherine Hakim's Feminists Myths and Magic Medicine

And my series of posts on "Why I Am Not a Feminist"


gillt said...

On one side are the fussy sounding Preference Theorists and on the other, the rabid Fempire, for which I found the ideal image.


Dennis Junk said...

Ha. It even has a cherry. I considered a pic of the Death Star.