Contrasting Views on the Gender Disparity in Science

Daniel Drezner links to two articles with alternative interpretations to the gender gap in science. Both are looking at a female exodus from hard sciences, but explain it in different ways. First, Lisa Belkin in the NYTimes takes the angle of institutionalized discrimination and a macho male culture:

Based on data from 2,493 workers (1,493 women and 1,000 men) polled from March 2006 through October 2007 and hundreds more interviewed in focus groups, the report paints a portrait of a macho culture where women are very much outsiders, and where those who do enter are likely to eventually leave.

The study was conceived in response to the highly criticized assertion three years ago, by the then-president of Harvard, that women were not well represented in the science because they lacked what it took to excel there.

The purpose of the work-life center's survey was to measure the size of the gender gap and to decipher why women leave the science, engineering and technology professions in disproportionate numbers.

The problem isn't that women aren't making strides in education in the hard sciences. According to a National Science Foundation report in 2006, 46 percent of Ph.D. degrees in the biological sciences are awarded to women (compared with 31 percent two decades ago); 31 percent of the Ph.D. degrees in chemistry go to women, compared with 18 percent 20 years ago.

And, women enter science engineering and technology (known as the SET professions) in sizable numbers. In fact, 41 percent of workers on the earliest rungs of SET career ladder are women, the study found, with the highest representation in scientific and medical research (66 percent) and the lowest in engineering (21 percent).

They also do well at the start, with 75 percent of women age 25 to 29 being described as "superb," "excellent" or "outstanding" on their performance reviews, words used for 61 percent of men in the same age group.

An exodus occurs around age 35 to 40. Fifty-two percent drop out, the report warned, with some leaving for "softer" jobs in the sciences human resources rather than lab bench work, for instance, and others for different work entirely. That is twice the rate of men in the SET industries, and higher than the attrition rate of women in law or investment banking.

The reasons pinpointed in the report are many, but they all have their roots in what the authors describe as a pervasive macho culture.

Second, Elaine McArdle argues the primary factors is women's preferences:

To help answer the question, Rosenbloom surveyed hundreds of professionals in information technology, a career in which women are significantly underrepresented. He also surveyed hundreds in comparable careers more evenly balanced between men and women. The study examined work and family history, educational background, and vocational interests.

The results were striking. The lower numbers of women in IT careers weren't explained by work-family pressures, since the study found computer careers made no greater time demands than those in the control group. Ability wasn't the reason, since the women in both groups had substantial math backgrounds. There was, however, a significant difference in one area: what the men and women valued in their work.

Rosenbloom and his colleagues used a standard personality-inventory test to measure people's preferences for different kinds of work. In general, Rosenbloom's study found, men and women who enjoyed the explicit manipulation of tools or machines were more likely to choose IT careers - and it was mostly men who scored high in this area. Meanwhile, people who enjoyed working with others were less likely to choose IT careers. Women, on average, were more likely to score high in this arena.

Personal preference, Rosenbloom and his group concluded, was the single largest determinative factor in whether women went into IT. They calculated that preference accounted for about two-thirds of the gender imbalance in the field. The study was published in November in the Journal of Economic Psychology.

Elaine McArdle also reports on studies by Susan Pinker that suggest that it might not be fertility decisions which cause women to leave science:

They have a provocative echo in the conclusions of Susan Pinker, a psychologist and columnist for the Toronto Globe and Mail. In her controversial new book, "The Sexual Paradox: Men, Women, and the Real Gender Gap," Pinker gathers data from the journal Science and a variety of sources that show that in countries where women have the most freedom to choose their careers, the gender divide is the most pronounced.

The United States, Norway, Switzerland, Canada, and the United Kingdom, which offer women the most financial stability and legal protections in job choice, have the greatest gender split in careers. In countries with less economic opportunity, like the Philippines, Thailand, and Russia, she writes, the number of women in physics is as high as 30 to 35 percent, versus 5 percent in Canada, Japan, and Germany.

"It's the opposite of what we'd expect," says Pinker. "You'd think the more family-friendly policies, and richer the economy, the more women should behave like men, but it's the opposite. I think with economic opportunity comes choices, comes freedom."

(Pinker's work contrasts this study which suggests that it is all about fertility decisions.)

I will allow you to draw your own conclusions based on these experiments. However, before people jump on me as a raging chauvinist, I would like to add some caveats.

  • One, I have repeatedly argued on this blog that innate differences in mathematical or technical ability do not explain the gender disparity in the sciences. There may be many and complex reasons for the gender disparity in science, but innate ability is not one of them.
  • Two, I -- like Drezner -- do not necessarily see these different explanations as mutually exclusive. The largely chauvinistic culture in the hard sciences affects the womens' preferences such that the previously technically inclined choose other jobs. A group populated largely by men is more likely to be chauvinistic because there is no one there to call them on their bullshit. Thus, the situation can become self-perpetuating. (I am not attempting to justify this behavior, just to understand it.) The gender disparity is almost certainly multifactorial in nature and disentangling those factors is very difficult to do.
  • Third, I remain highly skeptical of those studies like Rosenbloom's that suggest that there are innate differences in "preferences" and "personality." First, personal preferences are the consequence of numerous interactions in young life. If you don't believe that they are socially constructed, then you are forced to explain how systematic differences in preference are implemented in the brain. This is hard to do, and no one to my knowledge has done it. (It is a hell of a lot more complicated than the almost reflexive "testosterone did it" twaddle that people peddle out.) Second, one of my primary arguments against innate differences in ability between men and women is that you are dealing with traits that have distributions and those distributions largely overlap. Making a statement about any individual man or woman is largely useless. The odds of a women or man selected at random being better or worse at math are not particularly different. This argument applies just as well to differences in preference. Maybe there are differences on average, but they are still distributions that overlap. The key question becomes: to what degree do those distributions overlap? How different on men's and women's preferences on average?

Read the whole thing, and make your judgments accordingly.

UPDATE: Megan McArdle and Matthew Yglesias have more.


More like this

Sheril and other science bloggers are talking about the fact that the top 3 in the Intel International Science and Engineering Fair are female. I suppose that means we'll stop talking about gender disparities in science & engineering? Yeah, I doubt it.... At the same time, this just popped…
Related to the question of why there is a gap between the genders in math and the sciences is whether there are possible means of remedy. With respect to possible remedies it is often a good idea to look internationally at which countries don't have this problem -- to see what they are doing right…
I have tried to show that the gender gap in the sciences is not the result of cognitive differences, but that begs the question about what else to which it can be attributed. It could be that it is the result of conscious or unconscious discriminatory behavior. However, it could also be the…
Pseudoscience is effective. If it weren't, people wouldn't generate so much of it to try to justify opinions not supported by the bulk of the evidence. It's effective because people trust science as a method of understanding the world, and ideological actors want that trust conferred to their…

You are really skeptical that men and women as a whole have different innate preferences?

"The odds of a women or man selected at random being better or worse at math are not particularly different. This argument applies just as well to differences in preference."

The argument about ability being randomly distributed across genders I totally believe. Having seen it in practice, and proven many times over in studies.

That however does not magically apply to preference, especially if the averages are different, which sounds like they are.

The gender disparity is almost certainly multifactorial in nature and disentangling those factors is very difficult to do.

if this is true i wonder as to the immediate recourse to robust public policy response from many quarters. that implies that many don't think it is "multifactorial in nature," rather, they know what the problem is. Patriarch :-)

re: distributions, if you are selecting from the tails of the distributions the intersection is far less. if your subsets are selecting from an intersection of multiple distribution tails then the chances that your sample spaces are going to be the same size across the subsets will be the same is not good. in plain language, the intersection of traits and skills which would convince someone to devote their life to electrical engineering research are deviated from the norm. so you look at the means only as much as it can tell you have the nature of the tails. more concretly, i would be like assume that men and women should run in the same sprinting competitions because the distributions of speed overlap.

I recently retired as an engineering faculty member having served 37 years in both small colleges and state universities. When I started, the fraction of women in my classes was around 2% or so. There were no women faculty. Then, sometime in the 70s or 80s the percentage of women students quickly rose to around 15%, and stayed there for 20 years or so. Women faculty, after an appropriate lag, rose to a similar percentage. The failure of the percentages to rise to 50 is not due to overt (as opposed to hidden, subtle) discrimination, as women get better starting salaries than men, even as faculty, and the National Science Foundation favors women over men in in Career Awards, which go a long way to improving the chances of tenure. It seems to me that there is some sort of cultural effect at work. In my generation, girls were actively pushed away from science/engineering careers by family and friends, and I suspect that is still the case. The women I have had in class are as bright and as interested as the men, but maybe they are the tail of the distribution. I don't think the disparity between the percentage of men and women is a very big deal. We would like it to be less, but this is still a free country, at least until next January.

"...largely overlap..." -- performance at the high end of any skill is not about means and overlap, its about extrema. The fact that the distributions overlap in the "moderate skill" level is irrelevant.

If you want to talk about randomly selected men/women condition on their being from the upper tail of the pooled distribution. That's where the differences start to show up. It is also where the assumptions of symmetry and normal tails break down the fastest, so measures like Cohen's d less relevant.

I have got to wonder whether they have actually asked women who left the careers (not those who were already in a career of choice) why they left.

About societal pressure: it is not exactly easy to take the academic road when people think 1) you are crazy to waste your youth in a lab, 2) you are crazy to earn $25,000 or less a year, when with the same brain in business you would be starting at double that amount and 3) when you get psychologically harassed on the job.

By the time you are 30, IF THEY HAVE A CHOICE, most people would just give up and go for something where 1) they earn more (and given that on average women earn less than man, this is important if you want to have an independent life) and 2) work hours and work environment are more "normal". Something tells me that males get some emotional compensation (more guidance? being considered as thinking adults rather than "oh, you got this grant because of your XX chromosomes") that women do not get. And this "compensation" is higher in places where most faculty and students are...male. Of course, I am just speculating here.

People who leave science do not hate science. They like it, but they have spent too much time in an environment where they are being pressured beyond belief, and not compensated for it. And at some point, something cracks and they let go - to move on to better paid jobs and regain some mental and physical sanity.

At least, that is what I have observed around me in these past years.

You write: "The odds of a women or man selected at random being better or worse at math are not particularly different. This argument applies just as well to differences in preference." This is not an argument, however, but a statement of fact. You might as well say, "I have three oranges. This argument applies just as well to apples." We can measure ability and conclude that it is roughly equal between men and women. Can we measure preference and make a similar conclusion?

I won't resort to the testosterone twaddle, but there are undeniable, significant, and complex differences between male and female brain chemistry; why is it unreasonable to suppose that these (on average) result in different patterns of preference and behavior? Men and women act and think differently for a variety of social and biological reasons which are difficult to untangle, but I don't understand your skepticism about the idea that some of the difference may be innate. Our preferences depend on how our brains function, how our brains function depends in part on gender.

I was a physics major and now work in atmospheric science. Throughout my educational and professional life, most of my colleagues have been male. However, I have never witnessed hostility toward women; in fact, women are prized for their diversity quotient and have a better change of winning scholarships and good jobs than men do. I am also frankly at a loss for what aspect of scientific culture can accurately be described as "macho." I mean, other than that one time we had Hulk Hogan host the egg-catapult competition...

As a former Oak Ridge National Laboratory employee, I can say that I was never treated as an equal to my male peers. After almost six years I was still earning what I had as a new college graduate, while my male peers had been promoted on paper, and in wages, if not in responsibilities. Several times when I gave voice to my frustrations I was brushed off and told that I had a 'bad attitude'. Yes, my attitude was bad after being asked to babysit my bosses' children and run personal errands for him. I finally gave up and moved on to a new career.

Like that emo-glasses-wearing hobbit [Yglesias] knows donkey dick about science or can count past the fingers on his hand.

Can you please explain why this comment remains on your blog, but you have censored my comment about Megan McArdle? Is it because Yglesias is a liberal and McArdle is one of your favorite right-wing wackaloon bloggers?

Yes, my attitude was bad after being asked to babysit my bosses' children and run personal errands for him. I finally gave up and moved on to a new career.

Wow. That is just unbelievable. Well, it's not really, but I can't imagine my boss asking me to watch his kids or run his errands.

I hope you told HR about it before you left. That is completely out of line.

I agree that the causes for the gender gap are multi-factorial and not easily boiled down to one cause. But "macho" male attitudes are the least convincing explanation (and like other posters, I wonder what exactly what this means). Why then were women able to break into a formerly male bastion like biology and not engineering?

Nathan (and others along this vein),

It is really easy for you to say that you didn't see any discrimination or notice a macho culture when you are embedded in that culture. In an all male environment, I am often ignored or blown off in meetings, only to have my ideas repeated by a male colleague to praise and accolades. Unexpected results from me are considered a sign of a mistake in the lab rather than an exciting lead to follow up on. I am not often invited to happy hours or to spur of the moment lunch outings, which keeps me out of the loop for networking. Male colleagues of mine often don't notice stuff like this until it is brought up. And that is only the minor stuff!

It is very tiring being the only women. I end up speaking for my whole gender and not just for me. People make sexists remarks and jokes that make me the party pooper if I were to complain. I can see why some people just quit. And I'm sure the same is all true for ethnic minorities as well.

And this is as an adult. As a teenager the social pressure to not be so good in math and science was immense. And the teachers were just as bad. I was forced to tutor/babysit people in science classes because it was assumed I would want to help. My own teacher suggested that one of my (female) friends give up the math award our senior year in high school, since the runner up (a male) would use it more in college. How do things like that lead to "free choice"?