Why are there so few women in science? Is it because the STEM fields are biased against women?
We can all agree that sexism exists, is bad, and needs to end. But in a world where all are equally welcome in every line of work, would every occupation be 50 percent female? If not, should we be measuring our progress in fighting sexism in science by the sex ratio of the scientists? My answers are no, and no.
If sexism is the only cause of women’s under-representation in science, then the environment in STEM fields today must be very bad – even worse than it was in medicine and law in the 1960s and ’70s, a time when women’s representation in these professions was increasing dramatically.
This seems unlikely. Law is claimed to be sexist even today, even though the overall sex ratio is quite balanced. And sexism is the last thing one would suspect of academic sociologists, yet there is a marked preponderance of men in the more mathematical subfields of sociology (Section #3 here). Sexism exists, but something else is going on.
Another popular hypothesis is that the distribution of mathematical abilities has a higher variance for men than for women, so that men form a majority in the upper tail. The problem is that most of us are nowhere near the upper tail in any distribution, and yet we all manage to find work. Men form a majority of electrical engineers, but also a majority of electricians. Different variances might matter at the very top (and bottom) but, again, something else is going on.
In my opinion, the key element missing from this discussion has been an understanding of occupational choice. Many people seem to assume that in a world without sexism, if two people are equally good at math they will be equally likely to enter a STEM field. This belief is false.
The basic economic principle guiding occupational choice – comparative advantage – was discovered by David Ricardo in the early 1800s. It is simple but very powerful. It helps to explain patterns of occupational choice, it gives normative guidance on what those choices should be, and it suggests policies that can help achieve these goals.
Let me illustrate with an example from my first-year economics course. Suppose there are only two subjects in high school – English and mathematics. Performance in either indicates intelligence and industry, but math performance also indicates the potential to be a good scientist, while performance in English also indicates (let’s say) the communication and empathetic skills necessary to be a good doctor.
High school boys lag behind girls in every subject but math. So let’s suppose Susan is getting 90 percent in math and 95 percent in English, while Steven is getting 90 percent in math and 85 percent in English. Who is more likely to major in science?
The choice to enter a field does not depend just on your expected success compared to others in that field. What matters is your expected success relative to what you can expect to achieve by doing something else. You are making a choice. So perhaps the best reply to our question about why so few women are in STEM is simply another question: “What else are the men in STEM going to do?”
Comparative advantage matters. A simple regression will show that students are somewhat more likely to choose a STEM field if they have good grades in math. But the ratio of their math grade to their overall average adds a lot to this regression. The same pattern is seen when the results of aptitude tests are used instead of grades.
Should we be encouraging more women to enter STEM fields? In our example, Susan is as good as Steven at math, so she will be equally good as a scientist. But what if favouring Susan for a math or science field forces Steven to seek work elsewhere? In general, the pattern of job assignments that wrings the most from our limited human resources is the one that respects comparative advantage. In our example, that means Steven should be the scientist and Susan the doctor.
Of course productivity isn’t everything, and we might still wish for greater gender balance in all fields. An understanding of comparative advantage is useful here as well. Current educational policies focus on encouraging girls in math and science. What if we put the same energy into helping boys do better in all the other subjects? This would increase the proportion of women in STEM by giving boys an alternative. At the same time it might begin to address the other major equity problem in our universities – the severe under-representation of males.
Surely this is a policy we can all support.
Dr. Carmichael is professor of economics at Queen’s University.
One thing that this misses (but that all but the most facile feminist critiques don’t) is that the real question is why Susan and Steven get different grades. Is it because Steven’s brain is evolved to be inevitably better at math than Susan’s? No. It is because society constantly hammers home the message to boys that STEM is a boy thing and that other things are girly things. It won’t be easy, but we need to build a society where Susan grows up thinking that being an engineer is just as realistic and admirable for her as being a fashion designer. And yes, we need to give Steven more choices than being a unintelligent jock and being a STEM nerd.
Also: “yet we all manage to find work” ? I wish.
Also: Susan’s being in STEM doesn’t mean one fewer STEM job for Steven. More jobs means more economic growth, and more economic growth means more money for companies to hire people. This is especially true is supposedly innovative fields like STEM.
I have a general reply below, but just to be clear – Steven’s brain is not better at math than Susan’s.
In fact, in my example, Steven could be worse than Susan at math and it would still be best to have Steven be the scientist and Susan the doctor. To make it really stark, suppose Steven is getting 80% in Math and 50% in English. Susan is a better scientist than Steven, and yet she should still be the doctor.
But Steven didn’t do better than math at Susan. That was the whole point. The whole point is that they are equally good at math, but Susan is even better at English than she is at math.
And why is such less number of females choose to make a choice in a career in STEM? Is it because it’s a male majority? Is it because females are people-oriented while STEM are competitively male and culturally masculine? Is it because they don’t have enough female role models in STEM nor enough reinforcement into STEM by their female peers? Is it because females lose confidence in math and science abilities from biased standardized testing prior to college? Most likely. It’s easy to simplify it as a matter of choice based on biology instead of choice based on environmental AVERSION. Maybe men do indeed have an advantage when it comes to math but being disadvantaged does not mean women are incapable or disinterested. In China alone, 40 percent of Engineers are female so then why is the case different for women here in the United States? Maybe women in China aren’t as averse to entering to STEM as women in the US
A more general reply is below.
I think cross country comparisons are interesting. The following idea is speculative, but I think it is worth consideration. If you look at the countries with the most participation of women in STEM (eg., China, Iran), these are countries where I think most of us would say women have less choice about their careers. The same was true of Eastern European countries before the fall of the Soviet Union. The more egalitarian countries where women have more choice (eg., the Nordic countries) have relatively fewer women in STEM.
If your society values STEM success above everything else then everyone with mathematical ability will be guided toward a STEM field regardless of their other talents. Lots of women have mathematical ability. If your society allows for individual choice, some women with broad skills including mathematical ability will nonetheless choose to do something else.
I deeply apologize for taking away a STEM job from a poor, basically otherwise unemployable, man. It’s so selfish of me to pursue this career knowing I have unlimited options for jobs while men have so little. We definitely need to put more education resources into making sure males have more career choices, not groups that encourage women how to fight against proven discrimination (http://www.pnas.org/content/109/41/16474.abstract), lower wages (https://www.policyalternatives.ca/publications/reports/narrowing-gap), and a talking Barbie that says “Math is hard!”. Thank you Dr. Carmichael for putting the focus back where it belongs – on the disadvantaged males.
*I also love that one of the comments was so much better referenced and researched than the actual article.
Again, I think good people can disagree about important issues, and the ways to deal with them.
Your decision to enter a STEM field was an individual choice. In a world where everyone makes individual choices, the person who was displaced was someone who was close to indifferent between entering a STEM field or some other endeavour. It was not the individual who has no alternative. This person will continue work in STEM. Even as he or she might lose the opportunity you received, he or she will seek out another.
It is when we get to large scale policies that we need to be more careful. If we were to dictate that half of the workers in STEM be women then we would be displacing a lot of men who would be ill suited to their new occupations.
I don’t think anyone is really suggesting that we do this, even over the very long term. I think what most of us want is a world where men and women feel welcome wherever they go. So, let’s start working on this directly. That means we should also be measuring our success directly, regardless of the different occupational choices men and women end up making.
Thanks for the comments. All of you are missing the points of my article, although you have points of your own.
My main normative point is methodological. It’s just basic arithmetic. If half the engineers and scientists are going to be women then half the doctors, nurses and primary school teachers are going to be men. The costs and benefits of that situation need to be evaluated alongside those of having more women in STEM. That means you have to look at the whole list of skills and strengths of men and women – not just relative skills at math.
The empirical puzzle is to understand why over the last 50 years STEM fields have remained male dominated while law and medicine have not, in spite of explicit educational and financial attempts to affect change. There was a time when there were very few women in any of these fields. It could be that the crusty old male engineers of the last century were even more unwelcoming than the crusty old male doctors and crusty old senior partners in law firms. That could be it. But I doubt it. Direct evidence of sexism is necessarily anecdotal, but I’m sure I could find horrifying descriptions of the “macro-agressions” faced by pioneering female doctors and lawyers that would match or dominate anything faced by the women engineers of today.
My hypothesis is that there are a number of intelligent males who are decent to brilliant at math but who are also socially awkward and would dislike the constant human interaction that is a part of work in so many non-STEM fields. OK, call them nerds. There are women who feel this way too, but not as many. These folks enter STEM fields because it is the one thing they can do well and it is a place where they fit in.
I do not think these folks were socialized to be this way. Growing up, I think most would have greatly preferred to be counted among their more socially adept peers.
Sexism against women in science exists. Sexism against women in law and medicine exists. Sexism against men exists as well, in nursing and primary school teaching. We all want it to end. But sexism against women cannot be the sole explanation for the lack of women in STEM fields. Explanations based on personal choice can account for the data without requiring us to believe that sexism is worse in STEM today than it ever was in medicine and law.
It follows that we should not be measuring our success in fighting sexism by the sex ratio in STEM.
I find your point interesting. It is possible that a woman in STEM may feel that she has other viable career options whereas a man of comparable ability may not. What troubles me is that if she is deterred by sexism in STEM, she may be sacrificing earning potential by moving in another career direction that is not as lucrative. In fact, careers that are dominated by woman tend to have lower pay.
Also, if we accept your premise that the sex ratio in STEM to best serve society’s needs is not 50/50, and if this ratio is highly skewed, the imbalance will make it more difficult for women in STEM. I can attest to this from personal experience working in aerospace in the 1980’s.
I think good people can disagree about important issues, and the ways to deal with them.
I agree that women in STEM face sexism. I agree that women are as good at math as men.
The “leaky pipeline” – the fact that women start out in STEM fields but leave in greater numbers than men – is certainly consistent with the fact that women feel unwelcome in STEM fields. Like you, I hope for a world where women and men are welcomed in every field, be it science, medicine, or elementary school teaching. We are not there yet.
But I have to point out that the “leaky pipeline” is also consistent with comparative advantage and individual choice. Suppose Steven and Susan are in an engineering program and they are each finding it very difficult. Steven says to himself: “At least I’m in a program where I don’t have to write any essays.” Susan says to herself: “I could be in a program where all I have to do is write essays.”
Choice is always about the evaluation of personal alternatives. Women have more alternatives. From the abstract of the paper I cited above,
“Results revealed that mathematically capable individuals who also had high verbal skills were less likely to pursue STEM careers than were individuals who had high math skills but moderate verbal skills. One notable finding was that the group with high math and high verbal ability included more females than males.”