Last August, a male Google employee, James Damore, was fired for circulating a memo that claimed women should not be hired by tech companies because they are biologically unsuitable. In the ensuing uproar, I was reminded once again how persistent the belief is that women have a hard-wired brain difference that impairs their ability to do certain mental tasks.
This belief is usually shored up by an aggressive claim that the brains of men and women have been “scientifically proven” to be fundamentally different. That difference is usually expressed as some explicit or implicit deficiency on the part of the women: they can’t do math, they are “intuitive” instead of “logical,” they are “emotional” instead of “rational,” or, as Mr. Damore claimed, they care about people more than things. (NB: the word for individuals who care more about things than people is “psychopath.”)
I decided, for my talk last week at SAP’s International Women’s Day celebration, to address the neuroscience and performance data on sexual differences in brains. In 2016, I was asked by the International Finance Corporation to do a review on the topic and they kindly allowed me to publish the white paper on the Double X Economy site. Ironically, I was putting the post up the same day that the Science Museum in London opened an exhibit that presented to the public that there was a “pink” and a “blue” brain. There was also some serious pushback about that, especially from scientists, and the museum had to issue an apology. Once I put up the white paper (very short summary is here), I received an unusual amount of feedback from readers in finance, technology, and management consulting telling me how grateful they were for the information and that they were sharing the link with their friends and co-workers.
So, I thought the topic might be of interest to the SAP audience. I am going to summarize my remarks in this post and add a few more things that were not mentioned on stage. I am only going to sprinkle a few links to references in this post. If readers want to see the full references and discussion, the full IFC white paper is here.
Rather than going directly to the brain science, I began the discussion at SAP by describing the performance record on the gender gap in math abilities. Contrary to popular belief, there is no gender gap in math performance, at least in most countries. Especially among the developed countries, government investment in making higher math courses accessible to female high school students closed the gap about 20 years ago. Let me repeat for emphasis: the gender gap in math closed 20 years ago.
As national governments began to show more interest in this topic and collecting data on it, the possibility opened up for comparing national performance on standardized math skills tests. In a Science article published in 2008, Guiso, Monte, Sapienza, and Zingalez showed that female test performance varied directly with the World Economic Forum’s Global Gender Gap Index. That is, where gender equality is measurably high, female math performance is also high, but where gender equality is low, female math performance is also low.
The last gender gap left standing, as far as basic capabilities relevant to tech, was spatial rotations. No educational program seemed to affect it. Then, a 2007 study by Feng demonstrated that the gender gap in spatial cognition closed after girls played only ten hours of video games. Yeah. Ten hours. Not much to build a claim to biological superiority and global dominance on, do you think? Today, primary school children do not demonstrate this gap at all, most likely because of early exposure to games, phones, tablets, etc.
Another fable about math and logic skills also bit the dust. In the US, there were more males in the highest percentiles of math performance. This fact was also used to say that women just were, for biological reasons, not as good. However, comparative studies showed that the phenomenon did not occur among Asian-Americans. And, in Iceland, Thailand, and the UK, girls appeared in the highest percentiles in equal or greater number than boys. Further, it was shown by mathematics scholars that the higher number of “superstars” that were boys was a function of the larger pool of candidates who were male. So, by extension, if you had more women in tech, you would have more female versions of Bill Gates, just from the greater probability that a big talent would emerge from a larger pool. (I apologize for the unfortunate mental image.)
Because I am so irretrievably wonky, I feel I must also add that the original studies claiming that testing differences were biologically-based had sample sizes much too small to support such gross generalizations. They were also conducted lightyears ago. Newer meta-analyses have exponentially larger samples and contradict the earlier tests.
In sum, performance data not only shows that math skills gaps have closed, but also that previously-observed differences were caused by (1) inequalities in access to instruction, (2) socialization issues like toys and stereotype threat, and (3) sample size.
Once you know that performance differences have been closed, especially for these reasons, the burden of proof on neuroscience to show that there is a biological basis for math ability, logic, rationality, and so forth gets a lot bigger. But nobody is really trying to do that anymore. That’s because, thanks to advances in research technologies, neuroscience now works on an entirely different model of the brain than the one presumed by these statements.
Here’s the buzzword: cerebral plasticity. We now know that humans must generate 100 billion synapses in the brain in order to function properly as adults. Only 6,000 genes are available to do that at birth. That number of genes is sufficient to produce only 10% of those 100 billion connections. The remaining work is done through socialization (i.e., it’s learned).
We make these connections from birth by learning to recognize colors, determine the direction of sound, look at faces, and so on. We are constantly absorbing information and building synapses in each second of our individual lives. So, the overall pattern of connections in each person’s brain (called a “connectome”) is as unique as a fingerprint or a snowflake.
The connectome, BTW, is not a “hard-wired” phenomenon, though it is sometimes talked about that way. Importantly, it grows and recedes as you learn and forget. For instance, if you are a London cab driver, you have a map in your head of London that is way more detailed and accurate than anyone else has. As long as you are driving a cab, you will keep it in your head and there may even be certain physical evidence in your brain of those skills (like a thickening in a lining or something). But when you stop driving the cab, the memory and the skill set recedes—and the connections as well as the physical evidence disappear. This effect has been demonstrated for cab drivers, as well as for more discrete skills like juggling and playing piano.
But, because cultures push girls to like, want, do, and be certain things (pink, dolls, dance, sweet) and boys to like, want, do, and be a whole different set of things (blue, trucks, sports, boisterous), we end up with a pattern of differences by sex at the meta-level. Studies have shown an amazing number of interests, traits, and affinities sum up to a broad difference between the sexes when there is a very, very large sample size. These differences do not hold up at the individual or small group level and so such generalities should never be used at work for any reason. Remember the snowflake is still there, even in a blizzard. Remember, too, that even these differences can change if re-learning takes place.
What about the raw meat, if you will, of brains? Just plain old physical differences? Nope. Sorry. Brains do not exhibit “sex dimorphism,” as scientists call it—referring to physical manifestations of sex. In other words, brains are more like livers than they are like testicles or ovaries.
At this point in time, therefore, there is consensus among neuroscientists, as well as other scientists in related disciplines, that male/female brains are simply more similar than different. You do still see studies using various methods to show sex differences that the authors claim are important. Unfortunately, there is such an appetite among the public for evidence to buttress their existing beliefs that the media will push any study that claims to show sex differences in the brain. In the white paper, I discuss such a study that, despite a rather dubious premise, got a lot of press attention for “proving” that female brains are “intuitive” and male brains are “rational.” Yet contradicting studies with better methods and larger samples appeared both immediately before and immediately after—and were ignored by the press.
Public attention to this kind of misleading information is causing an ethical crisis in neuroscience. Some are arguing that the studies themselves should be banned. Before you jump to your feet in defense of open inquiry, let me give you two things to think about. There is already a very nasty history of scientists doing studies that purported to show biological differences, including in brains, between humans of different races and ethnicities. Such studies have been used for vile political actions, including genocide—and are currently being used to justify racist immigration policies. Ethically and scientifically speaking, those studies are no different from the ones that claim inborn mental deficiencies in females. I don’t think the work should be banned, but I think the bar needs to be set higher for such research to be published. In ordinary discourse, I think we need to question why some people set out to prove these things.
Then I want to remind you of the plasticity issue. Neuroscientists are worried, legitimately so, that the propagation of such beliefs has the potential to harm females, especially girls still in school. That’s because, actually, we now know that kind of socialization has a negative effect on both performance and the brain itself.
Already, studies have shown that when people are exposed to these statements as “science,” they go back to the workplace and make bigoted decisions. So, actually, it really matters a lot when people like James Damore express this false belief in the workplace. It’s offensive, sure, but it also has the potential to be damaging, in very concrete ways. Not only do such expressions hurt the women themselves, they potentially reduce performance on the job, negatively affecting business outcomes. In short, the threat posed by bigotry is much greater than the perceived danger of estrogen.
To conclude, it is simply false that women have biologically-grounded brain differences that would affect job performance at a tech firm, a bank, or a consulting company. Of far more concern is the damage caused when people in the workplace insist on believing that those brain differences are real.