Larry Summers, the president of Harvard, suggested the other day that innate differences between the sexes might help explain why relatively few women become professional scientists or engineers. For this, he has been denounced—metaphorically, of course—as a Neanderthal. Alumni are withholding donations. Professors are demanding apologies. Some want him fired.
Everyone agrees Summers' remarks were impolitic. But were they wrong? Is it wrong to suggest that biological differences might cause more men than women to reach the academic elite in math and science?
To begin with, let's clarify what Summers said. He spoke after the morning session of a conference called "Diversifying the Science and Engineering Workforce: Women, Underrepresented Minorities, and their S. & E. Careers." He offered three possible reasons for this gender gap. The biggest, he suggested, was that fewer mothers than fathers are willing to spend 80 hours a week away from their kids. The next reason was that more boys than girls tend to score very high or very low on high-school math tests, producing a similar average but a higher proportion of scores in the top percentiles, which lead to high-powered academic careers in science and engineering. The third factor was discrimination by universities. Summers said repeatedly that Harvard and other schools should work to eliminate discrimination. But he theorized that it was less a decisive factor than the others, since women were already underrepresented by the time they got to the pool of candidates eligible for top science jobs.
By some accounts, Summers referred to "innate ability" or "natural ability" as a possible explanation for the sex difference in high-school test scores. This is what set off the furor. One professor walked out in disgust. Others expressed their outrage to the media. ''Here was this economist lecturing pompously [to] this room full of the country's most accomplished scholars on women's issues in science and engineering, and he kept saying things we had refuted in the first half of the day," a fellow speaker told the Boston Globe.
The conference agenda for that morning, available online, includes two slide presentations and nine recommended readings. The first presentation concludes that "most of the gains" in female representation in science and engineering careers "can be explained by increases in Bachelors' [degrees]—potentially normal supply response." That's exactly what Summers argued. The second presentation indicates that degrees earned by women have increased more rapidly at the masters' level than at the bachelors' level, calling into question the breadth of discrimination at that stage.
One recommended reading, a 2004 Government Accountability Office * report on "Women's Participation in the Sciences," concludes, "A variety of studies indicate that experience, work patterns, and education levels can largely explain [gender] differences in salaries and rank." Another reading, based on a national study, adds, "There is general agreement that few women typically apply for academic positions in science and engineering departments at research universities."
Only one reading comes anywhere near challenging Summers' hypothesis. Claude Steele, a Stanford psychologist, writes that in his 1997 study, female students in a math test "performed equal to men when the test was represented as insensitive to gender differences." It's a fascinating study, probably just the sort of thing Summers had in mind when he called for further research into genetic and non-genetic factors in test performance. But the study compared average scores, not the distribution of high and low scores, which was Summers' point. Moreover, it was a test of college students, not high-school students, and the participants "were selected for being very good at math." In other words, it took place after the genetic bias hypothesized by Summers would have skewed the pool.
What's the evidence on Summers' side? Start with the symptom: the gender gap in test scores. Next, consider biology. Sex is easily the biggest physical difference within a species. Men and women, unlike blacks and whites, have different organs and body designs. The inferable difference in genomes between two people of visibly different races is one-hundredth of 1 percent. The gap between the sexes vastly exceeds that. A year and a half ago, after completing a study of the Y chromosome, MIT biologist David Page calculated that male and female human genomes differed by 1 percent to 2 percent—"the same as the difference between a man and a male chimpanzee or between a woman and a female chimpanzee," according to a paraphrase in the New York Times. "We all recite the mantra that we are 99 percent identical and take political comfort in it," Page said. "But the reality is that the genetic difference between males and females absolutely dwarfs all other differences in the human genome." Another geneticist pointed out that in some species 15 percent of genes were more active in one sex than in the other.
You'd expect some of these differences to show up in the brain, and they do. A study of mice published a year ago in Molecular Brain Research found that just 10 days after conception, at least 50 genes were more active in the developing brain of one sex than in the other. Comparing the findings to research on humans, the Los Angeles Times observed that "the corpus callosum, which carries communications between the two brain hemispheres, is generally larger in women's brains [than in men's]. Female brains also tend to be more symmetrical. … Men and women, on average, also possess documented differences in certain thinking tasks and in behaviors such as aggression."
Let's be clear about what this isn't. It isn't a claim about overall intelligence. Nor is it a justification for tolerating discrimination between two people of equal ability or accomplishment. Nor is it a concession that genetic handicaps can't be overcome. Nor is it a statement that girls are inferior at math and science: It doesn't dictate the limits of any individual, and it doesn't entail that men are on average better than women at math or science. It's a claim that the distribution of male scores is more spread out than the distribution of female scores—a greater percentage at both the bottom and the top. Nobody bats an eye at the overrepresentation of men in prison. But suggest that the excess might go both ways, and you're a pig.
The only implication I'd draw immediately is that it may prove easier to equalize gender representation in math and science in high school than in college, and easier to equalize it among students than among professors. Equal representation should be a goal that prods us toward equal opportunity, but the two mustn't be confused. Last year Harvard offered only four of 32 tenured positions in the arts and sciences to women. A genetic difference between the sexes doesn't mean four was anywhere near the right number. It just means the number doesn't have to be exactly 16.
Already Summers is being forced to apologize, in the style of a Communist show trial, for sending "an unintended signal of discouragement to talented girls and women." But the best signal to send to talented girls and boys is that science isn't about respecting sensitivities. It's about respecting facts. The only people who don't belong in science, male or female, are those who would rather close their eyes—and yours—than see what's there.
To err is human. Send critiques or corrections to firstname.lastname@example.org. E-mail may be quoted by name unless the writer stipulates otherwise.