By Ruth Marcus
Wednesday, December 3, 2008
Was Larry Summers right about women and science, after all?
As the mother of two girls, I hope not. In fact, Summers himself said in his infamous comments about intrinsic differences between the genders, "I would like nothing better than to be proved wrong."
But Summers may have been on to something, recent research suggests. Math and science test data, he noted, show gender differences at each end of the performance spectrum. In other words, men are overrepresented at the very top and bottom.
This small but significant variance, he hypothesized, suggested differences in innate aptitude -- "whatever the set of attributes are that . . . correlate with being an aeronautical engineer at MIT or being a chemist at Berkeley" -- that help explain the dramatic underrepresentation of women in tenured jobs on elite science and engineering faculties.
These remarks, of course, led to Summers's ouster as president of Harvard. Whether they cost him the Treasury secretary job in the Obama administration, they surely didn't help. Instead, Summers will head the National Economic Council, avoiding a confirmation hearing that would no doubt have dredged up the recent unpleasantness.
It's worth, though, going back to Summers's remarks to understand what he described, in a masterpiece of understatement, as "some attempts at provocation." He posited three factors to explain the faculty gender gap. The most significant, he suggested, was the all-consuming nature of the high-powered job.
In Summers's words, "the most prestigious activities in our society expect of people who are going to rise to leadership positions in their 40s near total commitments to their work. . . . And it is a fact about our society that that is a level of commitment that a much higher fraction of married men have been historically prepared to make than of married women."
The academic life has always seemed rather relaxing to me, but Summers's larger point is valid -- and, by the way, relevant to the president-elect's hires so far. Gratifyingly diverse, yes, but more of the men than women have children at home.
Summers did not discount, although he placed third, the role of overt and subconscious socialization and discrimination. Where he got into big trouble was the "intrinsic aptitude" part.
Here's the interesting thing, though. A group of researchers (all women, as it happened) looked at annual math assessments required by the No Child Left Behind law from 10 representative states that supplied details about gender and ethnicity, a total of 7 million students.
Their study, published in the July 25 issue of Science, found no differences between girls and boys in average performance -- not even, as earlier studies had found, once they entered high school. The gap between girls and boys on math SATs, they said, could be explained by the fact that more girls than boys go to college and therefore take these tests.
But, echoing Summers's point, there was small yet significant variance between the genders -- the degree to which the scores of girls or boys differed from the average. At the very highest level, the 99.9th percentile, the difference meant 2.15 males for every female. This difference was large enough that, in an occupation requiring math skills at that level, the job ranks could be expected to be filled 68 percent by men, 32 percent by women -- enough to explain, as Summers suggested, part of the gender gap.
Studies comparing girls and boys in different countries add to the puzzle, both underscoring gender differences and suggesting that the influence of cultural factors may be greater than Summers thought.
In performance on a standardized math, science and reading test given to 15- and 16-year-olds in 40 countries, girls in every country performed far better than boys in reading. Conversely, boys in all but three countries did better, but by not nearly as much, in math. In all but three countries -- Britain, Thailand and Iceland -- more boys than girls scored in the 99th percentile in math.
Yet this study, published in the May 30 issue of Science, also showed a correlation between girls' performance on math tests and countries where there is more "gender equality," as measured by things such as the share of female elected officials or women's participation in the workforce.
Summers was boneheaded to say what he said, in the way that he said it and considering the job that he held. But he probably had a legitimate point -- and the continuing uproar says more about the triumph of political correctness than about Summers's supposed sexism.