Sociologist Cited By Summers Calls His Talk ‘Uninformed’

Two sociologists whose research University President Lawrence H. Summers cited at an economics conference Friday said yesterday their findings do not support Summers’ suggestion that “innate differences” may account for the under-representation of women in the sciences.

University of California-Davis sociologist Kimberlee A. Shauman said that Summers’ remarks were “uninformed.” The other researcher, University of Michigan sociologist Yu Xie, said he accepted Summers’ comments as “scholarly propositions,” although he said his own analysis “goes against Larry’s suggestion that math ability is something innate.”

Xie and Shauman presented their findings at the National Bureau of Economic Research Friday afternoon, shortly after Summers’ remarks.

In an interview with The Crimson last night, Summers stressed that he only cited Xie and Shauman’s research as evidence that females are underrepresnted among the top 5 percent of test-takers on standardized assessments. Summers said the evidence for his speculative hypothesis that biological differences may partially account for this gender gap comes instead from scholars cited in Johnstone Family Professor of Psychology Steven Pinker’s bestselling 2002 book The Blank Slate: The Modern Denial of Human Nature.

‘PRIVILEGING BIOLOGY’

While Shauman was a graduate student at Michigan, she and Xie began analyzing massive sets of data containing junior high and high school students’ scores on standardized tests dating back to the early 1970s. Xie said that males are more likely to score in the 95th percentile or above on these tests. “Boys tend to be higher achievers and underachievers, while girls are more concentrated,” he said in an interview yesterday.

High-achieving males are more likely to enter science and engineering than females with the same top-tier test scores, the researchers found.

But Xie said Summers’ suggestion that this apparent gender gap accounts for the under-representation of females in science and engineering is “too simplistic.”

Shauman noted that societal factors might produce gender disparities in achievement even among eighth graders. “What was disappointing about [Summers’] comments was the kind of privileging of biology over socialization,” Shauman said.

While Shauman said that “people who go on to be Nobel laureates in the science probably are in that higher tail” of scores on standardized tests, both she and Xie emphasized that individuals can succeed in science even if they don’t perform amazingly well on achievement exams as high school students.

“We have very high-achieving women who choose not to go into science. We also have very low-achieving men who choose to go into—and succeed in—science,” Shauman said.

And while Summers said in his remarks Friday that mothers are often unwilling to devote the 80 hours per week necessary to achieve tenured posts at elite academic institutions, Shauman said that the 80-hour workweek is itself a social construction that discriminates against women.

“Why women should have to make this choice between being a mother and being a productive scholar in science and engineering is a societal question,” Shauman said. “We shouldn’t be willing to throw up our hands and say, ‘Women aren’t willing to make this choice.’”

Summers said last night he agrees that the 80-hour workweek for academic scientists is “a social convention” and “that it clearly has the effect of favoring men.” He said he hoped his comments would prompt researchers to consider ways to make academic careers more compatible with family life.

Shauman also cited a 1999 report that found elite science departments tend to marginalize female faculty members. The report, which studied MIT, was published by a committee chaired by renowned biologist Nancy Hopkins ’64, the scholar who walked out of Summers’ speech on Friday and lambasted the president’s remarks in several newspapers earlier this week.

Summers read a substantial portion of Xie and Shauman’s acclaimed 2003 book Women in Science: Career Processes and Outcomes, published by Harvard University Press, in preparation for Friday’s conference.

“For the biological interpretation [of the gender gap] to hold, it is necessary that both of the following assumptions be true,” the authors write on page 41. “[First,] the relationship between the measured aspects of brain functioning and math/science achievement is causal. [Second,] gender differences in thee aspects of brain functioning are biologically biased.”

“Neither of these two assumptions is supported by the scientific evidence,” Xie and Shauman conclude.

—Staff writer Daniel J. Hemel can be reached at hemel@fas.harvard.edu.