In a recent Chronicle of Higher Education "Career Network" piece (Woolston, 2001), a freelance science and medical writer wrote that the Ph.D. pipeline in the sciences is "equipped with a powerful female filtration system." The second in a series of four articles about women scientists in academia, this paper explores the extent of female filtration from U.S. academic science.

For many years, academicians blamed a nearly empty "pipeline" for the low numbers of women seeking faculty positions in U.S. universities and for the fact that women hold far fewer tenured full professorships than men at U.S. universities, particularly in the sciences. But as the flow of women into the pipeline has gradually increased, it has become clear that the supply of women entering the pipeline isn't the problem--that is, it isn't the only problem. The data reveal that the pipeline is leaky, hemorrhaging women from the time they enter as college undergraduates until the time a select few attain full professorship.

Attrition 101: Women as Undergraduates

Young women hold their own in high school in science, math, and technology, achieving higher grades than men but scoring lower on math SATs. But attrition begins soon after, at the undergraduate level, though the amount of attrition varies widely among the science and engineering (S&E) disciplines (Thom, 2001). According to the National Science Foundation (NSF) (1999, 2000), women reached parity with men in biological and agricultural sciences in 1996, earning 50% of the undergraduate degrees. That's a big improvement over 1975, when only 29% of degree recipients in these fields were women (NSF, 1999). Women are also approaching parity at the baccalaureate level in math, statistics, and chemistry. But women continue to trail in other sciences, earning only 19% of bachelor's degrees in physics and 18% in engineering.

The discrepancy in those fields arises in part because 70% of women abandon science before obtaining a baccalaureate degree (compared to 63% for men). Elite schools, both liberal arts colleges and research universities, do a better job overall of retaining undergraduates in the sciences: At these institutions 54% of women but only 39 percent of men drop out of science before attaining a degree.

Attrition 202: Women in Graduate School

Both the number and the proportion of women earning master's and doctoral degrees have risen steadily over the past 30 years; nevertheless, the discrepancy between women and men remains much larger than at the baccalaureate level. In 1966, women earned 5,469, or 13%, of all S&E master's degrees. In 1996 they earned 37,453, or 39%, of S&E masters degrees. Women's progress--and lingering disparities--are similar at the doctoral level. Women earned 924, or 8%, of S&E doctorates in 1966. In 1999 they earned 9,084, or 35% of S&E Ph.D.s. In the most recent data, from 2000, women earned 41% of Ph.D.s in biology and agricultural science, 26% in math, and 23% in the physical sciences (NSF, 2001).

It isn't known how many women leave science immediately after receiving their bachelor's degree and how many go on to graduate school but drop out before receiving a graduate degree. But the available data do allow us to calculate the proportion of bachelor's-degree recipients who go on to receive an advanced degree. The attrition of women and men from academic science programs is broken down by gender in the table below.

Attrition Rates for Women and Men in Science, Engineering & Technology

Women MenPre-bac. 70% 63%Bac. -M.S . 79% 71%M.S.-Ph.D. 77% 70%Bac.-Ph.D. 95% 86%

The data also tell a story about variables that influence the success of women who earn doctorates in S&E: they are more likely to have attended a liberal arts college rather than a university as undergraduates. Seventy-five percent of African-American women who earned doctorates in biology between 1975 and 1992 attended historically black colleges, most notably Spelman and Bennett--both women's colleges (Thom, 2001).

Postgraduate Attrition: The Postdoc Period

At the post-doctoral level, men outnumber women more than four to one in engineering and by almost as much in the physical and mathematical sciences. The number of female postdocs lags in the life sciences, but they outnumber men in the social and behavioral sciences.

Unlike most younger scientific trainees, many postdocs earn salaries, and salary comparisons provide another basis for a quantitative comparison of the situations of men and women who remain in academia. Even in the social and behavioral sciences, where they outnumber men, women earn $3000 less each year, on average. In engineering, men earn 20% more than women. Under these circumstances, it isn't surprising that some women are reluctant to stick it out, especially considering the personal sacrifices academic career women are often called upon to make.

Faculty

At the faculty level, the situation is more complicated, since faculty positions are stratified according to prestige. And the more prestigious and research-oriented a position is, the less likely it is that a woman will fill it.

Put another way: On faculties as on student bodies, the filtration of women out of academia is more effective the higher up you go. Women are relatively well represented on teaching-oriented faculties. In 1997, women represented 39% (3200) of the 8300 S&E faculty members with doctorates at 2-year colleges (where, incidentally, they are often adjuncts), but only 21% (30,300) of the 142,800 faculty members with doctorates at 4-year institutions; at research universities in 1995 women held only 19% (19,200) of 101,300 faculty positions among all scientists with doctorates (NSF, 2000).

Over a 30-year period, the percentages of ranked women S&E faculty members with doctorates at 4-year colleges and universities has grown considerably decade by decade, but are still not nearly what the percentages are for men. In 1997, among ranked S&E faculty with doctorates at 4-year colleges and universities, females composed 30% (13,100) of the 43,400 faculty members who received their doctorates between 1987 and 1997, 22% (9,200) of the 42,600 faculty for whom their doctorate is 10 to 19 years old, 12% (4900) of the 41,700 faculty who are 20 to 29 years out, and 5% (700) of the 15,500 faculty who are 30 years or more out of their doctoral programs. On the one hand, this suggests that progress is being made; on the other, it suggests that women are being removed selectively from the academic mixture.

Women S&E faculty members are far more likely than men to teach part time (40% versus 25% for men), and are also more likely to have fixed-term contracts--54% of women S&E faculty members were on a 1-term or 1-year contract, compared to 34% of men, in 1993 (NSF, 1996). Female faculty members are more likely to report spending more time on teaching and less time on research than male faculty members. This is partly a result of the fact that, on average, men have attained a higher academic degree (42% of female and 24% of male faculty members have a master's degree as their highest) and partly a result of where they work (more women work at 2-year institutions and colleges, where teaching is the primary mission).

And then there is tenure: Using 1997 as a base year, Tables 1 and 2 (see sidebar) show that women are less likely than men to hold any academic rank (among all ranked faculty, 19% are women), to be full professors (among all full professors, 10% are women), and to hold a tenured position (among all tenured faculty, 14% are women). Women are more likely than men to be working on a non-tenure-track line (16% of women versus 9% of men) or to not have tenure available to them (25% of women versus 16% of men).



Finally, a higher percentage of women than men is currently on a tenure track, but not yet tenured (24% versus 15%). While this might be seen as good news, in that it could suggest that these women are likely to advance to tenure and higher rank, this isn't the whole story. Women move through the faculty ranks more slowly than men and, even when productivity is controlled for, achieve tenure more slowly than men do. As a consequence, the assistant professor rank includes not only young women but also older women who should already have been promoted to the next rank (Valian, 2000). The main problem is not the supply of women into the bottom end of the science pipeline; roughly the same number of women as men enter college intending to study science. The problem, rather, is that the pipeline leaks, and it leaks women preferentially. Science's gender-filtration mechanisms kick in as soon as a potential young scientist enters college and continue at every level. The next article in this series will explore the mechanisms of gender filtration and consider what can be done to remove or neutralize them.

References

National Academy of Sciences, "Enhancing the Postdoctoral Experience for Scientists and Engineers," (Washington, D.C., National Academy Press, 2000).

National Science Foundation, Science and Engineering Doctorate Awards: 1999. Arlington, VA (NSF 01-314), January 2001.

National Science Foundation, Women, Minorities, and Persons with Disabilities in Science and Engineering: 2000. Arlington, VA (NSF 00-327), September 2000.

National Science Foundation, Women, Minorities, and Persons with Disabilities in Science and Engineering: 1998. Arlington, VA (NSF 99-338), 1999.

National Science Foundation, Women, Minorities, and Persons with Disabilities in Science and Engineering: 1996. Arlington, VA (NSF 96-311), 1996.

Mary Thom, Balancing the Equation: Where Are the Girls in Science, Engineering, and Technology? (New York: The National Conference for Research on Women, 2001).

Virginia Valian, Why So Slow? (Cambridge, MA: The MIT Press, 2000).

Chris Woolsten, "The Gender Gap in Science," Career Network, online Chronicle of Higher Education, 22 October 2001.

Cathy Ann Trower, Ph.D., is the research director of Collaborative on Academic Careers in Higher Education (COACHE) at the Harvard Graduate School of Education.