Barnard Center for Research on Women Advance at the Earth Institute at Columbia University
December 9-10, 2004
Women, Work and the Academy: Strategies for Responding to 'Post-Civil Rights Era' Gender Discrimination
Contact Conference Description Program Video Executive Summaries About the Participants

Women, Work and the Academy > Executive Summaries > Ellen Spertus

What We Can Learn from Computer Science's Differences from other Sciences

Ellen Spertus

Like many fields with "science" in their name, computer science isn't really a science. In many ways, it has more in common with engineering. Both qualitatively and quantitatively, it is different from any other science or engineering field. In most of the disciplines discussed at the workshop, the percentage of female students has been increasing, while the percentage of female faculty has not. In computer science, the reverse is true. This paper reviews the data and suggests some possible theories.

Decline of women earning bachelor's degrees

As Figure 1 shows, the percentage of bachelor's degrees in computer science going to women has been consistently lower than in other scientific fields and higher than in engineering [NSF 2004]. The peak occurred in 1984, when 37% of computer science bachelor's degrees went to women. While other fields of science and engineering have increased or at least maintained their female bachelor's degree production since 1984, computer science showed a marked decline. In 2000 (the most recent year for which data is available), 28% of bachelor's degrees in computer science went to women, a decline of nearly one-third from the 1984 peak. Women's under-representation is even greater at top engineering schools, such as MIT, where computer science had the smallest proportion of women for any undergraduate program, including those in engineering [MIT 1995].

figure 1

There are many theories as to why women's participation in computer science has declined since 1984 [Gürer & Camp 2002]. One theory is the rise of the personal computer. Before 1980, few people had access to computers at home or in K-12 schools. Thus, young men and women were equally inexperienced with computers when they entered college. Since 1980, computers have become common in (some) schools and homes, where they are disproportionately used by boys, giving them an advantage in introductory college computer science courses [Gürer & Camp 2002, Margolis & Fisher 2001].

Another factor is "The College of Engineering effect", in which the percentage of female students decreased as computer science departments were moved from science to engineering divisions [Camp 1997], consistent with Figure 1 above. Similarly, in the public's mind, computer science, more than any other academic discipline, became associated with antisocial male misfits [Gürer & Camp 2002, Margolis & Fisher 2001].

Increase in female PhDs and faculty

Just as the "pipeline" metaphor does not describe women's participation in the other sciences (with increasing degrees failing to lead to increasing faculty), it fails in computer science, although in a happier way. While women's undergraduate enrollment has decreased or stagnated, the percentage of women earning PhDs and joining the faculties of research universities has increased.

Figure 2 [CRA 2005] shows the percentage of computer science degrees awarded to women at the bachelor's, master's, and PhD levels. For decades, the higher the degree, the fewer women; however, since 1997, a greater proportion of master's degrees than bachelor's degrees were awarded to women. Observe too that PhD representation has been rising consistently, if unevenly, over time.

figure 2

Figure 3 shows the increase in tenured and tenure-track faculty in PhD-granting computer science and computer engineering departments in the United States and Canada.

figure 3

Why did women make gains at the faculty levels after and while losing ground at the undergraduate levels? In the following section, I offer some preliminary speculations.

The "Tech Boom" hypothesis

There was a peak in PhD (Figure 2) and faculty (Figure 3) representation during the mid-nineties, when the Internet boom took place, during which some graduate students and faculty temporarily or permanently left academia for industry. Perhaps male computer scientists were more likely than their female peers to seek greener pastures. On a personal note, I completed my dissertation on Internet search during this period. I saw many of my male peers work at or found start-ups, either instead of or while completing graduate school. (I did not have enough female peers to be able to comment on them.) When I looked for a faculty position in 1997, there were many more jobs available than people to fill them, both because departments were growing and because existing and potential faculty were choosing industry instead.

The improved academic environment hypothesis

During the same period that female undergraduate enrollment was increasing, female computer scientists and their male allies were aggressive in improving the environment for female graduate students and faculty. The watershed event occurred in the early 1980s, when female MIT graduate students and research staff got together, shared their experiences, and wrote a highly influential report, entitled "Barriers to Equality in Academia: Women in Computer Science at MIT" [MIT 1983], which is still being discussed more than twenty years later (see, for example, http://www.cs.washington.edu/homes/lazowska/mit/). On another personal note, this report, shown to me when I started doing computer science research at MIT, was a major factor in improving my view of women as computer scientists, as it demonstrated that their unequal representation did not arise on a level playing field, as I had presumed.

Another watershed event was Anita Borg's founding of the "Systers" email list [Systers 1993]. It began in the women's bathroom of an operating-systems conference in 1987 where the few female attendees met and recognized the benefits of forming an online community, and now has more than 2000 members. Otherwise isolated women could get practical information, such as how to manage a "two-body" job search (in which a woman and her partner were both seeking academic jobs) or issues in the timing of motherhood. I found Systers valuable not just for practical information (such as what to wear when presenting a paper at my first conference, a topic on which my otherwise excellent male advisor was useless), but for changing my perception of female computer scientists (and myself) from anomalous freak to abundant community. It also greatly influenced a report I wrote, entitled "Why Are There So Few Female Computer Scientists" [Spertus 1991].

Another major player has been The Computing Research Association (http://www.cra.org) and its Committee on the Status of Women in Computing Research (http://www.cra.org/Activities/craw/index.php), which have led many excellent programs, such as the Distributed Mentor Program, which provides undergraduate women with summer research opportunities helpful for preparing themselves for graduate school, and offers highly effective career workshops to graduate students and faculty. (The Computing Research Association is the source for much of the data [CRA 2004, CRA 2005] and one of the papers [Borg 1993] I cite.)

Conclusions

Just as increasing female undergraduate degree production in other fields does not guarantee increases in the professoriate, declining undergraduate enrollment does not preclude growth at the PhD and faculty levels. While computer science continues to struggle with its failure at attracting and retaining undergraduate women, its successes at later stages should be understood and replicated in other scientific and engineering fields where possible.

References

[Borg 1993]

Borg, Anita. Why Systers? Computing Research News, September 1993. Available online in its original form at http://www.cra.org/CRN/issues/9304.pdf and updated at http://athena.systers.org/about.html.

[Camp 1987]

Camp, Tracy. The Incredible Shrinking Pipeline. Communications of the ACM, vol. 40, no. 10, pp. 103-110, Oct. 1997. An expanded version is available online at http://www.mines.edu/fs_home/tcamp/cacm/paper.html.

[CRA 2004]

Computing Research Association. CRA Taulbee Trends: Women Students & Faculty, May 6, 2004. Available online at http://www.cra.org/info/taulbee/women.html.

[CRA 2005]

Computing Research Association. Women Among All Degrees Granted, 2005. Available online at http://www.cra.org/info/education/us/women.html.

[Gürer and Camp 2002]

Gürer, Denise and Camp, Tracy. Investigating the Incredible Shrinking Pipeline for Women in Computer Science: Final Report - NSF Project 9812016, 2002. Available online at http://women.acm.org/documents/finalreport.pdf.

[Margolis and Fisher 2001]

Margolis, Jane and Fisher, Allan. Unlocking the Clubhouse: Women in Computing. MIT Press, Cambridge, MA, 2001.

[MIT 1983]

Female graduate students and research staff in the Laboratory for Computer Science and the Artificial Intelligence Laboratory at MIT. Barriers to Equality in Academia: Women in Computer Science at MIT, February 1983. Available online at http://www.cs.washington.edu/homes/lazowska/mit/Contents.pdf

[MIT 1995]

MIT EECS Ad Hoc Committee on Women Undergraduate Enrollment. Women Undergraduate Enrollment in Electrical Engineering and Computer Science at MIT, February 17, 1995. Available online at http://www-swiss.ai.mit.edu/~hal/women-enrollment-comm/final-report.html.

[NSF 2004]

Science & Engineering Indicators - 2004 (NSB 04-1), National Science Foundation, May 2004. Derived from data in Appendix Table 2-22. Available online at http://nsf.gov/sbe/srs/seind04/start.htm.

[Spertus 1991]

Spertus, Ellen. Why Are There So Few Female Computer Scientists? MIT Artificial Intelligence Lab Technical Report 1315, August 1991. Available online at http://www.mills.edu/ACAD_INFO/MCS/SPERTUS/Gender/why.html.

Back to top

© 2004-2005 Barnard Center for Research on Women