Nonprofit, independent journalism. Supported by readers.


This doesn’t compute: As more women enter scientific fields, their numbers in computer science are declining

While the number of women in other fields of science and technology have increased over the past 40 years, the computer science area has seen a dramatic decline that has no easy explanation.

An early marketing campaign targeting women as computer users.
Courtesy of the Babbage Institute
An early marketing campaign targeting women as computer users.

Walk the halls of the computer science buildings on college campuses across the United States and you’ll notice a peculiar thing: there are very few women. At a time when women are swelling enrollments in many other university departments, computer science is conspicuous for its lack of female students.

Worse, percentages of female bachelor degrees earned in computer science are falling-down to 25 percent in 2004, the latest available figures, from a high of 37 percent in 1984. And all this is occurring at a time when National Science Foundation (NSF) funding to encourage women in the computer sciences — about $20 million annually — has never been higher.

Comparisons with other disciplines bring the trend in computer science into stark relief: Nearly all other scientific fields have seen a marked increase over the past 40 years in the percentage of bachelor degrees awarded to women, according to NSF data. For example, in 2004 the percentage of bachelor degrees awarded to women was about 45 percent in math; 42 percent in physical sciences, as well as in earth, atmospheric and oceanographic sciences. In all sciences combined, women earned 58 percent of bachelor degrees in 2004, compared to 43 percent in 1966.

Only computer science shows a decline-and a steep one at that. What’s behind this mystery in a field fundamental to advances in science, technology and medicine?

Cracking the case of women gone missing
An international group of computer science educators and historians convened in late May in Minneapolis to systematically try to answer the question. Sponsored by the Charles Babbage Institute, a research center for the history of information technology, the conference “History, Gender and Computing” looked to the historical record for insight.

“We think there’s a crucial ‘missing piece’ when lots of smart people, backed by handsome NSF funding, are conducting savvy projects to increase women’s participation in computing — but then the numbers go in the wrong direction,” said Thomas Misa,  historian of technology who directs the Babbage Institute. 

“From the peak of nearly two women for every five computer-science graduates in the mid-1980s, it’s slid — just like water draining out of a tub, across two decades — to the most recent number of around one woman in five. Both the accelerated rise of women into computing in the 1970s, and the disastrous drop since the mid-1980s call out for historical investigation,” Misa said. “Our hunch is that young people need better, more positive images of computing — doing real things in the world, not just solving logic problems. Also, we are very concerned that if these  measurable drops are occurring for women, it’s possible that other, less visible groups are not entering the field either.”

Hara Konsta, a Ph.D. student from the University of Athens, Greece, talked about the not-so-subtle images of women as lesser workers in advertisements used to promote computers in the workplace. She presented a visual record that typically placed women in passive positions, seated and doing clerical work with mice and keyboards. By contrast, men assumed iconic positions of dominance, standing above the seated women, pointing in instructional, supervisory gestures.

In another visual division of labor, Konsta offered evidence of women typically shown doing the manual labor of printing while men do the actual creation. Computer printers and women seem to have a curious connection. Printers are advertised with photographs of attractive and scantily clad women, some even being printed out in beautiful colors from the printers. Konsta suggested a social reason rooted in power relationships behind this visual conceit: it serves to keep women at the periphery of the computing industry by limiting them to peripheral devices. The heart and power of the computer — the CPU or the hard drive — is usually portrayed as the male domain.

“[The] gendered introductions of computing in popular and technical publications and the development of computing as a gendered technology have fed back on each other,” explained Konsta. “There is a crucial interaction between the growth of computing and the increased sophistication of…gendered computing advertisements.”

Hilde Corneliussen, an expert in humanistic informatics at the University of Bergen in Norway, also examined disparate cultural messages in computing: one communicates competence with computers to men, and the other communicates incompetence with computers to women. In the cultural discourse surrounding computer science, Corneliussen noted a curious pattern of visibility and invisibility.

“Images and discourse showed males as ‘self-educated wizards’ while women were portrayed as disinterested or frustrated by computers,” said Corneliussen: “For non-using males, computer literacy was portrayed as if it were a choice: ‘I could learn it if I want but I don’t need to/choose not to.”’ By contrast, for non-using females, computer literacy was portrayed as being linked to natural disinterest or even incompetence in computer science.

Clues to investigate
In addition to mass media visual images that communicate a male role for computer science fluency, other leads surfaced at the conference that might explain why the percentage of women with computer science bachelors’ degrees dropped precipitously after 1984. They include:

• Emergence of the male nerd-hacker stereotype in mass media. The popular stereotype of the male nerd/hacker computer scientist emerged in the 1970s and was the centerpiece of “Revenge of the Nerds,” the comedy film released in 1984 — the very time when women’s enrollment in computer sciences bachelor’s degree peaked. Prior to this time, computer science and programming were relatively new professions. “Given the lack of widespread and established stereotypes in the early days, a woman contemplating a BS may have seen few reasons not to be a computer scientist,” explained the University of Minnesota mechanical engineering professor Caroline Hayes. In fact, during the late ’70s and early ’80s, the representation of women grew faster in computer science degrees than in nearly any other discipline.

• Rise of personal computers in homes. The colonization of personal space by personal computers occurred in the early 1980s. The explosion of PC use resulted in larger numbers of students arriving at college with programming experience — and some college programs adding entry requirements of programming experience. Data show that college-age men tend to have more prior programming experience than do women. This may fuel sociological developments, such as the creation of a male-dominated “computer clubhouse.” Girls may not feel welcome or encouraged in programming classes; or feel comfortable in the informal hang-out groups common to computer labs at universities. A male-orientation toward the computer may be established early in the home when male siblings “hog” the computer.

Growing pains
Hayes noted that there has been an explosion in employment opportunities for computer science graduates over the past 35 years, so it isn’t opportunity that is keeping women out of the field.

“In 1974 it [computer science] was an esoteric field graduating less than 5,000 Bachelors of Science. Thirty-plus years later it has become a popular profession having international economic importance, graduating almost 60,000 bachelors in 2004,” Hayes said. “However, computer science has experienced growing pains, particularly with respect to the representation of women.”

As participants in a global culture of science, we need to better understand women’s declining participation in computer science, and work to reverse the trend, Hayes and Misa said, because  women are missing a significant opportunity to make creative contributions to society. “Mastering computer science is good for women, good for culture, good for science, good for families,” Hayes concluded.

Anne Brataas, a former reporter for the Star Tribune and Pioneer Press, is president of The Story Laboratory, a St. Paul group that specializes in the coverage of science. She wrote this article for the Inside Science News Service, which is supported by the American Institute of Physics, a not-for-profit publisher of scientific journals.