Why Aren't More Girls 'Geeks'?

Coauthored by Maria Charles, professor of sociology at the Universityof California, San Diego, with Karen Bradley of Western WashingtonUniversity, the study was presented Aug. 13 at the 100th annual meetingof the American Sociological Association in Philadelphia.

"Restrictive government practices that minimize choice and prioritizemerit may actually result in more gender-neutral distribution acrossfields of study," the researchers write.

Charles and Bradley analyzed data compiled in 2004 by the Organizationfor Economic Cooperation and Development (OECD) on higher educationdegrees awarded in 2001. Examining seven fields of study, includingengineering and math/physical sciences, Charles and Bradley calculatedrepresentation factors for each country by comparing male-to-femaleratios in a program to those same ratios in other academic programs(which controls for international differences in women's enrollments).

They found, as expected, that on the whole women predominate in suchtraditionally female-typed fields as education and health and lagbehind in stereotypically masculine fields. In computer science,females are underrepresented in all 21 of the industrialized countriesconsidered.

They also found that extent of the difference in male-to-female ratiosvaries a great deal. Males are overrepresented among computer sciencegraduates by a factor of 1.79 in Turkey, on the low end, to a factor of6.42 in the Czech Republic, on the high. That is, maleoverrepresentation in computer science in the Czech Republic is morethan three times more extreme than in Turkey.

In the United States, the "male overrepresentation factor" is 2.10 andin the United Kingdom, 3.10. (See figure for all 21 nations.)

"The ubiquity of women's underrepresentation attests to the persistenceof deep-seated and widely shared beliefs that men and women arenaturally different and that they are suited for differentoccupations," Charles said. "But the fact that there's so muchcross-national variability suggests there's lots of room forcountry-specific cultural and social influences to play out."

There is little evidence, though, Charles said, for standard argumentsof social evolution: The most economically developed countries do notproduce the greatest numbers of women in computer science. Nor is therea strong correlation with more women in the workforce or in high-statusjobs or in higher education generally.

Girls' higher math achievement does not equate with better representation in stereotypically male fields, the researchers find.

Broad cultural support for equal opportunity is also not a goodpredictor. None of the study's highest-scoring nations -- Turkey, SouthKorea and Ireland -- the authors note, is particularly known forgender-egalitarian attitudes or practices.

"The principle of being free to pursue your preferences is compatibleand coexists quite comfortably with a belief in essential genderdifferences. This essentialist notion, which helps to create what itseeks to explain, affects girls' views of what they're good at and canshape what they like," said Charles, who is also co-author of theaward-winning book Occupational Ghettos: The Worldwide Segregation ofMen and Women (Stanford UP, 2004; in paperback Aug. 2005).

In response to women's perceived interests and preferences, Charlessaid, modern societies diversify their educational systems and set upniche programs targeted to women -- vocational degrees intourism-hospitality, for instance, or early childhood education.Service-dominated economies, another feature of industrializeddemocracies, likewise result in more "pink-collar ghettos."

"There is no doubt that collective beliefs holding that men arenaturally 'better' at math and science are major factors that influencewomen's choices of college majors -- and determine the climate in mathand science programs worldwide," Charles said.

"When we emphasize choice and hold up self-realization as aneducational goal, girls will often freely choose poorly paid,female-typed fields of study that are in line with a conventionalfeminine identity and stereotypes about what girls are good at," shesaid, adding that such tendencies appear to be especially pervasive inthe most affluent, industrialized societies.

What countries with the best female representation in computer scienceseem to have in common, Charles and Bradley observe, are governmentsthat "exert strong control over curricular trajectories" and requiresubstantial math and science coursework. In South Korea, for example,math has to be studied through grade 12 and science through grade 11.The Irish require math and science throughout secondary school.State-mandated exams then control entry to universities and specificacademic programs.

The policy implications, Charles said, are clear: "Rather than lettingpeople take what they expect to love (or expect to be good at),educational systems should insist on more math and science for allstudents. As other research has repeatedly shown, choices made duringadolescence are more likely to be made on the basis of genderstereotypes, so we should push off choice until later."

The research is supported by grants from the National ScienceFoundation, the Spencer Foundation and the American EducationalResearch Association. It will be published as a chapter in aforthcoming book from MIT Press, Women and Information Technology:Research on the Reasons for Underrepresentation.

Figure:
Male "Overrepresentation Factor": Computer Science Programs, 2001
Values give the factor by which men are overrepresented in computer science programs
in their respective country, relative to their representation in the other fields of study.