The discipline of computer science studies the principles, designs, applications, and impact of computers and problem-solving processes. Professional-level occupations in this field include jobs such as systems analysts, network systems and data communications analysts, and computer and information scientists. Computing professionals work in every industry, from health to entertainment, and they tend to be well-paid compared to other occupations.
This intellectually engaging discipline and its creative and rewarding occupations exhibit substantial and increasing gender imbalance in most, but not all, countries around the world. This essay illustrates cross-national variation in computing’s gender composition with snapshots of women’s representation at the typical educational entry point and in the workforce.
Gender Composition of Degrees in Computing Varied Across Countries
Women’s share of tertiary computing degrees varied across countries in 2008, as evident in Figure E-6-1, which graphs data from the Organization for Economic Cooperation and Development (OECD). These numbers are somewhat misleading, however, because they fail to account for context.
1 Joanne Cohoon, associate professor, Science, Technology, and Society Department, School of Engineering and Applied Science, University of Virginia.
2 Caroline Simard, associate director of Diversity and Leadership, School of Medicine, Stanford University.
3 Juliet Webster, director, Gender and Information Technology Program, Internet Interdisciplinary Institute (IN3), Open University of Catalunya, and director, Work and Equality Research, London.
4 Cecilia Castano, professor in Applied Economics, Complutense University Madrid.
5 Juliana Salles, program manager, External Research and Programs, Microsoft Research.
6 Jane Prey, senior research program manager, Microsoft Research Connections, Microsoft Research.
7 Jacques Wainer, associate professor in computer science, the Instituto de Computação da UNICAMP, Brazil.
FIGURE E-6-1. Women’s Share of Computing Degrees
SOURCE: Organization for Economic Cooperation and Development (OECD) Education Database.
To put computing into context, one must recognize that women’s participation in higher education also varied from country to country. In many nations, women were quite well-represented, often exceeding men’s level of participation in higher education. In countries such as Korea and Turkey, however, there were only 65 to 75 women for every 100 men earning a college degree (39 percent to 43 percent). Likewise in India, which was not included among the OECD countries listed, women were underrepresented at a comparable percent. They comprised 41 percent of the 2007 total enrollment in Indian higher education.8 In contrast, countries such as Brazil, Spain, and the United States reported overrepresentation of women in their higher education institutions. In Spain, women were 55 percent of undergraduates; in Brazil they were 56 percent; and in the U.S., they were 58 percent. 9
This country-level contextual information is crucial if we wish to avoid confounding gender balance in tertiary computing programs with issues of women’s access to higher education. To account for in-country educational conditions, we averaged women’s share of degrees for all disciplines in a country and calculated the standard deviation from that average
representation. Then we can measure the difference between women’s representation in computing compared with their representation in the average discipline. Figure E-6-2 shows for the same countries as in Figure E-6-1, the extent to which women’s representation in computing deviated below their representation in the country’s mean discipline. The longer the bar, the more computer science in that country stands out as having unusually poor representation of women.
FIGURE E-6-2. Women’s Representation Relative to Non-Computing Disciplines
SOURCE: OECD Education Database.
As evident in Figure E-6-2, computing was gender imbalanced to women’s disadvantage in every one of these OECD countries. In most countries, women’s share of tertiary computing degrees was exceptionally low. Only in Turkey was that imbalance less than one standard deviation below the mean, because 24 percent female was not so different from women’s share of the average discipline in that country. Iceland, in contrast, exhibited particularly poor female representation in computing—9 percent—compared with the typically high level of Icelandic women in other disciplines. In Estonia, where women at first seem relatively well represented in computing (more than 25 percent, as shown in Figure E-6-1), the contextual information about
women’s typical representation in other tertiary degree programs made clear that computing’s gender balance was quite unusual. Virtually no other discipline in Estonian higher education exhibits this level of female underrepresentation.
In India (not shown), some engineering colleges now reserve 30 percent of enrollment for women (Mumbai Human Development Report 2009). Prior to implementation of these affirmative action policies, women were 26 percent of India’s 2005 Engineering/Technology-B.E. Level courses (calculated with data from Selected Educational Statistics, Government of India, Ministry of Human Resource Development, Department of Higher Education, computing-specific data were not available). Compared with the OECD countries in Figure E-6-1, India exhibited one of the highest representations of women. Indian women’s low participation in higher education (38 percent of all Bachelor degrees in 2005) also means that technical degrees were not exceptionally deviant; women’s representation in technical fields like computing and engineering was just 12 points less than their overall representation.
The big picture is that women in every country investigated were underrepresented in computing, but the relative severity of their underrepresentation varied widely from country to country. A glance at the numbers of students graduated, listed at the right of Figure E-6-2, also makes clear that this variation was independent of computing’s overall enrollment.
Gender Composition in the Computing Workforce Varied across Countries
Comparative data such as that presented for higher education are less available for the computing workforce. We therefore present information from four selected countries: Brazil, India, Spain, and the United States. It is not always possible to find the same type of data for the same year to represent all four countries. For context, the available data were women’s employment numbers for the workforce overall and by occupational categories that include professional-level computing and other occupations.
The educational systems in the first three countries produce new entrants to the computing workforce who are about 20 percent women. India’s computing graduates are about 26 percent women. It seems a reasonable assumption that women’s representation in the computing workforce correlates with the supply produced by the educational system, although other factors such as trends and women’s participation in the labor force probably also exert an influence.
Women’s Workforce Participation Provides Context
As with education, countries vary in the extent to which women are in the workforce. In Brazil, 60 percent of women aged 15 and older were in the workforce in 2009 (World Bank). Brazilian women recently composed 42 percent of the overall workforce, held 43 percent of executive positions, and made up an even higher percent (67 percent) of professionals in the Arts and Sciences. Looking at just the “Professionals” sub-classification of the Brazilian workforce in 2007, women composed 59 percent (calculated with International Labor Force data). Based on these statistics, it seems that Brazilian women are well represented in occupations that call for a college education or better.
In Spain’s workforce, women were less well-represented than in Brazil. Of all women aged 15 and older in 2009, only 49 percent were in the Spanish workforce. They composed 44 percent of the overall workforce in 2009 (European Commission [EC] 2010), and 34 percent of
business leadership positions in 2007 (EC 2009). Women in Spain were 54 percent of employed persons holding a Bachelors or Masters degree in 2009.10
In the United States, 58 percent of all women aged 15 and older were in the workforce in 2009. In the same year, women composed 47 percent of the entire U.S. workforce and 57 percent of the “Professional and related occupations.”11 Although substantial horizontal segregation persists, many U.S. women participate at high levels in certain occupations.
In contrast, women’s representation in the overall Indian workforce was much lower than in the three Western countries we considered. Only 33 percent of women aged 15 and older were in the workforce in 2009.12 (They composed 31 percent of the entire Indian workforce in 2008, and 32 percent of the “Professional, technical & related workers” in 2005.13 Another source puts women at 35 percent of the urban population holding “diploma/certificate” degrees in 2007 (National Sample Survey Organization 2009).
Women’s Participation in the Computing Workforce
Despite women’s high level of participation in many areas of Brazil’s, Spain’s, and the U.S.’s workforces, they comprised low and often declining shares of computing jobs that called for a college-level education. As shown in Figure E-6-3, Brazilian women comprised only 20 percent of the computing professional-level workforce. In Spain, women comprised 20 percent of computing professionals in 2009.14 In the U.S., women composed 24 percent of the computing workforce in 2009.15
In India, women’s level of participation in computing is disputed. Reports put women’s representation in the software industry at 36 percent women in 2008 (Nascom & Mercer 2009), but these numbers may include women not working in technical positions. A survey of software professionals in two large Bangalore tech firms found that women comprised 18 percent of those working as developers, module leaders, project leaders, or project managers in 2003 (Ilavarasan 2007). Because this finding is comparable to a previous industry report of women as 21 percent of the software workforce in 2001 (Noronha & D’Cruz 2006), it seems reasonable to estimate Indian women’s level of participation in professional level computing occupations at about 20 percent.
11 Calculated based on data retrieved from Current Population Survey, Bureau of Labor Statistics, 2010.
14 Calculated based on data retrieved from European Union Labor Force Survey (EU LFS) 2010. Available: http://epp.eurostat.ec.europa.eu/portal/page/portal/microdata/lfs.
15 Calculated based on data retrieved from Current Population Survey, Bureau of Labor Statistics, 2009.
FIGURE E-6-3. Country Comparisons of Employed Women
SOURCES: Calculated based on data retrieved from European Union Labor Force Survey 2010, the current Population Survey, Bureau of Labor Statistics, 2009, and Nasscom & Mercer. 2009. Gender Inclusivity in India: Building Empowered Organizations. New Delhi, India.
Figure E-6-3 makes clear that among the selected countries, there was little variation in women’s representation in the computing workforce. The range for each country in the latter half of the first decade in this century was 20 to 24 percent. This workforce representation may parallel women’s share of computing degrees in each country. Another interesting observation is that, like the situation in higher education, Indian women are relatively better represented in computing in the context of their low participation in the professional workforce. India exhibits a 15 point gap between women in the professional workforce and women in computing, compared with a 39 point gap for Brazil, a 34 point gap for Spain, and a 33 point gap for the United States. So, while women are far from well-represented in Indian computing occupations, computing jobs do not stand out as especially deviant in the context of their workforce.
Possible Reasons for Cross-National Variation in the Gender Composition of Computing
Cultural beliefs about gender and about the nature of computing occupations can influence the gender composition of a field if social structures facilitate indulging those stereotypes. For that reason, women’s participation in computing tends to be low in societies where computing is viewed as particularly suitable for men, where the culture and economic conditions encourage career choice as a form of self-expression, and where the educational system offers early opportunities for opting out of math and science (Charles and Bradley 2006). When choices are more constrained and careers are chosen pragmatically, there may be more women in computing. India may be an example of the latter circumstances—women’s relative economic participation and opportunities are particularly low. India was ranked 127th out of 134 countries, while Brazil was 75th, Spain was 90th, and the United States was 17th (Hausmann, R., Tyson, L. Zahidi, S. 2009). Indian women have few other good options, so they are less likely to “indulge their gendered selves” (Charles and Bradley 2009).
Charles, M., and K. Bradley. 2009. “Indulging Our Gendered Selves? Sex Segregation by Field of Study in 44 Countries.” American Journal of Sociology. 114(4):924-976.
Charles, M., and K. Bradley. 2006. “A Matter of Degrees: Female Underrepresentation in Computer Science Programs Cross-Nationally.” Women and Information Technology: Research on Underrepresentation. J.M. Cohoon and W. Aspray, eds. Cambridge, MA: MIT Press.
European Commission (EC). 2010. Employment in Europe 2010. Luxembourg: Publications Office of the European Union.
EC. 2009. Report on Equality between Women and Men. Luxembourg: Office for Official Publications of the European Communities.
Hausmann, R., L. Tyson, and S. Zahidi. 2009. Global Gender Gap Report. World Economic Forum.
Ilavarasan, V. 2007. “Is Indian Software Workforce a Case of Uneven and Combined Development?” Equal Opportunities International. 26(8):802-822.
Mumbai Human Development Report. 2009. New Delhi: Oxford University Press.
Nasscom & Mercer. 2009. Gender Inclusivity in India: Building Empowered Organizations. New Delhi, India.
Noronha E. & D’Cruz P. 2006. Organising Call Center Agents: Emerging Issues. Economic and Political Weekly. Mumbai, India. May 27, 2006. pp 2115-2121.
National Sample Survey Organization. 2009. Report No 531: Employment and Unemployment Situation in India: July -2007-June, 2008.
Rustagi, P. 2010. International Labour Organization (ILO). Asia-Pacific Working Paper Series: Employment Trends for Women in India. ILO Subregional Office for South Asia, New Delhi.