The National Center for Women and Information Technology publishes its annual Scorecard as one of the most systematically cited documents in the US women-in-computing policy conversation. Each edition compiles data on women's participation at multiple career stages — from computing bachelor's degrees to entry-level technical positions to senior leadership in technology companies — and places the figures in historical context. The 2015 edition is notable both for what it documents about the state of the field and for the methodological questions it surfaces about how we measure progress in women's computing participation.
The headline figure from the 2015 Scorecard that attracted the most attention was the continued decline in women's share of computing degrees relative to the peak of the early 1980s. Women earned approximately 37 percent of computer science bachelor's degrees in 1985; by the mid-2010s, that figure had fallen to approximately 18 percent. This was not a new finding — the downward trend had been documented for a decade — but the Scorecard's longitudinal presentation makes the persistence of the decline visible in ways that single-year snapshots do not.
What the 2015 Data Shows: Improvements and Stagnations
The 2015 Scorecard identified genuine improvements in some areas alongside persistent gaps in others. Women's representation in computing occupations in the US workforce had remained relatively stable in absolute numbers even as the total computing workforce grew, meaning women's share was effectively declining. In computing-intensive industries, women held approximately 25 percent of jobs, a figure that had not improved substantially in the preceding decade.
At the senior level, women held approximately 11 percent of executive-suite positions at Silicon Valley companies — a figure that the Scorecard placed in the context of women's general corporate executive representation of roughly 14 percent at that time, suggesting that the technology sector was underperforming even the not-very-impressive corporate average. The venture capital pipeline was starker still: women-founded technology companies received a small fraction of venture funding.
Against this picture, there were positive signals. Women's participation in computing education at the secondary level was showing some improvement in states that had implemented specific curricular interventions. The AP Computer Science participation data, while still male-dominated, showed some states achieving substantial increases in girls' participation through targeted programme efforts. These were early-stage signals rather than resolved trends.
Methodology and What the Scorecard May Overcount
Any careful reading of the NCWIT Scorecard requires attention to what it measures and what it does not. The Scorecard's definition of "computing" is broad and evolving: "computing occupations" includes a heterogeneous set of roles with very different technical depth, from software engineers and computer scientists to IT support technicians and data entry operators. Changes in the composition of this category over time can affect trend readings in ways that are not always explicit in the Scorecard's presentation.
The Scorecard is also primarily a US document. Its data on computing degrees draws on US federal surveys (primarily the Integrated Postsecondary Education Data System and NSF survey data), and its occupational data draws on US Bureau of Labor Statistics Current Population Survey figures. The Scorecard makes no claim to international scope, but its figures are often cited internationally as if they were representative of a broader trend. The women-in-computing situation in Germany, India, or Malaysia is not captured in NCWIT data, and the pattern of women's computing participation varies significantly across national contexts in ways the Scorecard does not illuminate.
A further methodological issue is the Scorecard's limited ability to distinguish between women who left computing voluntarily — who chose other careers that they preferred — and women who were pushed out by hostile climate, discrimination, or structural barriers. These are both real phenomena with different policy implications, but aggregate headcount data cannot separate them.
The Anita Borg Institute and Grace Hopper Context
The NCWIT Scorecard is one of two major annual data releases in the US women-in-computing landscape; the other primary institutional reference point in 2015 was the Anita Borg Institute for Women and Technology, which hosted the Grace Hopper Celebration of Women in Computing — at the time the largest technical conference focused specifically on women in computing. The 2014 Grace Hopper Celebration had drawn more than 8,000 attendees, and the 2015 conference was expected to grow further.
The relationship between NCWIT's research-and-policy work and the Anita Borg Institute's conference and community work represented two complementary modes of intervention in the same problem. NCWIT's approach was systemic and data-oriented: defining the metrics, documenting the trends, identifying the evidence-based practices, and working with companies and universities to implement them. The Anita Borg Institute's approach was community-oriented: building professional networks, providing peer support, creating a conference experience that countered the isolation many women felt in computing environments.
The CRA Taulbee Survey, conducted annually by the Computing Research Association, added a third data source specifically covering doctoral programmes and faculty at PhD-granting research universities. The 2014-15 Taulbee data showed women earning approximately 20 percent of CS doctoral degrees — a figure that had been slowly improving but remained far below women's share of other STEM doctoral fields.
What the Scorecard Does Not Measure
For an international audience, the most significant limitation of the NCWIT Scorecard as a diagnostic tool is its US-centrism. Computing industry and research participation in the US is not representative of the global picture. Several countries, including many in South and Southeast Asia, have higher women's representation in computing degree programs than the United States. Whether this translates into equivalent workplace participation is a separate question, but the assumption that the US pattern is the global pattern is not supported by available international data.
The Scorecard also focuses primarily on industry and commercial computing rather than on computing research and academic computer science. For WIGSAT's audience, the relevant question is often women's representation in computing research — as graduate students, postdoctoral researchers, and faculty — rather than in commercial software development. The data for research computing is tracked separately by CRA and NSF and tells a somewhat different story than the industry data.
Frequently Asked Questions
What is the NCWIT Scorecard?
The NCWIT Scorecard is an annual publication by the National Center for Women and Information Technology that compiles data on women's representation at multiple stages of the US computing pipeline — from education to workforce entry to senior positions — and places the figures in historical context.
What is the most significant finding from the 2015 Scorecard?
The continued decline in women's share of computing bachelor's degrees relative to the early 1980s peak is among the Scorecard's most cited findings. Women earned approximately 37 percent of CS bachelor's degrees in 1985; the 2015 figure was approximately half that, despite years of intervention programmes.
What are the main limitations of the NCWIT Scorecard as an analytical tool?
The Scorecard's primary limitations are its US scope, its broad definition of "computing occupations," and its inability to distinguish voluntary career transitions from discrimination-driven exits. It is a valuable US industry-focused benchmark, but should not be treated as a global measure or as a full account of women's computing participation in research settings.
← HeForShe and the Limits of Solidarity Campaigns | Kenya's Women-in-Tech Moment →