Reshma Saujani, a lawyer and former congressional candidate from New York, has launched a programme this summer to teach computer science to girls from underserved communities in New York City. The programme, Girls Who Code, has attracted attention from technology journalists and from the women-in-STEM advocacy community in its first weeks of operation. Now that it has begun, it is worth examining what the programme's design promises and — from the perspective of those of us who think about girls and women in STEM globally — what it does not.
The curriculum as described focuses on high-school-age girls in New York City, with an intensive summer immersion structure. Saujani's public statements emphasise access: the programme is explicitly targeted at girls from communities that do not have strong computer science education in their schools, aiming to introduce programming, robotics, and web development in an environment that counteracts the cultural messaging that positions computing as a masculine domain. This is a coherent educational theory and there are solid reasons to think it could be effective for the students it reaches.
The Curriculum Design: What We Know
The programme as described incorporates project-based learning — students work on applications and websites rather than completing abstract programming exercises — and includes exposure to women working in technology as mentors and role models. The academic literature on girls' engagement in computing consistently finds that project-based and socially-relevant framings improve girls' persistence relative to more abstract, technique-focused approaches; to the extent the Girls Who Code curriculum honours this, it reflects current best evidence.
The mentorship component is more difficult to evaluate in advance. The research on near-peer mentorship in computing education is positive — women slightly older than participants who have navigated similar educational pathways are effective mentors in ways that more distant professional role models are not. Whether Girls Who Code can recruit mentors at the scale needed to maintain quality in an intensive summer programme is an operational question that will be answered in practice rather than in advance.
The summer immersion format — which concentrates learning over a short, intensive period — is one model among several. It has the advantage of building cohort relationships that can persist beyond the programme, which is important for girls re-entering school environments where computing is not culturally valued. It has the disadvantage of limited exposure time; what is learned in a summer programme cannot substitute for sustained, high-quality CS education integrated into the regular school curriculum. Girls Who Code, as described, appears to understand this: Saujani has framed it as a pilot for a model that could be embedded more broadly, not as a complete solution.
Who Is Being Served — and Who Is Not
The most important honesty about Girls Who Code, as designed, is who it is for. The programme is for girls in underserved communities in New York City. This is a specific and well-defined population with specific needs, and there is nothing wrong with designing for that population precisely. It is, however, important for the women-in-STEM community globally not to mistake a New York City summer programme for an international solution, or even a national one.
The majority of girls who face barriers to computing education globally are not in New York City. They are in secondary schools in rural India where the computer laboratory has six computers shared among three hundred students. They are in Nigerian universities where the curriculum has not been updated in a decade and the faculty-to-student ratio makes individual attention impossible. They are in Central American schools where the teachers have not had meaningful professional development in information technology. Girls Who Code, whatever it achieves, is not designed to address these situations, and it would be a mistake to treat it as if it were.
This is not a criticism of Girls Who Code. An organisation designed for the specific context of US urban girls from underserved communities should be evaluated on how well it serves that population. It is a criticism of the tendency in international women-in-STEM discourse to treat US models as universal and to export US metrics as global benchmarks.
The International Context the Programme Omits
WIGSAT's perspective on women in global science and technology means that any assessment of a programme like Girls Who Code must ask about transferability and about what the international equivalent would look like. The curriculum model — intensive, project-based, mentorship-supported — is in principle exportable, but the specific conditions that make it viable in New York City are not uniform across contexts. Recruiting corporate technology mentors works in a city with a major technology industry; it is less straightforward in cities where that industry is less developed. Sourcing funding from US technology companies is also a model that does not transfer without adaptation.
The international women-in-computing landscape has its own infrastructure — organisations like the Association for Computing Machinery's Committee on Women in Computing (ACM-W), the Asian Development Bank's gender and ICT programmes, and the various regional chapters of IEEE Women in Engineering — that have been working on variants of these questions longer and in more diverse contexts. How Girls Who Code will relate to that international infrastructure, if at all, is not clear from the programme's description. It is worth watching.
A Note on What This Is Measuring
The outcome that Girls Who Code, as described, is designed to produce is increased interest in computing among high-school-age girls from underserved New York communities. That is a legitimate and worthwhile outcome. It is not the same outcome as increased numbers of women in computing research, or increased women's representation in doctoral computer science programs, or increased numbers of women computing faculty at research universities — all of which are also important.
A programme that successfully introduces fifteen-year-olds to programming may or may not translate into any of those later outcomes, depending on what happens in the intervening years: the quality of CS education available in their high schools, the availability of undergraduate programmes they can access and afford, the climate they encounter in those programmes, the availability of mentorship at each stage. The "pipeline" from a summer programme to a research career is long, and many things can happen in it. Girls Who Code deserves credit for addressing the entry end of that pipeline; other interventions will be needed to address the rest.
Frequently Asked Questions
What is Girls Who Code?
Girls Who Code is a programme founded by Reshma Saujani, initially as a summer immersion in New York City targeting high-school-age girls from underserved communities, with a curriculum emphasising programming, robotics, and web development in a project-based, mentorship-supported environment.
What does a critical, international-STEM-access perspective find lacking in the Girls Who Code model?
The programme is explicitly designed for a specific US context and population; it is not designed to address the computing education needs of the majority of girls globally who face barriers to CS. From an international perspective, the model's dependence on US corporate technology partners and US urban contexts limits its direct transferability, though the pedagogical principles (project-based learning, near-peer mentorship) are broadly applicable.
What outcome metrics should evaluate a programme like Girls Who Code?
Metrics should match the programme's design goals: increased interest in and engagement with computing among participating girls, measured over time, including persistence into high-school CS courses and post-secondary CS study. Pipeline metrics at the doctoral or professional level are legitimate but represent longer-term outcomes that depend on many factors beyond any single programme's control.