Do Girls Not Code?
"Girls aren't interested in coding. That's why the tech industry skews so male."
Girls coded at equal rates to boys until roughly 1984, when women were 37% of US computer science degrees. The share then dropped to 18% by 2010. That drop coincides exactly with the marketing of personal computers as 'boys' toys.' Countries without the same marketing pattern have better CS gender balance today. 'Girls don't code' describes an outcome Americans produced, not a natural preference.
What the data says
-
Women's share of US computer science bachelor's degrees peaked at 37% in 1984 and fell to 18% by 2010, rising slightly to 22% today. No other STEM field showed a comparable drop in the same period.
NCES Digest of Education Statistics · 2024 · Federal education statistics
-
In Malaysia, 52% of CS majors are women. In Iran, 70%. Cultural variation dwarfs any hypothesized innate preference difference.
UNESCO 'Cracking the Code' (2017) · 2017 · UNESCO global review
-
GitHub pull-request acceptance study: women's code was accepted at higher rates than men's when gender was not visible (78.6% vs 74.6%). When visible, the pattern reversed.
Terrell et al., PeerJ Computer Science (2017) · 2017 · Analysis of 3M+ pull requests
-
Code.org 2023 data: girls and boys enrolled in high school CS courses at equal rates where access was equal. The gap appears sharply in environments where cultural cues position CS as male-default.
Code.org State of Computer Science Education (2023) · 2023 · National enrollment data
Where it came from
The first computer programmers were women (ENIAC, Ada Lovelace, Grace Hopper, the NASA computers). CS's gendering as male emerged in the mid-1980s, specifically around the marketing of personal computers. Jane Margolis's 2002 'Unlocking the Clubhouse' documented the mechanism in detail — the 'computer room' became coded as boys' space, girls lost the early practice time, and the gap compounded through college.
What this means
The US CS gender gap is a cultural artifact of the 1980s, not a biology-rooted preference. In countries without that specific marketing history, women are majorities in CS. The 'girls don't code' framing treats a chosen outcome as a natural fact — and that framing is part of what sustains the outcome.
Frequently asked
Were women always a minority in computer science?
No. Women were 37% of US CS bachelor's degrees in 1984 and trending up before then. The decline is specific to the personal computer marketing era.
Is the gap a US phenomenon?
Largely, yes. Malaysia, Iran, India, and several other countries have majority-female or parity CS enrollment. The US/UK pattern isn't universal.
Is there evidence of innate preference differences?
No strong evidence. Cross-cultural variation dwarfs any hypothesized sex-linked preference effect. Outcomes track culture, not chromosomes.