NPR did a Planet Money podcast in 2014 posing the mystery of “When Women Stopped Coding”. The writeup online includes a striking graph of the percentages of women in different fields of study, plotted out over the last few decades. Female representation in medical school, law school, physical sciences, and computer science squiggles upwards steadily from the 70s through the 80s, and for everything besides CS, onwards to today; but the percentage of women in CS peaks at 37% in 1984 and slumps downwards to the right.
NPR’s investigation turned up this correlation: “The share of women in computer science started falling at roughly the same moment when personal computers started showing up in U.S. homes in significant numbers.”
Curiously, they didn’t study another very relevant data series, the raw enrollment numbers. But without those, from sheer percentages alone, it’s impossible to make the case that “women stopped coding”. Another mathematically plausible explanation for the graph above, though not necessarily a likely one, would be that women in fact kept coding, but increasingly more men flooded the field, bringing the percentage of women down.
So I went trawling through data from the National Center for Education Statistics.
The first thing that stood out to me when I drew out the stacked bar graphs was that the raw numbers of women in computer science have risen and fallen, and risen and fallen again, carried by the tides of overall enrollment and excitement about the tech industry; it’s not a clearly monotonic decrease. On further inspection, I found another interesting feature in the graphs: a correlation between the bursts in total enrollment numbers and the steeper drops in percentage, shortly after.
A series of Google searches led me to two ACM papers, a 1999 piece titled “Conserving the Seed Corn: Reflections on the Academic Hiring Crisis” and 2002 piece titled “Encouraging Women in Computer Science“, both authored by Stanford CS professor Eric Roberts. He wrote matter-of-factly to explain the attrition of women from the field:
[The] decline in the number of computer science degrees was largely the result of explicit steps taken by academic institutions to reduce computer science enrollments when it became impossible to hire sufficient faculty to meet the demand. These steps included, for example, imposing more stringent requirements for admission to the major, adding new required courses in mathematics, and transforming introductory courses into filters designed to limit entry into the field. Such strategies have a disproportionately negative effect on enrollment by women and minorities.
Although it is nothing inherent to computer science and no longer necessary for the practicalities of university instruction, that culture of exclusivity, laced in alpha nerd competitiveness, geek mythology, and implicit sexism, has persisted. To draw women back into the field, we have to fix that culture and popular perception of what it means to be a CS major, to be a coder, to be a software engineer.