This post is more than two years old. As a result, it may contain information that's out of date or that I no longer agree with.
Posted 21 Jun 2016

The slides for the talk are here. Here’s the description from the Open Source Bridge site.

My goal with this talk was to give gender bias in tech a more historical and contextual perspective. Often I think we are so tied up in resisting inequality, it’s hard to see how it has changed over time, or what allowed that inequality to be created.

Some additional things from my notes:

There’s been more research lately on why the participation of women in the technology industry has declined over the last 30 years. The graph in slide 10 comes from a piece by Planet Money.

Modern computing grows directly from the need for ballistics calculations during WWII. At this time, “computer” was a job description: women who performed sets of calculations for this work. These women were young, from recent college graduates to even high school students. They were hired for their math abilities, and often viewed as interchangeable. Men who might have taken these jobs were occupied in other parts of the war effort. The work wasn’t seen as intellectual so much as repetitive and focused. The men who designed and created the machines these calculations were performed on received most of the credit for the results.

In 1946, six women were recruited to work on the ENIAC, the first general-purpose computing machine to be built—designed for calculating ballistics tables. There was no programming manual to tell them how to do this. They had to learn the system and create programs for it—not with a written programming language, but cables and switches! At the time, they were not publicly credited for this work.

Post-war, this computing work developed into a technology industry, primarily focused on data processing. Women continued to be hired as operators and coders of computing systems, work that was consistently categorized as clerical and feminine. A more formalized division formed between this and “programming”, i.e. program design, in the job titles and roles assigned, along gendered lines. Women’s roles as coders included the debugging and trouble-shooting of computer systems–in doing so, they revealed that computer programming and coding was a much harder and more complicated task than the men who designed the systems had anticipated.

Women continued to be encouraged to join the computing industry into the 1960s. This is a Cosmopolitan article from 1967 encouraging young women to consider applying for programming jobs to meet a growing need for labor.

Slide 17 is from a book called Your Career in Computer Programming, also 1967, reviewed here:

The roles and narrative shifted over the next decade, though. Two things happened:

  • an expanded need for computer programmers that focused on a masculinized set of beliefs about who was skilled and qualified to do the work
  • the rise of a second narrative about the rebellious, creative computer hacker

The growth of the computing industry created a need for companies to expand their programming workforce rapidly. There weren’t enough existing programmers to cover the amount of work companies could line up, so new ones would need to be trained, and they struggled to figure out how to identify and train new programmers quickly. —image from Business Week, 1966.

In order to address this shortage, companies turned to various kinds of screening methods, especially aptitude and personality tests. They had no particular evidence that this was an effective way to select potential developers—the main motivation was that many new people would need to be trained in programming skills, and they needed a way to filter candidates.

Rather than increasing gender equality by widening the search for candidates, the use of these tests tended to reinforce certain biases: the idea that being a programmer required formal mathematics training, and that it was most suited toward anti-social masculine personalities. Women were less likely to screen successfully. Also, emphasizing these sorts of traits played into the power struggles that occurred when existing business systems in an organization were computerized. Slide 21 image via The Computer Boys Take Over (Proceedings of the Fourth Annual SIGCPR Conference on Computer Personnel Research, ACM, 1966).

The growing mythology around computer hackers grew out of groups and activities like phone “phreaking”, the Tech Model Train Club at MIT, and the Homebrew Computer Club in California. These emphasized a slightly different kind of masculinity than the corporate screening tests: one that involved creativity, cleverness, and rule-breaking. Slide 22 photo: Steve Jobs and Steve Wozniak in their Homebrew Computer Club days.

In the late 1970s and early 1980s, the computing world expanded to include personal computers and game consoles. The marketing efforts for these systems became heavily focused on market segmentation—specifically promoting them as desirable and suitable for boys. More reading on why this approach makes sense to marketers:

This segmentation is also reinforced in the other direction, by making sure that computers “for girls” have visible gender markers.

As a young woman, the way I learned to resist these pressures was through the idea of “girl power!” and direct counter-argument: “anything you can do, I can do better.” I’ve come to find that approach unsatisfying, and instead I want to work on the idea of re-normalizing gender diversity in tech, and addressing the structural impacts of these biases.

Stepping back a little, we have this field, programming, that was created and defined by women… [image of Margaret Hamilton]

…and yet, this is what technology companies look like today. [image of a startup staff page that only includes white men]

We hold invisible biases against even seeing that this inequality exists (remember that percentage of women receiving CS degrees? it’s about 18%) Slide 28 quote from

The effects of these biases and marginalization trickle down, from women in general, to women of color, to older workers, to those with disabilities, and so on. The slide 29 photo is of Christine Darden, who worked at Langley Research Center from 1967-2007.

Who benefits from inequality? It’s great to be part of the majority! You’re respected, you fit in—and because your skills are seen as important, complex, and in short-supply, you’re paid very well. Do companies benefit from it too? (homogenous teams could be easier to manage)

What’s the downside to creating more inequality? Pay rates go down, and so does prestige: once women start doing a job, “It just doesn’t look like it’s as important to the bottom line or requires as much skill.” Research on this:

What’s the new narrative we need? Starting points:

  • Technology is created by people of all genders
  • Gender diversity > gender neutrality
  • There’s nothing wrong or off-balance when a technical team adds more women

What do we have to give up to get there? It’s going to be tempting to hold onto the wealth and power that comes from being a valued elite (of course!) What will this look like and feel like for us?

We had a great discussion from the people in the room on these last questions. I’m excited to have this conversation with more groups.

Further reading

Past | Future | Random