Few people know that the first computer programmers were women, but it’s certainly no secret that there is now a huge gender gap in the technology industry.
Silicon Valley has a reputation for being male-dominated. And it’s a well-deserved reputation, as the majority of people working in the technology industry are both white and male. According to Google’s own data, only 23% of their tech employees are women, with Microsoft coming in even worse at a miserable 20%. And why is this? Could men really just be inherently more intuitive, more naturally skilled when it comes to coding than women?
Of course not.
In fact, coding was originally seen as a women’s game. Women were actually the original ‘computers’, performing long and complicated calculations by hand for the military. When the first digital machine ENIAC (Electronic Numerical Integrator And Computer) was developed in 1946, it was women that helped to program it – receiving little to no credit. Throughout the 1940s and 50s, women continued to be pulled into programming, remaining the dominant sex in the industry.
If we’re to go back even further, Ada Lovelace, genius mathematician and writer, designed the first computer algorithm in 1843 for Charles Babbage’s proposed Analytical Engine. So, if women were the creators of computer science, then how come the number of them working in the field now is so low?
It all goes back to stereotyping
Coding was originally considered to be clerical work, and therefore ‘women’s work’ that men weren’t interested in doing themselves. In 1967, Cosmopolitan ran an article called ‘The Computer Girls’, in which computing pioneer Grace Hopper said that women were ‘naturals’ at programming.
“It’s just like planning a dinner,” Hopper explained. “You have to plan ahead and schedule everything so that it’s ready when you need it”. Hopper was actively exploiting gender stereotypes to try and encourage women to enter the field, by telling them that they already had the skills necessary to succeed.
However, around this time men began to realise that programming was not a low-skilled clerical job after all, but something that required talent and advanced programming skills. Despite all the women that were already doing the job, and were clearly very capable of continuing to do so, industry leaders began to recruit men.
Not only did they push men towards the field, but they actively pushed women away from it, creating ad campaigns and using smear marketing to discourage women applicants. Aptitude tests (created by men) included the type of maths that men were more likely to have studied in schools at that time. The answers to these tests were also circulated around male fraternities, so all that was really being evaluated was your networking skills.
The media also started to personify coders as the typical ‘geeky loner with no social skills’, with the release of films such as WarGames and Weird Science helping to establish the stereotype. Personality tests began to be widely used, basing the ideal profile of a programmer on someone with a “disinterest in people” and “a dislike of activities involving close personal interaction” said Ensmenger.
An algorithm favouring men
Companies began to seek people out with these characteristics, and as a result, they became top programmers, self-fulfilling the prophecy. This still happens today. An online tech-hiring platform called Gild finds the perfect candidate by searching through their ‘social data’ and examining the trace that people leave behind online. According to Gild’s data, a solid indicator of strong coding ability is an affinity for a particular Japanese manga site.
However, Cathy O’Neil, American data scientist and author of Weapons of Math Destruction pointed out that “if, like most of techdom, that manga site is dominated by males and has a sexist tone, a good number of the women in the industry will probably avoid it.” Therefore, women are much less likely to be frequenting the manga site, regardless of their programming talent.
Gild undoubtedly didn’t set out to make a biased algorithm, but because it doesn’t use neutral data, it still discriminates against women. By not considering how men and women’s lives differ, old injustices are coded into the algorithm itself, resulting in hidden bias. This is an example of a much greater problem.
Today, algorithms are being used to make decisions in all aspects of life from healthcare, to transport and even the justice system. If the data used to ‘teach’ these algorithms reflects existing prejudices, then the output of the algorithm will echo them too. What is especially worrying is that as most companies keep their code a closely guarded secret, society has no way of telling how discriminatory these algorithms are.
The unconscious bias leading to faulty algorithms can be traced back as early as childhood. In the 1980s the first video game consoles were released, with their makers deciding to market them specifically at boys (and presumably halving their own profits). A study at Carnegie Mellon University in 1999, undertaken to find out why the intake of women for their computer science course was so low, found that the home environment usually only encouraged boys to work with computers.
If a computer was bought for the house, it was far more likely to be put in the son’s room than the daughter’s, increasing boy’s exposure to the machines. Typically, mothers were also less engaged with computers than fathers, so girls would soon learn to curtail their enthusiasm for the subject, following social cues. The clear message that ‘technology is for boys and not girls’ carried through childhood and into adulthood, infiltrating hiring processes and promotion decisions.
A hostile environment for women
The environment on computer science courses was also often hostile towards women. A 1983 study at M.I.T found that women were often talked over by other students and ignored by professors when raising their hands.
Comments such as “You sure are bitchy today – must be your period” were rife on campuses, with the report concluding that behaviour “sometimes approximated that of the locker room”, as men rated how ‘cute’ their fellow classmates were.
By the third year of the course, women were shown to have easily caught up with male students who had coding experience prior to college, although there were still large numbers of women dropping out, often due to the toxic nature of the course.
As coding became more popular and the number of students applying increased, courses were made harder to weed out anyone who didn’t understand immediately. If you had previous coding experience, you tended to thrive in this environment. Unfortunately, a lot of women didn’t, and so the culture of computer scientists being men carried on.
Closing the door to women
This takeover by men turned into a self-perpetuating cycle, with men in leadership positions only hiring those that looked and talked like themselves, leaving women out in the cold. Unfortunately, Silicon Valley finds this difficult to recognise, usually choosing to believe in meritocracy instead, and often ruling biology as the reason that so few women roam the halls of Google, Apple and the like.
But if this is true, then why were women so prominent in the early years of programming, when the lack of Internet made it a far harder discipline? And why, in India, are roughly 40% of computer science students women?
The truth is that the technology industry in the West has closed its doors to women and doesn’t seem too interested in changing that. Companies reticence to disclose sex-aggregated data is a clear sign of this, with Google even going to court in 2017 to avoid handing over its gender pay data. Before women can be welcomed back into the world that they created, the technology industry needs to admit that it has a problem, one that it certainly has the means and money to fix.
To learn more about how the world is designed for men, read Invisible Women by Caroline Criado Perez.