Looking Deeply at the Gender Gap in Computer Science

By Jesse Cox / Winter 2020

Until about a month ago, I used to coach middle school boys’ and girls’ soccer at a local school. As a good coach should, I tried to keep up with the kids in their studies, what they were interested in, what they wanted to be when they grow up etc. It was weird to hear that as early as sixth grade, the kids were being introduced to writing computer code; as someone who went through middle school in 2004, I can barely remember having to take typing class. What struck me further is that when I asked the kids if they were interested in learning more, or if they’d like to work on computers one day, only some of the boys said yes, none of the girls. A few of these kids will be the world’s next computer scientists and software engineers, and the question arises: why should they be only boys? What do we stand to lose from cultivating a homogenous community of workers, especially one that plays such a crucial role in society?

We use computers to solve our most complicated and pressing problems. We use them to protect our data and our identities. We communicate and play together. We can’t deny how central computers are in our lives, and if we want to do an analysis of human life in the United States, it would be impossible to disregard the impact of technology at almost every turn. It stands to say, then, that modern computer scientists shape our lives with their products, and as such have some pretty weighty responsibility in producing something for everyone. What is troubling is that this cohort is actually a pretty bad representation of our society as a whole, marked most notably by the fact that women currently only hold ~18% of Computer Science jobs in America. This creates a major problem in that in the business world, a diverse workplace is crucial in creating a product for a diverse consumer base. For example, facial-recognition software made in the United States has shown huge misidentification rates of Asian-Americans, compared to software made in China or Singapore (Clare Garvie). There’s certainly no such thing as a perfect program, and to label developers as ‘racist’ is, for me, a leap too far, but it certainly supports the idea that a homogenous producer will tend to create products for a homogenous consumer. What we need, then, is diversity at least enough to be creating something for everyone, and I believe that the trend in the last thirty years shows we’re actually moving further from that benchmark as time goes on. Let’s try to understand where the gap is coming from.

We have to acknowledge that the problem of misrepresentation is one that’s starting earlier than college. There is an expectation for what an engineer or computer scientist is, and there’s more pressure from both inside and outside school to get boys involved in Computer Science. This represents a systemic problem with how we view these jobs, and moreover the people we deem worthy enough to hold them. None of these twelve and thirteen year old kids have any classroom exposure to hard Computer Science ideas, and yet there are forces encouraging more boys to be interested than girls, including pressure from parents, or societal stereotyping. Stuart Reges, a lecturer at the University of Washington, was put on probation for arguing in a controversial 2018 article that girls don’t code because they are simply less likely to want to code (Reges). If he’s right, can we start to look at why this is? Can we start to introduce programs that make computing more accessible and approachable to girls? None of these are addressed in Reges’ article, pointing to what I believe is the central problem with the disparity: We see a powerful male figure at the center of the gender-gap debate who openly acknowledges that women are misrepresented in Computer Science, but suggests no road forward. This is acceptance in favor of keeping a position of power. This says: “yes the system is flawed, but that’s just how it is,” and it’s just convenient that the system has placed him in a position to shout this from a mountaintop. Unfortunately for Reges and other intellectuals of his ilk, there is a strong argument and an ocean of research that suggests that having a diverse workplace will actually drive profits up.

It’s easy to recognize the priority of profit for large companies, with no exception being given to the technology and software sector. It might even be easy to operate under the assumption that the gender gap we’re analyzing has some underlying motive in driving up the bottom line. I would like to move away from this idea in favor of the stereotype argument above. While it stands that there is also a pay-gap between men and women in nearly every industry in the United States, that gap is comparatively low in technology, and women hold a percentage of “C-Suite” jobs (CEO, CTO, etc.) that’s comparable, if not higher, than the rest of the field (Stych). Further, it’s easy to find studies proctored by Forbes, or the Harvard Business Review which conclude that diversity has a measurable positive effect on profits (Powers) The impact of diversity-driven curiosity and differing perspectives to difficult problems show up as big green numbers on company balance sheets. The reality is that companies know this too, and it becomes apparent when we see figurehead companies like Google and Facebook creating initiatives to hire more women in their offices. So where is the disparity actually coming from? And why are these initiatives only getting us to 18%? Hard as I looked, there were no readily available opinions stating that women are less productive, or less profitable in the tech environment. And while those might definitely be the opinions of some behind closed doors, I would like to instead focus on rebuking the overall public reaction from prominent figures in the tech industry: “women don’t want to code, and that’s ok.” They suggest the way women work and the way women think make them less likely to take Computer Science jobs. I would like to address these ideas as prejudicial, ill-informed and overall dangerous. To suggest that Computer Science is a “man’s job” breeds stereotyping, distorts reality, and ultimately keeps bright minds out of the industry. This needs to change.

Since the advent of computer software in the early 1990’s, computer scientists have overwhelmingly been men. “Computer geek” is such a comfortable stereotype that of the first 50 google images when searching for the phrase are decidedly male figures. What’s surprising, though, is that the percentage of women in computer science is about half of what it was in 1992. Why the 30-year downward trend? I believe the two above ideas are linked; a community dominated by one demographic tends to resist change, as it simultaneously becomes easy to view that community through its stereotypes or hallmark characters. In other words: a male-dominated workforce will naturally remain male-dominated, which creates a stereotype, which discourages women from entering the workplace, and a sort of vicious cycle starts to form. Stereotypes, even if ‘harmless’ or ‘joking’ or as passive as referring to the typical “computer geek,” increase our separation and atomization by creating differences between us. It becomes natural to see yourself as a part of a stereotype that you fit in to, and it becomes even harder to infiltrate a group that, at first glance, you may seem like an outsider to. We can see this start to compound as the classes and subjects get harder, and as many young women in Computer Science start to ask themselves if they’ve found themselves in a place that they might not belong.

There is an overall lack of effort to address the gender gap in computer science, especially at younger ages, illustrated by the lack of girls who are interested in the subject. There are surely clubs and classes and organizations to be involved in, but a lack of effort to address the disparity between girls and boys involved in those programs means that the ratio will continue to be skewed, as the stereotype of computers as a “boy activity” perpetuates. Certainly there will be outliers, the maverick young women whose overwhelming passion for technology pushes them past any social barrier, and those should be celebrated. Who we are forgetting, however, are the girls who are curious; the girls who might wonder what robotics or coding is all about, peek their head into a room full of rowdy and (let’s be honest) nerdy boys, and shirk the opportunity for fear of being mocked or not fitting in. Not everyone is a natural. In fact, almost none of us are. We need to remember that creating an environment where people feel warm and welcome is just as important as the actual delivery of information. Unfortunately, the extremely modern nature of Computer Science as a subject means that we’re still learning what to teach and how to teach it. In all honesty, I think that’s what draws a lot of people to technology: that discoveries are being made daily, and even our instructors are hungry to learn. What we’re forgetting among this hunger, though, is that a relatively “young” curriculum has created a steep learning curve, making it difficult for any newcomers to find an inroad. Having discussed the difficulties of being involved at younger ages, let’s analyze what that means for women who decide they might be interested in Computer Science when they arrive at a university.

Computer Science as a broad discipline is the youngest of its kind. Ideas being shared at schools and universities are newer than those of, say, English, Math, or even something as complex as Chemistry. Further, we don’t really have an agreed upon method to teach the ideas required to be a successful Computer Scientist, and we can see that most universities have instructive tracks that vary in form and practice. I would like to suggest that we’re not encouraging enough women to take part in Computer Science before college. If we can encourage more girls to want to code, maybe we can start to see a shift in the number of women who enroll in college and eventually contribute to the industry. As we learn how to teach Computer Science in higher learning, we need to remember that we’re not just proselytizing to a captive audience of undergrads, we are creating a community. These are, after all, the men and women we are preparing for the workforce, so it should be important that they feel part of a team. Instead we have classrooms where the curriculum is relatively new and while instructors attempt to iron out the kinks, some of us struggle and some of us thrive.

Among the shared experience of Computer Science university students is the dreaded “impostor syndrome” which leads you to wonder if you belong where you are when things get difficult, and is recorded to affect women at a higher rate than men. This is a difficult major, and we see a lot of turnover; people have dropped or changed majors in every Computer Science class I have been in at UCSD. Sometimes things have been difficult by their nature, and sometimes have been difficult because professors are still figuring out the pace of the class. It makes a lot of people doubt if they’re smart enough to be here, and a lot of the way through this doubt is reaching out for help, which has its own difficulties. It’s hard to ask for help! And it’s even harder when the people you need to ask aren’t exactly like you. I have asked myself many times if I belong in this major, or if things are too hard for me now as a returning student, and it’s intimidating to go and ask someone for help who has been immersed in the subject for two or three years, especially if they’re eight years younger than me. What I can imagine is that, for a lot of women, when the impostor syndrome inevitably rears its head, it becomes way more difficult to fight through when you stop and see an industry that is more than 80% men. Suddenly the question of “do I belong here?” is harder to answer, and the tragedy is that there are some who tell themselves “no,” even though there are plenty of us out here who would be yelling “yes!”

What we tend to see is the student who was introduced to computer science at a younger age kind of knows a lot of the material before hand, and thus needs less help, while those of us who haven’t had the time, or are new to the general process of things find ourselves feeling that aforementioned imposter syndrome. What we find happening here is a kind of insulation from the rest of academia where you’re lauded for being the most knowledgeable, and forgotten if you need a little extra help. If we recall that the gender gap in technology isn’t being properly addressed at younger ages, it becomes apparent that a lot of these college students with less experience are the girls of yesterday who weren’t sure if computer science was their thing yet, or figured that the after school club was better left to the boys. Failures at the middle, high school, and college levels are responsible for the low percentage of women in technology, and when added together in succession, they result in that percentage continuing to drop as the years go on.

The conclusion I’ve come to is that this problem is one of support. Namely, there’s not enough out there to make Computer Science a worthy place of interest for young girls. There’s not enough out there for women in the field to get the support they need from people like them. I’d like to draw a comparison between women in computing and women in soccer. Thirty years ago, there were no female professional soccer players and there was no US Women’s National Soccer team. In 2020 we have the strongest team in the world by a long shot, but it’s still not good enough. In comparison to the Men’s National Team, the women were paid less on a per-game basis, while practicing at inferior facilities. Despite this, we have to acknowledge and learn from their progress. These women went from nothing to bringing us a trophy from the highest possible stage. It took hard work, dedication, and a long campaign to get more women and girls involved because soccer, like Computer Science, is a beautiful thing and deserves to be shared with everyone. If we can treat Computer Science as the same sort of crusade, we will be heading towards a problem-solving industry that accurately represents the nation it serves. We will be creating technology and software that provides for everyone, and at a higher level. Let’s not let it take 30 years this time.

Works Cited

Clare Garvie, J. F. (2016, April 7). Facial-Recognition Software Might Have a Racial Bias Problem.

Reges, S. (2018, June 19). Why Women Don't Code.

Stych, Anne. “Percentage of Women in C-Suite Roles Inching Up.” Bizjournals.com.

Powers, Anna. “A Study Finds That Diverse Companies Produce 19% More Revenue.” Forbes, Forbes Magazine, 28 June 2018.