Gender Diversity in Tech: A Conversation about Responsibility

Tech is only 20% women and many of them leave. In order to make progress, we need to discuss how much responsibility different members of the community have in fixing the gender problem. A significant part of the discussion involves the role of men: what responsibility they have—and what responsibility do women have to educate men and include them in the discussion. The two of us have different perspectives, so we present a discussion that reflects both of our views.

This piece is adapted from the notes for a talk we gave at a Tech and Ethics Salon in a conversation format. You can find partial audio from our live conversation here.

Jean: If we want tech to include women, we need to change the values of the culture. Harassment is an issue: last year there were the high-profile stories of the GitHub employee and of the Tinder cofounder. There’s also this culture that worships masculinity and demeans femininity—when I worked at Google, for example, people would ask whether interview candidates could “code like men.” To change this culture it’s important to get everyone to care, especially the people who are perpetuating these behaviors. We have to brand the problem in a way that’s palatable. We have to argue for why this problem is relevant and propose actionable solutions that everyone can be a part of. We need to get everyone into the conversation.

Neha: You’re right but we need to think about what kind of effort that takes and who we expect to do that. It’s a lot of work, and we should not set the expectation that its women’s responsibility to do this. Instead we should have the expectation that individuals have a responsibility to learn (even if we doubt it will always happen). That’s one of the problems I had with Sheryl Sandberg and the whole “Lean In” missive. We shouldn’t tell people who already have to work twice as hard that they need to work even harder. I don’t believe we should put the onus of education, branding, and PR of diversity on the part of the community that is already suffering from discrimination.

Jean: You’re right that women shouldn’t be doing all the work. There’s evidence that men are trying to educate themselves, but in ways that are disconnected from the greater conversation. MIT professor Scott Aaronson’s Comment 171 eloquently illustrates this. As part of a justification about his views on gender, Scott describes how he spent his teens reading about feminism only to come out feeling paralyzed by privilege rather than empowered to promote equality. The follow-up discussion, much of which vilified Scott for his supposed insensitivity, demonstrated just how much these conversations about gender don’t have men in mind. (I wrote my own response here.) Other men confirm this. Male friends have told me they get “shouted down” when they try to participate in conversations about gender. Maybe the men don’t get it as much as we would like, but self-education can only go so far. Even brilliant guys like Scott try and fail. We need better ways of bringing these men into the conversation.

Neha: The Scott Aaronson thread is a great example of how coming to a conversation uneducated can set things back and be harmful, and we saw that happening in the thread. More so from other commenters in the thread than from Scott. For example, one commenter questioned whether STEM was more hostile than any other field, even though it has been shown that women are discouraged at a higher rate from STEM than from other fields. Women were gaslighted and being told they are being “feminazis” or “militant”. The same conversations are repeated over and over, with people just talking past each other. As an example, we’re not even on the same page about our goals — Every single article on diversity in tech I read has a commenter saying “gosh darnit well my wife/sister/daughter just doesn’t like programming, so why are we aiming for 50%? We don’t try to improve the percentage of men who choose nursing or teaching kindergarten.” This is a strawman that’s been debunked a long time ago; the point is not necessarily to get to 50% (though personally I see no reason why we shouldn’t), but right now women are actually being excluded and discouraged, and we want to fix that very real problem. Constantly having that questioned is annoying.

Jean: I agree that we need to get on the same page about goals. This discussion needs to include men—and it can’t just be about how men are oppressing women. Men are going to care more if they get to discuss what they don’t like and the changes they might like to see. For instance, apparently women are disproportionately judged as “abrasive” when they behave like the men in tech. Should we change our expectations of how women should behave or should we perhaps expect everyone to be a little less “abrasive?” Men have told me that they also feel alienated by the “macho” culture of tech. After watching a video where I spoke about how programming culture almost caused me to leave computer science, a man in Germany wrote to me about how the programming culture at his university had intimidated and discouraged him for years as well. If we listen, we might find that men are invested in change as well.

Neha: That’s a key idea — both genders suffer under this system. Not all men in tech flourish in this aggressive environment. So how do we encourage the men who feel left out while educating everyone who is holding the conversation back? Unfortunately, a lot of people do not consider it relevant to educate themselves on the studies and data until they come face-to-face with a problem directly. It can often be uncomfortable to even bring this topic up in a professional setting. A huge problem is that we don’t have a common vocabulary and framework to discuss this stuff. Many stories are anecdotal, and people become emotional and generalize when they share their personal experiences. More than a few male and female colleagues have expressed to me that in their opinion, women are actually overly favored in tech, via affirmative action. They see women-focused recruiting and women-only scholarships as unfair, because they haven’t personally experienced sexism in action. I’m glad that they had good experiences, but that doesn’t mean it’s ok to insist that others didn’t have bad ones.

Jean: The lack of common ground is holding us back. In order to make progress, we need to achieve a common understanding of the problems, goals, and metrics for success. At MIT, for instance, there have been initiatives that have established that the leaky pipeline and implicit bias as problems. There is a baseline level of awareness that people are now expected to have about gender issues being important and about what the main problems are. It’s easier to do this at MIT because there is a clear notion of who is in charge as well as a fact-driven, engineering-based mentality. It will be harder, but not impossible, to establish a common understanding throughout tech.

After listening to our conversation, one friend observed that when it comes to gender issues, men and women are like a romantic couple in need of counseling: resentment on both sides makes progress difficult. But we need to start somewhere—and a good starting point is to develop a common understanding of the problems and goals. In computer science research, we state our assumptions up front, describe our methodology, and contextualize our ideas with respect to previous work. Sharing a common vocabulary and framework helps researchers navigate disagreements. A fact-driven, engineering-minded approach may be what we need to move the tech diversity discussion forward.

Thanks to Adam Marcus for his feedback on this talk.

Neha Narula (@neha) and Jean Yang (@jeanqasaur) are both final-year PhD students in MIT’s Computer Science and Artificial Intelligence Laboratory. They have been discussing these gender issues with each other for years.

Neha is a member of PDOS, the Parallel and Distributed Operating Systems group. Neha’s research focuses on databases and distributed systems, particularly having to do with consistency and performance. Previously, she worked for Google as a Software Engineer.

Jean’s research focuses on designing programming languages and tools that help people more easily write the programs they intend to write. She created Jeeves, a programming language for automatically enforcing policies for security and privacy. Jean co-founded Graduate Women at MIT in 2009 and co-directs NeuWrite Boston, a workshop for scientists.