Diversity is Our — the Majority’s — Responsibility

Last year I attended the Diversity in Computing Summit. Today I want to reflect on what I learned there and have learned since. Crucially, diversity is in our best interest individually and collectively, and it is our responsibility.

Inclusiveness and diversity aren’t just nice: they increase the strength of the applicant pool, retention, creativity, performance, and impact. However, diversity by race/ethnicity, gender, and disability status in computer science (CS) is low, while diversity by religion and sexual orientation is under-reported. Moreover, overall gender diversity has been declining for decades. The exclusion of women and racial and ethnic minorities in particular, whether intentional or unintentional, has recently resulted in several high-profile errors. Meanwhile, the prevailing culture of computing is polarizing, unnecessarily turning away many people who would enjoy and excel in the work. I found CS hyper-masculine when I went to college and chose not to major in it for that reason; some of the labs I frequent feel similarly hostile, and I hear appalling comments occasionally at conferences and other social events.

Some undergraduate institutions are starting to achieve equal gender representation; for examples see Harvey Mudd, Berkeley, and Stanford. If we believe that trend will reach the graduate level, then we as a field — and members of individual academic departments — have an interest in increasing diversity and inclusiveness as much as we can now in order to maximize the strength of our applicant pool and incoming classes (relative to those of competing institutions) in the near future.

At the Diversity in Computing Summit, the plenary speaker provided a list of efforts within Google to promote diversity in CS education; there are similar initiatives at Microsoft, Apple, Amazon, and others. In their research on underrepresentation in CS education, Google found essential factors of success for minority groups in computer science are lacking. These include support from parents, teachers, and peers, career perception, media perception, formal academic access, and informal exposure. There is a web of social and structural barriers to the success of underrepresented groups in computer science, and these barriers are often unknown to or overlooked by the majority.

Moreover, it seems implicitly assumed that diversity, specifically increasing the representation of underrepresented groups, is the responsibility of those underrepresented groups. I think most people in a major research university would agree that diversity is a good thing, but few take personal responsibility for it. Delegating the correction of underrepresentation to the underrepresented group itself is both limiting on the rate of progress and inappropriate; diversity work currently is a tax on minorities. Meanwhile, due to their leverage, tenured faculty with privilege are positioned to make the largest impact. (Why is contribution to diversity not more prioritized during academic review?)

At the summit lunch I heard about the outcomes of the University of Maryland CS gender forum. What struck me most was one student’s description of how much they appreciated their other-gendered peers showing up to diversity events (during a hackathon, for example) to support them. We must be mindful that not all social contexts are the same, but to assume we have no business at diversity events is to prevent empathy and foster inequality. Our privilege blinds us; we have so much to learn. If we want change — and I argue change is in our best interest — then we must work to make it happen.

Moving forward, let’s acknowledge that diversity is our — the majority’s — 
responsibility. Let’s proactively learn about our own biases and myriad forms of discrimination against minority groups, learn how to recognize and respond to problematic arguments about diversity, and listen attentively to all voices. Let’s then improve institutional culture in day-to-day interactions, adopt inclusive teaching practices, conduct inclusive faculty searches, practice and promote advocacy and sponsorship, and measure our progress and reevaluate along the way. But we must be sure we neither dismiss nor assume minorities’ experiences, nor place each other in shame. To best promote diversity we must acknowledge our privilege, bias, and limited perspective and learn to practice empathy — with everyone.

Thanks to Margaret Mitchell for her formative discussions and references.

Thanks to Hanna Wallach for her thoughtful and detailed feedback.


Further Reading