June 21, 2018

Dear Allen School Community,

Some of you may have read a recent editorial written by an Allen School faculty member about gender diversity in tech. Regardless of whether you saw the specific article or not, this is a good time to reaffirm our values around diversity and inclusion as a School.

The Allen School actively supports diversity and inclusion, as articulated in our Inclusiveness Statement. We disagree with the conclusions drawn in the editorial and are optimistic that our substantial, ongoing efforts to build an inclusive community will make us a better school — and make the tech industry, as a whole, a better place in which people of diverse backgrounds and experiences are empowered to do great work that impacts people's lives. Our efforts include (but are not limited to):

  • An extensive K-12 outreach program focused on inspiring students to pursue CS — especially girls and students from underrepresented groups.
  • We are proud to be one of 11 leading computer science programs in the NSF-funded FLIP Alliance: Diversifying Future Leadership in the Professoriate, collaborating to launch and demonstrate the effectiveness of strategies for recruiting and retaining diverse doctoral students.
  • A quarterly workshop on Building An Inclusive Community, which provides participants practical strategies for promoting inclusion in their daily interactions.
  • A holistic undergraduate admissions process that aims to admit students who have strong academic and personal potential to contribute to the Allen School community.
  • Enthusiastic participation in the UW STARS program for students from underserved high schools, including CS coursework to help STARS students explore and succeed in CS.
  • A commitment to empowering people with disabilities to pursue computer science, through the AccessComputing Alliance and the nationally recognized work of Professor Richard Ladner.
  • Participation in national organizations promoting diversity, such as NCWIT and BRAID.
  • The Student Advisory Council, created to foster a strong undergraduate community and amplify the student voice.
  • Q++, another organization created and led by undergraduates to promote community among LGBTQ students.
  • Active ACM and ACM-W student organizations leading activities ranging from industry prep to workshops on imposter syndrome and sexual harassment.
  • Sponsoring students to attend national conferences such as the Grace Hopper Celebration of Women in Computing and the Richard Tapia Celebration of Diversity in Computing. This year, the Student Advisory Council and ACM-W will both present at these conferences.

Notably, many of these efforts have been led by students. Along with the work of faculty and staff, we fully support students’ ability to determine their own needs and solutions and we empower students to be change agents within the Allen School.

We acknowledge that we have a long way to go, but these efforts work. Enrollment in our CS undergraduate and Ph.D. programs is around 30% women, after a steady increase from about 20% a decade ago. Recent admissions trends are positive: women account for 38% of our incoming CS direct freshman admits and 37% of transfer students for the coming academic year.

As you can see in these numbers, women are interested in CS and women do code! It is central to the mission of the Allen School to pursue initiatives that broaden participation in CS. Furthermore, we do not believe that where we are today is the best we are likely to achieve. We continue programs at all levels that focus on building interest among women in CS. We are currently expanding our efforts to include a focus on underrepresented minorities, students with disabilities, and economically disadvantaged students. Beyond recruiting new people to the field, it is crucial to create and nurture an environment where all students — especially members of underrepresented groups — feel supported.

All members of the Allen School are entitled to share their ideas freely, and no one among our leadership has any interest in silencing or censoring people even when they express controversial ideas. However, our leadership also has the right and the responsibility to affirm our values and to discuss the many ways in which we are supporting and will continue to support those values.

The most important thing we can do to promote inclusiveness is to listen to the people we want to support and to respond in tangible ways to the needs of students. The leadership of the Allen School believes that all students have the potential to be interested in computer science, and we believe that our ongoing efforts to promote an inclusive community help students of all backgrounds and identities succeed. If you have feedback on these initiatives or anything else related, please talk with an advisor or send an email to me or other members of the school leadership.

Hank Levy
Director, Paul G. Allen School of Computer Science & Engineering


Follow-on Statement

July 2, 2018

About a week ago I wrote to you about an essay written by Principal Lecturer Stuart Reges. As I wrote then, I disagree with the essay’s conclusions, and I remain committed to our diversity efforts. I’d like to add a little more context to my previous message, both about the essay and the Allen School’s commitment to changing the current diversity imbalance. To be clear: I and other members of the Allen School leadership think Reges and his editorial are wrong about the causes of the gender disparity in computing and the futility of aiming for greater diversity. The existence of gender differences and of different choices being made as a result is not an adequate explanation for why more women are not going into computer science. Bias, whether implicit or explicit, is an issue for our industry. And 20% women in tech is not the best that the industry can hope to do.

Reges posits that the reason there are not more women in computer science is that they choose not to be — that their avoidance of tech careers in favor of other positions is due to personal preferences. But he did not adequately answer why women make that choice — which in instances such as outright discrimination or sexual harassment (which Reges completely dismisses because “women I talk to are enjoying their experience as software engineers”) really isn’t a choice at all. Even if women are following their passions by entering another field, the “why” is still a compelling question to me, as it should be to everyone who cares about the future of computing.

Why do so many women avoid computing altogether or leave the field once they are in it? There are many factors at work — factors which we as a field and as a society can and should address to make computer science more attractive and welcoming: parental expectations and encouragement (or lack thereof), early exposure to technology, socioeconomic status, stereotypes about programmers and programming, perceptions of work culture in the software industry (some very much deserved), encouragement by teachers and counselors, biases and inequities of many kinds, a failure to adequately communicate the empowering role of computer science, failure to teach computing in a way that is accessible to students with diverse backgrounds and experiences — I could go on.

Humans are complex, all of these factors clearly have an impact on peoples’ lives, yet Reges concludes that women’s choices are essentially entirely due to gender-based differences. This would imply that they cannot change, which I believe is a fundamental flaw. In 1975, only 16% of medical school graduates were women. This grew to 40% in 1995 and today, as Reges notes, approximately half of medical school students are women. Reges quotes Unlocking the Clubhouse, using the statement that “Concern for people, family, ‘balance in life,’ novels and a good nights’ sleep should not come at the cost of success in computer science” as an indication that women and men differ in their values and interests, and as a reason why women prefer fields other than computing. Yet these are the very same issues that exist in medical schools (residency is 3 years of isolation and sleep deprivation!), but somehow over the last 40 years women have chosen to become doctors in large numbers. How could that be? Obviously something changed over that long period – most likely both the medical system (which was male dominated and treated women poorly in 1975) and women’s interest in the profession. The point is, we can clearly work to change some of the factors listed in the previous paragraph, and young people (of all genders) can and do make different choices over time in the fields they wish to pursue: those choices are not predetermined at birth based on the colors of our baby blankets.

A second flaw in Reges’ logic, in my opinion, is his narrow concept of computer science itself. Computer science is not one thing – and in particular it is not just programming and algorithms. Arguably, CS has reinvented itself faster than any other science in history. Five years ago we described a model for modern computer science, which a number of other schools have adopted as well. The most important part of this model is the recognition that computer science has become extremely broad and highly applied to societal grand challenges; e.g., in our school we have students who have developed smartphone systems to detect jaundice in infants, mobile data systems used to improve health care in Africa, machine learning techniques for choosing the best therapies for leukemia, and platforms for analyzing cultural differences across different countries. Applications such as these make computer science more interesting and accessible to a diverse population of students because today’s students want to have impact. We are not just educating software engineers for the software industry – we are educating computer scientists for the world. Clearly, different subfields in in our technology program have different appeal. The bottom line is that neither women’s interests nor the field of computing are static and unchanging, and any analysis that misses that fact is flawed.

Our goal is not to force people into computing, but to make computing appealing to a broad community, to remove barriers that prevent people from discovering that they like computing, and to ensure that everyone is treated with dignity and respected for their contributions once they join our field. Instead of remaining silent about the problems women and other minorities encounter in certain segments of the industry, in a misguided attempt to avoid scaring them away, we should focus on fixing those problems, which will make a better community for all.

The one assertion in Reges’ editorial that I will agree with is that “it’s complicated.” But that is what we do in computer science — we develop solutions to complicated problems! To solve some of the truly thorny problems facing humankind — in sustainability, security, privacy, health, education, and many others — will require diverse contributors. Computing is a growth industry, offering economic opportunity to those working within it in addition to the opportunity to generate a tangible, positive impact on people’s lives around the globe. Computing also permeates many other sectors, including health care, education, financial services, agriculture, entertainment, retail, and more. Computing is the future — how can we, as educators and innovators, be content to have entire swaths of our population underrepresented in creating that future?

Here at the University of Washington, as elsewhere, the growth of interest in computer science shows that more students are aware of the expanding role that computing plays not just in their daily lives, but also in our economy and society. As we often observe here, “Every field is increasingly an information field.” To do nothing about the conditions that drive women from computing would deprive countless visionaries of the opportunity to participate in shaping the future of their society and to reap the economic rewards of a lucrative field. We would also lose out on a diversity of knowledge, creativity, and perspectives that research shows fosters higher performing teams and better business outcomes.

I refuse to accept Stuart Reges’ “difficult truth” that we aren’t likely to make further progress on gender diversity in computer science beyond an arbitrary ceiling of 20% women. Nor do I believe that increasing the percentage of women in the field has to come at the expense of men; practically speaking, the industry can’t hire qualified people fast enough — therefore, it makes good business sense, as well as being the right thing to do, to draw talent from all genders and backgrounds.

So, when I think about diversity and where we go from here, what I believe is that we need to do more, not less; that we should continue chipping away at artificial barriers to entry to increase opportunities for populations who have historically been denied them; to listen to underrepresented communities when they take pains to point out where our efforts are lacking; and to model the change we would like to see in our field.

I believe that the Allen School has walked the talk up to this point, including in our introductory courses. Our faculty, staff, students, advisers, and teaching assistants live and breathe our commitment to diversity every day. With the help and advice of our colleagues and friends, I am confident that we will make even greater progress in the future. If we don’t, it won’t be for lack of trying. Only by continuing to reduce barriers will we learn what the real limits are.

Hank Levy
Director, Paul G. Allen School of Computer Science & Engineering