Why Don’t Women Want to Code? Ask Them!
A few weeks ago, I declined to participate in a radio conversation on the topic of who “belongs” in computer science. This decision has been weighing on me ever since. I therefore take this opportunity to make my voice heard on the topic of female representation in the computer science community.
To be more specific, I refused to engage in a radio conversation with Stuart Reges and a host. Reges, author of the now notorious post titled Why Women Don’t Code, discounts the toxic environment described in books like Brotopia based on his beliefs about others’ experience in the field, stating that ‘the women I talk to who are working in Silicon Valley are enjoying their experiences as software engineers.’ He goes on to discuss that ‘negative stories about men behaving badly in tech… will do more harm than good.’ It would appear he neither believes the negative stories nor wants to hear more of them.
Yet uncovering negative stories is a fundamental way we learn about structural inequalities that impact the very topic of his article — why some women choose not to code. I want to emphasize that the personal vignettes I share below are, for the most part, not egregious or unusual. Unlike other colleagues who have bravely spoken out, I have not faced sexual harassment at work, thanks to a combination of luck and having had great colleagues and mentors throughout my career. But my anecdotes illustrate the structural inequalities that women face, even in environments that are supportive and awake to the risks and importance of supporting women.
Middle School: Negative Influence of Interpersonal Behavior
My daughter, 12 years old, learned Java a few years ago and mostly on her own, from online courses. I shouldn’t have to boast about her precociousness, but you might not assume she is so good at computer science if I don’t tell you. She easily absorbs and masters the material, and her work is creative and high quality. Her current project has over 500 lines of Java code and five classes. It is an alarm clock which randomly selects and plays audio files based on who set the alarm. She wrote it herself from scratch, and she is currently adding a state machine and support for a physical interface (LEDs and buttons). It runs on her Raspberry Pi. Of course she has every advantage — an expressed interest in programming, good role models, help solving problems, and access to resources and equipment.
I am trying my best to raise her in a supportive environment that encourages her interest in programming. However, while fostering her interest has been easy, creating a supportive environment has been surprisingly hard. She has raised a number of very valid complaints that I have had difficulty fixing, and that are pertinent to the question of why women do (or don’t) code.
My daughter is discouraged that her friends and family are much less likely to ask or talk about her programming than about her brother’s. This happens even when I encourage them to ask her about her projects and she notices and is discouraged by it. Programming is his main focus, so it’s pretty natural that talk about/with him turns to this area. This is amplified by the fact that he has few other hobbies, plays video games with most of his friends and received from his grandparents a new computer to assemble and hack on years ago. As a result, he exudes confidence and comfort with computing technology. These amplifiers are not unique to or universal among boys, but they are certainly more common among boys than girls.
Additionally, my daughter is often silenced by conversations about technology. My son can get so excited about what he’s doing that he will monopolize a conversation. In contrast, my daughter takes her time when discussing her ideas. I tend to quickly jump in and presume I understood her when she may be just pausing before completing her thought; therefore I occasionally correct her when the misunderstanding may have been mine. This can be frustrating, leading her to leave the conversation, and feeding self-confidence issues, which are associated with gender [1].
Many of these differences could, of course, be experienced by boys, as well. This does not diminish their importance — women’s issues are often about human issues that can improve everyone’s lives if addressed. Regardless, they represent some of the subtle, and overt, interpersonal interactions which, at age 12, might discourage a capable programmer from pursuing her interest in computer science.
College: Problematic Presumptions of Peers and Professors
As a college student, the choice of major often closely relates to career decisions later in life. Computing may be one of several competing interests for a college student, and small pushes in one direction or another caused by presumptions about interest and ability might suffice to shift one’s focus.
One example is the presumption that favoritism, rather than ability, explains a woman’s successes. For example, I was told by a fellow intern that I got my first research internship only because I was female. Although such behavior occurred in the 90s, it continues today. For example, while the responses of Reddit posters to Reges’s article are varied, one poster, in particular, resonated with this theme. The poster stated ‘He mentioned giving extra points to women on assignments because sometimes they need a motivational push to keep women in computer science.’ If it is true, this statement shows not only a shocking error in judgement and treatment of students, but a complete disregard for topics that are appropriate to raise in a classroom setting. Statements of this sort can cause male resentment and undercut a woman’s self-confidence and success. For example, simply making a positive statement about a minority group just before an exam, can negatively impact exam outcomes for the subgroup that is called out [2]. This effect, called “stereotype threat” has been replicated for multiple minority groups.
Another problem is presumed lack of interest. My college advisor assumed I’d be going to medical school without asking if I intended to continue in the field of computer science, which was my major. As a result, he had not guided me toward the courses I’d need to have a comprehensive technical background, e.g., ‘compilers’. When he discovered his incorrect assumption, there was little time left to fill in the blanks in my education during my senior year. It is easy to assume the wrong thing about someone, and this is where implicit bias can creep in. It is far better simply to ask.
Positive influences can be equally powerful at this stage, and in fact many women in computer science, myself included, are ‘interest changers’ who discover partway through college that they want to focus on this major.
Graduate School and Beyond: Reputation and Promotion
Once the choice of career has been made, additional schooling provides the skills to succeed one’s specific career. However, schooling and post school activities do not only provide and exercise skills. They also establish one’s reputation. Unfortunately, subtle pressures can impact reputation making and thus career advancement. The following discussion focuses on my own experiences, touching on events that influenced my success in the process of moving from PhD student to full professor.
One important aspect of reputation in academia is publishing. One of my first research projects was used as an example in a paper I was not invited to contribute to. This falls in a grey zone: it was a lost opportunity to do enough work for publishing credit. Engaging in research and authoring a paper based on it may still not be sufficient for appropriate credit attribution. For example, in a later situation, a student wrote a project history for a draft of a casual publication, stating that another faculty member who jointly advised the project was responsible for an idea that was actually mine. Provenance of the idea had never been discussed; the student simply jumped to an erroneous conclusion. When I pointed out the error, he revised the publication by removing the story about credit. Mistakes can certainly happen, but it is more common for men to get credit for collaborative work than women, as shown in a recent study of the impact of collaborative publications on tenure for men and women [3]. Further, degree of contribution is often encoded in authorship order, which is a very simplistic representation of the research process and therefore difficult to interpret. Additionally, women are less likely to occupy the prestigious last author position, usually given to senior authors, than are men, even controlling for seniority [4].
Another important reflection of reputation is promotion. While promotion to tenure is supposedly a fixed process, tied to a fixed length ‘tenure clock’, it is in fact more variable than it appears. The ‘tenure clock’ is seen as fixed and inflexible, but it can be lengthened during an institutional move — such as by following a spouse — or extended due to the birth of a child, or even shortened when someone is viewed as a ‘star’. These positive and negative adjustments to the tenure clock may be influenced by gender. For example, standout adjectives like ‘star’ are more commonly applied to men than women [5], while women academics are more likely than men to take on a larger burden of childrearing [6]. My clock was both extended and shortened at times. When I became disabled with chronic Lyme Disease pre-tenure, a senior colleague told me I should take a leave of absence rather than continue to work. This too is a women’s issue: chronic illness diagnoses are more common for women than men [7]. Instead of encouraging someone to leave, a supportive colleague could ask them what they or their department can do to best accommodate their needs and support them. In my case, in addition to learning how to be a successful academic with my diagnosis, that ultimately involved shortening a clock that had previously been lengthened due to childbearing so that I could move forward with managing my illness post-tenure.
In contrast to tenure, promotion to full professor is not a well defined process. As it was explained to me when I first asked about my case, others would put my name forward when I was ‘ready’ for promotion. When I next raised the issue, a few years later, I was asked, ‘What have you done since tenure?’ by a well intentioned mentor. This discouraging response caused me to wait again. When I next investigated, I discovered that I was well outside the average time in my school from tenure to this promotion. My final query triggered an off cycle rush, when it was recognized that my reputation indicated I should have already been put forward. The timing, six years after my tenure, is not the longest that a man or woman might wait for this promotion. However, what is important in this anecdote is the lack of clear guidance or mentorship, the discouraging reactions, and the scramble to fix things. These are all counter-examples of best practices that can help to equalize differences in self confidence and reputation.
So why don’t women code?
Foremost, I think this is the wrong question to be asking. As my colleague Anna Karlin argues, women and everyone else should code. In many careers that women choose, they will code. And very little of my time as an academic is spent actually coding, since I also write, mentor, teach, etc. In my opinion, a more relevant question is, “Why don’t women choose computer science more often?”
My answer is not to presume prejudice, by women (against computer science) or by computer scientists (against women). I would argue instead that the structural inequalities faced by women are dangerous to women’s choice precisely because they are subtle and pervasive, and that they exist throughout a woman’s entire computer science career. Their insidious nature makes them hard to detect and correct.
Worse, similar misunderstandings and biases can impact the evaluation of women, similar to the trouble I ran into moving forward with my full professor case. These very small differences can add up to significant differences in success over a career. In simulation, random differences in assessment of women of 1–5% were sufficient to create a hiring gap of 2:1 men to women over the length of a career [8,9]. Disadvantages accumulate and affect whether people stay in the career [10]. Racial and ethnic minorities face even more issues [11], as do people with disabilities.
Bias is only implicit when we fail to see it, or to ask the questions needed to make it explicit and therefore changeable. So the next time you wonder why a girl or woman is leaving your room, your class, your major, or your field, I encourage you to consider the possibility that maybe she is simply tired of being in an environment where she must navigate these types of situations. Even better, don’t assume: ask what you can do to support her.
One last point. This post was not easy to write — it required hours of work, and was improved by the generous feedback of many people. I note this because the labor of telling this story needs to be acknowledged. Many thanks go to my sister in law, spouse, friends, and fellow faculty for their time and helpful feedback.
[1] Donovan, Leslie A., and Peter D. MacIntyre. “Age and sex differences in willingness to communicate, communication apprehension, and self‐perceived competence.” Communication Research Reports21.4 (2004): 420–427.
[2] Steven J. Spencer, Claude M. Steele, and Diane M. Quinn. 1999. Stereotype threat and women’s math performance. Journal of Experimental Social Psychology 35, 1 (1999), 4–28.
[3] Heather Sarsons. 2017. Recognition for group work: Gender differences in academia. American Economic Review 107, 5 (2017), 141–45.
[4] Early, Kirstin, Hammer, Jessica, Hofmann, Megan, Rode, Jennifer A., Wong, Anna and Mankoff, Jennifer. To Appear. Understanding gender equity in author order assignment. CSCW 2018 (conditionally accepted).
[5] Schmader, T., Whitehead, J., & Wysocki, V. H. (2007). A linguistic comparison of letters of recommendation for male and female chemistry and biochemistry job applicants. Sex roles, 57(7–8), 509–514. Chicago
[6] Jacobs, J. A. (2004, March). Presidential address: The faculty time divide. In Sociological Forum (Vol. 19, №1, pp. 3–27). Kluwer Academic Publishers-Plenum Publishers.
[7] Denton, Margaret, Steven Prus, and Vivienne Walters. “Gender differences in health: a Canadian study of the psychosocial, structural and behavioural determinants of health.” Social science & medicine 58.12 (2004): 2585–2600
[8] Jonathan R. Cole, Burton Singer, et al. 1991. A theory of limited differences: Explaining the productivity puzzle in science. In The Outer Circle: Women in the Scientific Community, Harriet Zuckerman, Jonathan R. Cole, John T. Bruer, et al. (Eds.). Norton, New York, 277–310. Kimberle Crenshaw. 1989.
[9] Richard F. Martell, David M. Lane, and Cynthia Emrich. 1996. Male-female differences: A computer simulation. American Psychologist (Feb 1996), 157–158.
[10] Allison K. Shaw and Daniel E. Stanton. 2012. Leaks in the pipeline: Separating demographic inertia from ongoing gender differences in academia. Proceedings of the Royal Society of London B: Biological Sciences 279, 1743 (2012), 3736–3741
[11] Joan Williams, Katherine W. Phillips, and Erika V. Hall. 2014. Double jeopardy?: Gender bias against women of color in science. http://www.uchastings.edu/news/articles/2015/01/williams-double-jeopardy-report.php
