Gender and Tech Hiring
In late September 2015, I arrived early at my interview for a popular midsized startup with over 200 million dollars in funding. It was my first technical interview of the recruiting season and I was nervous and excited. I showed up early and waited for 10 minutes, growing increasingly anxious that I was in the wrong place. Finally at six minutes past the hour, a disgruntled young guy appeared and then proceeded to walk right by me. We then had an interesting dialogue that went as follows.
Me: “Excuse me? Do you know where room 210 is?”
Him: “Nope.” Keeps walking towards room 210
Me: “What company are you here for?”
Him: “Foo bar 2.0.” Keeps walking towards room 210
Me: “Umm, I think I’m your interviewee…”
Him: Looks me over. Oh.
I walked in.
Can I know for sure that this happened because of my gender? No. Perhaps my interviewer was oblivious to everyone. But wondering — worrying — is one of the things that makes these subtle negative encounters so difficult to explain. And there is a pattern. Most women in tech have a story like this.
The experience threw me off in that interview after a shaky start. My interviewer showed visible signs of disapproval in the first three minutes before I had even started programming, only raising his eyebrows momentarily when I told him that I had placed in the top 300 on the Putnam national math exam. “What about national CS competitions?,” he responded.
Gender bias in tech is a complex issue. Much of the conversation lacks data, and so I decided to collect some. Over the past weeks, I collected responses from 48 engineers over the phone and online about their experiences in tech hiring processes. Of the respondents, 30 identified as women, 17 as men, and 1 as nonbinary. I found that my story wasn’t unusual among women in tech.
38 percent of women stated that they felt that their interviewer had lower expectations of their interview performance due to their gender. An additional 10 percent said that they weren’t sure whether their interviewer had lower expectations. Nearly every women that I spoke to on the phone (as opposed to filling out an online form) had a story to go along with this.
One young woman interviewing at a tech giant reported being on the right track with a hard problem, only to be told it wouldn’t work, and was forced to switch approaches mid-way. Another woman said that her interviewer refused to believe that her solution in fact ran in O(n^2) time and O(n^2) space, despite her experience as a TA for her school’s complexity theory course. “My solution was clever, and I don’t think they were expecting me to be that clever,” she said.
Perhaps due to the negative experiences that many women reported having, women are more sensitive than men to the actual tech interview process. Women were far more likely to state that the quality of the interview process was a deciding factor in choosing between multiple offers. This shows that the specifics of an interview process really matter in getting more diversity. Working to eliminate preventable instances of bias like those reported above are a good start.
Women also expressed a strong preference to be interviewed by women. 70 percent of women felt at least somewhat more comfortable being interviewed by a female engineer. However, there was a wide spectrum of opinion on the matter. Artificially pairing female interviewers with female interviewees is not a solution.
Men and women also differ significantly in terms of which interview format they most prefer:
79 percent of the women I spoke to most preferred the in-person interview format, as compared with 29 percent of men. Counter to my hypothesis, the men we surveyed tended to prefer take-home programming projects more than women. That being said, women preferred take-home programming projects and phone interviews roughly equally as an initial step.
Solutions
Based on the aggregate experience of those who I interviewed and existing research, here a few concrete things that companies can do to increase diversity in their hiring processes:
Offer a choice of interview formats. My data shows significant variance of interview format preferences by gender. Offering candidates a choice of interview formats can allow them to showcase their strengths and reduce performance anxiety.
Standardize interview processes. The quality of an interview process can vary drastically. These differences are perceived by women, evidenced by my findings. Standardization of interview processes have been shown to reduce room for bias. The Clayman Institute for Gender Studies found that the more clear-cut the criteria for interviews, the more women and minorities are hired. By creating a rubric that interviewers can run through as they go, one can make the process more objective and fair for everyone.
Use a variety of sources. Referrals are biased towards those with connections and already in Silicon Valley, and have been shown to perpetuate the existing demographics of a company. Relying on sources other than purely referrals can help ameliorate an existing gender imbalance. By opening up the pool of applications beyond referrals, and/or explicitly ask employees to refer those from nontraditional backgrounds, one can attract diverse candidates.
Gather data. Gather more data on your candidates’ experiences in order to understand what is working and what is not. For example, is it that minorities in tech are not applying to your company in the first place, or are they dropping out at a particular stage in the process? The data on the problem is sparse.
The tech industry’s gender problems are subtle and self-perpetuating, with male-dominated companies at a disadvantage in hiring women. Rethinking the fine points of hiring offers a path forward. Companies that are thoughtful in their hiring processes can put themselves at an advantage in attracting a more diverse candidate pool. Such companies are less likely to pass over talented candidates that don’t happen to be in one’s network or are non-standard in some way. In addition, the problems of lack of standardization and network effects are not unique to minorities in tech. These deliberate hiring practices ultimately benefit everyone, leading to a more equitable and excellent interview process.
Closing Anecdotes
“Years ago, I had a quick interview with PayPal at Grace Hopper. At the end, the guy said ‘you have too much personality to be a programmer!’”
“After this most recent round of interviews, I feel so much more interested and engaged with companies who had female interviewers. At a subconscious level, it really made me feel more comfortable and even inspired that there were other women at the company leading the way and showing their technical capabilities. I feel an indirect kind of support with female interviewers, and I think this feeling ultimately lets me relax more and do better during interviews.”
“I was told that the interview was going to be a super chill culture fit interview. But then they gave me some weird looks. At the end they said, ‘I hate doing this but I think we should: what’s the difference between an abstract class and an interface? What’s throw/catch?’ All these easy Java questions. I didn’t think I did anything strange but at the end they sprung all these questions on me. That’s why I didn’t accept the offer.”
“The interviewers had lower expectations…but my scores beat the boys I was competing against. :)”
