On February 19 former Uber employee Susan Fowler wrote an explosive blog post describing her time as an engineer at Uber. She described severe sexual harassment, a hostile HR department that would not or could not help her, and a toxic work environment that eventually pushed her to leave the company. Her essay has received nationwide attention — and alarm. The reaction of many could be summed up as: It’s 2017 and we’re still dealing with this stuff?
Sadly, we are. And it’s important to note that it isn’t only Uber that (allegedly) has a problem retaining and supporting female engineers. While some of the details described by Fowler sound extreme, many female engineers have reported that hostile work environments are still unfortunately common.
Last year the Center for WorkLife Law released a report, in conjunction with the Society for Women Engineers, called “Climate Control: Gender and Racial Bias in Engineering?” Based on a nationwide survey of over 3,000 engineers, we found that the climate for engineers is tougher for women than it is for men.
Female engineers feel they have to constantly prove themselves, over and over again, something we call “prove-it-again” bias. 61% of women of all races but only 35% of white men reported having to prove themselves repeatedly in order to get the same levels of recognition as their colleagues. Said one woman, “Women have to look more professional and demonstrate technical prowess at all times to receive the same respect as a male engineer who is just an average engineer.” Part of this is due to our ingrained stereotypes — when we think of a brilliant engineer, who pops into our heads? For most of us, it’s a white guy. So women are left having to work twice as hard, proving themselves twice as much, in order to be seen as equally competent. Fowler was told by higher-ups that she didn’t have an “upward career trajectory,” even though, as she pointed out to them, she had published a best-selling book, spoken at tech conferences, and done everything that an engineer on an “upward career trajectory” was supposed to do — a classic description of prove-it-again bias.
Another type of bias is what we call tightrope bias, which describes how a narrower range of behavior is accepted from women than from men. This can affect everything from how colleagues perceive you personally to how assignments are handed out. Our report found that women report being less able to behave assertively than men (51% and 67%, respectively) and less able to show anger without pushback (49% and 59%). Female engineers also reported having to do the “office housework,” such as taking notes and planning office parties, more than men, and having less access to desirable assignments than men. At Uber, when Fowler started responding to the constant barrage of misconduct by reporting everything she could to HR, she was told, by HR, that it was inappropriate to do so and that she — Susan — might be the problem.
Our research indicates that these two types of bias are only a fraction of the challenges present in engineering. Our report also found increased bias against women who are mothers, gender bias in workplace systems such as compensation and performance evaluations, and bias against people of color and older engineers.
The engineering field isn’t doing itself any favors by letting these toxic climates continue. People who have been hired for their technical expertise can’t assert themselves freely, and get stuck doing things like taking notes for a meeting. Does that sound like a good use of expensive talent? In some cases, climates that are extremely hostile toward female engineers may even be creating the risk of expensive litigation or embarrassing public scandals.
Organizations need to address these kinds of problems, take practical steps to eliminating bias against women, especially in the STEM fields, such as engineering, and fix their hostile climates. Companies that don’t address these problems will continue to hemorrhage talent. (Fowler reported that when she started working at Uber her department was 25% women; by the time she left it was 3%.)
Let’s be real: Just talking about needing more women in tech and talking about how sexual harassment is bad isn’t enough. And endless sensitivity and diversity trainings clearly aren’t helping. To really reduce or eliminate bias, what these organizations need are evidence-based tweaks to business systems, tweaks that are practical and easily implemented. We call them “bias interrupters” because they reduce or eliminate bias without the need for changing deep-rooted conscious or unconscious beliefs (which, while a worthy goal, can be a glacially slow process). Bias interrupters are based on metrics, so organizations can measure their success, document their progress, and put their words into action.
For example, the debacle that Fowler describes with regard to her performance evaluations largely could have been avoided through a few simple adjustments. Fowler was told that “performance problems aren’t always…about work” and that her personal life affected her “performance problems”; received vague comments about “not being technical enough”; and was dinged for not having an “upward career trajectory.” The research shows that women are more likely to be criticized for personality traits than men are, more likely to be labeled “difficult” for traits that are seen as acceptable in men, and less likely to get the kind of specific, skills-based feedback that all employees need to grow in their jobs. How could bias interrupters eliminate this kind discrepancy in feedback? The company could simply require that comments about skills be appraised separately from comments about personality. Performance critiques could be backed up with specific evidence from a specific evaluation period. Managers who make comments like “Susan is not technical enough” could be directed to give specific, constructive critiques. Potential and performance could also be evaluated separately, given the tendency for men to be judged on their potential and women on their performance.
Making small but effective reforms like these can interrupt bias so that we can move past saying we need to do better — and actually do better.
By Marina Multhaup and Joan C. Williams