The New York Times’ David Gelles reported this week Salesforce CEO Marc Benioff has, for several years, been reaching for gender diversity and equality. “The goal was to achieve 100 percent equality for men and women in pay and promotion, and to make sure that at least a third of all participants at any meeting were women.”
A multi-year effort toward gender parity is far from common in today’s tech industry. Last year, Microsoft’s Satya Nadella stood before an audience of female tech computing professionals and told them that not asking for raises might be their superpower, prompting sharp criticism from many women (including me). The Register wondered aloud whether Nadella used the “female superpower” to get his $84 million dollar paycheck and called him a “female-baiting overlord.” Looksmart’s Evan Thornley gave a talk about how he came to be one of the few tech leaders who hires many women. The talk might have been well received had he not posted a slide that read: “Women: Like Men, Only Cheaper.” Getting hired is one thing, getting equal rewards is another.
Computing stands out from other technical fields: it’s the only one where women’s participation has been declining over the long term. And a series of recent disclosures revealed what many already suspected: that significant tech firms including Google, Apple and Facebook don’t come close to gender parity. Many major tech firms won’t even release data on employment diversity.
I’m an active booster of women in tech and particularly women in my own field of analytics, and have found that even writing about accomplished women often attracts resentful responses from men. Well-meaning men tell me their process isn’t biased, women just don’t want those roles. Angry men accuse me of complaining. There’s truth in that, I do complain about bias, and I believe it’s my duty to speak out and act against it. These responses raise an important analytical question: what measures can we use to assess bias and combat it?
Benioff started with an informal observation that he wasn’t seeing women in his management meetings. Simple numbers are also the easy place to begin measurement:
- How many women (or members of any particular worker segment) work here?
- What jobs do they hold?
- What’s the proportion of women in each job role?
Informal evidence was good enough for Benioff until he was confronted with a claim he did not believe. When his new, female, head of human resources and another female executive suggested that women were not paid as much as men, he balked. But Benioff was willing to invest in a salary study to get that data, and willing to act on the results. This introduces a more subtle, but very important layer of measurement: pay.
- What’s the overall average pay and pay range for women? Men?
- Within job categories, how do average pay and pay ranges compare for men and women?
Looking into pay will surely raise questions about merit – should women expect pay equal to men? What if they don’t have the same roles? What if their qualifications/hours/results are not the same? Well, those are good questions! What if more organizations put serious effort into measuring those things? What if they pushed measurement all the way back to examining what happens to job applications and even how they attract applicants in the first place? What if?
These are challenging things to measure, and doing so calls for careful formulation of questions and methods for getting answers. The effort required to measure may be holding some executives back from asking whether women and men get equal treatment in their businesses, but it’s surely not the only reason.
If more organizations invested in that kind of measurement, they might find hard evidence of the discriminatory patterns in hiring and promotion that outside researchers have been documenting for years. They might find patterns like these:
What in a name? Maybe an interview.
- Resumes with male names tend to be rated higher than identical resumes with female names.
- Resumes with “white” names draw more interviews than resumes with “black” names.
Merely seeing that someone is a man or a woman affects the chance of being hired
- “Without … information about candidates other than their appearance, men are twice more likely to be hired for a mathematical task than women. …. Providing full information about candidates’ past performance reduces discrimination but does not eliminate it.”
- As Jezebel’s Jasper Hamill put it, hot guys get more interviews. “Employer callbacks to attractive men are significantly higher than to men with no picture and to plain-looking men, nearly doubling the latter group. Strikingly, attractive women do not enjoy the same beauty premium.”
Women get paid less from the start.
- Female graduates in roles as varied as MBAs and academic scientists are paid less after graduation than male peers.
- Gender differences in college major do not fully explain the gap in women’s pay.
Why don’t more tech executives look at the data? Maybe it’s just because they aren’t prepared to act on what they find.
