SERIOUSLY. This was implied by a survey a client of mine recently completed. The survey, from a reputable organization that is dedicated to creating better workplaces and businesses, asked about a wide range of business practices including compensation. One of the questions focused on the pay differentials between male and female employees.
The topic was basically framed as follows. “Women at our company make the following compared to men: 0% – 25%, 25%+ - 50%, 50%+ - 75%, 75%+ - 100%.” You may immediately see the problem, or maybe not. No, it is not a lack of granularity. It is not a problem of defining “compensation” or “pay”. Do you see it yet?
The analyst doing the work came back to the head of HR and pointed out the issue. There was no way to indicate what was happening at their company. In looking at the data, they found that the average female employee actually made more than a similarly positioned male employee. Of course, there is nothing objectively wrong with that, but the survey didn't even consider it as a possibility. The survey, in effect, had implemented its own glass ceiling.
Imagine if you had a survey that asked the same question in the opposite direction, with a bit plainer English. “How much less do the men make at your company, when compared to the women: 75% less, 50% less, 25% less or, they make roughly the same.” If most of us saw a question like this, we would first laugh and then we would ask where the rest of the possible answers were (“25% more, 50% more, please don't ask me to answer that”.)
It is impossible to review accurate data when the data collection process is flawed. Bad questions lead to bad data. More importantly, it is upsetting that a question like this can be asked by smart well-meaning people without anyone noticing the bias (or assumptions) that the given answers imply. My client pointed out the problem. They also provided both the answer requested and the accurate answer as a backup. I wonder how often we get the answers we ask for rather than the answers we really need.
And, on a closing note regarding today’s issue of chromosomes in the workplace, the New York Times discovered, as reported on HelloGiggles, that CEOs are more likely to be named John (or David) than they are to be women. (There was no information provided on CEOs who are women named John.) Not really sure how this directly relates to compensation, but I guess we should ask one of the executive compensation data firms to research CEO names as related to compensation data and provide a “best names” list. Have fun with that!
Dan Walter is the President and CEO of Performensation a firm committed to aligning pay with company strategy and culture. Dan, Ann Bares and Margaret O’Hanlon of the Comp Café were recently honored to have their book “Everything You Do in COMPENSATION IS COMMUNICATION” called one of the “best comp books I’ve seen”, by Steve Browne of HRNet email fame. Dan has also co-authored of several other books you may find useful including “The Decision Makers Guide to Equity Compensation”, “If I’d Only Known That”, and “Equity Alternatives.” Dan welcomes connections on LinkedIn. Follow him on Twitter at @Performensation and @SayOnPay.