Ocean Springs, Mississippi Another article in the holiday reading pile was a piece in The Atlantic called They’re Watching You at Work. The reporter, Don Peck, reviewed the current utilization of “big data” to create something that big company human relations departments were calling “people analytics.” In a nutshell these outfits are crunching the information they have received in the recruitment process from tens of thousands of applicants, comparing it with other information they have on workers they judge to have been successful by their standards, and then trying to devise tools to fit these two pieces together to create the perfect decision in hiring a new employee.
At some levels, who is surprised? And, alarmingly, using the examples from Michael Lewis’ book Moneyball about the ways that general manager Billy Beane reconstructed the Oakland Athletics into a winning team with less money in a smaller market using a wider array of statistics than just the veteran scouts’ eyeball looks and long experience is seductive, because at its heart that book was about how biases in baseball influence the recruitment of talent regardless of production and performance.
In fact, Peck ends up making the case that people analytics might be better for workers given how much bias continues to be ingrained in the employment process. The cited studies that revealed that there was bias based on whether names sounded white or black; others that revealed bias when employers knew the applicants were men versus women; and, very interestingly, bias in the way those hiring tended to hire people who they felt had a personality, background, or interest, “similar to mine.” The most hardcore Luddite among us would be hard pressed to defend any of these hiring “methods” as anything but biased, but in some cases we would have to concede that almost inevitably any of us could have unwittingly been guilty of some of biases, especially the one of tending to hire people who we think might have elements of our own experience. It may not be the “old school tie,” but it’s bias none the less and springing from the same roots. It’s certainly no secret that other studies have found that the most effective way to find a job for seekers is through personal networks that connect people to job openings.
Nonetheless it was scary to read about big companies that would like to hire people solely on their scores based on taking various digital games that might reveal this data, since you can almost throw away folks from the job pool who have been trapped on the other side of the digital divide immediately. The number of HR recruiters who expressed the desire to completely discontinue any person-to-person interviews and just hire from the tests was startling as well, even give the fact that the direct interviews would likely trigger the systemic biases.
This was one of those articles where you got to the end more depressed than informed. The brave new world of employment based on people analytics, even if less biased perhaps on some levels, seems to erect some permanent, unscalable barriers virtually at birth for millions posing some ethical and societal challenges that remain unanswered.