Have a bad manager? Science explains why.

“People who are really good at being individual contributors automatically get thought of as management material,” says David Deming, a labor economist and dean at Harvard College. That instinct, he argues in a new study, is often exactly wrong. Deming and his colleagues set out to answer a question that haunts nearly every workplace: can you scientifically identify who will be a good manager before they ever manage anyone? Their findings challenge some of the most common hiring and promotion practices, and the results are not flattering.

Why companies pick the wrong managers

Deming begins with a blunt diagnosis: most organizations are terrible at selecting managers. When a management slot opens, companies typically default to one of two strategies. The first is promoting a star individual performer — the salesperson with the highest numbers, the engineer who writes the cleanest code. The assumption is that excellence at a task translates into excellence at leading others to do that task. Deming says that assumption is unsupported. The second common approach is the job interview, which Deming calls “a pretty terrible technology for figuring out what people can do.” Interviews, he argues, load too heavily on whether the interviewer likes the candidate or feels an unconscious sense of affiliation — a shared hobby, a similar background, a comfortable vibe. “If you and I bond