I was reading through the New York Times “Year in Ideas” online the other day, and I stumbled across a fascinating article about the most effective way to promote employees within a large company.
The article references something called the “Peter Principle,” named after Canadian psychologist Lawrence J. Peter (or is it Peter J. Lawrence? No one will ever know). Peter/Lawrence posits that we wrongly assume that people who are really good at their job will also be really good at the next job on the ladder. I can definitely see that–just because Jane is a good prostitute doesn’t mean she’d also be a good madame.
The article mentions a recent study of a simulated 160-person organization. The study ran a couple different experiments: In one, they promoted the best performers. In the other, they randomly promoted people. In the end, the company simulation with the random promotions ended up on top.
As crazy as this sounds, it makes sense. If someone is really good at their job, why would you promote them to a new job? Why not just give them a raise and more related responsibilities. Again, just because Patrick is a good copyeditor doesn’t mean he’ll be good at managing copyeditors. Sure, you could try to identify good managers among all copyeditors, good and bad, but there are so many other factors–work ethic, punctuality, professionalism, composure, and so on. Unless you have a brilliant model to weigh all these factors, why not promote randomly?
The overall message is that if you have someone who is clearly the right person for the promotion, promote them. If not, put a bunch of names in a hat and see how things work out. After all, if the person you promote is incompetent, they probably won’t last long anyway.