Has it ever occurred to anyone that having employees design incentives is a bit like letting the foxes run the henhouse?
Lessons to learn:
- Incentives don’t work if they are insulting, stupid, seemingly arbitrary or creepy.
- Even during a recession, the threat of getting fired is really not a great incentivizing force. At best, it gets people to put in exactly the minimum effort to not get fired (and really screws with morale to even threaten to fire people), and it’s surprisingly hard to actually fire someone for cause in most companies.
- This is the same sort of circular reasoning that makes it hard to just hire a manager to monitor employees…don’t you then have to hire another manager to make sure that the first manager is doing his job? And a third to make sure that the second is doing his job? etc.
- Simply allowing the creators of an incentive to get the incentive itself is not a solution to the problem- wouldn’t the incentive designers just set up the criteria so that they would benefit the most from the incentive? Duh.
This article gives a similar perspective as the Dilbert cartoon. Though titled “Incentive Pay Considered Harmful,” the article is mainly referring to silly awards like the one referenced above. To quote the opening paragraph:
“Mike Murray, a surprisingly hapless HR manager at Microsoft, made a number of goofs, but the doozie was introducing a “Ship It” award shortly after he started the job. The idea was that you would get a big lucite tombstone the size of a dictionary when your product shipped. This was somehow supposed to give you an incentive to work, you see, because if you didn’t do your job– no lucite for you! Makes you wonder how Microsoft ever shipped software before the lucite slab.”
While I don’t agree with the overall conclusion that incentives can never work, I do like the point regarding happiness versus disappointment with performance reviews. The general idea is that people are overconfident or optimistic enough that a good performance review can’t really come as a positive surprise. As a result, people are either neutral or disappointed by their reviews, which is counterproductive to a large extent. I am thinking about how to design an experiment to get at these sorts of effects.
On a sidenote, am I the only one that gets particularly upset when money is spent to buy stupid recognition items that no one really wants? It irks me mainly on principle, since I am rarely cooperative enough to get such an item. Nonetheless, I cannot get over the inefficiency due to the fact that the recipients of the lucite slabs likely don’t even value the award as much as it cost to buy the physical incarnation of it. If a manager wanted to do better, he could give cash instead…but then it would be painfully obvious exactly how cheap the award really was. To illustrate, I remind you of a scene from The Office (a recurring favorite of mine):
Angela: Gift baskets are… the essence of class and fanciness. They are the ultimate present that a person can receive.
Dwight: What about cash? With cash you can buy whatever you want, including a gift-basket, so… it’s kind of the best gift ever.
Jim: What about a gift basket full of cash?
Andy: Yes! Cash-basket! Nice work, Tuna.
(For the record, I think B.J. Novak’s college roommate was an econ major. That could explain a lot.)