Why might you get worse performance out of someone you’re paying than out of someone who’s doing the job for free? If you put on your economist’s hat, you’d say this is plainly irrational. If someone is willing to do a job for free, certainly they’d be willing to do a little better job for a little bit of money. After all, some money, no matter how small, is better than no money, right?
This is the question Uri Gneezy and Aldo Rustichini asked, and the answer appears to be no. In a variety of different experiments, Gneezy and Rustichini showed that small performance based incentives actually have detrimental effects on production. In one study, 160 students were paid $60 to show up to take an IQ test. The students were then divided into four groups: group one got no additional money, group two got an additional 10 cents per correct answer, group three got $1, and group four got $3. Surprisingly, the group that was paid 10 additional cents per correct answer got far fewer questions right than those who were paid nothing extra at all. There wasn’t much difference between the $1 and $3 groups: both scored higher than the group that received no extra money, and of course the group that received 10 cents per answer.
In a second experiment, a group of Israeli high school students went door to door, asking for donations for a charity, which is an annual event in Israel. The first group was promised no part of the donations they collected, the second group received 1% of the donations collected, and the third group received 10%. As we would expect, the group receiving 10% collected more than the group only receiving 1%. But shockingly, both groups received less than the group who didn’t get anything.
This phenomenon seems to work the other way too. A day-care center was having problems with late parents picking up their children after closing time. At the beginning of the experiment, the day-care center implemented a $10 fine for a delay of 10 minutes or more. The number of late parents suddenly shot up. After the day-care center stopped issuing fines, the number of late parents went up even more, plateauing and remaining constant at a rate far higher than the control group that never had a fine implemented.
Huh?
Gneezy and Rustichini present an interesting explanation for why this sort of irrational behavior occurs. They posit that once you enter into an initial agreement, you form an idea about your side of the bargain. Once you agree to pocket $60 in exchange for showing up to take the IQ test, your job is to answer the questions on the IQ test. But, if you are subsequently promised some rate for answering questions correctly, you correlate that rate with the amount of effort expected from you. If the rate is low, the amount of effort you’re willing to put forth is low. As the day-care center experiment indicates, once that change in perception is realized, it’s very hard to reverse.
So, Who Cares?
You should. Whether you’re trying to figure out how to pay your employees, how to collect the most donations for your charity, or just trying to get your buddy to help you move, understanding incentives is key. If you need help moving, and you call up a friend, you’re far better off just asking him to do it as a favor rather than offering to pay him $10. First off, the $10 is really easy to turn down. Once you turn that situation into a job offer, it becomes easy for your friend to decide that half a movie ticket isn’t worth sweating through his shirt all day and coming home with sore knees. Plus, even if you do get your friend to help, you’ve changed his perception of the effort you expect him to put in. Whereas he might have helped you schlep your boxes full of Japanese cartoons all day long if you asked him to do it as a favor, offering to pay him a measly $10 might net you a half hour of real help and six hours of standing around, drinking beers, and yelling at you to lift with your knees.
.