Rewards Are Still Bad News 25 Years Later

October 28, 2018

Rewards Are Still Bad News (25 Years Later)

By Alfie Kohn

[This is a slightly expanded version of the published article, which was titled “Science Confirms It: People Are Not Pets.”]

The field of social psychology is sometimes accused of doing no more than ratifying common sense, so it’s worth paying attention when its findings are genuinely surprising. Experimental results that challenge entrenched beliefs and practices are even more noteworthy. Case in point: the discovery that when we are rewarded for doing something, we tend to lose interest in whatever we had to do to get the reward. This outcome has been confirmed scores of times with all sorts of rewards and tasks, and across cultures, ages, and genders.

Does this mean that rewards don’t actually “motivate” people? That depends on how you define the word. Psychologists point out that there are different kinds of motivation, and the kind matters more than the amount. Extrinsic motivation (for example, doing something in order to snag a goodie) can be contrasted with intrinsic motivation (wanting to do something for its own sake).

It’s not just that these two are different but that the first tends to undermine the second. Intrinsic motivation (loving what you do) is also the best predictor of high-quality achievement, which is why — brace yourself for another counterintuitive discovery — people promised a reward for doing something often end up doing it more poorly than people who weren’t.

I began tracking, and trying to explain, these lines of research back in the late 1980s. This year, for the twenty-fifth anniversary edition of my book Punished by Rewards, I reviewed as many studies as I could find that were conducted since its original publication. The conclusions that rewards frequently kill both interest and excellence have, if anything, grown more solid in the intervening decades. Yet many teachers, parents, and managers persist in using various versions of what has been called “sugar-coated control.”

A number of studies, for example, have shown that children are apt to become less concerned about others’ well-being if they were rewarded earlier for helping or sharing. Students, meanwhile, become less excited about learning once they’ve been given a grade (or some other artificial inducement) for doing so. And even though the average American corporation resembles a giant Skinner box with a parking lot, no controlled study has ever, to the best of my knowledge, found a long-term enhancement in the quality of work as a result of any kind of incentive or pay-for-performance plan.

Over the years, researchers have investigated some intriguing questions that stemmed from these basic findings. To wit: What if the reward is really large and luscious? (Answer: In that case, it’s apt to do even more damage to intrinsic motivation.) Are rewards destructive because they distract people from the task? (Apparently not, because other distractors don’t have the same negative effects.) Which is worse, giving people a set reward for doing something or making it contingent on how well they do it? (The latter, by a long shot.)

Give a bunch of adults or children a puzzle to solve, a poem to write, or something else to do that requires creativity. Promise half of them a reward if they’re successful — and then watch as they end up being less creative (and less interested in the task) than those promised nothing. Such studies have led some observers to conclude that rewards should be avoided for interesting tasks but that they may be harmless or even appropriate when people with more power want to make people with less power do boring stuff. (If there’s one thing this field has taught me, it’s that rewards, like punishments, are ultimately about power.)

But newer research finds that rewards may also backfire when they’re offered for doing things that aren’t especially interesting, particularly if you watch to see what happens after the rewards stop coming. For example, a 2008 study by Thane Pittman of Colby College and his colleagues found that when people put off doing something — which often happens when a task seems unappealing — a reward offered for finishing early either didn’t help or actually led to increased procrastination.

Or what about rewarding people just for showing up? In 2015, researchers at Hong Kong University and New York University studied nine year olds in a very low-income area of India whose school attendance was spotty. These children were promised a reward if they came to school at least 32 out of 38 days. During that period, not surprisingly, many kids’ attendance improved. Afterwards, however, it promptly dropped — either back to the earlier low levels or, in the case of students on whom the reward hadn’t had even a temporary effect, to a level much lower than it had been to begin with.

Another study, conducted by Carly Robinson at Harvard and her colleagues, and released as a working paper this past summer, cast a much wider net. They followed more than 15,000 students in fourteen California school districts, watching to see whether those who received a reward for exemplary attendance in the fall would come to school more often in February as compared to those who hadn’t been rewarded. Again, the rewards either had no effect or actually led to poorer attendance.

I should pause here to note that what’s even more striking than studies that challenge popular assumptions about human behavior are studies whose results surprise the researchers themselves. (Hint: Watch for the phrase “contrary to hypothesis.”) That was the case with both of these attendance studies — even though their findings really were predictable in light of earlier research. In fact, by now it should be clear that the trouble doesn’t lie with the type of reward, the schedule on which it’s presented, or any other detail of implementation; the problem is with the outdated theory of motivation underlying the whole idea of treating people like pets — that is, saying: Do this, and you’ll get that.

Indeed, various researchers over the last half-century have admitted to being surprised by the ineffectiveness or destructiveness of rewards when money was offered to adults for succeeding at a tricky task, when movie tickets or praise was offered to children for tasting an unfamiliar beverage (kids in both groups ended up liking the beverage less than those who received neither a tangible nor a verbal reward), when merit pay failed to improve teachers’ performance, and when incentives didn’t increase seat belt use or help people lose weight and keep it off.

The best that carrots — or sticks — can do is change people’s behavior temporarily. They can never create a lasting commitment to an action or a value, and often they have exactly the opposite effect….contrary to hypothesis.

The new California attendance study challenged one other expectation that had been shared even by many psychologists. Offering students a reward unexpectedly, after the fact, turned out to be even worse than telling the students about it in advance. Earlier studies are actually mixed on this question, with some showing that unexpected rewards typically didn’t help, while expected rewards had a negative effect.

And even when presenting subjects with a surprise reward didn’t hurt, that was only in contrived, one-shot lab studies. If you get an unexpected reward in the real world, you’re likely to expect another one next time. Then you may feel resentful if you don’t get one — and manipulated if you do.

In case you’re wondering, attendance rewards probably don’t fare any better — certainly not over the long haul — when used with adults. Timothy Gubler at the University of California, Riverside, and his colleagues published a study in 2015 with employees at an industrial laundry cleaning company in the Midwest. The rewards’ effect was disappointing across the board, particularly with workers who already had a good attendance record. They showed up at work less often after they had received what amounted to a patronizing pat on the head from the boss.

Working with people to help them do a job better, learn more effectively, or acquire good values takes time, thought, effort, and courage. Doing things to people, such as offering them a reward for jumping through someone’s hoops, is relatively undemanding for the rewarder, which may help to explain why carrots and sticks remain stubbornly popular despite decades of research demonstrating their failure.

In the case of attendance, it’s a lot easier — and much less threatening to those in positions of authority — to reward students or workers for showing up than it is to reconfigure schools or workplaces so that people are more likely to want to show up.

To be notified whenever a new article or blog is posted on this site, please enter your e-mail address at