The Praise Problem¶
Definition¶
Praise is widely assumed to be a costless motivator — unlike money or prizes, it seems to carry no downside. Kohn argues this assumption is unsupported: studies on classroom praise found no reliable correlation with achievement gains, and several experiments found praised subjects performed worse afterward than subjects given neutral comments. He identifies four mechanisms: praise can signal that a task was easy enough to need only encouragement, implying low ability; it raises pressure to keep living up to the compliment; it teaches people to avoid difficult tasks where they might risk failing and losing the praise; and, like other rewards, it can undermine intrinsic motivation in the activity being praised.
In the Book¶
Kohn cites an Israeli study of fifth and sixth graders on a creativity task: those praised for their performance went on to do lower-quality work on a similar task than neutrally-treated peers — worse even than their own prior performance. A series of studies with American college students found praise "consistently led to impairment in skilled performance," concluding that complimenting someone right before a task is an effective way to disrupt it. He quotes Sylvia Plath describing acceptance as something that "ironically freezes me at my work," and painter Andrew Wyeth on fearing he'll "lose it" once a work-in-progress is praised. Kohn also examines the case for praise being less harmful than other rewards — it's less salient, feels less controlling, and usually comes as a surprise rather than a promise — and shows that whenever a specific instance of praise is salient, controlling ("You're doing fine, as you should be"), or expected, it produces the same negative effects as any other extrinsic reward.
Why It Matters¶
This overturns a nearly universal management and parenting instinct: that praise is the safe, unlimited-supply substitute for more manipulative incentives. It gives a testable criterion instead of a blanket rule — ask whether a given instance of praise is salient, controlling in tone, or anticipated by the recipient — and predicts that praise passing that test will undermine performance the same way money or prizes do, while praise that doesn't (offered as genuine, unexpected, non-comparative information) is less likely to.