Overjustification Effect

51 preschoolers who freely drew with magic markers were assigned to three conditions. Expected reward — promised a certificate before drawing. Unexpected reward — given a certificate as surprise after. No reward — drew freely. Two weeks later, free-play observation showed expected-reward children spent only 8.6% of time drawing, vs. 16.7% for unexpected-reward and 18.1% for control. The reward had to be expected to undermine motivation. The contingency, not the reward itself, was the problem.

Two complementary accounts. (1) Self-perception (Bem 1972) — people infer their own motivation from observing their behavior in context; reward provides an obvious external explanation that crowds out internal attribution. "I must be doing this for the certificate, not because I love it." (2) Self-determination theory (Deci, Ryan) — controlling rewards undermine autonomy, a basic psychological need that supports intrinsic motivation. Both effects compound.

No. Conditions matter. Robust when. (1) Task is initially interesting. (2) Reward is tangible (money, prizes). (3) Reward is expected and contingent on performance. (4) Reward is salient. Less robust or absent when. (1) Task is boring (extrinsic motivation is needed regardless). (2) Reward is verbal praise. (3) Reward is unexpected. (4) Reward signals competence rather than control. (5) Ongoing reward continues. The headline "rewards backfire" oversimplifies a moderated effect.

Carefully. For routine, low-interest work, performance pay can sustain effort without undermining anything. For creative or self-directed work, extrinsic incentives often hurt. Ariely et al. (2009) — large performance bonuses can degrade complex cognitive tasks via choking, even before any motivational shift. Daniel Pink's Drive (2009) synthesized these findings for management. Many tech and creative companies have moved away from individual cash bonuses for this reason.

Bribing children to do things they already enjoy is risky. Common pattern: parent rewards reading, child reads less when reward stops. Better practice — provide structure and choice, praise effort and process, save tangible rewards for chores or tasks the child finds intrinsically aversive. Praise that conveys autonomy and competence ("You worked hard on that") supports motivation; praise that controls ("You did exactly what I told you") doesn't.

Partially. Operant conditioning predicts reinforced behavior should increase. The overjustification effect shows this is true while reward continues but can reverse below baseline once reward stops — for already-interesting tasks. The findings expanded behaviorism by showing that the meaning of a reward matters, not just its delivery. Modern motivation theory integrates both, recognizing that rewards interact with prior engagement and perceived autonomy.

Many gamification efforts add badges, points, and streaks to activities (exercise, learning, work) hoping to boost engagement. Short-term engagement often rises; long-term and post-removal engagement often drops below baseline. Best practices recommend designing reward structures that signal competence rather than control, fade rewards as mastery grows, and let the activity itself become the reward. Duolingo's streaks work in part because they align with users' existing language-learning goals; many corporate gamification rollouts fail because the underlying activity isn't interesting.