Know the three conditions that produce overjustification
The effect requires: initial intrinsic interest + expected reward + reward contingent on doing the task.
Key takeaways
- What it is: The effect requires: initial intrinsic interest + expected reward + reward contingent on doing the task.
- Why it works: Lepper and colleagues demonstrated that the overjustification effect depends on a specific conjunction of conditions. The reward must be expected (known about in advance), tangible (not just verbal), and contingent on performing the activity. Change any condition and the risk profile changes significantly: unexpected rewards, quality-contingent rewards, or verbal praise for competence generally do not show the same undermining pattern. Understanding the three-condition structure lets you use incentives strategically.
- Evidence: Backed by randomized trials / meta-analyses.
- Avoid: Applying the overjustification caution to all reward situations, including low-interest tasks where there is no intrinsic motivation to undermine — which leads to withholding useful incentives unnecessarily.
Why it works
Lepper and colleagues demonstrated that the overjustification effect depends on a specific conjunction of conditions. The reward must be expected (known about in advance), tangible (not just verbal), and contingent on performing the activity. Change any condition and the risk profile changes significantly: unexpected rewards, quality-contingent rewards, or verbal praise for competence generally do not show the same undermining pattern. Understanding the three-condition structure lets you use incentives strategically.
How to do it
- 1Before offering or creating a reward, check: Is there initial intrinsic interest in this task?
- 2Check: Will the reward be known in advance (expected)?
- 3Check: Is the reward contingent on doing the activity, regardless of quality?
- 4If all three are yes, redesign the reward (make it unexpected, quality-contingent, or verbal).
What the evidence says
RCT / meta-analysisThe original Lepper, Greene & Nisbett (1973) study demonstrated the effect with children; subsequent meta-analyses have confirmed the pattern holds specifically for expected, tangible, task-contingent rewards applied to tasks with initial intrinsic interest.
Honest caveat: Cameron & Pierce (1994) ran a competing meta-analysis disputing the generality of the effect. The current consensus is that the effect is real for the specific conditions above, but "all rewards harm all motivation" is not supported.
- — Lepper, Greene & Nisbett (1973), undermining children’s intrinsic interest with extrinsic reward, Journal of Personality and Social Psychology
- — Deci, Koestner & Ryan (1999), Psychological Bulletin meta-analysis of 128 studies
Common mistake
Applying the overjustification caution to all reward situations, including low-interest tasks where there is no intrinsic motivation to undermine — which leads to withholding useful incentives unnecessarily.
IX Coach asks about your intrinsic interest level before any reward structure is added, and flags when the three conditions for overjustification are met so you can redesign the incentive accordingly.
Practice this with IX Coach →