The overjustification effect: when rewards backfire

Evidence grade: STRONG, with well-mapped boundaries. The core effect — expected, tangible, contingent rewards reducing subsequent intrinsic interest in an already-interesting activity — is meta-analytically established. The equally important boundary conditions (unexpected rewards, informational feedback, uninteresting tasks) are part of the same settled literature.

In 1973, Lepper, Greene and Nisbett ran one of psychology's most consequential simple experiments. Preschoolers who enjoyed drawing were split three ways: promised a "Good Player" certificate for drawing, given the same certificate unexpectedly afterwards, or neither. Days later, in free play, the promised-reward children drew roughly half as much as the others. The reward hadn't added motivation on top of interest — it had replaced the interest with a transaction, and when no transaction was on offer, the drawing went with it. They called it overjustification: the child, observing themselves drawing-for-a-certificate, concludes drawing must be the kind of thing one does for certificates.

What held up

This finding attracted decades of dispute (notably from behaviorist quarters) and survived it in refined form. Deci, Koestner and Ryan's meta-analysis of 128 experiments (1999) settled the shape of the effect:

Expected, tangible, contingent rewards — offered in advance, for engaging with, completing, or performing an interesting activity — reliably undermined later free-choice interest. Completion- and engagement-contingent rewards were among the worst offenders.
Unexpected rewards did not undermine (no transaction was ever framed).
Verbal praise tended, on average, to support intrinsic motivation — when informational ("your shading choices are getting deliberate") rather than controlling ("good boy for drawing").
For dull tasks, there's little intrinsic interest to undermine — rewards there are less dangerous, though building internalized value (rationale-giving) still beats renting behavior.

The mechanism sits comfortably in self-determination theory: rewards experienced as controlling shift the perceived origin of one's behavior from "I choose this" to "I'm being made/paid to" — and behavior rented is behavior that stops when the rent does.

Why this haunts education and edtech specifically

Learning is the domain where we most need motivation to outlast the incentive — the entire point is a child who reads, types, checks sources, and debugs after the sticker chart is gone. It is also the domain most saturated with engagement-contingent reward machinery: token economies, points-per-minute, prize wheels, "play to earn" learning apps. The overjustification literature doesn't say all of it is harmful; it says the burden of proof sits on any design that announces a trinket in advance as the reason to learn — precisely the pattern most gamified products lead with (our fuller account of the gamification evidence covers the classroom trial where badges-and-leaderboards lowered exam scores).

The workable line, drawn from the boundary conditions: rewards as byproducts, not bargains. Progress that produces recognition ≠ recognition dangled to produce progress. The same object — a star, a badge, a cosmetic — sits on either side of the line depending entirely on framing and timing.

What the evidence doesn't say

It doesn't say never reward anything — unexpected recognition, informational praise, and celebration of mastery reached are repeatedly safe in this literature.
It doesn't say grades and incentives have no place — it prices them, so you spend them where interest doesn't yet exist rather than where it does.
It doesn't mean children shouldn't enjoy earning things — the earning-as-side-effect structure is exactly what survives scrutiny.

In the classroom

Never pay for what's already loved. Find what each child freely does — reading, drawing, coding — and keep transactions away from it.
Swap "if-then" for "now-that": "now that you've mastered the ledger, here's your certificate" carries none of the freight of "finish the ledger and you'll get one."
Praise process, informationally — specific, true, about the work.
Audit reward apps with one question: does it lead with the prize or with the progress? The order tells you which side of the line it lives on.

How Wiz Kids applies this

Our economy is built on the now-that side: sparks and stars result from practice and mastery (never offered in advance as inducements — "do X to win Y" phrasing is banned in our content style guide), cosmetics are earned-only and purchasable never, celebration fires at mastery moments, and the product's own fiction gives reasons ("practice makes the spell yours forever") rather than bargains. Where a child already loves a realm, nothing steps between them and it.

References

Lepper, M. R., Greene, D., & Nisbett, R. E. (1973). Undermining children's intrinsic interest with extrinsic reward. Journal of Personality and Social Psychology, 28(1).
Deci, E. L., Koestner, R., & Ryan, R. M. (1999). A meta-analytic review of experiments examining the effects of extrinsic rewards on intrinsic motivation. Psychological Bulletin, 125(6).
Lepper, M. R., & Henderlong, J. (2000). Turning "play" into "work" and "work" into "play": 25 years of research on intrinsic versus extrinsic motivation. In Intrinsic and Extrinsic Motivation.

© Glu IO Pty. Ltd. — Wiz Kids (wiz.kids). Link freely; republication requires permission — see terms. Found an error in our reading of the research? We correct fast: tell any teacher piloting Wiz Kids.