Points can erode existing intrinsic motivation

What it is

Adding extrinsic incentives (points, rewards) to behaviour people already do for intrinsic reasons (enjoyment, meaning) can reduce motivation—the crowding-out effect, known in psychology as the overjustification effect. The person's attribution for the behaviour shifts from "I like doing it" to "I do it for the reward," the reward becomes the behaviour's main justification, and when the reward disappears or devalues, the behaviour loses support. The most counterintuitive rule of gamification design: adding points to something people already love may destroy the love.

Why it happens

Crowding-out works through attribution and self-perception. People infer their own motives by observing their behaviour and its accompanying conditions (self-perception theory): when behaviour comes with a salient external reward, "I do this because of the reward" is the more available explanation, and intrinsic reasons (interest, meaning) lose attribution weight. Rewards can also shift the behaviour's interpretive frame—from "my choice" to "a transaction"—damaging autonomy. Meta-analytic work refines the conditions where the mechanism fires: it is strongest for activities already high in intrinsic interest, for expected tangible rewards, and for rewards contingent on task performance; it is weak or absent for dull tasks, unexpected rewards, or rewards framed as recognition rather than transaction (verbal praise, symbolic awards). These conditions map precisely onto gamification's danger zone: the most intuitive design of "adding points to behaviour users already enjoy" sits inside it.

Studying it

• Paradigms: the classic experimental design (Deci's Soma cube studies and decades of replications and extensions) measures free-choice-period engagement in the target behaviour before and after reward introduction; field experiments in education, health, and open-source communities compare long-term behaviour maintenance between incentivised and control groups.
• Variables: independent variables include reward expectancy, tangibility, performance contingency, and the task's initial interest level; dependent variables are post-withdrawal intrinsic-motivation indicators (free-choice behaviour, self-reported interest, engagement in later unrewarded periods).
• Methodological cautions: short experiments miss crowding-out (it appears after reward withdrawal), and only long-running field tracking (months and beyond) is reliable; cultural differences are well documented (some reward forms crowd out less in collectivist cultures), so cross-cultural extrapolation needs care; crowding-out does not apply to all behaviour—for tasks with no prior intrinsic motivation, extrinsic incentives are usually a net gain.

Where it stops holding

Crowding-out does not apply to every incentive scenario. For behaviours that lack intrinsic motivation in the first place (compliance requirements, dull but necessary processes), extrinsic incentives are usually a net gain—there is nothing to crowd out, and the incentive supplies motivation that was simply missing. Individual and cultural variation in crowding strength is significant: it runs stronger among participants with high task interest and in autonomy-emphasising cultures, so laboratory conclusions need re-verification in each concrete product. Incentive form also draws the boundary: unexpected rewards and recognition-form rewards (verbal praise, honour marks) barely trigger crowding-out, while expected, exchangeable rewards are the high-risk form. Finally, the evidence base comes mostly from free-choice, low-stakes tasks; the relation for high-stakes occupational behaviour (pay and work) is far more complex, so "bonuses destroy love of the work" should not be extrapolated to every workplace.

Applying it

• Assess the behaviour's existing motivation structure before gamification: do users already want to do it (high crowding risk) or need pushing (net incentive gain)? Trade-style points deserve extra caution in the former case.
• Shift incentive form from transaction to recognition: feedback emphasising progress and achievement (rather than exchangeable value), informational rewards ("you're improving") replacing controlling ones ("do this, get that").
• Verification: record a baseline of unrewarded behaviour before launch, then measure "incentive attribution" periodically (self-reported "I do this because I enjoy it / because of points"). Attribution drifting from intrinsic to extrinsic is crowding-out's early signal—dialling back incentives at that point costs less than withdrawing later.

Related

• Same group: W11.02.2 Behaviour falls back when rewards stop · W11.02.3 Rewards must strongly relate to the target behaviour
• Nearby: W11.01 When gamification applies · C3.09 Decision-making and loss aversion · P2.02 Autonomy and sense of control
• Search terms: overjustification effect · motivation crowding · intrinsic motivation · self-perception theory