AI tools helped gather research and write the first draft of this article. I edited it and checked every study figure against the paper or its official abstract, listed at the end. The cover illustration is AI-generated. Details at the end.
Ten day-care centres in Israel had a familiar problem: some parents came late to collect their children, and a teacher had to stay after closing time. In six of the centres, chosen at random, the researchers Uri Gneezy and Aldo Rustichini introduced a small fine for arriving more than ten minutes late [1].
Lateness went up, not down. Across the six centres, the average number of late arrivals per centre per week rose from 8.0 before the fine to 16.4 over the twelve weeks with it, by my calculation from the paper's own table [1]. In the four centres without a fine, nothing much changed. Then the fine was removed. Lateness stayed high: 18.7 a week over the next four weeks [1].
One of the authors' explanations is in the title of their paper: "A fine is a price" [1]. Before, a late parent was imposing on a teacher. After, the parent was buying extra time at a posted rate. The money did not add to the reason people had for being on time. It replaced it.
That is the effect this article is about. Psychologists call one version of it the overjustification effect, economists call it motivation crowding out. The common idea is simple: an incentive can change how people understand what they are doing, and sometimes the new understanding is weaker than the old one.
This is the fourth part of the series Influence & Incentives, on how other people, rules and rewards shape what we decide, write and publish. Earlier parts looked at crowds, gifts and scarcity. This one looks at rewards, which crypto communities hand out more freely than almost anyone.
The first experiment: a reward for something children already liked
In 1973 Mark Lepper, David Greene and Richard Nisbett ran a field experiment with children who had already shown, during baseline observation, that they enjoyed a particular activity [2]. The children were split into three groups:
- Expected award. They agreed to do the activity in order to get a reward.
- Unexpected award. They received the same reward afterwards, without knowing about it in advance.
- No award.
Later, when the children could choose freely, the expected-award group showed less interest in the activity than both of the other groups [2]. The reward itself was not the problem: the children who got it as a surprise were fine. The problem was doing the activity in order to get something.
The researchers' interpretation drew on self-perception: if I am doing this for a prize, perhaps I am not doing it because I like it [2]. Once the reason has shifted, removing the prize does not automatically shift it back.
What 128 experiments found
One study with children proves little on its own, and the topic produced decades of argument. In 1999 Edward Deci, Richard Koestner and Richard Ryan pooled 128 experiments in a meta-analysis [3]. Their main results, as the official abstract states them:
- Rewards given for doing a task, for finishing it or for reaching a performance level all significantly reduced free-choice intrinsic motivation, meaning how much people kept doing the activity when nobody was paying. The effect sizes were d = −0.40, −0.36 and −0.28 [3].
- The same was true for tangible rewards as a group and for expected rewards as a group [3].
- Positive feedback worked the other way. It increased both free-choice behaviour (d = 0.33) and self-reported interest (d = 0.31) [3].
- Tangible rewards tended to do more harm to children than to college students [3].
These are moderate effects, not a collapse. But the pattern is consistent: a reward that people expect, that is material, and that is tied to doing the task tends to reduce the interest that was there before. Information that tells people they are good at something tends to raise it.
Two of the authors, Deci and Ryan, founded self-determination theory, which links motivation to three psychological needs: competence, autonomy and relatedness [4]. Read through that lens, a reward that feels like control can work against autonomy, while feedback that confirms skill supports competence.
When a fine becomes a price
Back to the day-care centres, because the details matter. The fine was NIS 10 per child for a delay of ten minutes or more, added to the monthly bill [1]. That is small but not trivial; the authors note that a babysitter at the time earned NIS 15–20 an hour [1]. The money went to the owner, not to the teacher who stayed late [1].
Two things stand out. First, the change was not instant. Lateness rose steadily over the first weeks and then settled at almost twice the initial level [1]. Parents seemed to learn what the new rule meant. Second, removing the fine did not restore the old situation [1]. Once the relationship had been described as a transaction, it stayed one.
The authors are careful: they do not claim every fine works this way, and they note that a very large fine would probably stop lateness altogether [1]. Their point is narrower. A penalty or a payment enters a relationship that already had unwritten rules, and it can change what people think those rules are [1].
Pay enough, or do not pay at all
In a second paper the same year, Gneezy and Rustichini tested the size of the incentive [5]. Their summary: the effect of money on performance "was not monotonic" [5]. Among people who were paid, more money produced better performance. But people offered a small payment did worse than people offered nothing [5], [1].
So the choice is not only between a bigger or a smaller reward. Introducing a reward at all changes the frame. A small payment can be enough to switch people from "I am helping" to "I am working", and too little to make the work worth doing on those terms.
It is not a law of nature
None of this means that incentives do not work. In 2014 Christopher Cerasoli, Jessica Nicklin and Michael Ford reviewed four decades of research, with 183 samples and more than 212,000 people from school, work and sport [6]. They found that intrinsic motivation was a medium to strong predictor of performance whether or not incentives were present [6]. Incentives predicted the quantity of performance better; intrinsic motivation predicted its quality better [6].
They also found a crowding-out pattern of their own: intrinsic motivation mattered less to performance when incentives were directly tied to performance, and more when the link was indirect [6]. Their conclusion is that incentives and intrinsic motivation "are not necessarily antagonistic" [6].
That gives a more useful rule than "rewards are bad". Rewards are good at producing more of a countable thing. They are weaker, and sometimes harmful, when what you want is care, quality or a behaviour people already did for their own reasons.
Rewards on social platforms and in crypto
Social media already runs on a reward. A 2021 study in Nature Communications by Björn Lindström and colleagues analysed more than a million posts from more than 4,000 people and found that posting behaviour followed the principles of reward learning: people spaced their posts to maximise the rate of social rewards such as likes [7]. An online experiment with 176 participants supported a causal link [7].
A field experiment on Twitter by Olivier Toubia and Andrew Stephen points in the same direction. By adding followers to non-commercial users and comparing them with a control group, they concluded that for most users "image-related utility", the wish to be seen, was a larger motive for posting than the intrinsic pleasure of it [8].
Crypto adds explicit, tradable rewards on top: points for activity, airdrops for on-chain actions, tokens for posts or quests. I have not found a controlled study that measures crowding out in these programmes, so what follows is an interpretation, not a measured result. The research above suggests what to watch for. Points are expected, tangible in effect and tied to doing a countable task, which is the combination that reduced free-choice interest in the meta-analysis [3]. They are likely to raise the quantity of activity, as incentives usually do [6]. Whether people stay once the rewards stop, and whether the activity has any quality, is a separate question.
Four signs that a reward may crowd out motivation
Taken together, the studies point to four warning signs:
- People already did it for their own reasons. Crowding out needs something to crowd out. Paying for a chore nobody liked is a different case [2], [3].
- The reward is expected and tied to the act. Rewards announced in advance and paid per action did the most damage; surprise rewards did not [2], [3].
- The reward is small or turns a relationship into a price. A small fine or payment can replace a social or moral reason with a cheap transaction [1], [5].
- You care about quality, not volume. Incentives drive quantity; intrinsic motivation drives quality [6].
Five questions before you pay for participation
These are for anyone designing a programme: a community lead, a protocol team, an editor, a publisher.
- Why do people do this now, without a reward? If the answer is interest, belief or helpfulness, a reward may replace that reason rather than add to it.
- Will people expect the reward and count it per action? If yes, expect more actions and ask what happens to the reason behind them.
- Is the amount large enough to justify the work on its own terms? If not, you may get the worst case: the old reason gone, the new one too weak.
- What happens when the reward ends? In the day-care study, removing the incentive did not bring the old behaviour back [1]. Plan for the exit before the launch.
- Can you reward with information instead? Specific feedback, credit and visible recognition of skill increased interest in the meta-analysis [3]. They cost less and do not set a price.
A worked example (fictional)
The community, the token and all figures below are invented to show the questions in use.
A small DeFi project has a forum where about forty users write detailed guides and answer newcomers' questions. Nobody pays them. The team decides to "reward the community" with 50 tokens per post, roughly the price of a coffee.
Run the questions. People were already writing for their own reasons: the incentive has something to crowd out. The reward is expected and paid per post, so the number of posts will likely rise, and many will be short replies written for the token. The amount is too small to make a careful guide worth writing as paid work, but large enough to make it feel like paid work. When the budget runs out, the long-time writers may not return to unpaid guides; the day-care parents did not return to being on time.
An alternative with the same budget: thank the best guides by name each month, ask the authors to review the project's documentation as credited contributors, and give an occasional unannounced grant for exceptional work. The research suggests these keep the original reason intact. It does not guarantee they will work; it says they are less likely to replace it.
What honest rewards look like
For anyone who publishes or runs a community, including this site:
- Prefer feedback and credit that say something true about the quality of the work.
- If you pay, pay properly and call it work. Do not use a token tip to turn a volunteer's help into piecework.
- Be wary of rewards per post, per like or per reply. They measure volume, and they get it.
- Disclose paid participation next to the content, so readers know which voices were compensated.
What this article cannot tell you
The experiments here were run with children, students, employees and parents in specific places and decades. Their effect sizes do not transfer directly to a token programme or a social feed. I found no controlled study of crowding out in crypto points or airdrop programmes; the section on crypto is an interpretation of general research, not a measured result. The day-care study covered ten centres in one country, and the averages quoted are my calculation from its published table. The size of crowding out is still debated, as the many meta-analyses on it show [6]. Nothing here measures the motivation of Dapp.Expert readers or authors. This is not legal, financial or investment advice.
How this article was made
| Task | What the AI did | What a person did | How it was checked |
|---|---|---|---|
| Research | Collected studies on motivation, rewards and online behaviour into research notes | Chose the question and the studies to rely on | Figures checked against the papers or their official abstracts [1]–[8]; the day-care averages were recalculated from the paper's Table 2 |
| Draft | Wrote the first draft from the notes | Edited the text, the structure and the worked example | A separate review compared figures and claims with the sources |
| Images | Generated the cover photograph-style illustration | Chose the illustration; designed the explanatory diagrams | Labelled as AI-generated in the disclosure line and alt text |
| Not done by AI | — | The decision to publish and responsibility for the text | — |
Conflicts of interest: none. The author has no commercial relationship with the researchers or organisations named; the community and token in the example are fictional.
Responsible for this article: Dmitriy Lastovka, author. Corrections: contact form.
Sources
Labels: paper (peer-reviewed study, read in full), abstract (peer-reviewed study, read via its official abstract).
- Uri Gneezy, Aldo Rustichini, "A Fine Is a Price", Journal of Legal Studies 29(1) (2000), 1–17. https://doi.org/10.1086/468061 Paper (author copy: https://rady.ucsd.edu/_files/faculty-research/uri-gneezy/fine.pdf).
- Mark R. Lepper, David Greene, Richard E. Nisbett, "Undermining children's intrinsic interest with extrinsic reward: A test of the 'overjustification' hypothesis", Journal of Personality and Social Psychology 28(1) (1973), 129–137. https://doi.org/10.1037/h0035519 Abstract.
- Edward L. Deci, Richard Koestner, Richard M. Ryan, "A meta-analytic review of experiments examining the effects of extrinsic rewards on intrinsic motivation", Psychological Bulletin 125(6) (1999), 627–668. https://doi.org/10.1037/0033-2909.125.6.627 Abstract.
- Richard M. Ryan, Edward L. Deci, "Self-determination theory and the facilitation of intrinsic motivation, social development, and well-being", American Psychologist 55(1) (2000), 68–78. https://doi.org/10.1037/0003-066X.55.1.68 Abstract.
- Uri Gneezy, Aldo Rustichini, "Pay Enough or Don't Pay at All", Quarterly Journal of Economics 115(3) (2000), 791–810. https://doi.org/10.1162/003355300554917 Abstract.
- Christopher P. Cerasoli, Jessica M. Nicklin, Michael T. Ford, "Intrinsic motivation and extrinsic incentives jointly predict performance: A 40-year meta-analysis", Psychological Bulletin 140(4) (2014), 980–1008. https://doi.org/10.1037/a0035661 Abstract.
- Björn Lindström, Martin Bellander, David T. Schultner, Allen Chang, Philippe N. Tobler, David M. Amodio, "A computational reward learning account of social media engagement", Nature Communications 12, 1311 (2021). https://doi.org/10.1038/s41467-020-19607-x Abstract.
- Olivier Toubia, Andrew T. Stephen, "Intrinsic vs. Image-Related Utility in Social Media: Why Do People Contribute Content to Twitter?", Marketing Science 32(3) (2013), 368–392. https://doi.org/10.1287/mksc.2013.0773 Abstract.










