The rule was to choose the brightest box. The problem is that those further to the right made more money. And so, faced with the possibility of earning something more, several participants began to ignore the instructions. When they saw that others were doing it too, some went even further. A dynamic observed in an experiment on 750 American adults, where even political affiliation had its weight: seeing someone from your party break a rule made it more likely to do the same.
The study, published on Proceedings of the Royal Society Balso identified a particularly enterprising minority: just over 4% of participants went beyond the violations they observed. In subsequent computer simulations, a few individuals with this tendency were sometimes enough to drag the group into increasingly serious infractions. Bad example, it seems, can find particularly zealous imitators. With a clarification: for now we are talking about an online experiment, without sanctions for those who transgress.
A shooting star and the temptation to earn more
Researchers at the University of Amsterdam, led by psychologist Lucas Molleman, recruited 750 people in the United States, shortly before the 2024 presidential election. About half identified with the Democratic Party and the other half with the Republican Party. A precise choice, because the group also wanted to understand how important it was to feel close to those who set the example, whether good or bad.
Participants were shown a shooting star passing through a row of 100 boxes, lighting them up with different intensities. The task consisted of identifying the brightest one, generally located in the central part of the row. So far, a little observation game. The reward system, however, had a peculiarity: the further the chosen box was to the right, the greater the gain. Even when that box wasn’t the brightest. To obtain maximum compensation, therefore, the delivery had to be ignored.
Already in the first answers, without knowing the decisions of the others, approximately 23% of the choices fell on the box that guaranteed the highest prize. The economic temptation worked even without external encouragement. Then the researchers added a variable. Participants could observe another person’s choice, know whether he complied or violated the rule, and learn about his political orientation. At that point they could change their answer.
If one of our own breaks the rules, following him becomes easier
Those who observed someone strictly following instructions tended to change their behavior little. Things were different when the example showed a clear violation, especially if committed by a person from the same political party. In one of the experimental conditions, choosing the most profitable box rose from 23% to almost 40% after observing another politically close participant commit that same transgression.
Political affiliation, in this case, functioned as a signal of social closeness. The choices of those we consider part of our group can weigh on our decisions more than those made by a stranger with whom we feel we have little in common. This also applies when we know perfectly well that that person is doing something against the rules.
The experiment also suggests a certain asymmetry: more severe violations influenced behavior more than examples of strict compliance with instructions. Those who behaved correctly contributed little to convincing others to do the same. Those who aimed directly for the maximum prize, however, were more easily able to shift the observers’ choices. Not everyone, of course, was convinced. Some participants continued to follow the rule even after seeing others ignore it.
Those 4% who decide to exaggerate
Then there is a result that has particularly attracted the attention of researchers. Most people influenced by the choices of others tended to reproduce a violation similar to the one observed or less serious. A small minority, just over 4%, managed to exceed it.
If someone had chosen a box just to the right of the correct one, these participants decided to move further forward, increasing their gain and the severity of the infraction. A minority behavior, of course. But the researchers wanted to understand what would happen if people with this tendency interacted repeatedly within a group.
They then built computer simulations using the responses collected in the experiment, creating digital representations of the 750 participants and having them interact in groups of 100. In some scenarios, even a very small share of individuals inclined to increase violations could push the entire group towards extreme levels of transgression.
The mechanism was fed through subsequent imitations. Someone broke the rule slightly, another followed, and a third person went further. The latter in turn became an example for others. By repeating the process, infringements could progressively increase in the simulations.
It didn’t always happen, though. When the simulated groups retained a sufficient presence of people firmly intending to respect the rule, the progression stopped at intermediate levels. A few individuals could therefore contribute to a collective escalation, but their presence alone was not enough to make it inevitable.
From gaming to real life, with some caution
The research helps understand one aspect of social influence: People observe what others do and can change their own behavior accordingly. Even when the initial choice seemed to have been made independently.
Of course, in the experiment no one risked a fine, a conviction or any concrete social consequences. Participants had to choose a box on a screen and could earn some money by ignoring an arbitrary rule. Situations such as a street protest, a traffic violation or incorrect behavior at work involve decidedly different responsibilities, motivations and consequences.
Furthermore, the experiment only showed the choices of other individuals, without reconstructing all the relationships and dynamics that can develop within a real group.
The authors themselves invite us to avoid an overly simple conclusion: breaking a rule does not necessarily mean carrying out a harmful action. Challenging norms and conventions can also contribute to social change, creativity and learning. It depends on the rule, the circumstances and what breaking it entails.
Compliance with the rules, moreover, had already been examined by Molleman and other researchers in a study published in 2025 on Nature Human Behavior. Experiments had shown that many people respect the rules even when they could gain economic benefits by breaking them, without being observed or punished.
The new research adds a piece: that individual behavior can change when the choices of others come into play. And those who go beyond the received example can have a disproportionate influence.
In the simulations, however, there were also those who continued to respect the delivery, regardless of what the others did. The presence of these people prevented the group from sliding towards the most extreme violations. No heroic gestures were needed: they simply kept ticking the right box.
