Individuals are swayed by their peers, leading to more severe punishments, study finds

Punitive peers can increase how much fellow decision-makers support punishment, by up to 40 percent.

PROVIDENCE, R.I. [Brown University] — When acting as one part of a group charged with deciding how to punish someone — a jury, for example — individuals are swayed by their peers to punish more often than they would if deciding alone, a new study found.

“People can get together in a group and be intensified by the other people in their group to behave in ways they wouldn’t typically when alone, including becoming more punitive,” said Oriel FeldmanHall, an assistant professor of cognitive, linguistic and psychological sciences at Brown University and senior researcher on the study. “Even in a fairly sterile laboratory setting, when you’re just exposed to the minimal preferences of a few other people, it is enough to amplify your punishment recommendations of perpetrators by 40 percent.”

The findings were published on Monday, Aug. 12, in the journal Scientific Reports.

FeldmanHall, who is affiliated with Brown’s Carney Institute for Brain Science, and her research team conducted five experiments involving almost 400 participants. Four looked at individuals’ willingness to punish people who behaved selfishly in economic tasks, and another involved determining punishment recommendations for hypothetical perpetrators of crimes of varying severity.

Across all experiments, participants decided — either as a member of a group, or alone — whether or not to punish the offender. The study also measured differences in partiality: Some experiments were set up such that the decision-maker was tasked with serving as an impartial juror; in others, the decision-maker was directed to imagine they were the victim of an unfair offer or mock crime.

The team found that as the number of pro-punishment people in the group increased, other participants become up to 40 percent more willing to recommend punishing a perpetrator, FeldmanHall said. That trend held true whether the experiment was framed such that the participant was an impacted victim or an impartial juror. 

However, they also found some differences. Victims were more readily swayed by their peers’ decisions to punish. In contrast, jurors conformed to group decisions at a lower rate than victims and also took into account the severity of the perpetrator’s offense when deciding whether to punish. 

Using a computational model that describes how people use contextual information to make decisions, the researchers found that participants used both their peers’ preferences as a guidepost for how much they should value punishment and were less cautious about making decisions when they believed they were only one voice among many. 

“When punishment is delegated to groups, there’s the benefit of pooling people’s preferences and perspectives, but it also introduces the danger that people will conform to the group’s preferences,” said Jae-Young Son, first author on the paper and a doctoral student in FeldmanHall’s lab. “In real-world contexts, such as a jury, there's a possibility that being part of a group will make everyone within the group less cautious about their decisions — that may be sufficient to convince some people to conform to the majority opinion, and that creates increasingly large majorities that eventually convince everyone else.”

Although these results may seem alarming in certain contexts, FeldmanHall added that conformity can also be adaptive — it helps humans survive. 

“People use each other as a reference points all the time because it is adaptive and helpful for gathering information,” she said. “Looking to other people, and how they approach a justice dilemma, can — although not always — be a useful thing.”

However, more research is needed to understand the extent to which people are willing to be flexible about moral decisions, she added. 

The participants were recruited from the Brown community and online through Amazon Mechanical Turk, two common ways for recruiting participants for this type of experiments. FeldmanHall said she prefers to use both methods of recruitment to ensure that the team’s results are robust.

In addition to FeldmanHall and Son, Brown investigator Apoorva Bhandariwas an author on the paper. The research was supported by Brown internal funding.