Children as young as six years of age are willing to sacrifice Skittles to create a more just society.

Previous research has documented that even in infancy, children prefer members of their own social or cultural groups, while adults will punish members of these in-groups less harshly than those in out-groups for antisocial behavior. Now, for the first time, a team of researchers from Yale and Harvard has examined the developmental trajectory of this bias for the first time. The study found that while humans are naturally inclined to be biased in response to selfish behavior, the bias can be partially overcome during development, said Jillian Jordan GRD ’19, a doctoral student in the Yale psychology department and coauthor of the paper.

“As a social psychologist, I would just say that the findings are striking for how early this complex form of in-group favoritism occurs,” said Susan Fiske, a professor at Princeton University and one of the paper’s editors.

Previous studies have shown that starting in infancy, humans learn to recognize an in-group and out-group. For instance, children prefer puppets that have the same physical characteristics or speak with the same accent as themselves. Other research has demonstrated that adults often show in-group favoritism when punishing unfair behavior, even if the punishable act does not directly affect them and a cost must be paid to intervene.

This study was the first to bridge the gap between infant in-group bias and adult norm enforcement by investigating the development of in-group bias in children’s reactions to selfish behavior.

“Past research shows that children will take actions to prevent unfairness from happening to themselves, but the open question was whether they might also pay to prevent unfairness from happening to other people,” said Katherine McAuliffe, a postdoctoral fellow in the Yale Psychology Department and coauthor of the paper.

In the study, children were assigned to either a blue or a yellow in-group according to their favorite color. The researchers confirmed this simple assignment created a bias where the young children displayed a preference for those wearing the same color party hat as their group assignment.

The children then witnessed one child treating another unfairly by keeping six Skittles to himself and sharing none with another child. Based on whether the perpetrators and victims were members of the blue or yellow group, the researchers hoped to determine how children’s willingness to intervene at a personal cost depends on in-group bias. The children participating in the study could condone this unfair allocation by placing one of their own Skittles into a green box, or reject the allocation by sacrificing their own Skittles by placing them into a red box. At the end of the study, the child was allowed to bring all Skittles from the green box home, but all Skittles in the red box would be thrown away.

The study found that six-year-old children were more likely to engage in costly punishment in response to one child treating another child unfairly when the actor was an out-group member and the recipient was an in-group member. Eight-year-old children were also more likely to punish out-group members for unfair behavior, but they were equally likely to punish unfair behavior on behalf of in-group and out-group recipients.

The study suggests that children’s default sense of morality is biased from the start but their partiality can be overcome with age, Jordan said.

The results of this study are inconsistent with the “black-sheep effect,” which predicts that people usually punish in-group members for violating social norms or traditions more harshly than out-group members. The researchers attribute these contradictory findings to the fact that selfish behavior is not a particular norm violation.

The study did not investigate the developmental changes occurring between ages six and eight that caused this to the slight decline in in-group bias.

Jordan also said that this study only focused on two different social groups. Since there are many more groups outside the laboratory, the results of the study may not directly translate.

“Research shows that [two groups] primes a competitive mindset, but in everyday life, there are all sorts of groups,” she said.

The study was published in the journal PNAS on Aug. 18.