An explanation for genocide? Testing aggression against other groups

By Yann (talk) 14:23, 28 March 2009 (UTC) (Own work) [GFDL (, CC-BY-SA-3.0 ( or CC BY-SA 2.0 (], via Wikimedia Commons

It’s a cliché that people act in self-interested ways. Yet this is generally seen as being rational (or enlightened) self-interest. Unless we’re sociopaths, we do things that may harm others only if they benefit us, not simply for the sake of harm.

A recent study from a team of Swiss researchers may challenge this view. The provocatively titled “Do we harm others even if we don’t need to?”, published in Frontiers in Psychology, explores willingness to disadvantage others for the sake of group affinity.

The researchers randomly divided test subjects into groups, whose group identity was signaled by shirts of different colors. The members of each group were then tasked with allocating money within their group and to the other group. There were three scenarios: participants could invest money without harming anyone else, could take money from the other group to benefit their own group, or could take money away from the other group without it having any effect on their own group. They could also do nothing. This wouldn’t allow them to increase their money through investment, but nor would it affect anyone else.

The third situation is the most interesting. 23% of participants in this scenario chose to deduct money from the other group. As these participants indicated, they acted out of a desire “to harm the other group”. What’s key here is that this kind of aggression wasn’t strategic or self-interested. People were being antagonistic for the sake of antagonizing a different group.

This apparently irrational act can be reframed if, like the researchers, we consider another motive: building group solidarity. The participants who took money away from the blue-shirt group didn’t benefit individually from hurting the blue shirts, but in a way they did benefit socially by acting jointly. They fostered their own group’s cooperation at the expense of the other group.

The group identity is key here. Other research has suggested that members of one group are likely to help another group if it benefits their own group’s image, not their own individual reputations. “When groups have a long history of intergroup conflict, outgroup helping may not be interpreted as an act of kindness, but instead as a means of asserting social dominance.” So whether harming or subtly intimidating another group, what seems antisocial may actually be extremely social: a way of cementing social relations within one’s own group.

Perhaps all this isn’t surprising. After all, we see this kind of behavior all the time, if at a minor level. At school, banding together to pick on the new kid makes us feel more secure in our own status. At work, we feel more connected to the members of our own department by complaining collectively about those elsewhere in the office. As a country, we may spread colonial practices, even overseas aid, in a way that reinforces our continued power in the region.

Another interesting finding from the “Do we harm others” study was that those who had been victimized (had money taken away from them) were more cooperative. It wasn’t even necessary for them to be harmed in one of the multiple rounds of the experiment. The simple threat of victimization was enough for them to act more cooperatively than they would have otherwise. The real-world corollary here might be the new kids at school (or another department at work) forming a defensive faction even before others have a chance to bully them.

Of course, all the standard caveats about this kind of research design apply. It can be difficult to extrapolate from a lab setting to real life, and researchers can’t be certain about reasons for behavior in these kinds of psychology games. In addition, information about the research subjects would be helpful to contextualize these results. It’s been found, for instance, that men are more generous when it comes at little cost to them, while women are much more altruistic: they give even when it has a major impact on themselves.

There are likely to be intertwined social and physiological factors at play here. There’s a large body of research documenting the neurological links to fairness, empathy, and prosocial or antisocial behaviour. Different parts of the brain are activated in group settings, and when determining whether to help or harm others. This neural activity can be manipulated; oxytocin, for instance, makes test subjects more trusting and affectionate toward other members of their group, while testosterone increases aggression toward other groups.

The possible implications from all this work are fascinating — and potentially troubling. The Swiss researchers behind “Do we harm others” note that strengthening a community is generally seen as positive, yet the flipside of tighter bonds may be hostility toward others. They refer to the noxious radio station urging greater ethnic solidarity, which contributed to the Hutu-Tutsi divide, and in turn to the Rwandan genocide.

Therefore these findings suggest that feeling more closely connected, whether that’s based on ethnicity or apparel, has a dark side. There is a bright side, however. Recent research also suggests that a self-interested third party can help mediate conflict. In fact, just the possibility of intervention by a mediator, such as a country with political interests in regional stability, can increase the likelihood of cooperation. So if we’re willing to go to destructive places to maintain social connection, the presence of selfish third parties may help to limit the harm.