We’re more likely to steal from large groups than from individuals

Many of those at the top of large corporations seem content to exploit large groups of people for their own significant financial gain.

Strangely, this is somewhat at odds with previous research in behavioural economics, which tends to find that people are generally quite prosocial, honest, and overall unwilling to steal considerable amounts from others. From results like these, it's difficult to piece together exactly how we've arrived at such levels of financial inequality.

Psychologists have floated a number of potential explanations for this phenomenon. Perhaps those who end up rich have innately higher levels of psychopathy, for example. Or perhaps it's the case that the cultural norms within certain businesses change the way people make economic decisions. However, new research in Nature Human Behaviour from Carlos Alós-Ferrer and colleagues suggests an alternative explanation: when it comes to acting selfishly, we are much more likely to do it at a grand scale than a small one.

Previous tasks that seek to measure selfishness, such as the Dictator, Ultimatum, and Trust games, all probe how much of a sum a participant will transfer to another person under slightly different conditions and opportunities for personal gain. Responses to such tasks are generally pro-social, with participants often being generous to the other party, and mostly keen to trust them, despite risks to their own gains. But, though these results are generally consistent across previous literature, they don't resemble the sort of behaviour we see day to day from those with the opportunity to exploit large groups of people.

In order to investigate this mismatch, the research team at the University of Zurich decided to look at how acting upon a larger group of "victims" influences a robber's willingness to be selfish, using a new task called the Big Robber game.

During this game, participants are assigned the role of "robber" or "victim". Robbers are asked to choose what portion of the victims' earnings they wished to take from a group of 16 victims: 0%, 10%, 33%, or 50%.

This new task was administered to 640 student participants alongside Dictator, Ultimatum, and Trust games, where each participant played rounds as both sender and receiver.

After participants had made all of their transactional choices in these tasks, they were presented with the option to donate a freely specified percentage of their earnings to charity. The authors hoped that uptake of this opportunity would be indicative of the emotions participants felt after making their choices. If they felt guilty about their actions, they may choose to give away a higher percentage of their earnings, while lower donations may be indicative of selfishness and a lack of regret.

Analyses showed that the majority of robbers chose to steal from victims. Around 56% decided to take half of victims' earnings for themselves, and a further 27% chose to take a third of victim's earnings. Just under 15% elected to take a tenth of the amount, and a staggeringly low 2% declined to steal anything.

A closer look at the time taken to reach these decisions showed that those taking half of victims' earnings were faster to arrive at their choice than those taking a more moderate third of earnings. Decision times formed an inverse U-shape, suggesting that those choosing extreme decisions may have faced less of a moral struggle than more moderate robbers. Those who robbed half of victims' earnings also donated less in absolute terms than those who robbed a third. This, the authors believe, is indicative of a lack of guilt about their decision.

This is in stark contrast to the Dictator, Ultimatum and Trust games, during which these same participants were mostly generous and cooperative with the lone other party in the transaction. In all games, the majority of participants who had been robbers declined to take the options that would harm the receiver's earnings the most, despite it being more profitable to them personally.

Taken together, these results indicate that harming many individuals seems to be easier for us to do than harming just one.

The authors suggest that we mentally make a trade-off between our own personal gain and concern for others. If we take a small percentage of earnings from a large group, we're able to make large financial gains while minimally affecting each individual member of said group. But, if we were to reduce this to a one-on-one transaction, we would have to take magnitudes more of the other party's earnings to achieve the same profit. This is far more uncomfortable, psychologically speaking, than exploiting the earnings of larger groups, which may explain why we routinely see such behaviour from those at the top of larger organisations.

Future research could investigate whether this effect holds in other demographics, across different cultures, or indeed in less trying economic circumstances than we've experienced recently. Explorations of individual differences, such as personality, profession, or financial status, may help further tease apart the factors dictating willingness to steal large amounts during this game. This research represents a timely look at willingness to exploit a large group for personal gain, which may become increasingly relevant in years to come.

Further reading

– Generous with individuals and selfish to the masses

About the author

Emma L. Barratt (@E_Barratt) is a staff writer at BPS Research Digest