It Pays To Be Kind: The Science Behind Generosity

Most of us would agree that kindness is a more helpful trait than selfishness. But for decades, scientists created mathematical models of human interaction that seemed to predict the opposite:  selfish players often came out the winners.But all those models had discounted one crucial fact, which the latest research has now revealed. New mathematical models provide a specific reason why it makes evolutionary sense for us to be kind to strangers: it often doesn’t cost much to be generous, but a single act of stinginess could cost you a long-term friend. In other words, petty greed just isn’t worth the risk.
A study published in the Proceedings of the National Academy of Sciences demonstrated this by pitting computerized players against one another in a Prisoner’s Dilemma-type game. This classic game is based on a simple story: Let’s say the police arrest two suspects, Charlie and Al, but don’t have any solid evidence on either of them. So the cops separate the two prisoners, and separately offer each of them the same deal: If Charlie testifies against Al (“defects”) and Al doesn’t squeal on Charlie (“cooperates”), then Charlie gets to go free, while Al spends a year in the slammer. But if neither Charlie nor Al snitches on the other, then the judge – who doesn’t have any solid evidence – will convict them both, but give them lighter sentences than if just one of the suspects had snitched. In an odd legal twist, if both Charlie and Al betray each other, they both get a lighter sentence as well. Now, here’s the dilemma: Charlie and Al must each decide whether to betray the other (“defect”) or not. And by the way, the cops have already assured both Charlie and Al that any defection will be kept secret until the end of the trial. So, what should Charlie and Al do?

Mathematicians called game theorists often program computers to run game after game of Prisoner’s Dilemma, recording the acts of both Charlie and Al in each round. By programming each of the players to respond to defection and cooperation in a variety of ways, game theorists began to discover some definite response patterns that resulted in better survival.

One of these successful patterns was “tit for tat” – in other words, Charlie cooperates with Al unless Al snitches on Charlie; in which case Charlie retaliates by defecting in the next round, and only gradually puts more trust in Al as Al proves his trustworthiness. But if Charlie has the chance to snitch on a prisoner he may never see again – let’s call this guy Vito – then Charlie has no reason not to defect on Vito. The decision of whether or not to defect on old Al, on the other hand, would consider cooperation a much more reasonable option.

However, this new study provides a revolutionary Prisoner’s Dilemma strategy. A team led by psychologists Andrew Delton and Max Krasnow of the University of California, Santa Barbara showed mathematically that cooperation was a more evolutionarily stable strategy regardless of whether or not an agent reasoned that it would encounter the same partner again. Even when the likelihood was more than 90 percent that Charlie would never see Vito again, he still tended to cooperate.

This stands in stark contrast to those earlier mathematical models, which tended to predict that the best strategy is to be generous with one’s regular reciprocal partners, but selfish in one-time-only interactions. Instead, this research shows that the cost/benefit ratio for both these kinds of generosity is about the same.

If the researchers are right, this means that generosity has a strong tendency to co-evolve with social cooperation.

It’s also interesting to note that this model predicts the same sort of generosity regardless of the size of the group – what’s important isn’t how likely you are to meet the same person again, but simply that there is a chance, however slight, that they might help you in the future. But even if they don’t, it costs nothing to be nice.