Threatening to Punish People Will Never Lead to Peace, Study Shows

Scientists call the prison system into question.

Getty Images / Justin Sullivan

It’s unclear to scientists why socisn’t given way to chaos, as the self-interested humans of The Purge so horrifically depicted. The teachings of Charles Darwin, father of evolutionary biology, say that natural selection favors selfish people, so why aren’t we all out there indulging in maximum greed, lust, and gluttony?

Many have theorized that it must be the threat of punishment that keeps our worst instincts in check and allows a cooperative soci to thrive, but an international team of researchers recentlyto thrive, Many have discovered through a ‘prisoner’s dilemma’ experiment that th this is not quite so. In a study published in the December edition of the Proceedings of the National Academy of Sciences, they write that punishment is is actually “surprisingly ineffective in promoting cooperation.”

“While the implied message when punishing someone is ‘I want you to be cooperative,’ the immediate effect is more consistent with the message ‘I want to hurt you,’” the researchers write.

The dilemma: Testify or stay quiet?


The experimental design was based on the prisoner’s dilemma, a widely used game theory simulation that tests how much people can cooperate while still acting in their own best interest. In the classic version of the game, participants play the role of two robbers who have been caught by police and are waiting to be sentenced. If neither gives the other up, they’ll both be charged — but with a lesser crime than the one they actually did. If one testifies against the other, however, there’s a chance that person will get off free while their counterpart is put away behind bars. And if they both testify against each other, they could both be punished with the full weight of the sentence.

In this study, co-authors Marko Jusup, Ph.D., a mathematics professor at Japan’s Hokkaido University, and Zhen Wang, Ph.D., a game dynamics professor at China’s Northwestern Polytechnical University, modified the classic prisoner’s dilemma so that they could test whether the threat of punishment would force people to cooperate more.

In the first group, where the opponents were reshuffled each round, defectors prevailed over the course of 50 rounds. In the second group, where the opponents remained the same for the 50 rounds enabling them to identify cooperative neighbors, the cooperative cluster survived. In the third group, the option to punish opponents failed to boost cooperation.

Xuelong Li, et al., Proceedings of the National Academy of Sciences of the United States of America, December 19, 2017.

They started by separating 225 Chinese university students into three trial groups. In group one, trios of students played the game 50 times, but in each round the participants in each trio changed. The objective was for one student to gain points while interacting with two “opponents”. If they all chose to “cooperate” with each other, for example, the student got four points. If they all “defected” (that is, turned on each other), the student would get zero points. But if the student betrayed the opponents, and they chose to cooperate, then the student would get eight points. The second group played the game the same way, but the main difference was that groups of students stuck together to play the duration of the 50 rounds — ensuring that they’d be able to identify which individuals were cooperative, and which tended to rat others introducing the option to punish each other did not improve the levels of cooperation.

The third trial group also kept its trios together, but its main difference was that participants had the ability to “punish” each other. When they did so, the punisher would lose a few points, and related [e punished would lose a lot of points.

The researchers theorized that introducing punishment as an option would be a means to force people to cooperate. However, they found the opposite. While there was increased cooperation among the people who kept playing the game with each other — a 4 percent cooperation rate compared to a 38 percent cooperation rate — introducing the option to punish each other did not improve the levels of cooperation.

In a statement released Thursday, the scientists explained that punishment was demoralizing to the players, causing them to lose interest in the game and play with less strategy. The availability of punishment, they explain, decreased the overall incentive of players to cooperate and reduced their motivation to compete to win.

There is a caveat to the results: Other scientists have expressed doubt about whether the results from prisoner’s dilemma experiments actually translate into real-world scenarios, and related experiments have shown that it’s the balance of low-impact punishments and high-impact rewards that enable society to exist.

Still, results are results, and these students showed that, at least within this population of humans, punishment isn’t the best way to keep people in line. Jusup theorizes punishment may still exist because “human brains are hardwired to derive pleasure from punishing competitors” — which you probably already knew from playing board games with your smack-talking grandparents over the holidays.

Related Tags