There is a major psychological flaw in how society punishes people

Individuals break rules. They face consequences. Subsequently, they may alter their behaviors. This intuitive belief that human behavior is logical and responsive to punitive measures forms the foundation of Western justice systems, economic models related to crime, and theories of cooperation evolution. Unfortunately, a wealth of research indicates that punishment often fails to produce desired outcomes.

Studies consistently show that severe penalties, such as the notorious “three strikes” laws, do not effectively deter crime. In their evaluation of the death penalty, the US National Research Council could not substantiate its effectiveness. Meanwhile, the United States, despite having one of the most severe systems, grapples with alarmingly high incarceration and recidivism rates.

These practical insights conflict with numerous experimental findings. A renowned experiment by economists Ernst Fehr and Simon Gächter involved a game where players were allocated money with the option to contribute to a communal fund. The fund was increased and redistributed, optimizing benefits for all involved when contributions were made. However, individual players had the incentive to refrain from contributing while others did. Lack of punitive measures resulted in dwindling cooperation—but with the introduction of punishment, contributions surged again.

So, what discrepancies exist between real-world applications and experimental scenarios? In our recent study published in PNAS, we proposed that individuals in punishment roles often harbor incentives that compromise their authority and undermine public trust. In places like Ferguson, Missouri, local officials utilized fines to fund city operations, disproportionately impacting Black residents. Nationwide, police departments have seized substantial amounts through civil asset forfeiture, allowing confiscation of property from individuals suspected of crimes.

We theorized that such self-serving motives in punishment diminish the potential for cooperation as they obscure its ethical connotation. Unlike other species, humans possess “theory of mind”—we are acutely aware of others’ intentions. Punishment conveys disapproval that compels behavior modification. This messaging is effective only when the punisher’s intentions are perceived as fair. Humans are inherently social beings who ponder, “What is your motive?” When the reasoning appears self-serving, the authority of punishment diminishes.

To examine this hypothesis, we conducted various experiments employing similar games that initially demonstrated how punishment encourages cooperation. In these setups, one participant (the dictator) decided whether to share resources with another (the receiver), while a third (the punisher) had the power to deduct resources from the dictator. Here, we introduced an incentive: the punishers received compensation whenever they executed a punishment. Just as a police department may impose ticket quotas to increase revenue, our punishers received bonuses associated with their punitive actions. The result was the opposite effect—rather than enhancing cooperation, punishment decreased it. Trust in punishers waned, leading to reduced willingness to collaborate.

Our outcomes indicate a need to rethink strategies for crime prevention. When punishment is perceived as self-interested, it cultivates distrust and undercuts the cooperation it aims to foster. To cultivate safer and more collaborative communities, we must eliminate systems that compromise the ethical essence of punishment. This includes abolishing policies such as ticket quotas and profit-driven incarceration—practices suggesting that punishment stems from profit rather than justice.

Raihan Alam and Tage Rai are affiliated with the Rady School of Management at the University of California, San Diego.