Individuals engage in wrongdoing and face consequences. They begin to collaborate. This fundamental belief that humans act rationally and alter their behavior in response to consequences lies at the core of Western legal frameworks, criminal behavior, and economic theories. The evolution of cooperation. However, extensive research over several decades indicates that punishment may not be as effective as thought.

Analysis has consistently shown that harsher penalties, including the three strikes law, do not reliably deter crime. The U.S. National Research Council’s report could not establish the death penalty’s effectiveness. Meanwhile, the U.S. possesses the most punitive criminal justice system globally, with notable rates of imprisonment and recidivism.

These real-world observations stand in stark contrast to much of the controlled experimental literature. In notable research, economists Ernst Fehr and Simon Gechter designed a game where participants could allocate money to a communal pool, which would then be doubled and redistributed, maximizing benefits for all contributors. Yet, individuals found it more advantageous to withhold contributions while others carried the load. While cooperation diminished in scenarios lacking punishment for free riders, the introduction of penalties significantly boosted contributions to the pool.

What explains the gap between experimental results and real-world dynamics? We explored this question in a recent study published in PNAS. We noted that individuals tasked with enforcing penalties often have conflicting incentives that can erode their credibility and diminish the public’s trust in them. In Ferguson, Missouri, authorities relied on fines to fund city services, disproportionately impacting Black communities. Across the U.S., billions are confiscated through civil asset forfeiture, permitting law enforcement to seize property from individuals suspected of criminal activity.

We hypothesized that self-serving motives behind punishment could undermine cooperation by clouding the ethical implications associated with collaborative behavior. Unlike other species, humans possess a “theory of mind” and are acutely aware of others’ intentions and motives. Punishment carries a message of disapproval necessitating behavioral change; however, this only holds weight if individuals perceive the punisher’s motives as justifiable. Humans, inherently social beings, ask, “What’s your reasoning for this?” If the answer appears self-serving, the deterrent effects of punishment diminish.

To validate our hypothesis, we conducted a series of experiments utilizing the same game that demonstrated punishment’s role in enhancing cooperation. Here, one player (the dictator) decides on potential monetary sharing with another player (the receiver), while a third player (the punisher) can revoke funds from the dictator. A twist emerged where punishers were incentivized financially; akin to police utilizing ticket quotas to boost revenue, our punishers received bonuses each time they issued a penalty. This change reversed the expected outcome; rather than improving cooperation, the incentivized punishment led to decreased collaboration, as trust in the punisher waned.

Our results indicate a need to reassess crime management methodologies. When punitive actions are perceived as self-serving, they foster distrust and hinder the cooperation that such measures are intended to reinforce. To cultivate safer, more cohesive communities, it is essential to eliminate policies that compromise the ethical underpinnings of punishment. This includes abolishing speed trap quotas and profit-driven incarceration, practices implying that punishment is motivated by financial gain rather than justice.

Raihan Alam and Tage Rai are students at the Rady School of Business at the University of California, San Diego.