Â鶹ÒùÔº

UC San Diego researchers show that paying enforcers to punish makes people less likely to cooperate with others, which has major implications for law-enforcement quotas, asset forfeiture, and for-profit prisons.

Why do people cooperate with each other and follow society's rules—and what happens when those who enforce the rules stand to profit from doing so? A new from the University of California San Diego Rady School of Management researchers tackles this fundamental question and reveals that cooperation breaks down when punishment becomes profitable.

Published in the Proceedings of the National Academy of Sciences, the study by Tage Rai, assistant professor of psychology at the Rady School and Raihan Alam, a Rady School doctoral student in management, finds that when people get paid to punish others, it actually makes everyone less likely to cooperate—even those who would gain from working together.

The findings have broad implications for the structure of criminal justice policies, particularly those involving private, for-profit prisons, quota-based policing, and civil asset forfeiture—when law enforcement seizes property believed to be connected to criminal activity, even without charging or convicting the owner of a crime.

When punishment becomes a business

At the core of the study is a twist on a classic behavioral experiment. In a series of online economic games with more than 4,000 participants, the subjects were assigned one of three roles: a "decision-maker" a "community member," or a "third-party punisher." The "decision-maker" received a sum of money and could choose to share it with the "community member" or keep it all.

The "punisher" observed the "decision-maker's" choice to keep or share money and could penalize them by reducing their payout. In previous using the "classic" version of the game, introducing punishers increases cooperation.

"This has often been taken as a model for how law-abiding societies come to exist," Rai said. "But when researchers look at real-world crime data, they find that severe punishment doesn't deter . That creates a puzzle for scientists—so we asked 'what dynamic is happening in the world that isn't being captured in the lab?'"

And so the authors added a twist to the classic game. "In the real world, punishers often have incentives to punish that may erode trust in them, and so we moved that dynamic into the lab by giving punishers a bonus payment every time they punished," Rai said.

What the researchers found is that when they made punishment profitable, they reversed the classic effect—cooperation declined after punishment was introduced, and social norms began to break down.

"When punishment was profitable, people were less likely to cooperate from the start," Alam said. "For example, 'decision makers' were more likely to keep all the money, acting in their own self-interest."

Even when researchers optimized the conditions so that punishment only ever targeted selfish behavior, cooperation failed to recover, suggesting that the initial damage to trust persisted.

Alam added that this shift in perspective eroded the very cooperative norms that punishment is supposed to uphold.

"They viewed the whole situation differently—not as a social interaction about fairness, but as a game of maximizing how much you made," he said. "And they didn't trust the punishers' motives."

Implications for criminal justice reform

The study provides insight into real-world dynamics that plague the criminal justice system—particularly in communities that have long distrusted law enforcement.

"Punishment only works as a social signal if people believe it's being done for the right reasons," said Rai. "But in many real-world situations—like civil asset forfeiture, arrest quotas, or private prisons—the person being punished may assume that the enforcer is acting out of self-interest. And when that trust is broken, punishment backfires."

This breakdown helps explain why efforts to reduce crime through harsher penalties or increased policing often fail to deliver results. Instead, they may create feedback loops where distrust leads to non-cooperation, which in turn justifies even more aggressive enforcement.

Even more striking, the study found that people's attitudes toward punishment shifted depending on whether they imagined themselves as being at risk. When participants viewed punishment from a distance—seeing others as the ones who might be punished instead of themselves—they were more trusting of punishers and tended to support harsher penalties, even when the punisher stood to profit. But when participants imagined themselves as the ones who might be punished, they became more skeptical. They were less trusting of punishers and worried about being treated fairly.

"Even when we told participants that punishers were being paid, those who weren't at risk still believed harsher punishment would boost cooperation," said Rai.

"They trusted the punishers too much and failed to consider how it might feel to be on the receiving end. I think that speaks to why voters often back 'tough on crime' policies. They don't take the perspective of the communities most affected or realize how those policies may actually backfire, especially in communities where trust in authorities is already low."

The path forward: Rebuilding trust

One key takeaway from the study is that changing behavior isn't just about changing outcomes—it's about signaling intent.

"Even perfect enforcement doesn't work if the underlying trust isn't there," said Rai. "You need structural reforms that clearly and publicly remove profit motives from punishment. People need to see—and believe—that enforcers are acting morally, not opportunistically."

That could mean eliminating quota systems, ending civil asset forfeiture practices, or banning for-profit incarceration. It may also mean rethinking the role of punishment altogether.

"When people don't trust the institutions that punish them, they're less likely to follow the rules," Rai added.

Making the case in the lab

Though based on online experiments, the study with thousands of participants used real monetary incentives to simulate real-world stakes.

"These economic games allow us to strip away pre-existing attitudes people may have on law-enforcement and the legal system and look closely at how people respond when punishment becomes monetized," said Alam. "It's a powerful way to measure behavior under controlled conditions."

The authors conclude that the findings help explain not only the failures of some justice policies, but the persistent misperceptions that allow those policies to continue. As debates over police reform, incarceration, and community safety continue, they argue that this work provides compelling evidence that trust—not punishment—is the foundation of any cooperative society.