We live, for better or for worse, in a competition-driven world. Rivalry powers our economy, sparks technological innovation and encourages academic discovery. But it also compels people to manipulate the system and commit crimes. Some figure it’s just easier—and even acceptable—to cheat.

But what if instead of examining how people behave in a competitive setting, we wanted to understand the consequences of competition on their everyday behavior? That is exactly what Amos Schurr, a business and management professor at Ben-Gurion University of the Negev, and Ilana Ritov, a psychologist at The Hebrew University of Jerusalem, discuss in a study in this week’s Proceedings of the National Academy of Sciences. “How can it be,” Schurr asks, “that successful, distinguished people—take [former New York State Gov.] Eliot Spitzer, who I think was a true civil servant when he started out his career with good intentions—turn corrupt? At the same time, you have other successful people, like Mother Theresa, who don’t become corrupt. What distinguishes between these two types of successful people?”

Schurr and Ritov found that when people win a competition in which success is measured by social comparison rather than by a fixed standard, they are more likely to engage in unrelated unethical behavior—in the case of this study, to cheat their peers out of money. “We are the first to ask what happens to contestants and their behavior after a competition ends,” Schurr says, “and we found that competitions have long-lasting effects.”

The researchers performed a series of experiments to test these effects. For the initial competition, they had groups of students take part in an estimation task. The students were told that those who performed best at estimating the number of signs that flashed across a computer screen would be considered “winners”—and given a pair of earbuds as a prize. In reality, however, winners were randomly assigned in order to avoid selection bias.

The students were then assigned another task, with a different payoff. They were randomly split into pairs: one student was given two dice and a cup with a hole in the bottom, the other was told to simply watch. The pair then played a game over 12 shekels (the equivalent of 12 quarters), in which the first student put the cup over the dice and shook it so that only he could see the results of the roll. His outcome, between two and 12, would dictate how many shekels he could take; his partner would receive the remaining amount. No one except Schurr and Ritov knew who had won or lost the initial competition. Compared with a control group, in which the claimed payout was approximately seven, or the expected value (halfway between two and 12), students who had previously won the competition overclaimed the outcomes of their rolls and took in an average of 8.75 shekels. “You should note that they’re not stealing from me, the researcher,” Schurr says. “They’re taking from their fellow students, from their friends.”

Schurr and Ritov repeated the dice-under-a-cup game after students participated in a number of other tasks in which “winning” was defined by different parameters. The researchers found that competitive settings determine behavior. “You have two types of success,” Schurr says. “One involves social comparison [as in the case of being a better estimator] and the other does not. And when you measure success in terms of ‘how good am I’ in reference to other people, that’s when people may turn corrupt.”

For instance, participants were asked to recall either an experience in which they won a competition or a situation in which they met a specified goal. The students who recalled winning a competition cheated in the subsequent dice-under-a-cup game whereas the students who recalled meeting a goal did not.

Similarly, participants who had simply won a lottery did not end up cheating when they reported the outcome of the dice roll but participants who had outplayed their peers in a trivia competition (again, controlled for selection bias) did later overclaim their winnings. “When we win in competition, in particular when we establish we are above others in rank, we will feel more powerful,” says Dacher Keltner, a psychologist at the University of California, Berkeley who did not take part in the research. “And dozens of studies have found that the simple feeling of power makes people feel above the scrutiny of others and act in impulsive, self-gratifying and unethical ways. Feelings of power, whether it comes from wealth, a person’s position in a hierarchical structure or in this case competition, can indeed lead to various abuses like lying and stealing.”

Schurr and Ritov attribute the cheating that occurred in their study to a number of possible psychological mechanisms, particularly entitlement. Their study “ties into recent work that relates to the influence of social-economic status and its influence on ethical behavior,” says Shaul Shalvi, a behavioral economist at the University of Amsterdam who was not part of this study. “People who are of higher status would break the rules more often. So, for example, you’re more likely to see a very fancy car ignoring the red traffic light compared to the guy in the normal car because they apparently feel entitled. So it’s nice that this study links to that, because people of high status have probably had the experience of winning.”

Schurr plans to continue his line of research. “We could look at the other side of the coin and see how much competition winners contribute to society as a whole,” he says. “Instead of doing bad things, will people do good things after the competition ends?” He also mentioned the possibility of better understanding gender differences in winner behavior. And it would no doubt be valuable to analyze real-world competitions. “There’s always a trade-off between experimental control, which was high in this study, and the ability to generalize the findings,” Shalvi says. “So these experiments are very important and telling in clarifying underlying psychological mechanisms of the dark or dangerous consequences of competition. But what still remains to be seen is whether these effects cannot also be seen when analyzing data from sport competitions and business-related competitions.”

If confirmed by such analyses, the research suggests that discouraging social comparisons and focusing instead on fixed goals might be a good way for organizations—from sport teams to businesses to governments—to reduce corruption.