Spoiler alert: everybody cheats. It may be as minor as cutting in line or as venal as lying on a balance sheet before presenting an investment opportunity. So the question is not whether people will choose to betray each other’s trust—but rather, when.

Check out more from The Trust Project at Northwestern University here.

J. Keith Murnighan of the Kellogg School of Management and colleagues conducted research to get at this question. They discovered that “anticipatory regret” can make people think cheating is acceptable—especially near the end of a job or task.

“Think of interns who are leaving and have no chance of getting a long-term job with the same company, or a short-term worker at a plant where they’re leaving and will never be back,” or contract-work employees, who fuel the so-called gig economy, Murnighan says. “As you’re leaving, you might grab something you shouldn’t have on the way out the door.”

Knowing that one is unlikely to face potential consequences can certainly highlight the temptation to cheat. But Murnighan and his coauthors show that it is not the ill-gotten gains themselves that motivate this “cheating at the end.” Instead, the dishonest behavior is motivated by an urge to avoid future feelings of regret at passing up a last opportunity for personal gain.

Now or Never

To explain the idea of anticipatory regret, Murnighan says to imagine someone thinking about buying a piece of real estate. It may look like a wonderful property, but other circumstances — interest rates, job insecurity, etc.—might make the potential buyer decide to pass.

Enter anticipatory regret. “Before they pass on the idea, they almost inevitably think, ‘it would really be stupid of me not to pursue this opportunity,’ even though they don’t pursue it,” Murnighan says.

(Anticipatory regret need not be limited to finances. Murnighan offers an example from his youth of deciding not to ask out a woman because she was rumored to be dating someone else. “I still regret it fifty years later,” he says. “If I had anticipated those feelings, I might have acted differently—to avoid the future regret of passing up that chance.”)

“The takeaway isn’t that leaders should always expect people to act unethically. Instead, they should make it easy for people to act ethically.”


Get the latest from Kellogg Insight delivered to your inbox.

Murnighan and his coauthors, Daniel Effron of the London Business School and Christopher Bryan of the Booth School at the University of Chicago, predicted that anticipatory regret about losing a last-chance opportunity to enrich oneself can promote unethical behavior.

To test this idea, the researchers devised four experiments to measure the timing of participants’ willingness to cheat.

Flip a Coin

In two of the experiments, hundreds of participants recruited through Amazon’s Mechanical Turk were asked to flip a coin and report what side it landed on, ostensibly as part of a study on whether people can move objects with their minds. Since the coin flips were self-reported, participants had ample opportunities to fudge their reports, and there was a small cash reward for landing on one side versus the other.

After analyzing the results of over 25,000 coin flips, Murnighan and his colleagues showed that participants were more than three times as likely to cheat when they thought they would have no more opportunities to do so.

Furthermore, the researchers ruled out alternative explanations for this dishonest behavior. For example, participants might have been motivated to cheat more at the end, because resisting multiple temptations had exhausted their reserves of self-control, or because repeated exposure to these temptations gradually acclimated them to the idea of getting away with it. Or participants may also have engaged in “moral licensing,” a peculiar psychological phenomenon in which prior instances of good behavior can make people feel like they have earned permission to loosen their ethical standards.

By manipulating the timing of some opportunities to cheat, Murnighan and his collaborators showed that anticipatory regret was the likely explanation. For example, the researchers surprised participants who thought they were only flipping the coin seven times by offering an eighth flip, thus providing an extra chance to cheat. Participants’ dishonest behavior spiked on their seventh and “final” flip, then diminished for the eighth unexpected flip. If participants had been motivated by the other potential incentives, says Murnighan, they would have continued to cheat on the eighth flip.

Testing Cheating at Work

Another experiment was similarly structured but used participants who were actually employed by the researchers for a specific number of tasks.

Participants signed up online to be temporary research assistants and review a series of short essays. They were paid by the minute and had to self-report their time. What they did not know was that the researchers could, in fact, see how much time they actually spent on each essay.

As with the coin flips, participants were more likely to fudge their billing when they were working on their final essay.

Endgame Ethics

As that experiment shows, the findings have the biggest implications for managing risk among short-term employees, Murnighan says.

“We tend to want to squeeze the last drop of productivity we can, particularly from part-time workers who aren’t coming back,” he says, “but I think that’s a mistake.”

Instead, he recommends loosening contractor commitments near the end of an engagement to mitigate potential dishonesty. “If you’ve got a two-week job, consider letting people go home at one o’clock on the last day instead of keeping them until 5 p.m.,” he explains. “When you give them a little bonus like that, they’re less likely to try and take some other kind of bonus for themselves on the way out.”

The results also suggest that when companies have limited resources to monitor employee behavior, managers should focus their attention on natural stopping points in procedures and schedules, which may tempt unethical behavior. “If we learned anything from the financial crisis,” Murnighan says, “it was that people are not going to regulate themselves when the temptation [to cheat] is high, and that’s particularly true in endgame situations.”

Still, Murnighan does not think that managers should interpret his research cynically.

“The takeaway isn’t that leaders should always expect people to act unethically,” he says. “Instead, they should make it easy for people to act ethically.”

To that end, he adds, common sense already goes a long way: “Don’t leave stacks of uncounted money laying around your office. Particularly on Fridays.”