Life, for better or worse, is full of endings. We finish school, get a new job, sell a home, break off a relationship. Knowing that a phase is soon coming to an end can elicit the best in us, as we try to make amends for errors past and avoid last-minute regrets. We might try to visit that local museum, or make time for happy hour drinks with a longtime coworker, or be more generous with our praise to a partner.
But while the sense of an ending can draw out people’s finest selves, it can also, new psychological research suggests, bring out their darker side. This study concludes that, as people get closer to finishing an activity, they become more and more likely to deliberately deceive others for their own benefit. And they do this, the research shows, because they anticipate regretting a missed opportunity to cheat the system.

To demonstrate this “cheat at the end” effect, the experimenters conducted an ingenious series of experiments. In the first one, they recruited a large number of people through the internet to participate in a coin-flipping experiment. The task was to flip a coin a handful of times and, each time, to simply guess which side it would land on. So hundreds of people at their home computers duly dug a quarter from their pockets and flipped. Each time, they marked whether they had guessed right or wrong. And each time they got it right, they won a small cash prize.

The experimenters told people explicitly not to cheat. But of course, since people were performing the study in private, there was no way to catch anyone red-handed. It was, on the other hand, possible to detect cheating in aggregate. If no one cheated, the percentage of correct guesses is expected to be around fifty. Minor deviations from this value are normal, but statistics tell us that anything greater than a few percentage points is evidence that people stacking the deck in their favor.

The results were surprising. In the early rounds, the percentage of right guesses reported deviated very little from fifty percent—suggesting few people distorted the results. (This is reassuring for those optimists out there, as it suggests
that people are often honest even when they don’t have to be.)

In addition, the number of times people tossed the coin did not have a strong effect on how much they cheated. Even by the fifth flip, the number of correct flips reported wasn’t much greater than chance. What did matter, by contrast, was how many tosses people thought they had left. The experimenters told some people they’d get seven tosses, others ten. The “seveners” were fairly honest up until the seventh flip—then they cheated like mad, with two thirds reporting a correct guess—sixteen percent higher than expected. By contrast, the “tenners” were honest up until the tenth flip—and only then did they start cheating. This suggests that what makes people want to cheat isn’t how many opportunities they’ve had in the past—it’s how many chances they’ve got left.

The experimenters wanted to see if this phenomenon held up in a more true-to-life situation. In this experiment, they “hired” several hundred people to help grade essays that had ostensibly been written as part of another study. Similar to the coin-flip study, people were told they would be reading either seven or ten essays. They were paid according to the amount of time they invested in each essay. After each essay, they were asked to report how much time they’d spent.

Unbeknownst to participants, their time was also being logged by a secret timer. So the experimenters could simply see how much their reported time differed from reality. The results confirmed the previous study: while the seveners cheated wildly by the time they reached their seventh essay, reporting spending at least 25% more time than they actually did, the tenners waited ‘til their final round to do the same. Once again, cheating happened most when people thought they were nearing the end of the line.

The implications of this research extend beyond the walls of the laboratory. Political terms, job tenures, school years, golf games—all happen over a finite period of time. We’d be wise to keep an extra-vigilant eye, therefore, on lame-duck senators, students in spring semester, and golf partners on the eighteenth hole.

Even further, it demonstrates another side of a natural ability to anticipate the future. A strange mood comes over us when we have the sense of an ending. We get a little friskier; we live with more abandon. (Jon Stewart, for instance, was arguably never perkier than in the last few months of filming the Daily Show.) This fact jibes with studies showing that that dopamine, a chemical in the brain associated with pleasure and risk-taking, ramps up in rats as they near the end of a maze. This sense of anticipation can help us eke the most out of transient moments, as we try to squeeze a little more from the toothpaste tube, even if it doesn’t belong to us.