Dishonesty is a slippery slope: If you behave dishonestly once, you may become more likely to do so again in the future, a new study from England shows.

The reason may be that the brain grow less sensitive to self-serving dishonest behavior over time, the researchers said.

In the study, researchers asked 80 adults ages 18 to 65 to advise a second person about the amount of money in a glass jar of pennies. In several of the trials, conditions made it so that dishonesty benefited the participant. For example, researchers might promise the participants a higher reward if their partners overestimated the number of pennies in the jar.

People’s dishonesty escalated over the course of these trials, found the study, published online today (Oct. 24) in the journal Nature Neuroscience.

“This study is the first empirical evidence that dishonest behavior escalates,” Neil Garrett, the lead author of the study and an experimental psychology researcher at University College London, said at a news conference about the new findings.

With 25 of the participants, the researchers conducted the penny-jar experiments while a functional magnetic resonance imaging (fMRI) machine scanned the person’s brain. The results showed that the amygdala, a part of the brain connected to emotions, showed a marked reduction in activity in response to self-serving dishonesty over the course of the trials.

In fact, researchers found that the amount of the reduction in the amygdala’s activity for each trial could predict the amount that the participant’s dishonesty would increase in the next trial: The bigger the fall in amydala activity during one trial, the bigger the lie would be the next time.

“It is likely [that] the brain’s blunted response to repeated acts of dishonesty reflects a reduced emotional response to these acts,” Garrett said.

The study also offers support for the idea that the activity in the amygdala “signals aversion to acts that we consider wrong or immoral,” Garrett said. In other words, whenever a person lies for personal gain, the amygdala produces a negative feeling that helps curb that act but the more often a person lies, the more the response fades, leading to a slippery slope that may encourage an escalation of dishonest behavior.

People in the study actually lied the most when their lies benefited both them and their partners. This may be because it is easier to rationalize these lies, said Tali Sharot, the senior author of the study and an associate professor of cognitive neuroscience, also at University College London. In this condition, the amygdala did not show the same response pattern as when people lied solely to benefit themselves, she said at the news conference.

Interestingly, though, the researchers found that study participants never lied as much as they could have. Participants’ estimates of the value of the coins in the jar were always significantly lower than the ceiling, meaning that the individuals “always had an opportunity to lie more than they actually did,” the paper said. Sharot explained that people usually lie by just a little bit, perhaps so they can still hold a relatively positive perception of themselves.