Contagious Dishonesty: How One Bad Apple Can Ruin the Barrel

March 12, 2009
Behavioral Science, Leadership

DURHAM, N.C. –- Bernard Madoff’s arrest and guilty plea for scamming investors out of billions of dollars may serve as a cautionary tale for some, but could actually increase dishonest behavior by others, says Professor Dan Ariely of Duke University. 

In a study published in the current issue of Psychological Science, Ariely and colleagues found that social norms exert a strong influence that can override other factors in determining how people behave after they observe dishonesty.

“We see examples of unethical behavior all the time, but that doesn’t mean we all are influenced to act dishonestly,” said Ariely, the James B. Duke Professor of Behavioral Economics at Duke.  “But if you identify with someone, say you want to be a Wall Street-type like Madoff, in these cases our research shows that you are more likely to emulate their unethical behaviors.” 

In order to understand the settings in which people were most influenced by others’ unethical actions, Ariely, working with Francesca Gino of the University of North Carolina at Chapel Hill’s Kenan-Flagler Business School and Shahar Ayal of Duke, conducted a series of experiments testing the conditions under which students would cheat. 

The researchers gave groups of students from Carnegie Mellon University a series of math problems, $10 cash and a sheet on which to tally the number of problems they could solve in five minutes. In one session, the students were given no opportunity to cheat; they submitted their answer sheets to a proctor who tallied the scores and instructed the students to keep 50 cents for each correct answer, and return the balance of the $10 in an envelope.

In order to assess the number of students who would cheat, students in a second session  were not monitored. They were asked to tally their own scores then shred their tally sheets, and pay themselves 50 cents for each correct answer. Students in this group reported solving 50 percent more questions than the monitored students.

“What we saw in this experiment is that participants inflated their scores in order to take home a bit more cash,” Gino said.  “None of the students knew that the others were cheating; they were each making individual decisions about what behavior they would feel comfortable with in that situation.”

Next, the team tested what would happen if the students knew somebody else had cheated. The researchers hired an actor who dressed in a plain white T-shirt and looked like a university student. Sixty seconds into the test, the actor stood up and said, “I solved everything. My envelope is empty. What should I do with it?” He was told he was free to leave if he did not have any money to return.

Finally, the team ran a slightly different version of the experiment in which the actor wore a University of Pittsburgh t-shirt to participate in the experiment conducted at Carnegie Mellon. Again, the actor declared he was finished and had no money to return after one minute of the test. 

“We looked at how the other students responded to the actor’s announcements that he had completed the entire test in one minute, even though the test is virtually impossible to complete even in five minutes,” Ayal said.  “We found that whether the actor appeared to be a member of the students’ community (CMU student) or a member of another group (Pitt student) influenced their cheating.” 

When the actor wore a white T-shirt and thus appeared to be a Carnegie Mellon student, a quarter of the students followed suit and claimed to have correctly solved all of the problems. However, when the actor wore a University of Pittsburgh T-shirt that indicated he was not a member of the students’ community, only one student (3.6 percent) imitated this behavior and claimed to solve all of the problems correctly.

“This is a frightening example of just how easily our own behaviors can be swayed by our judgments of the people around us,” Ariely said.  “In the Madoff case, for example, these findings would indicate people who do not relate to Madoff in some way will not be negatively influenced by his behavior. But people who do identify with him, whether that’s through a desire to make more money, or a need to view themselves as successful Wall Street-types, may in fact be encouraged to act dishonestly because they see his behavior.” 

This story may not be republished without permission from Duke University's Fuqua School of Business. Please contact for additional information.

Contact Info

For more information contact our media relations team at