People often face multiple opportunities to cheat undetected for financial gain. For example, a series of invoices provides several chances to overbill, and a series of expense reports provides several chances to over claim reimbursable expenses. Earlier research studies show that, when facing such opportunities for dishonesty, most people will cheat a little bit, but not as much as they could get away with.
What this earlier research did not address, however, was whether people commit their little bit of cheating at predictable time points. I recently investigated this question with fellow researchers Christopher Bryan at the Booth School of Business and J. Keith Murnighan at the Kellogg School of Management.
We theorised that people are more likely to cheat when their window of opportunity is about to close. Faced with several opportunities to enrich themselves dishonestly with no chance of getting caught, people may be more likely to cheat on the last opportunity – a "cheat-at-the-end effect."
To test this idea, we conducted several experiments that presented more than 2,500 participants with more than 25,000 chances to cheat, undetected, for money.
The results showed that the odds of cheating were indeed higher when people knew that it was their last chance to cheat, compared to when they thought they had more chances left. The data also suggested that people cheat at the end because they want to avoid feeling regret about passing up a "last chance" for personal gain.
In one experiment, we hired 327 research assistants to perform several short work tasks over the Internet. The assistants were paid based on how long they claimed to have spent on each task. To determine whether the assistants overbilled for their time, we secretly measured how long each assistant actually spent on each task.
Different participants were assigned different numbers of tasks. The results showed that the assistants overbilled the most for whichever task was last. For example, participants overbilled for their seventh task by 42% more when they knew it was their last one, compared to when they knew that three tasks remained.
These results have important implications for organisations that want to reduce dishonesty but have limited resources to monitor employees. Such resources are perhaps best deployed when dishonesty is most likely: when the end of a contract or the close of a billing period makes people feel like it is their last chance to cheat.
Daniel Effron is assistant professor of organisational behaviour at London Business School
