A little cheating costs a lot

Behavioural economists are having a field day with the market meltdown, as traditional cost-benefit analysis and self-interest give way to more penetrating insights about how people can be systematically wrong — and how small acts can lead to broader failures.

“This whole thing with the market falling was very good for us,” says Dan Ariely, a Duke University professor, founder of the Centre for Advanced Hindsight and author of Predictably Irrational. He was speaking in Toronto at a seminar sponsored by SAS and the Richard Ivey School of Business.

“Alan Greenspan’s testimony when he came before Congress and said ‘Oops’, that was a very big day.” Indeed, behavioural economics reveals that people are myopic, vindictive and easily confused. But that also, Ariely says, leads to a free lunch, if people are irrational in a systematic way.

His involvement in behavioural economics began during his convalescence from burns he suffered during an explosion. He pondered perception of pain: in pulling off a bandage, which aspect should be minimized, its duration or its intensity? His nurses opted for duration. He learned later that it was the intensity that should be minimized.

This is part of a broader lesson on how intuitions betray us. Optical illusions, for example, occur because “your eyes still fool you in predictable ways.” Not only that, but we process information very superficially, taking in what we need to know, rather than everything that is relevant, and so make “systematic predictable mistakes.”

While the standard view in economics is of human beings as rational utility maximizers, for Ariely, they are “myopic, vindictive, mistaken, and don’t know what they really want.” These things come together in how people make decisions. And, he says, “the silver lining is that there are free lunches.”

It turns out that even slight changes in complexity can stump rational decision-making. Ariely points to the phenomenon of enrolment in an employer-sponsored pension plan. He suggests that a bad employer would give employees a list of options, a letter stressing that this is the most important financial decision the employee would ever make and thus requires reflection, and finally, a form asking the employee to opt in. A good employer could do exactly the same thing, with one difference: give the employee a form that requires opting out. Beneficial decisions frequently rely on the default option, Ariely notes.

“Some defaults are unavoidable,” he says. “The question is, how are we going to use them.”

Ironically, complexity can also force a decision by making it more explicit. Imagine an option between an all-expense paid vacation in Paris or in Rome. Now add an option for Rome without free coffee. The insertion of that third choice causes the participant to focus on Rome, with free coffee.

Another of his findings concerns the role of regret. In one study, people were offered money to take their medications. In a different iteration, they were offered a 10% chance to win $30. Those who were told they had won the lottery, but who hadn’t taken the medication, immediately started taking it.

“It turns out regret is a huge driver of human behaviour,” Ariely says. “Our happiness is often not determined by where are, but by where we could have been.” Thus, in Olympic competitions, bronze medial winners turn out to be happier than silver medal winners. That’s because the silver winner did take the gold. The alternative for the bonze winner was not to have won a medal at all.

The same insight applies to cheating, which is particularly topical now as misplaced incentives come to light on Wall Street. Yet, Ariely’s work began in the aftermath of the Enron failure, in an attempt to determine whether that was the work of a few rotten apples or something systemic.

In an experiment that involved answering 20 questions in five minutes, a $1 prize was awarded for every correct answer. On average, participants got five questions right. Ariely then varied the experiment. Participants could rip up their answer sheets and tell him how many questions they got right. The average rose to 7. Raising or lowering the stakes did not change the cheating rate. Ariely concludes that “lots of people cheat by just a little,” because they want to have their cake and eat it too; to “benefit from cheating and still feel good about oneself.”

This fudging is more pervasive than one might suspect. In another experiment, he had students sign a form saying they will abide by their university’s honour code when doing the experiment. Cheating stopped. Oddly, however, the university didn’t have an honour code. When the experiment was performed at a university that did have an honour code, and required a week-long boot camp, there was no impact. Ariely argues that “signing at the moment of temptation, even if it has no teeth, has a big impact.”

He tried to get both the Internal Revenue Service and an insurance company interested in the concept, based on the premise that lots of people cheat just a little. The only success he had was for a trivial project. The insurance company used his suggestion of having clients sign a waiver before they filled out a mileage form. Reported mileage rose 15%.

A related phenomenon is how people feel about cash. On one campus, Ariely planted six-packs of Coca-Cola in shared dorm fridges. They didn’t take long to disappear. He repeated the experiment using 6 one-dollar bills on a plate. No one took the money from the fridge, even though it could have been used to buy pop.

Similarly, when participants were put through the math quiz again and offered tokens that could be converted into money, cheating rates doubled. It is this experiment that he uses to examine stock options. While people wouldn’t dream of embezzling from their employer, they may show no such qualms about back-dating stock options, despite the fact that the options are convertible to cash.

Then there’s social acceptance. Repeating the math quiz, Ariely had an actor stand up after 30 seconds to claim the money. But then he had the actor wear clothing identified with a rival university. That discouraged cheating.

In running the experiments, Ariely says, “we find very, very few bad apples.” At most, he’s paid off $70 or $80 to big cheats. On the other hand, “we find lots of people who cheat just a little bit and cost us thousands of dollars.”

The lessons he draws are two: “We have multiple irrational tendencies,” he argues. And, “we have bad intuitions about them.”

(04/16/09)