Deal or No Deal...economists holding 25 suitcases?

David Glenn in the Chronicle of Higher Education (hat tip: Jay White) on economists at work-- studying why people do what they do...even in game shows!

This year’s annual meeting of the American Economic Association, which wrapped up here on Sunday, had its share of sober panels on poverty, disease, climate change, and the American mortgage crisis. But there was also, perhaps inevitably, a session devoted to a certain ubiquitous game show.

On Sunday afternoon, scholars from three continents gathered to discuss Deal or No Deal, which has become a worldwide phenomenon since it had its premiere in the Netherlands in 2001. The show — in which contestants must decide whether to accept a firm cash offer or pursue an uncertain, potentially larger reward — has been catnip for economists because it acts as a laboratory of human risk-taking.

“The show offers real and large stakes, much larger than we’re able to offer in our own lab experiments,” said Morten Lau, a lecturer in economics at Durham Business School, in England. The sole drawback for scholars, Mr. Lau said, is that it’s often hard to know much about the contestants’ personal characteristics. (Another caution was raised by Sandra Maximiano, a graduate student at the University of Amsterdam, who wondered whether social pressure from the studio audience makes contestants’ behavior different from typical economic responses.)

The central question that interests economists here is: How do people behave under conditions of risk and uncertainty? Does their level of risk aversion remain constant, as so-called expected-utility theory predicts? Or does a person’s risk tolerance fluctuate across situations, as prospect theory predicts? If it fluctuates, does it do so in a predictable way?

In a paper scheduled to appear in the March issue of the American Economic Review, four scholars argue that contestants’ behavior in Deal or No Deal fits prospect theory much better than expected-utility theory. For example, if players receive a shock of bad news — if they open briefcases containing the highest rewards, which means that their own potential reward has sharply dropped — then they will typically become much more tolerant of risk in later rounds of the game.

At Sunday’s panel, three of that paper’s authors presented a follow-up study that compares contestants’ behavior across national editions of the game. The potential highest reward varies heavily from country to country. (In the Netherlands, the top suitcase is worth 5 million euros, whereas in Australia the top prize is 2 million Australian dollars, which is the equivalent of only 1.2 million euros.) The scholars found that contestants’ behavior is not heavily affected by those differences; local, situational effects within each game, not the absolute level of the potential prize, shapes a player’s risk behavior....