As Jeff pointed out in an earlier post, David Levine thinks rational choice theory is remarkably successful and that behavioral economics may be doomed.  This message has made it into experiments and the New York Times:

[S]uppose, instead of scanning people’s brains as they’re sipping wine in a laboratory, you tested them in a more realistic situation: a restaurant where they’re spending their own money. That challenge was undertaken at an upscale restaurant in Tel Aviv by two behavioral economists, Ori Heffetz of Cornell and Moses Shayo of the Hebrew University of Jerusalem, who expected to be able to manipulate diners’ choices by changing the prices on the menu.

Unbeknownst to the diners or to their waiters, the economists monitored the choices of people who ordered from the prix fixe menu. The three-course meal cost included a choice of five entrees: shrimp gnocchi, pork shank, red mullet fillet, sausage or stuffed artichoke.

Next to each of these entrees on the menu, in parentheses, was the cost of what it would cost to order that entree from the à la carte menu. These prices didn’t affect the cost of the prix fixe meal, which was the equivalent of $30 no matter what the entree, but the researchers expected just the sight of the prices to make a difference. If the mullet were listed at $20 and the other entrees were $17, more people would presumably be enticed into ordering the seemingly more valuable fish.

But after three months of testing various combinations of prices, the researchers found they couldn’t sway the customers. Putting a higher price on the shrimp or any other entree didn’t make people more likely to order it.

This same stubbornly independent streak was manifest in another food experiment by the same researchers. This time they let people sample two kinds of candies — peanut butter bars and caramels — and varied the sticker prices for each one.

Superficially, the manipulation seemed to work, because people said they would be willing to pay more for a candy if it had a higher sticker price, but that was just in answer to a hypothetical question. When people were given a chance to pick a bag of candy to take home, they pretty much ignored the sticker prices and chose what they liked.

Why weren’t people duped into favoring the high-priced candies and entrees? Why did they follow their own tastes?

“Maybe, sometimes, old-fashioned economics is just about right,” Dr. Shayo says. “Maybe when it comes to food, people do have reasonably stable preferences. Some people like shrimp and some don’t, even if it’s worth a lot of money.”

Interestingly, the results also back up another hobbyhorse of economists: experiments with real payoffs give very different results to those relying on answers to hypothetical questions.  As economic decisions involve real payoffs, its the results with real consequences that are a better predictor of what decision-makers will do when faced with real decisions.  Economists insist that research papers with experiments use monetary rewards.  I always wondered if this really mattered – perhaps it does.