Experimental Economics

GeekPress links to a fascinating article about neuroeconomics, and what particularly interests me are the various experiments the author describes. Experimental economics is itself rather new and exciting, and I did some reading about it for one of the lectures a few weeks ago at the Cato conference. One of the most interesting experiments is the "ultimatum game", described thusly:

One is the "ultimatum game," which involves two subjectsâ€”researchers generally recruit undergraduates, but if you're doing this at home, feel free to use your own kids. Subject A gets 10 dollar bills. He can choose to give any number of them to subject B, who can accept or reject the offer. If she accepts, they split the money as A proposed; if she rejects A's offer, both get nothing. As predicted by the theories of mathematician John Nash (subject of the movie "A Beautiful Mind"), A makes the most money by offering one dollar to B, keeping nine for himself, and B should accept it, because one dollar is better than none.

But if you ignore the equations and focus on how people actually behave, you see something different, says Jonathan D. Cohen, director of the Center for the Study of Brain, Mind and Behavior at Princeton. People playing B who receive only one or two dollars overwhelmingly reject the offer. Economists have no better explanation than simple spite over feeling shortchanged. This becomes clear when people play the same game against a computer. They tend to accept whatever they're offered, because why feel insulted by a machine? By the same token, most normal people playing A offer something close to an even split, averaging about \$4. The only category of people who consistently play as game theory dictates, offering the minimum possible amount, are those who don't take into account the feelings of the other player. They are autistics.

Even more interesting than this seeming irrationality is the simple "dictator game", in which the person who starts with the money can keep as much of it as he wants with no consequences. Participants still tend to give the other player around \$4, despite the fact that they could keep all \$10. There are many variations on these economic games and I think we can learn a lot from them, but it's hard to find funding for any experiments that would actually create incentives on the same scales that people face in their real-life economic decisions (accepting or rejecting a job offer, for example).