Game Theory vs Behavioral Economics

Michael Mandel

The economics blogs had a very positive reaction, in general, to this morning's announcement of Robert Aumann and Thomas Schelling as the latest Nobelists in economics . For example, Tyler Cowen writes

I am happy to see Schelling -- a fruitful generalist if there ever was one -- and Aumann, a deeply philosophical thinker, get the nod. Aumann I don't know personally, but there are few scholars I admire more than Thomas Schelling.

I'm going to take a bit of a different position. I agree that Schelling and Aumann are deserving prize winners. But their subject matter--game theory--has big problems.

In my opinion, despite today's Nobel prize, game theory has hit a dead end. The more fruitful road forward leads through behavioral and experimental economics. (I say this as someone who wrote his doctoral dissertation on game theory).

In any real-life situation, game theory does not predict a single equilibrium. Instead, it typically offers a very wide range of possible outcomes. Moreover, game theory assumes a certain kind of rationality which has not been backed up by the facts.

Behavioral and experimental economics don't start off by assuming rationality. Instead, they actually look at people's real-world behavior, and then make predictions based on that. For example, see the 2002 Nobel Prize in economics, granted to Daniel Kahneman and Vernon Smith.

More about this later.

    Before it's here, it's on the Bloomberg Terminal.