Thu, 27 Sep 2007

Game Theory

I'm dubious of the value of game theory to economics. Games have winners and losers. Markets have participants. Free market only have winners; the would-be or potential losers choose not to play.

I'm simplying, of course, glossing over the existance of human error. People who think they're going to win will lose, but they do so only because they've made a mistake. Perfect markets are not an option, as any economist will readily acknowledge.

So where does the idea of "perfect markets" come from? Economists must use thought experiments to create theories about how markets work. One of these experiments is the perfect market; one without human error, without transaction costs, and where every participant has equal knowledge. Another experiment is the unchanging market; one without growth, decay, decisions, or any other number of human values. Within these simplifications, you can say what a market will do.

Of what value is that in the real world? It helps economists to say what will tend to happen at the margin. If transaction costs are lowered, markets will behave more like perfect markets. If human error can be reduced (hey, at great cost, we did it for the Apollo missions), markets will behave more perfectly. If real markets then turn out to behave differently, the theory is wrong, and the economist goes back to the drawing board.

Unfortunately, undergraduates have been imperfectly taught about perfect markets, and you can see the results.

Posted [00:54] [Filed in: ] [permalink] [Google for the title] economics [digg this]