Asymmetrical Information

From the desk of Jane Galt:

Welcome, Tech Central Station Readers

To what is, apparently, one of the primary homes of Sweetwater economics on the Web.

I'm not quite the uber-free marketer some Chicago folks are; for a true "markets are efficient" advocate, head over to the excellent Zimran Ahmed, who has studied at the feet of such Chicago luminaries as Kevin Murphy and Eugene Fama. My position on free markets might be aptly summed up as "Markets fail. . . and government intervention fails worse." But you can see for yourself by poking through the archives.

I've been maligned on one point: I studied the works of Kahneman and Tversky in business school, and I think they represent some important insights about human behavior. Where I think they fall short is what Arnold Kling points out: we haven't yet found a way to scale their findings up into a useful predictive model. Until we do, we're going to have to assume in large ways that human beings are rational actors in the economic marketplace. Why? Because to an economist, there's only one way (or at least, a very limited number of ways), that people can act rationally; ultimately we assume that they are doing their best to maximize value as they perceive it. Contrariwise, there are an unlimited number of ways to act irrationally, which makes it hard to build a predictive model around it.

When you're engaging in economic activity -- say, buying a car -- it's easy to assume that you will act rationally. For an economist, this means you maximize your utility, which is fancy econospeak for the value you get out of an object, by choosing the combination of price and features that suits you best. It's very hard, on the other hand, to assume that you won't act rationally. Will you pay too much? Get too few features? Sit down on the tarmac and shove chopsticks up your nose? The possibilities are limitless. Economists know that people don't always rationally maximize value. . . but unless we know exactly how they go astray, we have to continue to assume they're rational. The resulting models are imperfect -- but better than throwing up their hands and giving up.

What Kahneman and Smith tried to do was to set up controlled experiments to reveal the way in which people act irrationally. The problem is that it is very hard to take those experiments out of the lab, and translate them into the way people behave in real life. Just one example: how well people maximize value is tied to the magnitude of the activity involved; you spend more time trying to get a great price on a car than on toothpicks. In a lab test, where the amounts are trivial and the environment feels like playtime, people may behave very differently from the way they do when they're shopping, working, or do all the other stuff we call economic activity. So while their work is important, it is perhaps not quite useful yet.

But the same thing can be said of babies and butterflies. We still love them anyway.

Update Darn that John Irons, summing up my views better than I did:

I think that there is likely to be a nice synthesis between the two views - one that models people as "rational" in the sense that they optimize something. It's the "something" that is the trick; and behavioral and experimental economics are trying to work that out.

Posted by Jane Galt at November 13, 2002 01:14 PM | TrackBack | Technorati inbound links