March 22, 2005

Black box  

Via apostropher I see this article on neuroeconomics, apparently the latest paradigm for thinking about where basic economic assumptions about rational actors might be going wrong:

Neuroeconomics, while still regarded skeptically by mainstream economists, could be the next big thing in the field. It promises to put economics on a firmer footing by describing people as they really are, not as some oversimplified mathematical model would have them be. Eventually it could help economists design incentives that gently guide people toward making decisions that are in their long-term best interests in everything from labor negotiations to diets to 401(k) plans. Says Harvard University economist David I. Laibson, another leading researcher: "To understand the real foundations of our behavior and our choices, we need to get inside the black box."
I'm not against finding new ways to approach problems, but something bothers me about this. Maybe it's the implicit judgment in this idea that we could help people make "decisions that are in their long term best interests." One of the refreshing things about economics (when it's properly applied) is the assumption that if an individual chooses something, it's because that's what has the most value to her. It's called revealed preference, and it's value neutral (because the individual decides for herself what to value). This isn't to say that traditional economists don't make value judgments (!) -- but it removes a lot of the analytical distance between what we choose and what we should choose.

By the way, neuroeconomics isn't the first stab at these problems with rationality. Richad Thaler, who's briefly mentioned in the article, has done some interesting work on this, albeit with a more traditional economic approach.

MORE: A reader/economist writes in: "Perhaps you don't find it that interesting because it's obvious to you that we use different parts of the brain to make different economic decisions. But to 99% of economists, it's like a revelation. We're conditioned to believe that people make decisions that roughly correspond to maximizing payoffs. But when it's so obviously not the case, like the NHL debacle, we don't know what to say."

Frolic  {March 23, 2005}

Considering humans as irrational actors instead of rational actors should really improve the predictive power of economics.

Isn't that the problem of all the pseudo-sciences, their inability to predict future results? Not that there isn't value in studying past behavior, it's just dangerous to assume that social models can work like scientific ones.

paul  {March 23, 2005}

Economics doesn't have so much trouble predicting results as you might think. On the macro level the systems are certainly too complex, but on the micro level some of the predictions are startlingly accurate. I had a big argument about this yesterday -- what the distinction between social science and science is -- and I'm pretty sure it's not about predictive power (as the person I was talking with argued).

Maybe the main distinction has to do with human subjects -- in social science it would be unethical to run the "ideal" experiement because the subjects are human, so we have to regress out the unwanted factors statistically instead of through strict experimental design.

Post a comment

Remember personal