Econ and Data (Calming Down)

Ok, my worry about my attack on econ was that I would make the same mistakes that anti-evolutionists do, which all ultimately stem from not knowing the field. I certainly drew the same response from SLU's economist, Steve Horwitz, that PZ Meyers typically gives evolutionists anti-evolutionists. [Typo corrected after the first post.]

So maybe its time I talk more reasonably. I should concede that there are empirical data that are used to verify economic models.



Google quicly reveals that Springer-Verlag publishes a journal called Empirical Economics. Science has an article this week about an experiment which combines a standard game used in economics experiments with neuroimaging technology. You know the kind of game--where you are given a certain amount of money and you have choices like cooperating or not cooperating, and diffferenc choices lead to different payoffs. The very fact that I can refer to a "standard game" used in economics shows that there is data here.

What bothers me is that data in econ always comes from labratory situations, like the game in the neuroimaging examples, or from parts of society that have established rules based on economic principles, places like stock markets. Economic models work on the parts of human existence that are designed on economic models.

The problem is that people who use economic arguments often claim more than this. A commentor on the Bitch discussion began his post by reminding us that economics is not just about money, it is really a branch of sociology. Econ 101, as I understand it, begins with the claim the economics is the study of the distribution of scarse resources.

See, that's the problem. If economics were only about the money, than it would work. It is not very good as a sociological theory or as a theory of scarse resources in general. As far as I can tell you don't get succesfull predictions in contexts that are not directly about money. The same commentor on B PhD (haloscan is down, so I can't link to him) said that we should not shy away from applying economic models to the family. But this is precisely the point where econ makes no good predictions. Psych and Sociology can tell us things about the risk factors for domestic violence for instance (family background, alcoholism, loss of external support systems, economic stress.) Models that assume people are idealy rational preference maximizers have no comparible success.

This is why we can't let economic language frame policy debates--for instance by arguing for public support for children because children are a "public good". Economic arguments can play a role in determining the best means to ends once the goals are established, e.g. how to pay for public support for children once we agree that it is the only decent human thing to do, thus avoiding the dire consequences Steve warns of. But we shouldn't act like most of our lives are made up of economic transactions. Because they aren't, and we wouldn't want them to be.