Your use of the word "classical," by analogy to physics, evokes another major failing of economists. They do not understand nonlinear models; they think of the world as an interlocking system of partial differential equations.

The old notion was "simple models, simple behaviors." If you picked up a complex behavior, you needed a complex model to explain it. It turns out that this is far from true. You can't blithely assume that the "invisible hand" will guide a system to a stable equilibrium, because in a lot of cases, it won't: there are a lot of simple models that, for certain ranges of parameter values, either oscillate periodically or just behave chaotically.

If something like fluid dynamics, with its simpler feedback mechanisms, requires this to be taken into account, how much more must a working economy, with all its informational feedback loops, require it?

Posted at July 2, 2007 2:19 PM in response to Predicting with a Handicap: Why are Economists’ Predictions So Often Wrong?