Equilibrium Concepts in Economics

4/05/2013 06:55:00 AM
Noah Smith has an excellent post answering the question: What is an equilibrium?

There is something I wish Noah had mentioned: most economics theory papers these days explicitly write out the mathematical definition of the equilibrium concept they use, except maybe if they are on a topic that already has a well-established equilibrium convention. It's not uncommon for papers, especially in game theory, to apply three or four different kinds of equilibrium concepts and compare them. And of course there is a whole subsidiary line of research about when it is appropriate to use what kind of equilibrium in your model.

The issue is this: you can write down utility functions, budget constraints, and production functions, but those things by themselves don't give you a prediction for what would happen in this theoretical economy. So we stipulate that households maximize utility and firms maximize profits. Because, why would they do anything else? But even that doesn't necessarily give us a clear prediction. Often we make additional assumptions, like all firms act "symmetrically" which basically just means they all make exactly the same choices. Why, you ask? Well...they are all identical firms, so why wouldn't they behave identically? Basically, the definition of an equilibrium is all that extra stuff you need to turn a mathematical description of an economy--the utility functions, production functions, budget constraints, etc--into an actual prediction for the economy. If that sounds dubious, it is. But it is good enough most of the time.

I think that what Noah is getting at is that people who criticize economics because "equilibrium doesn't really exist" aren't making a valid argument. What they should be arguing is that "this equilibrium concept does not make reliable real-world predictions." That, of course, is an empirically testable thesis.