Microfoundations Revisited

Some response to feedback I got on yesterday’s microfoundations post. Some microfoundations fans seem to feel that the “Lucas Critique” has some special properties that lead to the conclusion that macroeconomics has a unique need for microfoundations that does not exist elsewhere in the physical, natural, or social sciences.

This seems wrong. If you detach the specific content of the critique from the rest of Lucas’ career and program you’re left with a macro-level critique of macro-level theories about macro-level phenomena. Lucas points out that if you look at past data and find statistical relationships between different kinds of variables (specifically inflation and unemployment) it doesn’t necessarily follow that you’ve discovered an exploitable policy tradeoff because once it becomes known that policymakers are attempting to exploit the tradeoff that changes the dynamics of the situation. This is logically persuasive and over the course of the 1970s it seemed to pass the “provides a useful insight about the world” test as policymaking driven by a particular line of thought about the Phillips Curve did not lead to the results that were wanted.

But that’s to say that the problem with the idea of the endlessly exploitable Phillips Curve inflation-unemployment tradeoff is that it’s wrong not that it lacked microfoundations.

Noah Smith while being seemingly annoyed by my physics analogies gets this right:

And I agree that productive theoretical analysis can in principle be conducted at any level of complexity. If this idea is used to argue that we shouldn’trequire macro theories to be microfounded, then I agree. If this idea is used to argue that looking for microfoundations in macro is a waste of time and effort, then I disagree. Having microfoundations is better than not having them.

Switching science metaphors, the way I would put it is that lower-level foundations make a good theory even better. As we go from Darwin’s insights to Mendel’s to deeper understanding of the workings of DNA replication, a perfectly nice theory of evolution by natural selection is getting stronger and more persuasive. But you don’t reject Darwin out of hand because he didn’t know how to ground his theory in Mendelian genetics. And you certainly don’t make inquiries about how the world works by arbitrarily insisting that all inquiry has to move from the smallest-scale phenomena upwards.