Prof. McKean clarifies the situation for us, again...
She writes
Rational choice modeling is only as good as one's
initial guesses about the actor's preferences, and it's
not helpful in a situation where the actor has
conflicting preferences (which is undoubtedly the case
with Noda and taxes).

Doesn't that pretty much put rational choice theory out of the running as a serious tool of understanding the complexity of almost any social, cultural or political phenomena? When analysts have to guess at actor's preferences (esp. when these are usually only revealed after the fact by actions or in self-serving interviews), and to isolate single or at least non-conflicting preferences, we have a pretty narrow range of applicability. More generally, doesn't the claim to explanatory power here require such a simplification of the flux and flow human behavior (including but not limited to the speculation of motive, impossibly elastic definition of "utility," and that we ignore actor's own imperfect understanding of the world and its various choices...phew...), well, does this not generate a rather arbitrary or self-evident or unprovable explanation? (This is to set aside the claim that a 'culturalist' or 'institutionalist' deterministic argument is any better or worse.)

Even if we imagine that any "explanation" indeed needs to do some of these things--that is, find a way to sort out the complexity into some orderly patterns--there must be some lower limit below which we cannot go without charges of reductionism outweighing possible insight generated through the attribution of 'rationality'?

David Slater
Sophia U.

--
David H. Slater, Ph.D.
Faculty of Liberal Arts
Sophia University, Tokyo