And now just after writing the below, I find myself reading George Akerlof’s 2006 AEA presidential address, where he argues, more or less, that the whole problem with modern macroeconomics is the assumption that only economic arguments enter into utility functions. Bring in norms, and it’s goodbye to the permanent income hypothesis, Modigliani-Miller, Ricardian equivalence, and the rest of it, and back to where positive, empirical, pluralistic macro left off in the ’70s.
(The problem of preferences that include non-economic arguments is also the subject of Amrtya Sen’s article The Impossibility of a Paretian Liberal, which John Holbo so catastrophically misunderstood last year.)
Different angle of approach, same destination, seems to me.