Jim Devine wrote,
> Some sort of abstraction is needed if you believe that the macroeconomy
> is more than the sum of its parts. The measure of aggregate demand is an
> abstraction, but without it one is stuck with a pre-Keynesian vision of
> the world (which basically saw the macroeconomy as the microeconomy writ
> large, applying what's now called the "representative agent model," and
> applied Say's bogus "Law").
I agree that abstraction is needed. I don't agree that a single
abstraction captures both the "more than the sum" and the "of its
parts" perspectives. To sum up our differences, it seems to me that you
are saying inflation arises from the aggregate as conditioned by its
heterogeneity. I'm saying it arises from heterogeniety as it is
influenced by the aggregate. The whole may be more than the sum of its
parts, but the parts have to come before the more.
> If one measures aggregate demand using nominal GDP, by the way, the
> relative weights of demand are irrelevant, since nominal GDP is simply
> the amount of money spent buying newly produced goods and services
> supplied through the market.
But if one is thinking about change in aggregate demand, the relative
weights aren't irrelevant.
> this doesn't make sense to me. Say's Law was shown to be inadequate by
> Marx and Keynes, among others. That is, the "Law" is wrong, untrue. It
> should be discarded. The Phillips Curve is not on the same level of
> abstraction. It's an empirical generalization that has several theories
> attached to it.
My impression was that Say's Law is one of those tautologies that is
trivial to the extent that it is true and misleading when it is stretched
beyond its short reach. There may well be good uses to which the Phillips
Curve may be put, but it seems to me it has mainly been used as an after
the fact and beside the point justification for good ol' fashioned
Treasury conventional wisdom.
Tom Walker
