Now that I have been pulled back into this discussion of AS-AD I
can't seem to get out. So here is my reply to Barkley Rosser (hi
Bark!):
1. Why do I get so agitated about the assumption that nominal MS
remains constant as AS shifts? Why isn't this just like the
pleasant fiction about demand curves remaining fixed as supply
curves gyrate in microland? Well, for the record, I look skepti
cally at the micro formulation as well when I teach micro (which I
no longer do...). I *do* include expectations as ceteris paribus
conditions of both curves and discuss circumstances under which
the shift in one curve will induce changes in the expectations
underlying the other. This is an important tool for analysis in
many markets. My particular gripe with fixed nominal MS is two-
fold: First, insofar as MS is endogenous there is a logical, and
not merely incidental, relationship between changes is AS and the
MS. That is, movement along an AD curve indicates a change in the
price level, which in turn induces agents to alter their money-
creating activities. Second, the relationship between time-scale
and the rate of change in (presumed) ceteris paribus conditions is
different in microland and macroland. In the former, the time
frame is usually on the order of weeks or months, and during that
period it is reasonable to suppose, for instance, that consumer
preferences are fairly stable. The time frame for AS-AD, on the
other hand, is quarters, and it is *always* the case that nominal
MS changes drastically from one quarter to the next. A misde
meanor in one context becomes a felony in the other.
2. The bottleneck theory of a relationship between the price level
and real output is somewhat plausible, although it should be
remembered that in a mass production economic with large fixed
costs there will be a large zone within which greater capacity
utilization leads to lower costs. But this reasoning pertains to
*realized* combinations of output and prices, not to a notional
curve that represents only one blade of the scissors.
3. Back in the days when I taught micro, I did exactly as Barkley
suggests: I presented the Keynesian cross in nominal terms and
then discussed to what extent movements were also real.
4. I took on the international substitution effect argument for
the AD curve in an exchange with Tom W. In short, I pointed out
that such an effect requires fixed nominal exchange rates, but
that in an otherwise ceteris paribus world, differences in na
tional inflation rates would result in offsetting exchange rate
adjustments. To argue that they wouldn't would imply nonrational
behavior by traders.
Peter Dorman
