1) Yes, there are numerous difficulties with AS and AD, one not
mentioned so far is the interdependence of the two, perhaps implicit
in Peter Dorman's comments on circular flow models. Colander's new
text basically has vertical (classical) AS curves that shift when AD
changes, no question of higher price drawing forth greater output.
(Like D,F, and S he also was forced by the publisher to stick in
"regular" AS and AD curves in an earlier chapter.) I am currently
using his book, and students like it, but one difficulty with this
model is, how do you know when an AD change will cause a price change
and when it will not?
2) The statement that "inflation is everywhere a purely monetary
phenomenon" is a classical view quoted in most texts but only in a
very few actually put forward as the final truth about inflation.
If V and M are constant (mild classical case), then AD is a rectangular
hyperbola.
3) To Peter: I do not see why you get so worked up about the ceteris
paribus condition of holding nominal MS constant. Of course it changes
when AD or AS changes. But I know you teach micro. Do you teach supply
and demand? Ceteris is not paribus at the micro level either when one
draws the supply and demand curves (how many out there actually tell
students that FREQUENTLY people buy more of things when their prices
go up because of speculation, changes in expectations, etc.?) As near
as I can tell the only alternatives to supply and demand are either
Marx or (Gil Skillman and Jim Devine, are you reading?) Prisoners
dilemmas are in YOUR future! that is lots of game theory.
As you well know, fixed nominal MS gives the Keynes effect of
real interest rates rising with P thus reducing I and the quantity of AD.
4) Although pretty hokey, the nonlinear AS is a story that makes sense
to student in the short-run, especially if one is talking about the
effects of increases in AD. Prices and wages are fixed in deep recessions
(the "Keynesian zone") but begin to rise as different sectors become
"bottlenecked" until AS goes vertical at the PPF.
5) This story can be easily integrated with the Keynesian cross if one
does what none of the textbooks do and tell that story with NOMINAL
variables on both axes. Then the change in equilibrium measured along
the horizontal axis equals the shift of AD in the AS-AD space. How
much of that change becomes changes in P and how much changes in Q
depends on where one is on the SRAS. (BTW, the multiplier story can
still be told, even in a classical world, but the extra rounds of
spending all show up as successive price increases.)
6) Besides the Keynes effect, downward-sloping AD can be justified by
Pigou effect and the international substitution effect. Of course there
are problems with all of these as I show in my Fall 1991 JPKE piece
where I present the possibility of a vertical AD corresponding with a
vertical AS.
Have a nice semester!
Barkley Rosser
James Madison University
