----------------- HES POSTING -----------------
I thought I had better add my 2 cents on AS-AD. First, I disagree a little with Prof.
Lipsey that the AD curve is no better or worse than the IS/LM model, and that it is better
done in intermediate courses. One can easily have the price level affecting aggregate
expenditures in the AE/AP model and thereby derive the AD curve as a combined goods
implicit money market equilibrium. Most principles books now do that.
The point is whether it makes sense to do so and to add an aggregate supply curve back to
the resulting diagram to determine equilibrium. If everything is done algbraically and
one only talks about eqiuilibrium points, one can have a
consistent analysis, but as soon as one starts talking about disequilibrium, you have to
do some fancy talking to explain how this combined goods market money market equilibrium
curve interacts with the aggregate supply curve that is somewhat ambiguous in its own
right. Both principles books and intermediate books are rather unclear about this issue.
Since the entire discussion of reasons for the downward slope of the AD curve are of
doubious empirical relevance for the degree of aggregate fluctuations we see, much of the
analysis of the downward sloping AD curve that they present is simply "irrelevant to the
checkerboard of real life" as I believe Pigou put it early on.
Professors focus on the AD curve because they don't want to get into those issues, and the
AS/AD model is usually used as a simple way to discuss policy, whereas in earlier
intermediate books there used to be an entire aggregate
model (production functions, wages, prices etc) that served as a foundation for it. Back
then it was a complete aggregate model. Now, for the most part, it is just a policy model
that comes out of nowhere, or almost nowhere in both intermediate and intro books. That's
because the clear (but empirically suspect) production function basis for the AS curve is
no longer there. Instead, there are a variety of assumptions that come from labor market
or product market regidities.
This lack of foundation is why a number of intro and intermediate books are now putting
inflation on the vertical axis and calling the downward sloping curve they put on it
(downward slope explained with a Taylor rule reaction function) an AD curve.
As for the vertical AD curve, in the next edition of my principles book I have a box that
tries to make the point that in the range of price level fluctuation that we are talking
about, the AD curve will likely be highly inelastic. Moreover, since the price level only
responds slowly to changes in AD, the elasticity doesn't matter so much, since shifts in
the AD curve will be the more important element.
The problem I see with the entire exercise is that we are trying to put a discussion of
multidimensional dynamic disequilibrium onto a dto dimensiional comparative static
framework, and it doesn't work well.
As to why we do it, from my point of view it is less the publishers than the reviewers who
drive us there. These reviewers reflect the way the majority of economists think about the
issue, and want to teach it. Publishers will let an author put in anything you want in a
book, but if one reviewer after another is telling you this is the way it should be, and
is the way the other books have it, and complain strongly about any attempts to change it,
then most authors succumb. I've tried explaining the complications in my texts many
different ways, but I simply must be lousy at explaining them, because reviewers generally
pan my attempts at more sophisticated explanations.
Dave Colander
------------ FOOTER TO HES POSTING ------------
For information, send the message "info HES" to [log in to unmask]
|
