Well, here we go again on AS-AD: The Theory that Would Not Die.
I won't reply to Barkley's comments on foreign substitution
except to say that I am not impressed by references to textbooks,
since I have seen them say dubious things in the past. This is
not to say I assume them to be wrong, just that I would need the
logic behind their claims as well as the claims themselves. In
any event, however, since the timing of the J-Curve is incompati
ble with that of the putative AD curve, the issue is decided on
other grounds.
Tom W. shifts ground by asking, not what the justification is for
a downward-sloping AD, but what would such a curve "explain" if
we allowed it to. A few reactions:
(1) As my earlier protests should make clear, I don't think the
AD curve "explains" anything, even if it predicts it. As I said
at the outset of this debate (a year ago), you can set up infla
tion/real GDP space, draw one upward-sloping curve, one downward-
sloping curve and reproduce any historical change as a movement
from one intersection to another. This is not the same as expla
nation, however. For that, you need stories about mechanisms.
(2) I don't think AS-AD is even potentially an appropriate frame
work for capturing supply shocks. A supply shock is a change in
one or more relative input prices. The reason the oil shock of
1973 was one of these was that the price of oil rose relative to
that of other commodities. One consequence was an inflationary
impetus, but even supposing a monetarist world in which manage
ment of the money supply could have kept the price level constant
(oil prices up, enough other prices down), we would still have
had a supply shock.
(3) I think Tom and I are *very* far apart on the question of
NAIRU, the "natural rate", etc. NAIRU in my view is simply the
level of unemployment which exactly offsets other pressures for
an acceleration of inflation. That such pressures seem to exist
in all mature capitalist economies appears to be a given, al-
though we may disagree as to why. Without a corresponding mea
sure of labor market slackness we tend to drift out from the
origin from one short run Phillips Curve to the next. Any time
the economy's "autonomous" inflationary pressures shifts, so does
NAIRU. (This is how I would explain David Gordon's results.) On
the other hand, I simply don't know what it means to say that
there is any long-run level of Y to which the economy tends in
the absence of active economic management. In each period there
is short-run demand for GDP in the usual Keynesian sense--what
else is there? And why should we expect NAIRU to be approximated
by market forces? Is there any evidence that market economies
tend this way without being manipulated?
Peter Dorman
