Regarding the remarks of James Ahiakpor and Pat Gunning, it would appear
that they
have moved the goal posts of this discussion. The original issue was the
efficacy or lack
thereof of changing real output (and employment) by means of changing
aggregate demand.
We were told that because of Say's Law and its general acceptance by
everybody who
was not a completely ignorant moron, along with the general acceptance of
the natural
rate of unemployment, vertical aggregate supply curves, and (at least
long-run) vertical
Phillips Curves, even discussing aggregate demand was something that should
probably
be removed from all textbooks, to the extent that this particular purge has
not already occurred,
and certainly it should not be mentioned further in polite conversation in
seminars on economics.
Now, after the discussion of Hitler's Germany, the issue has suddenly
shifted to the
efficacy of fiscal policy in the absence of a reinforcing monetary policy.
Hmmmm. Not the
same thing, and looking to me like suddenly these gentlemen realize that
government policy-
induced changes in aggregate demand might actually have the potential to
increase real
output and reduce unemployment, although only if carried out at least partly
by monetary
policy, and maybe only in certain cases such as nasty dictatorships.
On this latter point, I think the issue is empirical, and it may well
be the case that "monetarism"
beats "Keynesianism" in the sense that monetary policy changes may have
stronger impacts on
aggregate demand than do "equivalent" fiscal policy changes. I would
certainly agree that a fiscal
policy is going to have trouble changing real output if monetary policy is
actively trying to offset it,
as for example seems to have happened when Ronald Reagan's initial round of
stimulative fiscal
policy ran into the hard monetarist policy of Paul Volcker, and the US had
the rather sharp and
deep recession of 1982. OTOH, monetary policy by itself may only bring
about an inflation and
not an increase in real output, as Mason Gaffney's comment regarding the
"Bonn 1923" hyperinflation
should remind us. I would certainly agree that if you want to stimulate
aggregate demand with
also some hope of increasing real output, it is best to start from a
situation where there is some
unemployed labor and some excess capacity in the capital stock, and to use
both fiscal and
monetary policy together, as Hitler did in the mid-1930s in Germany.
Regarding Gunning's citation of Mises, the latter is discussing the
full-blown wartime economy
of Germany. That is not what it looked like in the more dramatic 1933-36
period that Adelstein
emphasizes (although I noted further declines in unemployment after that and
before the war). It
happens to be the case that the sort of economy Germany had in 1939-45 did
not appear overnight
in 1933 upon Hitler's accession to power, but only did so over a period of
years. It was only in
1936, after the main reduction of unemployment had occurred, that full-blown
central planning
was introduced. It was only over a period of time that the bilateral trade
deals that came to
characterize the international relations of the German economy arose. That
early period of 1933-36
looks much more like a classic Keynesian expansion, with no price controls
(they were introduced
in 1936 also), few extraordinary trade deals, and with only part of it being
due to a military buildup
and conscription, although some of that was going on. Much of it was the
building of autobahnen
and the like, with all of this certainly be strongly assisted by an
expansionary monetary policy, as
James Ahiakpor has accurately pointed out.
For those who insist that this sort of model must not be considered
because it occurred under a
dicatatorship, well, let us consider the US during WW II. Now, I already
said that I avoided this
example because I did not want to get into the discussion of whether WW II
ended the Great
Depression in the US or not. I am perfectly willing to go along with the
claim that by and large
most of the Great Depression was over for the US by the time it entered the
war in December, 1941.
That said, I would note that during 1941-45, there was a large increase in
real output in the US, and
an accompanying large increase in employment, so large that it involved
entry for the first time in
significant numbers by women ("Rosie the Riveter") and minorities into both
the labor force and into
employment where they had never been before, or certainly not in any
significant numbers. This
increase was also due to an enormously stimulative fiscal policy that was
supported by an expansionary
monetary policy, with price controls and rationing thrown on top to restrain
the accompanying
inflationary pressures. But, last time I checked, the US remained a
democracy and not a
dicatatorship during WW II, even if it became a temporarily command
capitalist economy, like
Nazi Germany's.
I am not putting these examples forward to say that this is the sort
of economies that we
should have in peacetime. I am simply pointing out that indeed policies
that stimulate aggregate
demand most certainly can stimulate real output, and fiscal policy can very
much reinforce this,
even if it may be necessary to have an accompanying monetary policy to make
it really work
On that last point, does anybody want to seriously argue that we would
have seen the real
output increases in the US that we saw in WW II if we had had only the
stimulative monetary
policy that occurred without the accompanying fiscal policy? I think this
is a ludicrous idea.
Regarding how seriously entrenched and accepted the reality of
long-run, vertical Phillips
curves are, I would simply comment that anybody who heard George Akerolof's
presidential
address at the AEA meetings in Chicago a month ago, heard him state that in
his view long-run
Phillips curves are downward-sloping at low rates of inflation and
upward-sloping at high rates
of inflation, although this point is not made in the accompanying paper. In
any case, I would be
a bit cautious about going on too much about how universally accepted and
absolutely and
unequestionably true the idea of vertical long-run Phillips curves are.
They are only true in
certain kinds of models, such as those with rational expectations, whose
assumptions appear
not to be generally empirically true.
Barkley Rosser
|