Re: Colander and New Keynesian Economics

[FAC_BROSSER@xxxxxxxxxxxxxxxx]

Response to Paul Davidson on New Keynesian Economics:
     This may be the butterfly responding to the elephant, but
what the heck.  There are as Paul well understands, at least
two varieties of New Keynesian economics.  The most prominent
is the least interesting.  This is the Akerlof-Stiglitz variety
that depends on asymmetric or imperfect information causing
menu costs and similar phenomena.  The upshot is a standard story
about wage rigidities (probably accurate in the short-run in the
real world) which gives at least a short-run upward slope to a
garden variety AS curve and allows for unemployment in response
to a decline in AD.  Paul's dismissal of this is reasonable.
     The other is the Colander-coordination failure version,
drawing on the work of Harry Garretsen and several other people.
         Such models can be generated using rational expectations
with nonlinearities in supply or demand curves, thus presumably
drawing forth Paul's scorn that these are merely stinky gingko
models rather than roses ("true Keynesian models"?).  I agree
with Paul that reality is nonergodic (we've had some go-arounds
on just exactly what that means, but let's put that aside for the
moment) and that rational expectations is therefore a lot of hooey.
Nevertheless I would contend that the fact that models incorportating
these assumptions rather easily generate multiple _long-run_
equilibria (yes, Paul, not only is reality nonergodic, but it is
nonlinear also!  The butterfly hovers about the elephant) represents
a profound undermining of the New Classical myopia and opens the
door for responsible governments to intervene in the economy in
order to influence which equilibrium will result.
     I think this is in the spirit of Keynes, if not directly in
the General Theory.  I also think the coordination failure view
is very Keynesian at its base.  Keynes's argument for the "socialization
of investment" fundamentally depended on the inablility of investors
(meaning real capital investment here) to coordinate with each other
lies at the heart of their inability to rationally forecast and their
unpleasant tendency to bunch investments at the same times thereby
generating macroeconomic fluctuations.  Much of the "coordination
failure" literature is connected to the "sunspot eqpuilibrium"
literature which emphasizes the concept of self-fulfilling prophecies
as determining real outcomes.  These can be driven by the autonomous
("extrinsic" to use the jargon of the literature) fluctuations of
good old fashioned Keynesian "animal spirits."  Paul, is this not the
basis of the Keynesian long-run unemployment condition, the collapse
of animal spirits by investors wandering in a fundamentally uncertain
universe?
     I note that some observers (David Romer, JEPS, Winter 1993) do
not accept the coordination failure approach as being "New Keynesian."
He categorizes schools of thought according to their attitudes to the
Classical dichotomy and Walrasianism.  According to him, "Traditional
Keynesian," Monetarist, and "Lucas Imperfect information" schools
reject the classical dichotomy, but are Walrasian.  New Classicals
accept both, New Keynesians (presumably Akerlof-Stiglitz approach)
reject both, and "Coordination failure" school accepts classical
dichotomy but in non-Walrasian.  I do not know what I think of this.
     I also note that in my "Chaos Theory and the New Keynesian
Economics" (Manchester School, Sept. 1990) I labeled the Akerlof-
Stiglitz approach "weak New Keynesianism" and what is being now
labeled "coordination failure" "Strong New Keynesianism."  I also
note that the most sophisticated versions of it have been developed
by the French, notably Jean-Michel Grandmont in his "Endogenous
Competitive Business Cycles," Econometrica, 1985.
Barkley Rosser
James Madison University
PS:  Have several good days!