Re: Colander and New Keynesian Economics

[FAC_BROSSER@xxxxxxxxxxxxxxxx]

Yet another regoinder to Paul Davidson:
     This one is on Grandmont-Malgrange and Azariadis, etc.
     In the quote you cited from G-M I note the crucial word
"if" regarding the prospects of reaching the long-run e. (when
"we are all dead").  Grandmont and company are extremely skeptical
about such adjustment processes.  In his 85 piece he has
adjustments going to limit cycles.  Benassy and Blad (1989,
Journal of Economic Dynamics and Control) show that to get convergence
to RE you must have it to start with.  Evans (1985, QJE) has
shown that many "learning processes" can go chaotic.  If you can
never get to the l-r e. where full employment holds and uncertainty
has disappeared, then one must focus on dealing with the short-run.
That is the policy argument for "Keynesian" interventions by
governments made by Grandmont and cohorts, if I am not misrepresenting
them.
     "Sufficient contingent markets" have an infinite time horizon
in Azariadis' model.  His result can be used to bolster the
difficulties given how far we are from having such in the Real World.
     I note that ergodicity vs, non-ergodicity is ultimately an
empirical issue, not one whose truth or falsity is to be determined
strictly on logical grounds, although the implications of such models
can be worked out logically (sometimes).  Gintis' point about what
kind of non-ergodicity is well-taken.  Increasingly many economists
recognize that many processes are non-ergodic (even some Herb may not
be willing to grant), the issue may increasingly become, which ones and
in what ways?
     I completely agree that full employment may be possible in
barter economies, even with uncertainty, when it might not be in a
monetary economy.
     Have a nice, even though non-ergodic, day!
Barkley Rosser
James Madison University