Re: Bob Solow on the post keynesians

[Paul Davidson <PB108928%UTKVM1.BITNET@xxxxxxxxxxxxxxxxx>]

FROM:  Paul Davidson
"      Economics Department
"      523 Stokely Management Center     974-4221
Dear Ric and others on the PK network. Let me apologize that my response to
Ric regarding Bob Solow's comment got garbled --- at an especially critical
point. Obviously, my laptop has still got something wrong with it when I go
into e-mail.
    The point that I was trying to make was that Bob Solow disparaged my 1983
argument that Keynes's revolution required rejecting the ergodic axiom of
classical economics (and , as my comments on Samuelson was meant to suggest,

the ergodic axiom is a foundation of 19th century classical physics as well).
But by 1985, Solow (AER Papers and Proceedings, May, p.328) wrote: " Economics
is a social science.... To express the point more formally, much of what we
observe cannot be treated as the realization of a stationary stochastic
processwithout straining credulity". Since nonstationarity is a suggifcient
condition for nonergodicity (but not a necessary condition), two years after
Solow disparaged my claim that macroeconomics had to deal with nonergodic
systems, he was claiming that economics was dealing with nonstationary and
therefore nonergodic systems. Moreover, on p.330, Solow complains that "the
best and the brightest in the profession proceed AS IF economics is the
[classical] physics of society". Solow's analogy of dropping a modern
economist from a time machine and his belief that independent of time or place
his economic model will "be applied", means that Solow recognizes that the
profession assumes ergodic systems -- and in Solow's view, the profession has
been mislead by the "best and the brightest", and according to Solow (p. 331)
"we need a different approach". Thus, Ric, Solow has already provided an
eloquent defense of the Post Keynesian methodology of rejection of the ergodic
axiom.
      My reference to the Moscow Mathematical school of probability was to
indicate that Keynes did not specifically use the term ergodic vs. nonergodic
because the theory of ergodic processes was just being developed in the Soviet
Union when Keynes was writing. Hence, he probably was , as most english
speaking economists and mathematicans of that day, unaware of the concept of
ergodicity. Nevertheless, Keynes's famous criticism of Tinbergen's League of
Nations econometric study was that Tinbergen assumed that the data "was
homogenous" with resppect to time, i.e, Tinbergen assumed stationarity. Keynes
insisted that one could not assume away that notion that most of the data that
Tinbergen was using (time series data of aggregates) was non-homogeneous. Thus
Keynes was arguing that macroeconomics must deal with nonstationary systems,
and that is a sufficient condition to assure that we are dealing with
nonergodic systems. (Mathematically it is possible for a stationary system to
be nonergodic-- e.g., limit cycles [yes Barkley stationary nonlinear systems
may be nonergodic]. An example of a limit cycle in the economic literature is
Hicks trade cycle model with ceilings and floors.
      One of "the best and the brightest" that Solow was refewrring to may
have been Smuelson who insisted in 196os that the "ergodic hypothesis" was
what made economics a "science" and "removed it from history". Does this
suggest that Solow was pointing his finger at Samuelson for shunting economics
down the wrong track?
       Accordingly, Ric, hasn't Solow providedus with the basis for a coherent
positive alternative theory to the one used by mainstream theorists. In
Solow'sone words, haven't we found the Achille's heel to orthodox economics?
Can't we get all the heterodox people to rally around Keynes's alternative (as
elaborated by some PKers) where nonergodicity is made the keystone of the
theoretical analysis???? My problem is understanding why the anti-mainstream
people refuse to become organized around a viable alternative to mainstream --
an alternative that even one of the major mainstream players has recognized
has at least faced the fact that economics is, in Solow's own words, "a social
science" once it is admitted that nonstationarity (and therefore
nonergodicity) is essential for a social science.
    Yet Jim Devine feelsthat mainstream economists can "add something to an
economist's kit of tools". Without being "patronizing" Jim, I find it hard to
understand how a useful theoretical tool can be forged on a logically
inapplicable axiomatic foundation. But then I try to maintain a logical
consistency when I try to teach my students economic theory and economic
reality. Of course, some may reply that "consistency is the mark of a small
mind".
       Herb Gintis thinks that the nonergodicity critique is "correct", but
there is nothing non-mainstream about this. Herb, if nonergodicity is correct
for investment theory then it is impossible to use any existing market data to
produce actuarial probabilities regarding expected future profit rates. Thus
the mainstream theory of risk and uncertainty is not the same. If you read my
JOURNAL OF ECONOMIC PERSPECTIVES (Winter 1991) paper, you will see where I use
Savage's own development of Expected Utility Theory to show that Savage
recognized that EUT is, as Savage wrote "is practical in suitably limited
domains" and that actual "behavior of people is often at variance with the
theory". In other words, EUT cannot describe actual behavior when decisions
are made in nonergodic circumstances.
     What permits mainstreamers like Solow, as well as heterdox economists
like Devine and Gintis, to think that Post Keynesianism is merely a
sociological club, is that we cannot organize around some single fundamental
principle. (Will Rogers once said he is not a member of any orgainbed
polictical party - he is a Democrat. I often feel the same way.) If we remain
unorganized, then it is easy for the Solowites in the mainstream, and the
Gintis-Devine people in the heterdox camp, to dismiss Post Keynesianism as
"silly" or "misguided" or as Gintis puts it not "very different from
mainstream in any significant way". But even Solow recognized if one does not
assume ergodicity one is working with a significantly different --- and are
your reading Jim -- and empirically applicable system.
     When Gintis says there hasn't been significant new insights in pk in
"recent years", I find that shocking. Has he been reading the minisymposiums
in the JPKE on HYSTERESIS, on REFORMING THE WORLD'S PAYMENT MECHANISM, on THE

Have a good day!