Re: Paul D. on this year's Nobel

[mongiovg <mongiovg@xxxxxxxxxxx>]

In response to Paul:

>===== Original Message From paul davidson <pdavidson@xxxxxxx> =====
>At 11:18 AM 10/10/02 , you wrote:
>> >
>
>What these Nobel prize winners are claiming is that for all sorts of
>psychological or other reasons, in the short run, some economic agents may
>not make rational decisions--- although the information for rational
>decision exists and is often given by the investigator (even i it is in an
>obtuse form) to the students (laboratory rats) being observed.  [If the
>rational choice was not knowable to the investigator -- and hence
>potentially available to the students -- how would the investigator "know"
>the students that the choice made by the students was not the rational
choice?
>

There's another way to think about this work, as I said in my original post.
These results imply that the economist's conception of rationality doesn't
jibe with how people actually behave. Mainstream economists might take that to
mean that economic agents aren't rational. I take it to mean that the
economist's conception of rationality is not very useful, because  (i) I don't
think people are generally irrational; and (ii) I don't see how general
propositions about social behavior can be derived if human beings are presumed
to be irrational. It seems to me that economists might have to revise their
notion of rationality to encompass the behavior patterns exposed by Smith et
al.

This work moreover implies that propositions built on the supposition that
agents are rational in the economist's narrow conception of rationality are
defective.  I don't understand why Paul, a long-time critic of orthodoxy,
finds this result uninteresting or unimportant or irrelevant.

>And if rational choice is possible, then there are five dollar bills lying
>in the street, and those who make these choice -- even if they do not  know
>the choice was rational --but merely act a"as if" they knew -- will collect
>all those five dollar bills!
>

This is a non sequitur. Behavioral economics suggests that the context in
which a choice is made can influence the choice in subtle and complicated
ways. I doubt if even the most anti-ergodic Post Keynesian on the planet would
maintain that, as a general rule, agents will decline to retrieve a
five-dollar bill in plain view on a sidewalk.

>
>
>And if the system was not ergodic, then there can not exist any possible
>choice (that required the outcome to be at a future moment of time then the
>point where the decision is made) to be described as rational.  In that
>case, then all the investigators must argue is they do not know what a
>rational choice decision is, under the [nonergodic] circumstances.
>

Behavioral economics gives insight into how agents make choices in certain
contexts and circumstances.  Whether we call those choices rational or not
depends on how we define "rationality" but is irrelevant to the soundness of
the BE results. Maybe they're not sound in nonergodic circumstances; that's an
empirical question. I don't think we can just presume, as Paul does, that the
BE propositions about choices must be wrong in those circumstances.
>

>
>>   To the extent
>>that behavioral economists share an essentially neoclssical understanding of
>>market processes, their work no doubt contains trace elements of the kind of
>>reasoning to which Paul objects.  But those elements seem to me to be
>>incidental to, and distinct from, the behavioral insights, which stand alone
>>(and which may or may not be sound in themselves).
>
>God help us from this type of reasoning-- all it does is perpetuate the
>basic Arrow-Debreu general equilibrium model -- and then permit "special
>cases" by adding ad hoc assumptions. It never will get us to the General
>Theory discovered by Keynes -- where money contracts for organizing
>production and liquidity is the essence of the entrepreneurial system.


There are lots of things wrong with neoclassical orthodoxy.  Nothing I said
gives aid & comfort to Arrow-Debreu GE theory. My point was that Smith's
empirical work gives us another reason to distrust orthodox propositions.  I
don't think there's anything ad hoc about their assumptions or their results.
Though of course you are right to point out that we need to be careful about
extending experimental results to other contexts.  I did not say otherwise;
nor has anyone else on this list.

>Do you not notice that in all the Kahenmann and Vernon Smith experiments
>about markets -- there are no costs of production (as Mat Forstater pointed
>out) and there is never a liquidity problem in their simulation of
>financial markets!!!
>

That's part of what's interesting: despite all that, they get results that
call conventional economic reasoning into question.  If I may turn your
question back at you: "Don't you see that?"


Gary