[lbo-talk] Wealth Distribution & Kinetic Theo

andie nachgeborenen andie_nachgeborenen at yahoo.com
Wed Apr 25 10:40:16 PDT 2007


If I were to look for an explanation of this phenomenon, if it is real, I would not look for it in the behavior of gases or on any analogy thereto. I would rule out a priori the bare possibility that social systems behave on some analogy to physical system. I am a pragmatist. I have my can't-helps (non-revisable beliefs, which might be wrong, but you won't persuade me of that), but I don't rule things out a priori. However, bare possibility and $3.79 will buy you a cuppa cappuccino. If someone can make an analogy or "physical" explanation of social phenomena do work and also fit with what else we know, they done something interesting. I have not heard of such a theory and this clearly isn't one. I myself would look for an explanation of whatever wealth distribution there is and whatever statistical properties it has starting with the insights of Marx, Veblen, Keynes, Robinson and others who have provided useful and plausible models of the behavior of people who behave in accord with the constraint of applicable political and economic institutions.


> But I still think it is an interesting and
> non-trivial observation if the
> Gibbs distribution (which is not an obvious
> distribution in the first place)
> consistently models wealth distribution in diverse
> societies. The
> explanation for the "findings" provided by the
> authors could be ideological
> and yet the initial study could also be non-trivial.
>
>
> There isn't even a correlation;
> > there could not be, because an idealized gas is
> > neither a dependent nor an independent variable of
> a
> > wealth distribution. There is only, perhaps, the
> fact
> > that a wealth distribution exhibits certain
> > statistical properties also exhibited by ideal
> gases.
> > But statistical properties are similar in lots of
> > systems that have no relationship. Bell curves.
> > Bimodal distributions. Etc. You find them all over
> the
> > place.
>
>
>
> Nicely stated. And yes the Gibbs distribution is a
> statistical model, but
> not one I would expect to provide a model for wealth
> distribution.
>
>
>
>
> It is total nonsense to suggest that there is some
> > sort of social physics involved such that you are
> > explaining social phenomena in physical terms, or
> that
> > physical laws or models are somehow explanatorily
> > illuminating merely because you find, if you do,
> > common statistical properties in the idealized gas
> of
> > stat mech and in some econometric measure of
> wealth
> > distributions.
>
>
>
> Well, it is possible that complex social systems do
> act in ways similar to
> physical systems and that similar models can be
> applied to each. I wouldn't
> rule it out a priori.
>
> The abstract or quote from the research quote posted
> a
> > bit later makes this quite clear. In fact what
> these
> > guys have to offer is boring and conventional
> > bourgeois economic commonplaces about the rich
> being
> > richer because they are thriftier, which is false,
> > uninteresting, and has no connection whatsoever to
> any
> > part of physics.
> >
> > I'm actually open-minded about reductionism as a
> > program for psychological and social research. I
> wrote
> > a doctoral dissertation and a number of papers
> > defending the possibility before I got bored with
> the
> > subject. This is the sort of claptrap, however,
> that
> > gives reductionism a bad name. If there's nothing
> more
> > useful to be said about physical bases for
> phenomena
> > like the distribution of wealth than this, best to
> > give it up for a bad job. As far as reductionist
> > programs go, sociobiology is far more promising.
> (And
> > yes, I think there's something to sociobiology.)
> > Properly framed it is at least coherent, possible,
> and
> > potentially explanatory. This stuff is just dreck.
>
>
>
>
>
> --- Jerry Monaco <monacojerry at gmail.com> wrote:
> >
> > > On 4/24/07, andie nachgeborenen
> > > <andie_nachgeborenen at yahoo.com> wrote:
> > > >
> > > > The idea that income
> > > > distributions obey the laws of classical
> > > statistical
> > > > mechanics is ludicrous. In any event, there
> are
> > > far
> > > > more useful and comprehensible explanations in
> > > terms
> > > > of the behavior of capitalist markets and
> power
> > > > politics.
> > > >
> > > >
> > >
> > > Yes, well this is exactly my knee jerk reaction.
> > >
> > > But take the study on its own terms and assume
> for a
> > > moment that you
> > > view the evidence in the most favorable light
> > > against the party whom
> > > you just delivered summary judgment against
> assuming
> > > that all the
> > > facts as presented by the researcher are
> established
> > > and drawing all
> > > reasonable references from those facts.
> Starting
> > > from here the
> > > evidence is that wealth distribution can be
> found to
> > > obey the models
> > > of statistical mechanics and specifically the
> model
> > > of the Gibbs
> > > distribution at least in Japan, the U.S., the
> U.K.,
> > > India, and 19th
> > > century Europe. If this is so even for a few
> > > societies and for a
> > > limited amount of time, what does this say about
> the
> > > structure of
> > > human societies and/or power distribution within
> > > human societies?
> > > This is the limited question.
> > >
> > > It is possible that it is a coincidence that the
> > > Gibbs distribution
> > > and wealth distribution correlate, but even so
> it is
> > > a strange
> > > correlation even for a few societies.
> > >
> > > But a bigger _if_.... If it is so also true for
> > > societies that don't
> > > have market systems what would this say about
> the
> > > structure of complex
> > > societies in general? (Maybe you should leave
> out
> > > my "bigger" ifs....
> > > )
> > >
> > > Jerry
> > > ___________________________________
>
=== message truncated ===

__________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com



More information about the lbo-talk mailing list