RE: Re: experimental economics

["Forstater, Mathew" <ForstaterM@xxxxxxxx>]

first, despite the problems with exper. stuff in general, the distinction
between human and rat experiments is important.  Vernon Smith, I believe, has
never done rat experiments and does not endorse them.  Kagel, et al. is the main
source for rats.

as far as Smith goes, one of the interesting things is his Marshallian stance.
He has to defend Marshall and partial analysis for his results to matter. So
right off the bat every problem with partial equilibrium analysis applies to
this stuff.  Ceteris MUST be paribus, because there is no ceteris in the context
of an economics experiment.

I actually got to meet Smith in a small group setting when I was at Gettybsurg
and he came to give a lecture there. I noted to him that in Marshall's
framework, price is equal to cost of production in long run equilibrium.  But in
Smith's world, the "equilibrium" is purely arbitrary in that respect.  I thought
it was a pretty basic question and was surprised when he appeared as though he
never had thought about it, no one had ever mentioned it before, and he had no
real response.

The next thing is that most of these experiments have rules that participants
MUST follow. This has a couple of implications. First, this limits the relevance
of any results to cases where this holds (where people also must follow these
rules).  Second, it does open up the idea that economic "laws" like the laws of
supply and demand might be looked at not as descriptive laws but *prescriptive*
ones.  Of course, this flies against the homo economicus vision of
maximizing/minimizing as part of human nature, as opposed to products of
socialization.

Then there is the issue of structure.  Like the example of rules, the
experimental world has a certain structure and so the likelihood of results of
the experiment being observed in reality depends on the existence and
persistence of a similar structure.  In this case, the experiments can serve
another purpose, like Hahn has argued general equilibrium theory serves: showing
how UNlikely it is that these results would ever obtain in reality, since the
conditions under which the results are obtained are so unrealistic.

Lots of the experiments are auctions.  Auctions are not how most everyday
markets are set up (although I suppose there are people out there claiming that
with the rise of the internet and e-bay type markets, experimental econ is
getting a boost).

Mat




-----Original Message-----
From: Jim Devine [mailto:jdevine@xxxxxxxxxxxxxxx]
Sent: Monday, October 02, 2000 1:37 PM
To: pen-l@xxxxxxxxxxxxxxxxxxx
>Subject: [PEN-L:2584] Re: experimental economics


At 09:53 AM 10/2/00 -0700, you wrote:
>Last night, I noted in the Journal of Economic Perspectives that one of
>the leading experimental economists claimed that their work verified
>Hayek's (my voice recognition read this as: high acts) theory of
>spontaneous order.  They also take credit for developing the spectrum
>auctions, which I suspect are being sold off at less than what they're
>worth, despite the auction theorists idea of the winner's curse.

the guy in the office next door to mine does experimental economics. He
interprets his results as saying, for example, that all people have pretty
much the same sense of what it means to be "fair" and are willing to be put
it into practice. (It's along the line of "to each according to his or her
effort.") Matt Rabin's survey of the psych literature in a recent JOURNAL
OF ECONOMIC LITERATURE also suggests that the standard economics model of
people -- i.e., homo economicus -- is radically wrong.

Jim Devine jdevine@xxxxxxx &  http://bellarmine.lmu.edu/~jdevine