13 November 2003
Alan --
cc: Other SimSoc Mailing List participants
At 11:07 AM 11/13/2003 +0000, Alan Penn wrote:
>Can we broaden the question can anyone think of excellent examples of
>simulation providing explanation in any field (let alone sociology). It
>seems to me that simulations (at their best) are built on top of
>pre-existing explanatory theory, they may act as tests of those theories,
>but in a true Popperian sense, can do little to confirm only perhaps help
>falsify them.
With some considerable trepidation, I offer a possible example (not
necessarily an "excellent" example) where computational experiments helped
me to understand better the possible explanation for an important observed
phenomenon.
I was interested in exploring in labor markets what is known as the "excess
heterogeneity" problem --- why workers with observationally similar
structural characteristics are observed to have distinctly different wage
earnings (see, e.g., John Abowd et al., Econometrica, 1999 lead
article). The standard explanation is omitted structural factors that
cause differences in on-the-job productivity (so the answer is to search
more diligently for these omitted structural factors). However, Abowd et
al. report that the unexplained variation in observed wage earnings drops
considerably if, in addition to the usual structural regressors (gender,
schooling, etc.), wage earning regressions take into account the names of
workers ("person effects"). Unfortunately, Abowd et al. do not explain why
the simple introduction of "names" as regression variables should reduce
the unexplained portion of observed variation in wage earnings.
Given the "person effects" findings of Abowd et al., I wondered if the wage
earnings differentials they observed in their data could be resulting from
some type of unobservable work-site interaction effects (personal chemistry
between workers and employers if you will). So I set up a simple economy
in which workers (all with observationally equivalent structural
characteristics in the initial period) and employers (all with
observationally equivalent structural characteristics in the initial
period) repeatedly seek out preferred work-site partners (one period
contracts) and evolve their work-site strategies over time (shirk or not
shirk on the work-site) in response to work-site experiences. And I was
able to see that, indeed, a persistently non-degenerate distribution of
wage earnings across workers resulted, generated from two distinct sources
-- the evolution of persistently idiosyncratic work-site behaviors
(behavioral hysteresis), and the evolution of persistently asymmetrical
network relationships (network hysteresis). These two distinct sources of
hysteresis in wage earnings seem to provide a *possible* explanation for at
least *some* of the observed "excess heterogeneity" in wage earnings, an
explanation that I was only able to express in vague intuitive terms in
advance of the experiments.
If you are interested, these findings are reported at length in the
following article:
"Hysteresis in an Evolutionary Labor Market with Adaptive Search," pp.
189-210 in S.-H. Chen (ed.), Evolutionary Computation in Economics
and Finance, Physica-Verlag Heidelberg, New York, 2002 [ (ps
preprint,216K), (pdf preprint,534K)].
Abstract: This study undertakes a systematic experimental
investigation of hysteresis (path dependency) in an agent-based
computational labor market
framework. It is shown that capacity asymmetries between work
suppliers and employers can result in two distinct hysteresis effects,
network and behavioral,
when work suppliers and employers interact strategically and evolve
their worksite behaviors over time. These hysteresis effects result in
persistent
heterogeneity in earnings and employment histories across agents who
have no observable structural differences. At a more global level, these
hysteresis
effects are shown to result in a one-to-many mapping between
treatment factors and experimental outcomes. These hysteresis effects may
help to explain why
excess earnings heterogeneity is commonly observed in real-world
labor markets.
http://www.econ.iastate.edu/tesfatsi/evlab.pdf (pdf preprint, 534K)
http://www.econ.iastate.edu/tesfatsi/evlab.ps (ps preprint,216K)
Best wishes,
Leigh
----------------------------------------------------
Leigh Tesfatsion Department of Economics
Tel: (515) 294-0138 Iowa State University
FAX: (515) 294-0221 Ames, IA 50011-1070
[log in to unmask] http://www.econ.iastate.edu/tesfatsi/
|