[OPE-L:3268] Re: one-sector models

Duncan K. Fole (dkf2@columbia.edu)
Fri, 4 Oct 1996 08:14:58 -0700 (PDT)

[ show plain text ]

A few more words on 1-sector and multi-sector production models in response
to Jerry:

1. I agree that in order to address some important problems of political
economy (for example, those that have to do with sectoral adjustment,
relative price movements, and trade) we have to formulate them in a
multi-sector production model. But "realism" in and of itself doesn't solve
the problem of abstraction and aggregation: once we decide to go
multi-sector should we look at 2-sector models (which still violently
simplify the complexity of economic reality) or I/O models with 80 or
several hundred sectors (which still violently simplify the complexity of
economic reality) or go to the thousands or millions of sectors that might
be necessary to reflect the detailed differences in production over the
whole world economy? Simply posing this question makes it obvious that it
has no simple and universal answer. What we do is to attack particular
questions at the level of aggregation and disaggregation (or abstraction)
necessary to deal with the issue at hand. In considering many problems of
capital accumulation 1 sector suffices, and once you've reached
methodological clarity in the 1-sector context, it is straightforward
(though sometimes notationally tedious) to generalize to as many sectors as
you like. I'd like to see us discuss context more in these methodological

2. You might want to consider why, despite many decades of people pointing
out the realism of fully dynamic and fully disaggregated models, there has
been so little progress in developing them and getting useful results out
of them. I myself do not believe that this is simply a matter of getting
higher-powered mathematical technology going: the most difficult problems
are conceptual problems that one encounters in the very simplest models.


Duncan K. Foley
Department of Economics
Barnard College
New York, NY 10027
fax: (212)-854-8947
e-mail: dkf2@columbia.edu