end.gr.--increasing returns

["Abu Rizvi" <srizvi@xxxxxxxxxxxxx>]

Heinz and Neri,

Enjoyed the insights in your paper very much.  Would it be possible
for you to expand on the network on the issue of increasing returns
in these models?  Your points on this in the paper are well taken but
they deserve repeating and emphasis since it seems a number of
people do not have access to the paper.

Specifically, claims are often made for these models, as well as the
related industrial organization and international trade models, that
they now usefully incorporate increasing returns (even if, as Barkley
points out, they do not address path-dependence).  Going along with
your 'nothing new under the sun' theme, Sraffa and others pointed out
in the 20's that increasing returns which are external to the firm
but internal to industry *are* compatible with the usual competitive
set-up, but they are precisely the sort of increasing returns which
one is hard-pressed to identify in any actual processes.  In other
words, if the increasing returns process is not the property of
any firm or subset of firms, those firms do not outgrow the others
and ruin the assumption of perfect competition.  Thus this sort of
increasing returns is convenient though artificial.

Do the NGT models rely on this sort of increasing returns?  If they
do, does this help explain why they do not address R&D and related
tech. change issues, which are generally an internal-to-the-firm
process?  Why they have to rely on "knowledge" and "human capital"
because these are more plausibly not the property of any single firm?
And why symmetry assumptions among firms, or the use of
representative firms, are so relied upon in the models, since using them
makes sure that no one firm can get ahead of the others?