Fri Dec 3 11:06:27 EST 1999
----------------- EH.RES POSTING -----------------
>Yes, this point seems to have been lost in the current discussion.
>If the lesson of QWERTY is simply that history matters -- a point
>painfully obvious to non-economists -- then it is hard to see what
>all the fuss is about.
To me, this statement is perfect evidence that economists do not really
grasp how "history matters." Reducing path dependence to "market failure"
makes it in fact less interesting, not more. But it also makes it less
threatening to economic orthodoxy. This, I suppose, is why we are
expending so much effort tapping out arguments over QWERTY as market
failure (and think how much we all would have saved already with a
different keyboard!). Shoe-horning path dependence into a familiar
framework dodges uncomfortable issues.
First, it is well to remember that path dependence is part of a broad,
growing literature that takes history seriously in social theory. Consult
the evolutionary theory of Nelson and Winter, the work in management of
technology, such as Abernathy and Utterback's "dominant design" paradigm,
the literature on organizational learning, on policy feedbacks, and some
(though not all) of the "New Institutionalism," introduced very well in
Smelser and Swedburg, The Handbook of Economic Sociology. This literature,
largely economic, is supported as well by work in the history of
technology, especially the social construction of technology paradigm.
If we put aside market failure for a moment, we see that this literature
collectively addresses a very broad range of economic phenomena. Indeed it
covers a good portion of modern technology, where feedback, lock-in, path,
and learning effects operate, where complementarities between parts and
where network externalities are strategically important. How big? The
answer can no longer be calculated as the social cost of market failure.
Big refers rather to how large an economic landscape does this literature
map. The answer is, we don't know, the domain is still growing. Consider
though that it would be impossible to understand the computer and
telecommunications industries without it. That's a pretty good start, I'd
say. You could add as well a portion of electricity and transportation.
The question of value then, is not are there "feasible government
interventions to mitigate this problem," the problem being "markets
converge only on local optima, not global optima." Theory is not
fundamentally about policy; theory is about gaining greater insight into
phenomena, from which better policy perhaps can be made. Business
strategist, for example, seem to find it very useful to know about
feedback, lock-in and network effects. Having a theory that can make sense
of otherwise inexplicable or seemingly random events is useful, without
theory being reduce to a policy recipe book. This would hold, of course,
for all economic theory, as theory.
The more radical, and frightening implications of path dependence, et al.
are in the theoretical realm, moreover, and that is why so much discussion
is about trying to reduce it to a simple policy question of market failure.
If correct, then path dependence is calling our attention to a limit of
markets (not perhaps in the strict, Pareto sense a failure) in assuring the
best technology ever gets invented, or sees the light of day. At each
point in the path, the market may well make the best choice among
alternatives, but the results may still be less then the best. If for
example, technology A at time 1 is cheaper than technology B, it may so
overshadow B, that B withers on the vine before we get to time 2. The
point of path dependence is, the way back (to B) is difficult, if not
impossible, not because markets fail but because in a market sitatuion no
rational actor would ever want to pay the costs of backtracking. But had B
been given a longer time to develop, it would prove superior to A. Besides
price, one could imagine other reasons why A overshadows B early on, even
though B would in the end be the better choice. These causes are discussed
in the literature.
I say this is threatening to economic orthodoxy for the following reason.
It it calls into question the boundaries of the discipline, which have
sectioned off these tricky inter-temporal issues, and slighted the
inextricable connections between efficiency and the messy social,
institutional, and cultural world. Effort to expand economic theory into
messy reality is fine (though it might cost you your job at the World
Bank); so long as those efforts are true to the phenomena being studied.
Unfortunately, most of the world exists in the stream of history. And in
fact, so does the premier economic institution, the market. Once history
gets in the back door, you see, it threatens the Panglosian market
metaphor; no longer can getting the market right assure us of living in the
best of all possible worlds. Or as I suggested in the "pessimist" view of
things, unfortunately we may very well live in the best of all possible
worlds.
How big a loss are we talking about here? I don't have an answer. But
research seems to show that the various historical effects operate over a
broad range of phenomena. So many, many little losses may well be adding
up to a significant decrease in welfare. Though again, this is not really
an allocative issue.
The questions of measurement and correction rest with the limits of
counterfactual history. Since B never got much of a chance, we can't
really compare it with A in real world operation. That's where good
history comes in. Oddly enough, Liebowitz and Margolis can be accused of
being economists doing *good history.* Not perhaps correct history, but
good, in that they try to see around what did happen in order to
investigate other possibilities, ones not written directly into the record.
I say this is odd, because they are also trying to deny history matters.
A good, (post?)modern narrative should work to reveal lost opportunities,
missed paths, alternatives, and the like. Of course, we can never know for
sure, but we at least free ourselves from the tyranny of "what is," to
imagine alternatives. This is done not just for the sake of the past, but
to open ourselves up to the future as well. A pessimist does not have to
be a nihilist, after all. In the real world, decisions are not so
mechanistic as in the stilted trope of the omniscient planner making things
better. Of course, such as straw man is easy to dismiss. In the real
world, though policy makers, and also business decisions makers, try to do
things a little better with the best knowledge they have. History helps
here. What more do you want? KL
*********************************************
Kenneth Lipartito Tel 305-348-1860
Department of History Fax 305-348-3561
Florida International University lipark at fiu.edu
University Park
Miami, FL 33199
------------ FOOTER TO EH.RES POSTING ------------
For information, send the message "info EH.RES" to lists at eh.net.