Comment from Arno Tausch (for God's sake, please forget the Embassy heading
at the top - these are my personal scientific comments)
I've read Gernots paper with great interest. I'd like to comment, though,
on some of the points raised by him. Although the paper is brilliant, we
should - please pardon me - not try to invent the wheel again, and put the
stuff in the right and proper framework.
In preparing my "Globalization and European Integration" (which by the way
I could present on last Friday here in Warsaw at a magnificent Polish
Academy of Sciences meeting, chaired by our friends and wsn subscribers
Prof. Jadwiga Staniszkis and Jery Gierus) I've come across very valuable
literature on the discrepancy between GNP per cap and PPP's, in the
framework of what most economists simply call ERDI (exchange rate deviation
index). The Vienna Institute for International Economic Comparisons has
done some magnificent research on that (if I had only their website address
right now at hand!; but at any rate YAHOO or other good search engines will
bring you to their Website), and you could advance the point in the
following empirical fashion:
1) calculating the regression between PPP per cap (x-axis) and GNP (y-axis)
2) subtracting real, observed GNPs from the estimated GNPs, i.e. taking the
residuals
This will be a powerful indicator for an artificially high or low exchange
rate.
Peripheries, in general (and hence Gernot's correct point on Mozambique)
will be tending towards low, very negative residuals, while most of the
centers will be showing positive residuals. For that reason, the hypothesis
is justified, that ERDI could be positively related to economic growth or
social development, keeping the Matthew's effect or the basic human needs
Plateau curves constant.
However: we must be careful. This might be a very spurious relationship,
that expresses nothing else than the fact that Mozambique is a periphery
and, say, the US a center. Does the relationship still hold, when we
include the new UNCTAD MNC Penetration data for over 150 countries (UNCTAD
World Investment Report), from 1980 to 1985, 1990, and beyond, into our
multiple regressions, predicting economic or social development or decay
since the 1980s? Does it still hold under inclusion of terms of trade data,
social expenditure data (to control for the influence of "Keynesianism" on
growth or development)?
That, Gernot, is the crucial empirical question, and I
hope you or your doctoral students will jump to their PCs and their EXCELs
or
SPSSs and use the UNDP or other data for post 1980s development to arrive
at the
empirical verdict in this debate.
Money laundering etc. is now a powerful force that inflates the GNPs of
semiperipheries dangerously upward from time to time. What Ruediger
Dornbusch and others have written in this context on the tequila or
caipirinha effect is valuable scholary knowledge, regardless of the fact,
folks, and don't massacre me, that neoclassics have written it. We also
should recall Michael Lipton's wonderful essay on the urban bias in world
development, published back in 1977. The urban elites and property owners
(and today, the money launderers) will always have an interest in high
exchange rate, high
interest rate, while the export industries, and rural society will be
benefiting in the long run from a low ERDI.
That is also one of the main reasons, why I am personally afraid that the
"Asian virus" will hit, with full force, the Italian Lira this spring or
summer, perhaps also some countries in East Central Europe, to roam on in
1999 to hit the superstar, the $, thus finalizing the Arrighean
rock the boat cycle of financial transfers to - yes ASIA, which will come
out from the mire in an Andre Gunder Frankian fashion by, say, 1999 or 2000
in the end.
The future experience of the EURO will also teach us, Gernot, what is good
for growth: a low or a high ERDI. Germany's present ills, to my view, are
to a large part caused by the too high ERDI during the late 1980s and
the 1990s; which ruins the export performance, threatens the current
account balance, and causes long-run unemployment.
So it might very well be, that for semiperipheries 1/ERDI will be
positively, and ERDI negatively related to long-run "development", again
keeping Matthews effects or Plateau curves constant.
This makes the case for true multivariate research on the determinants of
growth and development and or decay in the 1980s and 1990s all the more
important, at a time, when we have entered, long ago, around 1980, a new
Kondratieff-cycle set up of "flexible specialization", that puts into
question our research results about development from the 1960s to the late
1970s, which were valid for that waning Kondratieff cycle, but not
necessarily any more for the new logic of the post-fordist/corporatist
world.
Kind regards and happy further research, which indeed is needed
Yours
Arno Tausch