Michael P, responding to Jim D. writes:
Verdoorn was a Dutch economist who got the idea from Tinbergen's data.
Kaldor picked it up & then said that Arrow called it Verdoorn's Law
when writing about learning by doing. Arrow did not call it a law.
The literature on the law is mixed although it worked pretty well until
1970. The story is mixed up with lots of other factors, including
increasing returns.
[For those who don't follow these issues, Verdoorn's Law as Kaldor
originally stated: "typically" a 1% increase in the growth of output
produces not just a proportionate increase in productivity and in
employment - but rather a 0.5% increase in both the *rate of growth* of
productivity and of employment. This implies that growth can be a
cumulative process because of large economies of scale.]
1) Verdoorn worked under Kaldor in the late 1940's in Geneva at the UN
Economic Commission for Europe (where British diplomatic spying cost Kaldor
his post at LSE). It was here that Kaldor first developed his work on
productivity (which blossomed in the '60s & '70s) and this may be why he
felt obliged, years later, to give precedence to Verdoorn (whose article
was published in 1949 in Italian and never published any other language and
focuses on pre-war growth in Eastern Europe). Actually earlier work had
been done by others at the NBER (S. Fabricant) during the WWII.
BUT, for what I think "matters" - the break with neoclassicism - the two
economists really did different things. Verdoorn focused on "static"
economies of scale (e.g. choosing from existing technologies) that applied
to individual subsectors and with a Cobb-Douglas homogeneous production
function. Verdoorn's approach does not exclude the tradition of the Solow
Residual, Arrow, Paul Romer, Robert Lucas, Barro, Robelo, etc and often
lends itself to arguments for "market mechanisms".
In contrast, naturally, for Kaldor the whole point was the "dynamic"
economies of scale - where new possibilities are opened ("induced"
technical progress, as well as the usual external economies and "learning
by doing"). For Kaldor (unlike Verdoorn) this was only possible in
manufacturing - so industrialization was the engine of
growth. Furthermore, Kaldor's dynamic view (unlike Verdoorn) requires
transformation (not just industrialization) of an entire economy and so it
strongly implies a wider framework of goals and government planning (hence
Kaldor's role in the ECLA school and progressive development policy of the
time).
I think the link with Verdoorn helped undercut and confuse things (see for
example the Rowthorn-Thirwall exchange in the Economic Journal in '79-'80)
helping to open the door to the neoclassical policy blends that prevailed
as "development economics" in those decades. This, in turn, helped create
the climate for the problems that Latin America experienced (and that some
of East Asia sidestepped, for now). I find that, today, some PKers speak
of Kaldor's Second Law of Growth rather than Verdoorn's Law.
2) Kaldor's boss at the UN had been Myrdal and to me that is an
important link. By the 1950s Myrdal was publishing his work on "cumulative
causation" (which I gather draws heavily on Veblen). Myrdal had already
somewhat applied these ideas in his 1944 work on race in the US ("An
American Dilemma") and black reactions to poverty. In the 1950s Myrdal
lays the groundwork for Center-Periphery models of growing inequality to
exist alongside (or instead of) equilibrium models where inequality declines.
3) As Michael P. points out, rightly or wrongly people started taking
shots at the econometric analysis that Kaldor and his supporters had relied
on so much.
The supporters continue to maintain that they continue to have current
econometric studies that show Verdoorn's Law holds, even in China. But
they have not been as well equipped to respond as they might have been. In
the Post War years the Cambridge crowd were not known for their
broadmindedness towards each others ideas - never mind those outside of the
Cambridge School. Indeed, this contributed greatly to the premature loss
of the Cambridge Department itself to neoclassicals (a point I owe to Geoff
Harcourt, who is in a position to know). Each participant and sub-school
were reluctant to see their specific insights synthesized into a broader
contexts. This inflexibility, in turn, left them vulnerable when material
conditions appeared to "disprove" or at least discredit a specific insight
-- they could not fall back on a broader body of work and of support.
So, IMO, much of the work on Verdoorn's Law ignores Kalecki or
Robinson-like work on distributional issues and growth (even though Kaldor
wrote much about distribution). Often this leads Verdoorn-like analysts to
an emphasis on export of manufactures (and the "Balance of Payments
Constraint) as an engine of growth, at the expense of possibilities for
income redistribution to create an internal market. Likewise, broader
ideas like Goodwin Growth Cycles or Ed Nell's Transformational Growth
become excluded. (These are points Lance Taylor has made in the JEL.)
More broadly, Verdoorn-like analysts (and much of the Demand-Led
Growth School) often have an exclusionary approach to long period analysis
of issues like profit rates. This is despite Kaldor's close friendship
with Sraffa. The Italian Sraffian Antonella Palumbo has spoken to this
point in Contributions to Political Economy. The late David Gordon once
wrote a paper "Kaldor's Macro System: Too Much Cumulation, Too Few
Contradictions". David re-wrote a Kaldorian econometric model to include
profit rates and "SSA-like" factors. Not much like that has been written.
Paul
