The discussion now starts to become tricky. As
new names are added, the problem of understanding
what exactly these people were saying, and what
therefore was their role in the formulation of
the principle, becomes more and more difficult.
Tentatively I would say that they were all trying
to explain the same observed fact, that is, that
the fluctuations of investment are far larger
than fluctuations of consumption. But premises,
approaches and solutions were not exactly the
same. One nowadays, with the benefit of
historical hindsight, thinks of the acceleration
principle roughly as follows: starting from the
premise that, given the technique and the degree
of utilization of capital, the ratio of capital
to consumption is constant, net investment is
undertaken only to face an increased demand for
consumption goods. (I hope this answers Pat's request).
But if we look at what these people wrote, we
note that: Bickerdike was focusing not on capital
but on durable goods —two overlapping but not
identical categories. Bouniatian clearly laid out
the premise, but stopped short of concluding that
investment depends on the rate of change of
demand: like Hobson & Mummery, he was content
with claiming that net investment stops when
consumption stops increasing. Clark was instead
adamant on this, and indeed suggested the word
'acceleration' for the 'derived demand' having in
mind the mechanical analogy. Carver reasoned in
terms of values rather than quantity, which Clark
assumed out at the outset. Harrod changed his
mind on the meaning of the whole thing (but
claiming continuity) between 1936 and 1939, so
that we have two different versions in his book
and in his article. Mummery and Hobson reach the
conclusion that the increase of consumption
involves a far larger investment, but do not lay
out the premise in terms of capital/consumption
ratio, and do not express it in terms of rate of
change. As well as Carver and Bouniatian, but
not as clearly, they explained that the further
back one climbs in the ladder of more
'roundabout' production (Bouniatian explicitly
referred to the austrian tradition), the more the
problem is amplified, while Clark assumed that
away by imagining a one-good economy. The
conditions (at the very minimum, a constant
technique and degree of utilization of capital:
see Manne RES 1945 for more) were not always
expounded, and some of our writers were
criticised for that (Harrod by Keynes and
Bouniatian by Hawtrey, for instance). The nature
of the relation was also conceived differently.
Clark and Harrod treated it as a simple piece of
arithmetics (Harrod called it "the Relation"),
while other writers did not think of it in
arithmetical terms at all (Hobson's figures are
purely illustrative). The implication also are
very different: Aftalion and (less clearly)
Bouniatian relied on the lag of response of
investment with respect to consumption to explain
the cycle, while Harrod vehemently denied that
the lag has anything to do with it; and whether
and how the acceleration principle supplies an
investment function is another debatable point.
And, of course, the critics had a number of
points, in particular Frisch (see the debate with
Clark in 1931) and Tinbergen (Economics 1938),
which apply to some but no to all the writers mentioned here.
The list, of course, is far from complete. A
history of the acceleration principle, then, is
still to be written. Accordingly, also an agreed
definition is still wanting. Perhaps one would
need to inquire beginning from the history of the
premise, as the suggestion that the ratio of
capital to consumption has something to do with
panics was around before Hobson (for instance
Hayley 1879, although he sometimes writes
'capital' and other times 'investment': his
writings are interesting in this connection
whatever the case, because either he anticipates
the accelerator or denies it ante litteram).
I hope someone will inquire the matter in depth.
Also because other names would deserve a mention --Pigou's, for example.
Daniele Besomi
|
