232 LAW AND FREEDOM IN HISTORY

A.D. 1924*—and by a Russian scholar, N. D. Kondratieff, who published
his own findings in A.D. 1926* Kondratieff articulates his long-wave
cycles as follows :3

Total

Trough Crest Trough duration

(i) circa 1790 1810-17 1844-51 50/60 years

(ii) 1844-51 1870-75 1890-96 40/50 years

(iii) 1890—96 1914—20

An attempt to"correlate'these forty/sixty-years 'Kondratieff cycles' with
the nine/ten-years'Juglar cycles' and the three-and-a-half-years 'Mitchell
cycles' had been made by J. A. Schumpeter.4 His suggestion was that
each 'Kondratieff cycle' was a clutch of six 'Juglar cycles', and each
'Juglar cycle' a clutch of three 'Mitchell cycles'.

'We . . . postulate that each Kondratieff should contain an integral
number of Juglars, and each Juglar an integral number of Kitchins.5 The
warrant for this is in the nature of the circumstances which give rise to
multiplicity. If waves of innovations of shorter span play around a wave
of a similar character but of longer span, the sequences of the phases of
the latter will so determine the conditions under which the former rise
and break as to make a higher unit out of them, even if the innovations
which create them are entirely independent of the innovations which carry
the longer wave. . . . For every time series the sweep of any cycle is the
trend of the cycles of next lower order. . . . The three deepest and longest
depressions of the Industrial Age—1825-30, 1873-8, 1939-34—were de-
pressions in the cycles of all three wave-lengths alike.. . . Barring very few
cases in which difficulties arise, it is possible to count off, historically as
well as statistically, six Juglars to a Kondratieff and three Kitchins to a
Juglar—[and this] not [just] as an average, but in every individual case.'6

Schumpeter's hypothesis had not, however, won the support of W. C.
Mitchell.

'No arrangement of our monthly measures in groups of three consecu-
tive cycles will produce an approximation to "Juglar cycles" of from nine
to ten years.7 . . . The evidence is better that business cycles vary sub-
stantially within periods of "Juglar cycles" than that they do so within the
long-cycle periods.8. . . The trough dates of the "Juglar cycles" corre-
spond roughly to the trough dates of severe business depressions.'9

But, in the estimation of W. C. Mitchell and his colleague A. F.
Burns, it remained still an open question whether "the periods separating
severe depressions are genuine cyclical units,'10 while the same two in-
vestigators' judgement on the 'Kondratieff cycles' was that the evi-
ience told, on balance, against their claim to be realities."

Review of Economic Statistics, November 1935.
3 As summarized in Mitchell, ibid., pp. 337-8.

Schumpeter, J. A.: 'The Analysis of Economic Change*, in The Review of Economic
istics, May 1935, p. 8, s Alias Mitchells.—A.J.T.

Schumpeter, J. A.: Business Cycles (New York 1939, McGraw-Hill, a voia.)» vol. x,
p. 172-4.

i Burns, A. F., and Mitchell, W. C.: Measuring Business Cycles (New York 1946,
Rational Bureau of Economic Research, Inc.), p. 443. » Ibid,, p. 444,

» Ibid., p. 448. ™ Ibid., p, 464. Cp. p. 460. » See ibid,, p,- 465.