Long Cycles in Strike Activity: an Empirical Investigation
| Published date | 01 March 1987 |
| Date | 01 March 1987 |
| DOI | http://doi.org/10.1111/j.1467-8543.1987.tb00703.x |
| Author | Ernesto Screpanti |
Long Cycles in Strike Activity: an
Em
pi
rica
I
I
nvest ig
at
ion
Ernesto Screpanti*
INTRODUCTION
The long period of economic and social disease that began in the early
70s
stands
in
striking contrast to the golden age of the
50s
and the
60s.
Many
economists are now struck with the observation that the same long run
succession of good and bad times has occurred more or less regularly four
times in the last two centuries: the Industrial Revolution
(1790-1820),
the
Restoration
(1820-1850),
the Age of Capital
(1850-1870),
the first Great
Depression
(1870-1895),
the Belle Epoque
(1895-1915),
the second Great
Depression
(1920-1936),
the Golden Age
(1950-1970),
the Leaden Years
(1970-?).
Such an impressive alternation of splendid and gloomy epochs
tallies quite well with the results of much economic research
on
the
ex-
istence of Kondratieff cycles. These are waves in the growth rates of in-
dustrial output, consisting of a long phase of rapid growth, lasting
20-30
years, followed by an equally long phase of stagnation.' A periodisation of
long cycles which would be accepted by the majority of students (see Bies-
haar and Kleinknecht,
1983)
is the following:
Cycle
1st Kondratieff
ca. 1790-1848
2nd Kondratieff
ca. 1848-1894
3rd Kondratieff
ca. 1894-1945
4th Kondratieff
ca. 1945-?
TABLE
1
Phases
of
Kondratieff Cycles
upturns
1789192
1844151
1892196
1939148
downturns
1810117
1866175
1913120
1967174
A problem that has attracted the interest of social researchers is whether
long cycles involve other aspects
of
life besides the economic one. In par-
*
University
of
Trento.
100
ticular, some attention has been paid to the evolution of class relationships:
is there a tendency of class conflict to move in phase with Kondratieff cycles?
Kondratieff (1979) himself observed that major social upheavals tended to
occur in the upswings
of
long cycles
in
the nineteenth century. After the
second World War, the same theme was taken up by Dunlop (1948),
Hobsbawm (1964) and Imbert (1959), who, in an effort
to
support their
theses with historical data, posed the problem correctly and elucidated the
empirical and theoretical difficulties implied in its study. More recently,
various authors have revived interest in the phenomenon (Bouvier, 1964,
Phelps Brown, 1975, Gattei, 1981, Cronin, 1979, 1979b, 1980, Gordon,
Edwards and Reich, 1982, Mandel, 1980, Screpanti, 1984).
No
one,
however, has succeeded in producing evidence capable
of
satisfying the
agnostic student. Nor has there been any clarification of the nature
of
the
relationship between Kondratieff cycles and the evolution
of
class conflict.
One reason for this is that the notion of ‘class struggle’, as well as
of
related concepts like ‘intensification
of
class struggle’, ‘social explosion’, and
‘fundamental unrest’, are difficult to handle in empirical research; for they
are qualitive concepts, such that any index meant to measure their
manifestations will always be somewhat arbitrary. Does, for example,
Dunlop’s use
of
union membership adequately measure ‘fundamental
unrest’? On the other hand, the unwillingness to use any quantitative index
at all exposes the analysis to the risk
of
producing unfalsifiable or ambiguous
results. No doubt a great part
of
the discordant results mentioned above,
stems from problems
of
this kind.
To
these must be added the problem
of
choosing the countries and the historical periods to be investigated; for, in
principle, it is possible that different theses hold true in different countries
and in different periods.
In
the present paper, the intensity
of
class struggle will be measured by
three dimensions
of
strike activity: frequency
of
strikes,
F;
thousands
of
strikers involved in strikes,
S;
and thousands
of
working-days lost in strike
activity,
L.*
Put simply, it is claimed here that any changes occurring in the
‘qualitative’ nature of class struggle should also manifest themselves in the
intensity of strike waves.
As
far as
I
know, the only systematic studies
of
the
relationship between strike activity and Kondratieff cycles are those
of
Cronin (1979, 1979b). The main reason for this scarcity of research is
undoubtedly that ‘reliable’ data on strike activity in major capitalist
countries have been gathered only since the 1880s; a time series which is not
longer than a hundred years is too short for investigating cycles that are not
shorter than half a century.
To
overcome this difficulty it has been necessary
here to lengthen the time series by making use
of
unofficial data. Thus, with
some series
of
frequency starting in the 1860’s it has been possible to
embrace an historical period containing three peaks
of
the long cycle.
Of
course, none of the longer series are internally as homogeneous as those
based on only official data; nor are unofficial data likely to
be
as reliable as
official ones. This could distort the analysis that follows, especially when
secular trend curves are fitted through the data, or when subperiods
of
a time
British Journal
of
Industrial Relations
To continue reading
Request your trial