Trade Union Membership Concentration, 1892‐1987: Development and Causation
| Date | 01 September 1993 |
| Published date | 01 September 1993 |
| DOI | http://doi.org/10.1111/j.1467-8543.1993.tb00406.x |
| Author | Jeremy Waddington |
British
Journal
of
Industrial Relations
31:3
Sept
1993
0007-1080
Trade Union Membership
Concentration,
1892-1
987:
Development and Causation
Jerem
y
Wadding ton
Abstract
Existing studies
of
British trade union membership concentration have used
the variance of
logs
and concentration ratio measures to show
a
tendency
to wards increasing concentration. This paper proposes the Herjindahl index
as
a
more appropriate measure. The annual variation in concentration is
analysed by means
of
the Herfindahl index and shows some fluctuation,
particularly prior to
1947.
Divergent trends are also shown for manual and
white-collar membership concentration. An explanation of the aggregate
trend and of these different occupational trends is advanced in terms of
bargaining structure and the development of general unionism.
1.
Introduction
Between
1892
and
1987,
the number of unions in Britain declined from
1105
to
333,
while aggregate membership increased from
1,507,681
to
9,873,512.
The size of the average union rose, therefore, from
1364
to
29,650
members.
Associated with these trends has been a tendency for the distribution
of
union membership to become skewed, as a few large unions achieved
membership growth proportionately greater than rises in aggregate mem-
bership. In combination, these features
of
trade union structure have
contributed to a historical trend towards increasing concentration. The
process of concentration is defined here as the extent to which aggregate
union membership is increasingly organized by fewer relatively larger
unions. The focus, then, is on membership concentration rather than
organizational concentration (the number of organizations at a particular
level of union structure: see Windmuller
1981).
Such a focus allows
examination
of
the effects on union structure
of
differential rates
of
membership growth.
Jeremy Waddington works at the Industrial Relations Research Unit, University
of
Wanvick.
434
British
Journal
of
Industrial Relation5
Although membership concentration is rarely discussed explicitly in the
context
of
trade unionism, it has implicitly informed several
key
debates
concerning union structure and policy. Since the early 1920s, for example,
the TUC has advocated a more ‘rational’ union structure. While the TUC’s
support for industrial unionism has diminished since the 1920s, recent
attempts to co-ordinate recruitment are, in part, directed towards increasing
membership concentration in particular sectors of the economy. Concentra-
tion has also been promoted by the TUC’s advocacy of mergers between
affiliated unions and, at an industrial level, by the Bridlington principles,
which have restricted opportunities for affiliated unions to recruit where
others already organize. In addition, membership concentration has in-
fluenced the establishment of TUC policy. Congress decisions have been
determined by card votes dependent upon membership strength,
so
the
more concentrated is union membership, the greater the likelihood that
decision-making will be influenced by fewer organizations.
Issues relating to membership concentration are thus central to the
development of union structure. The purpose of this paper is to move
towards an explanation of the process of membership concentration in
British unionism. The paper comprises three main sections. Section 2
examines the requirements
of
an appropriate measure
of
membership
concentration. Existing studies use the variance of the
logs
of membership
size (Hart and Phelps-Brown 1957; Simpson 1972) and concentration ratios
(Buchanan 1981). In rejecting these approaches, a different measure
of
concentration
is
introduced: the Herfindahl index. Section
3
presents the
results from the application
of
the Herfindahl index. It
is
concerned with the
trends in concentration for aggregate and occupational membership, and
variations in the level of concentration between broad occupational groups
for the period 1892-1987. Section
4
seeks to explain the process
of
concentration. Reference is made to changes in company structure,
bargaining structure and forms
of
union organization and government. The
data sources are presented in an Appendix.
2.
Establishing an appropriate measure
of
membership concentration
Two constituent elements of concentration have been identified: the
number of existing organizations, and the inequalities in their sizes (Hart
er
al.
1973:1&21); Hannah and Kay 1977:41-63).’ Both these elements are
required
in
a measure
of
concentration, the central concern being the
relative weights attached to each. Davies (1979) examined this relationship
by means
of
iso-concentration curves and identified two desirable character-
istics
of
an appropriate measure
of
concentration. First,
if
concentration is
to
remain constant while the number of organizations increases, size inequali-
ties
must also increase. Second, the larger the number of organizations in a
population, the less sensitive an appropriate measure of concentration
becomes to further increases in the number of organizations (Davies 1979).
To continue reading
Request your trial