The Effect of Trade Unions on the Provision of Training: Australian Evidence
| DOI | http://doi.org/10.1111/j.1467-8543.1994.tb01051.x |
| Published date | 01 December 1994 |
| Author | Robert Drago,Mark Wooden,Judith Sloan,Sean Kennedy |
| Date | 01 December 1994 |
British Journal
of
Industrial Relations
32:4
Dec
1994
0007-1080
$3.00
The Effect
of
Trade Unions
on the Provision
of
Training:
Australian Evidence
Sean Kennedy, Robert Drago, Judith Sloan and
Mark Wooden
Abstract
This paper uses data from the Australian Workplace Industrial Relations
Survey conducted in
1989-90
to examine the net impact of trade unions on the
extent of formal types of employer-provided training.
A
net positive effect
is
found, but only where unions are active in the workplace and not merely
de
jure
representatives of the work-force. The results also indicate, with respect
to in-house training, that unions shijit the emphasis towards work-forces with
long tenure, ajinding that
is
consistent with ‘union voice’ arguments.
1.
Introduction
Standard monopoly union models predict that the effect of unions on
employer-provided training via increased labour costs, flatter wage
structures and experience-related seniority rules should be negative, and
such predictions receive support from the US-based research reported in
Mincer
(1983),
Barron
etal.
(1987)
and Lilliard and Tan
(1992).
In contrast,
other studies from both the USA (Lynch
1992)
and the UK (Booth
1991;
Claydon and Green
1992)
report positive effects. Similarly, in a cross-
national study of young males, Tan
et al.
(1992)
report that positive union
effects predominate in the USA and Australia (though in the UK sample
insignificant effects were predominant, at least once apprentice training was
excluded).
While there are many possible explanations for this divergence in
findings, we believe that part of the difference may stem from the fact that
Sean Kennedy is Research Officer at the National Institute
of
Labour Studies, Hinders
University
of
South Australia. Robert Drago is Associate Professor in the Department
of
Economics, University
of
Wisconsin-Milwaukee. Judith Sloan is Professor and Director of the
National Institute
of
Labour Studies, Hinders University
of
South Australia. Mark Wooden is
Associate Professor and Deputy Director of the National Institute
of
Labour Studies, Hinders
University
of
South Australia.
566
unions exhibit different approaches to and emphases on training across
countries, over time, and even within a particular country. In the analysis
that follows we therefore provide a new analysis
of
unions and training,
using data from the recent Australian Workplace Industrial Relations
Survey (AWIRS), which pays close attention to what unions actually do,
particularly in terms
of
bargaining with management at the workplace.
Owing to data constraints, the analysis focuses solely on formal training
programmes, including both in-house courses provided directly by the firm
and externally delivered training courses, attendance at which is financed by
the employer. The analysis does not therefore cover the more informal
training process that takes place on the job. While on-the-job training is
almost certainly more pervasive (see Baker and Wooden
1992
for some
Australian evidence), a focus on formal training arguably permits us to
obtain a better understanding
of
the impact
of
unions on managerial
decisions about whether to invest in the human capital of the work-force.
Further, although much and perhaps even most firm-specific training is
informal, it seems reasonable to believe that in-house training involves a
large firm-specific component relative to external training, which will be
more general. Therefore, the analysis should still provide some insight into
the effects
of
unions on both general and specific training.
A key feature
of
the analysis, which permits us to address differences
between unions, is the variable used to represent union influence. Like
most studies into the impact
of
trade unions, the earlier studies mentioned
above all use a union membership variable (union density, in the case
of
workplace data) to represent union influence. In contrast, we argue that
union density should be replaced with a measure more closely related to
what unions actually do in the workplace. The rationale for this argument
is
provided in the work
of
Drago and Wooden
(1991)
who contend that in
many industrial relations settings mere union presence does not necessarily
guarantee effective provision
of
union voice. In the case
of
Australia, for
example, they argue that the protections afforded trade unions (e.g. via
restrictions on the ability
of
workers to organize alternative forms
of
worker representation and the widespread incidence
of
closed-shop
arrangements), in combination with a system
of
trade unionism that is still
structured largely along occupational lines, undermines the incentives for
unions to provide voice. Consequently, variations in union membership
may be only weakly related to the extent to which unions provide effective
representation.
In what follows we first briefly outline the theoretical arguments con-
cerning the different channels by which unions might influence training.
After introducing the data, we then proceed to the presentation
of
results
from a simple model designed to estimate the impact
of
unions on training.
In particular, we explicitly test whether the effects
of
unions on training are
better represented by workplace unionization rates
or
by an indicator
of
the extent of union delegate involvement in workplace bargaining.
A
brief
discussion
of
the implications
of
these results completes the paper.
British Journal
of
Industrial Relations
To continue reading
Request your trial