Content uploaded by Stephane Le Queux
Author content
All content in this area was uploaded by Stephane Le Queux on Sep 13, 2016
Content may be subject to copyright.
10.1177/0730888405279077WORK AND OCCUPATIONS / November 2005Lévesque et al. / DEMOCRACY AS UNION REVITALIZATION
Union Disaffection and Social Identity
Democracy as a Source of Union Revitalization
CHRISTIAN LÉVESQUE
HEC Montréal
GREGOR MURRAY
Université de Montréal
STÉPHANE LE QUEUX
Griffith University
This article examines union members’evaluation of the relevance of unions and their identifica-
tion with a traditional collective value frame for union action. It seeks to take account of the
impact of increasing labor market heterogeneity, declining instrumentality, and the behavior of
unions and employers. Using Canadian data gathered from individualunion members and their
local union leaders, the study finds that new labor market identities are not linked to weaker
belief in the relevance of unionsbut are associated with weaker identification with the traditional
value frame. Although declining instrumentality and hostile employer behavior are associated
with greater identification with traditionalvalue frames, greater union democracy is associated
with less membership disaffection on both the relevance of unions and their collective modes of
action. Union democracy is therefore found to be a key tool to address membership disaffection
and to generate collective identities for a renewed union project.
Keywords: labor union; democracy; social differentiation; workplace; union renewal
Apart from the many external challenges arising from neoliberal eco-
nomic globalization and hostile employers and public policies, unions
face challenges to their institutional legitimacy from within. What is the
400
Author’s Note: We wish to thank the many trade unionists and colleagues who con-
tributed to this study and, more particularly, Luc Cloutier, Nicolas Roby, and Sylvain
Schetagne, who greatly contributed to the gathering of the data. We also wish to ac-
knowledge the financial support of the Social Sciences and Humanities Research
Council of Canada, the Fonds québécois de recherché sur la société et la culture, and
the Confédération des syndicats nationaux. Christian Lévesque, HEC Montréal,
Christian.levesque@hec.ca; Gregor Murray, School of Industrial Relations,
Université de Montréal, gregor.murray@umontreal.ca; Stéphane Le Queux, Depart-
ment of Management, Griffith University, S.LeQueux@griffith.edu.au.
WORK AND OCCUPATIONS, Vol. 32 No. 4, November 2005 400-422
DOI: 10.1177/0730888405279077
© 2005 Sage Publications
impact of deep structural shifts in the economy and society on the validity of
the traditional union project for union members? This traditional project was
rooted in an older model of production and predicated on assumptions
increasingly open to question: predominantly male, exclusively in the realm
of formal labor market activity, concentrated in larger workplaces and pri-
marily about full-time, permanent employees working for a single employer.
Moreover, it was assumed that the interests of these workers could be rela-
tively easily framed or aggregated into a collective project.
There are, however, many manifestations of structural change that affect
union identity: the massive rise in the active labor force participation of
women (Fudge & Vosko, 2001); the relative decline of employment in goods-
producing sectors (Frenkel, Korczynski, Shire, & Tam, 1999); the movement
away from a paradigm of full-time, stable employment (Carré, Ferber,
Golden, & Herzenberg, 2000); the aging of the labor force and the more pre-
carious labor market insertion of the post–Baby Boom generations
(Tannock, 2001); the increased access to higher education and the growing
importance of knowledge work (Valkenburg, 1996, p. 91). These structural
changes involving the lives of workers translate into an ambient discourse
about the inability of unions to come to terms with the realities of a changing
labor market. For many, new workplace identities remain immune, even anti-
pathetic, to the kinds of collectivism synonymous with unionism. Thus, not
only are unions buffeted by structural change and a hostile political environ-
ment but the emergence of new setsof labor market values mark apermanent
and even more worrisome change for their future.
An important series of contributions in the French sociological literature
(Boltanski & Chiapello, 1999; Touraine, Wieviorka, & Dubet, 1984) have
certainly articulated this view. Their argument is that unions are a spent social
force because their institutional niche (the blue-collar industrial worker) is
no longer the emblematic force driving the economy. Similarly, Charles
Heckscher (1988) has offered a diagnosis of the need for a new unionism
based on an analysis of the fault lines of union solidarity and action. This has
also been taken up by observers such as Michael Piore (1995), whose vision
is similar to that of Boltanski and Chiapello inasmuch as other social identity
groups are more likely to drive social change than are unions.
Disaffection from unionism is clearly a key issue. In our own qualitative
research work conducted during the past couple of years with local unions in
Canada, we consistently find that managing greater membership diversity
and rekindling membership identification with collective values are at the top
of the agenda of concerns about union renewal.
There are currently a number of strategies for union revitalization or renewal
(see, for example, Baccaro, Hamann, & Turner, 2003; Bronfenbrenner,
Lévesque et al. / DEMOCRACY AS UNION REVITALIZATION 401
Friedman, Hurd, Oswald, & Seeber, 1998; Cornfield & McCammon, 2003;
Fairbrother & Yates, 2003; Frege & Kelly, 2004; Turner, Katz, & Hurd,
2001). These strategies include investing greater resources and efficacy in
union organizing, labor union restructuring in an attempt to consolidate sec-
toral bargaining strength and organizing initiatives, renewing union politics
through a broadening of social alliances and community unionism, and
rebuilding workplace strength through greater membership involvement, to
name but a few.
Our objective is to contribute to the union renewal debate through an
examination of the links between union disaffection and social identity. We
use Canadian data gathered from individual union members in the context of
a research partnership with one of the major union federations in Quebec
province. It is our contention that although there is in fact little empirical sup-
port for a fatalistic view of union disaffection linked to the emergence of new
labor market identities, these identities do pose a real challenge for union
revitalization. We argue that individual union member disaffection is linked
to union actions. In particular, union democracy is a key tool—indeed, one of
the only tools—to address membership disaffection and generate collective
identities for a renewed union project.
The first part of this article considers some of the major hypotheses
regarding union disaffection. We then draw on our Canadian study to con-
sider the relevance of three particular explanations of weak support for
unionism. We will do so in relation to two dimensions of union support or dis-
affection: what respondents perceive as the relevance of unionism and also
what we identify as a particular frame of union values or its modes of action.
SOURCES OF UNION DISAFFECTION
It can be argued that contemporary social relations are characterized by a
radicalization of difference. Divergent social identities challenge traditional
notions of collectivity and lead to an erosion of the relevance of traditional
frames of reference. A variety of authors have made this point. Muller-
Jentsch (1988), for example, identifies three sources of union crisis: an
increase in heterogeneity that leads to a crisis of aggregation of interests, an
increase of decentralization of regulatory regimes that leads to acrisis of loy-
alty, and the inability of unions to reach the new labor force in growing seg-
ments of the economy, which leads to a crisis of representation. Richard
Hyman (2001) has made similar arguments. This discourse is especially
prevalent in countries, such as those of northern Europe, where individual
union membership has generally been tributary of a strongly institutionalized
402 WORK AND OCCUPATIONS / November 2005
unionism that appears to experience difficulty in contending with the new,
and not so new, social currents of changing gender roles, youth movements,
and ecological consciousness (Leisink, Van Leemput, & Vilrokx, 1996;
Madsen, 1996). Such a contention is far from limited to northern Europe.
Labor movements in the developed world are everywhere seeking to come to
terms with this changing dynamic between individual and collectivity. This is
compounded by a neoliberal agenda that in a number of countries, facilitates
the institutional conditions for union disaffection (see, for example,
McCallum, 2002; Peetz, 2002).
There are currently three basic approaches to the explanation of union dis-
affection. The first, and probably the predominant one, suggests that increas-
ing social differentiation leads to a weakening of traditional union identities.
The second, drawing on the classic formulations of business unionism, looks
to the ability of unions to deliver relative advantage: Disaffection is therefore
the expression of declining instrumentality in a context where unions are less
able to deliver the goods. The third founds the relationship between union
member and unionism on the social relations of production. Although mate-
rial context and instrumentality are important, they can only be understood in
terms of the relationships between actors at work.
The first approach—the individualization thesis—posits a relationship
between increasing differentiation in social identity and a weakening of tra-
ditional union identities. It contends that it is a problem for unions as collec-
tive organizations to aggregate and express increasingly different interests in
the context of an employment relationship that is ever more individualized.
Industrial unionism was founded on a certain cultural homogeneity linked to
the manufacturing sector in which male workers shared a common heritage
that depended on the union to improve their terms and conditions of employ-
ment. Hobsbawn (1994) appositely speaks of the centrality of collectivity in
the life of the worker—the predominance of us over I.
The second approach focuses on the impact of material context for union
identity; in particular, how the strength of union identities or disaffection by
union members flow from the capacity of unions to shape their particular
material context. One aspect of this is, of course, bread and butter or instru-
mentality. There is an overwhelming literature on the importance of instru-
mentality as a link between the union and its members (Barling, Fullagar, &
Kelloway, 1992; Waddington & Whitson, 1997). Union members’ link to
their union can be understood in terms of the union’s capacity to improve
their relative fortunes. In many workplaces, an argument can be made for
declining instrumentality as unions appear to be charged with managing
increased workload, declining job security, and greater precariousness at work.
Lévesque et al. / DEMOCRACY AS UNION REVITALIZATION 403
The third approach looks at the role of the social relations of production in
the construction of union identities. In effect, it considers how employer and
union behavior affect union identities. In a study of the relationship between
unions and their members, Gallie (1996) found that aggressive employer
behavior weakens the relationship between the union and its members. This
is certainly consistent with much of the recent literature on union organizing
and the dissuasive effects of employer behavior (Bronfenbrenner & Hickey,
2004). Unions can also affect union identity. Fiorito, Gallagher, and Fukami
(1988), for example, find that the quality of the relationship between a union
and its members exerts a structuring effect on members’ links to their union.
Newton and McFarlane Shore (1992) report that the communication of infor-
mation to new members and the quality of union feedback on collective bar-
gaining reinforce union identity. These findings support an approach that
suggests that union renewal projects must be linked to the ability of union
organizations to redefine their relationships with their members (see, for
example, Sharpe, 2004).
Each of these approaches offers a distinct explanation of union disaffec-
tion, which we will now subject to empirical scrutiny.
METHODOLOGY
The data used in this study is the result of a research partnership between
the team of researchers and one of the major union organizations in Quebec
province—the Confédération des syndicats nationaux—a union organiza-
tion with roughly 250,000 members in more than 2,000 local unions or
workplaces. It should be emphasized that a local union in this particular
union structure generally covers one workplace and that like in both the
United States and Canada, most workers in a unionized workplace in Canada
belong to the same union. Moreover, in accordance with the legal regime of
the Rand Formula under monopoly unionism, all workers pay dues to the
union and are represented by it, whether they wish to be members or not, and
are thereby de facto members. Indeed, when we randomly selected our sam-
ple of union members, the union was unable to distinguish between member
and nonmember dues payers.
In this two-phase study, we first sent a questionnaire to each of this organi-
zation’s local unions to which we received a 48.6% response rate (n= 977).
From these respondent workplace unions, we then made a random selection
of 308 unionized workplaces for which we created a stratified sample of
404 WORK AND OCCUPATIONS / November 2005
union members, by size and sector, of 2,288 union members. A polling firm
conducted telephone interviews with 1,457 of these union members. The
response rate was 63.9%, with only 11.9% refusing to participate in the sur-
vey. These interviews lasted an average of 22.5 minutes and covered a broad
range of questions concerning individual experiences of and perspectiveson
unionism at the workplace and beyond.
The final sample was broadly representative of the membership of the
union organization in this research partnership: 55.2% women, 31.7% non-
standard workers (i.e., not working full-time on a permanent basis), 25.6% in
private services, 24.1% in goods production, and 50.3% in public services,
14.5% in work units of less than 50 employees, 39.7% in workplaces
between 50 and 249 employees, and 45.6% in units of 250 or more
employees.
The weakness of much of the survey research on individual union mem-
bers is the strong bias toward methodological individualism. Studies based
on member evaluations are often decontextualized because of limits in the
ability of individual respondents to provide detailed information about their
union. Our study attempts to remedy this problem by drawing on two levels
of analysis: first, at the level of individual union members in terms of their
sociodemographic characteristics and assessments of working conditions, of
their union, and of their employer; second, at the level of their local union on
the basis of information provided by local union presidents (i.e., the senior
union representative in each workplace) and in terms of information on the
union, its relations with the employer, and its strategies for workplace
change.
The study seeks to take account of four types of variables. The social iden-
tity variables report the age, education, employment status, sex, industry or
sector, and workplace size of respondents. Working conditions pertain to
respondents’ evaluation of their satisfaction with their job and the evolution
of their job security and workload in the previous 3 years. Employer behavior
and union variables are taken from both union members and workplace rep-
resentatives. Individual respondents were asked whether the employer treats
the union as an adversary, whereas workplace representatives were asked to
evaluate to what degree their employer seeks to undermine the legitimacy of
the collective agreement and whether the employer creates a climate of fear
in the workplace. Union variables concern the extent to which individual
members feel that their union takes account of members’ views and explains
its decisions. Local union representatives reported on the type of position that
their union adopts in the context of work reorganization.
Lévesque et al. / DEMOCRACY AS UNION REVITALIZATION 405
ARE UNIONS STILL RELEVANT?
A first dimension of union identity concerns the relevance of unions in the
eyes of their members. According to Klandermans (1989), a minimal test of
loyalty to the union is the desire to remain unionized. We, therefore, asked
individual member respondents whether, if they had the choice, they would
prefer to be unionized? In response, 85.7% of members said that they would
prefer to be unionized, 11.5% indicated that they would prefer not to be
unionized, and 2.8% were unable or unwilling to provide a response. We also
asked respondents whether unions are necessary for defending the working
conditions of workers: Of individual respondents, 62.6% strongly agreed
that unions are necessary, 24.2% somewhat agreed, and 13.2% either some-
what or strongly disagreed. This strong membership support for unions,
which is consistent with much of the recent literature on union membership
in North America (see, for example, Freeman & Rogers, 1999; Lipset, Meltz,
Gomez, & Katchanovski, 2004, p. 79) certainly challenges facile notions of a
weakening of the link between unions and their members.
To obtain a more generic evaluation of the relevance of unions for their
members, we combined these two indicators (loyalty and necessity) to iden-
tify three profiles of union members. For a first group (62.6%), unions are
absolutely relevant. A second group (13.2%) believes that unions are not
really relevant, either because the respondent would prefer not to be union-
ized or does not believe that unions are necessary for improving working
conditions. A third group (24.2%) is more equivocal in its support for unions.
These respondents prefer to be unionized but only believe that unions are
somewhat necessary for the defense of working conditions.
The question remains, however, as to whether the fault lines of social iden-
tity, bound up with the types of workplaces in which those identities are
found, manifest weaker support for the relevance of unions. Or, again, is
there less support for unions according to the presence or absence of certain
types of contextual variables such as working conditions, employer behavior,
and the strategies and structure of the local union? Table 1 reports the results
of the multivariate analysis in opposing the different profiles of belief in the
relevance of unions.1
Overall, these results show that social identity variables such as age,edu-
cation, and working conditions are not significantly associated with the rele-
vance of unions. However, employment status, workplace size, sectoral loca-
tion, and employer and union behavior are significantly associated with
workers’ belief about the relevance of unions. More specifically, three types
of variable differentiate those members who see unions as absolutely relevant
from those who do not see them as relevant (column 1).
406 WORK AND OCCUPATIONS / November 2005
Lévesque et al. / DEMOCRACY AS UNION REVITALIZATION 407
TABLE 1: Probability Estimations of the Relevance of Unions
Multiplier Estimations
Not Relevant Not Relevant Somewhat Relevant
Versus Versus Versus
Variables Absolutely Relevant Somewhat Relevant Absolutely Relevant
Age
Less than 30 years 1.54 1.13 1.38
30 to 39 years 1.00 1.00 1.00
40 to 49 years 1.04 1.31 0.78
50 years and older 0.74 1.15 0.76
Education
High school or less 0.72 1.20 0.76
College 1.00 1.00 1.00
University 0.68 1.03 0.72
Employment status
Full-time regular 1.00 1.00 1.00
Other 1.39 1.54* 1.03
Sex or sector
Men in goods 1.00 1.00 1.00
Women in private services 1.64 1.26 1.86*
Men in private services 0.72 0.52 1.79*
Women in public services 2.13* 1.08 1.81**
Men in public services 1.20 0.74 1.88*
Workplace union size
Small (< 50) 0.88 0.92 0.88
Medium (50 to 249) 1.00 1.00 1.00
Large (≥250) 0.78 0.58* 1.30*
Working conditions
Satisfied 1.00 1.00 1.00
Dissatisfied 1.03 1.03 0.96
Job security 1.00 1.00 1.00
Job insecurity 0.86 0.99 1.02
Stable workload 1.00 1.00 1.00
Intensified workload 0.83 0.98 0.83
Employer
Nonadversarial 1.00 1.00 1.00
Adversarial 1.46* 1.49* 0.86
Does not undermine
collective agreement 1.00 1.00 1.00
Undermines collective
agreement 0.61* 0.67* 1.00
Does not create a
climate of fear 1.00 1.00 1.00
Creates a climate of fear 1.57* 1.54* 0.87
(continued)
First, in terms of the social identity of respondents, women in the public
sector are 2.13 times more likely to consider unions as not relevant as
opposed to men in the goods sector. It should be emphasized that we do not
observe significant differences between men in the public sector and men in
the goods sector, nor does this difference appear to be associated with partic-
ular pressures in working conditions that might characterize one sector as
opposed to another.
A second type of explanatory variable concerns employer behavior,
which is associated with both a weakening and a strengthening in the belief of
the relevance of unions. Where, as measured by local union leader percep-
tions, an employer seeks to create a climate of fear or, according to members,
cultivates an adversarial relationship with the union, there is a greater likeli-
hood that union members will view unions as not relevant. However, where
an employer seeks to undermine the collective agreement, the reverse is true
408 WORK AND OCCUPATIONS / November 2005
TABLE 1 (continued)
Multiplier Estimations
Not Relevant Not Relevant Somewhat Relevant
Versus Versus Versus
Variables Absolutely Relevant Somewhat Relevant Absolutely Relevant
Union
Takes account of
members’ views 1.00 1.00 1.00
Does not take account of
members’ views 4.08** 2.99** 1.23
Explains decisions 1.00 1.00 1.00
Does not explain decisions 2.48** 1.79* 1.35
Union stance on work
reorganization
Exerts pressure for
change on the employer 0.75 1.02 0.71*
Does not exert pressure
for change on the employer 1.00 1.00 1.00
Opposed to change 1.01 0.97 0.94
Not opposed to change 1.00 1.00 1.00
Works jointly with
the employer 1.55* 1.09 1.30*
Does not work jointly
with the employer 1.00 1.00 1.00
N
1,032 499 1,177
Pseudo
R
20.23 0.19 0.05
*
p
< .05.**
p
< .005.
and union members are more likely to see the union as absolutely relevant.
These results are consistent with other findings that employer behavior can
inflect both positively and negatively union member evaluations of their
union (Gallie, 1996).
The third and most powerful explanatory variable pertains to the union. If
members perceive that their union does not take account of their views, then
they are 4.08 times more likely to believe that unions are not relevant. Simi-
larly, if the member believes that the union does not explain its decisions,
they are more than twice as likely (2.48) to believe that unions are not rele-
vant. In other words (and this is at the very heart of current debates about
union renewal), democracy matters. It is entirely logical that those members
who do not believe in the relevance of unions tend to highlight the shortcom-
ings of union democracy. However, this finding goes beyond mere tautology
because a considerable proportion of those who do not see the relevance of
unions nonetheless have a positive evaluation of union democracy. More-
over, the perception of the accountability of local union leaders in the context
of work reorganization is associated with strong effects on the evaluation of
the relevance of unions. Where unions work jointly with the employer and
there is a perception that the union does not take account of member views or
explain its decisions, there is a much greater likelihood (5.63) that the respon-
dent believes that unions are not relevant. We also observe a much stronger
relationship between union democracy and the evaluation of the relevance of
unions than we do between employer behavior and the evaluation of the rele-
vance of unions. The strength of this finding, relative to all the other factors
considered, suggests that unions do have a space to work on union identity.
Democracy is a building block in the construction of union identity.
The multivariate comparison of union members who believe that unions
are not relevant as opposed to those who believe that they are somewhat rele-
vant (column 2) further confirms the importance of union democracy vari-
ables, which are again strongly associated with a negative or positive evalua-
tion of the relevance of unions. Similarly, in terms of employer behavior, a
climate of fear and adversarialism weaken the link to unions, whereas an
employer who seeks to undermine the legitimacy of the collective agreement
strengthens the link between members and their union. Workplace size and
employment status are also related to workers’ belief about the relevance of
unions. Respondents in large workplaces, relative to medium-size work-
places and those who do not have a full-time, regular job are less likely to
believe that unions are not relevant.
The final column in Table 1 reports the differences between those who see
unions as absolutely relevant and those who see them as somewhat relevant.
Neither union democracy nor employer behavior differentiate these two
Lévesque et al. / DEMOCRACY AS UNION REVITALIZATION 409
groups but the stance taken by the union in the context of work reorganization
does. Where unions adopt a proactive position in exerting pressure on the
employer, then respondents are more likely to see the union as relevant. Con-
versely, where unions work jointly with the employer, there is a greater likeli-
hood that respondents will see the union as only somewhat relevant. In other
words, the pursuit of an autonomous agenda, rather than jointness, reinforces
a more positive evaluation of the need for unions. The sectoral location of
respondents is also significant. Men in goods constitute a high water mark in
their evaluation of the relevance of unions that is not matched by either men
or women in public and private services. Finally, respondents in large
workplaces are less likely to find unions absolutely relevant.
WHO IDENTIFIES WITH TRADITIONAL
MODELS OF COLLECTIVE ACTION?
Industrial unionism, as developed during the past century in North Amer-
ica, is a social construction of sets of beliefs and practices. When new union
members query the way a union works, the response is likely to come in terms
of a representation of cultural artifacts or a narrative frame of reference. Our
second dimension of union disaffection concerns the degree to which union
members identify with a traditional narrative of collective worker action.
This approach is informed by the notion that the framing of values is an itera-
tive process whereby individuals continuously construct and reconstruct the
meaning of the social movements of which they are part (see, for example,
Benford & Snow, 2000; Mansbridge, 1992; Offe & Wiesenthal, 1985).
We used three questions to capture the traditional union value frame. The
first concerns the geography of solidarity: Should unions defend all workers
or only their members? In response, 37.1% of respondents strongly agreed
that unions should defend all workers, 32.3% were somewhat in agreement,
19.1% disagreed somewhat, and 11.5% strongly disagreed. The second
question concerned the importance of the strike as a tool for defending work-
ers’ rights. There was much less support for this idea: Of the respondents,
17.3% strongly agreed, 22.8% somewhat agreed, 34.6% somewhat dis-
agreed, and 25.3% completely disagreed. This weaker support for the strike
reflects a broader social trend as regards the reluctance of workers to go on
strike. It is also probably related to the importance of the public service in our
overall sample, where many workers are in a direct relationship with the ben-
eficiaries of their services who are often vulnerable, as in the case of health
and social services. The third question tested a classic notion of industrial
union solidarity, namely, that a worker who does not want to go on strike
410 WORK AND OCCUPATIONS / November 2005
should have the right to cross a picket line. In response to this question,
19.5% of respondents strongly agreed, 12.9% somewhat agreed, 20.3%
somewhat disagreed, and 47.3% strongly disagreed. These three questions
allow us to capture some of the archetypal values associated with a traditional
narrative of industrial unionism: Unions ultimately have a broader social role
in promoting the industrial citizenship of all workers and not just of their
members; the strike is a necessary tool in establishing the relative balance of
advantage between employer and union; to ensure an equitable power rela-
tionship between workers and their employers, collective rights must prevail
over individual rights, as in the case of the right to cross a picket line.
Although it is no doubt possible to find other values associated with this tra-
ditional union model, our qualitative discussions with many union activists
corroborate the importance of these values. It is difficult to imagine that a
traditional industrial unionist would be in disagreement with any of these
values.
Cluster analysis enables us to identify three groups according to the
degree of workers’identification with this traditional union model of collec-
tive action. The first group (n= 488, 35.8% of the sample) exhibits a strong
identification with the traditional basket of union values. As well as believing
that unions should advance the cause of allworkers, this group maintains that
strikes are necessary and that workers should not have the right to cross
picket lines. In the second group (n= 453), union members do not believe that
strikes are necessary nor should unions seek to defend all workers; however,
workers should not have the right to cross picket lines. Their identification
with the traditional union model of collective action is therefore limited to
solidarity during a strike. The third group (n= 453) extends its solidarity to all
workers but does not believe that strikes are necessary and maintains that
workers should have the right to cross picket lines. The prescience given to
individual rights leads us to consider that this third group has the weakest
identification with the traditional basket of union values. Indeed, both of
these latter groups demonstrate a much weaker identification with the
traditional union model of collective action.
Table 2 presents the probability estimates for the degree of identification
of our union respondents with traditional models of collective action when
both social identity and contextual variables are taken into account. Three
key findings emerge from this analysis. First, all social identity variables are
strongly related to the traditional model of collective action: employment sta-
tus, education, age, gender, and sectoral location. Second, both job dissatis-
faction and intensity of workload are associated with the evaluation of modes
of collective action. Finally, union democracy and, to a lesser degree,
jointness in the context of work reorganization continue to demonstrate a
Lévesque et al. / DEMOCRACY AS UNION REVITALIZATION 411
strong link with disaffection from the traditional value frame. However,
employer behavior and workplace size are not associated with the degree of
identification of our union respondents with traditional models of collective
action.
What emerges most clearly from these results is the reluctance of women
in the public sector to espouse the traditional union model. They are 5.5.
times more likely than men in the goods sector to adhere to a weak identifica-
tion than a strong identification. This is to a lesser degree the case with men in
the public sector (2.0), younger workers (2.7), union members with precari-
ous jobs (1.7), and those with a college or technical education. Of course,
some combination of these characteristics is likely to increase further the
probability that these groups will figure among the weak identifiers with tra-
ditional modes of collective action. For example, young women in the public
sector are much more likely to be in this disaffected group as regards the tra-
ditional model of collective action. In other words, many of the so-called
fault lines of the new labor market are strongly related to a weakening of
traditional union identity.
Workplace contextual and union variables do, however, mitigate this
effect. Members dissatisfied with their working conditions or for whom
workload has increased during the previous 3 years are more likely to
embrace the traditional model. Similarly, union democracy also exerts a posi-
tive impact on the identification with the traditional narrative of union values.
The probability of a strong identification with the collective model is greater
when the union takes account of members’ views (1.82) and explains its deci-
sions (1.71). Disaffection with this model is greater when the union works
jointly with the employer (1.53).
Although union democracy and its stance in work reorganization exert an
impact on the identification with traditional models of collective action, it is
quite clear that the differentiation of social identity either plays a more
important role (as in the case of sectoral location, gender, and age) or an
equally important role (as in the case of employment status and education).
The contrast between the two weaker forms of identification with the tra-
ditional model of collective action (column 2) yields very similar results,
except that neither the union stance in work reorganization nor the members’
perception of local union democracy is statistically significant. Otherwise,
the relationships observed confirm the importance of social identity, working
conditions, and union variables.
The contrast between medium and strong identification with the tradi-
tional collective model indicates that members exhibiting a weaker identifi-
cation (medium as opposed to strong) are similarly sensitive to sectoral and
gender variables (women in the public sector) and union democracy (not
412 WORK AND OCCUPATIONS / November 2005
Lévesque et al. / DEMOCRACY AS UNION REVITALIZATION 413
TABLE 2: Probability Estimations of Identification With Traditional Models of
Collective Action
Multiplier Estimations
Weak Versus Weak Versus Medium Versus
Strong Medium Strong
Variables Identification Identification Identification
Age
Less than 30 years 2.76** 2.04** 1.02
30 to 39 years 1.00 1.00 1.00
40 to 49 years 0.90 1.11 0.71
50 years and older 0.87 1.12 0.64*
Education
High school or less 0.72* 0.95 0.75
College 1.00 1.00 1.00
University 0.55* 0.67* 0.70
Employment status
Full-time regular 1.00 1.00 1.00
Other 1.77** 2.24** 0.79
Sex or sector
Men in goods 1.00 1.00 1.00
Women in private services 1.63 0.91 1.52
Men in private services 0.80 1.13 0.67
Women in public services 5.52** 2.36** 2.24**
Men in public services 2.02** 1.86* 1.00
Workplace union size
Small (< 50) 0.73 0.83 0.92
Medium (50 to 249) 1.00 1.00 1.00
Large (≥250) 0.89 0.86 0.93
Working conditions
Satisfied 1.00 1.00 1.00
Dissatisfied 0.60** 0.68* 0.86
Job security 1.00 1.00 1.00
Job insecurity 1.14 1.05 1.03
Stable workload 1.00 1.00 1.00
Intensified workload 0.70* 0.73* 1.08
Employer
Nonadversarial 1.00 1.00 1.00
Adversarial 1.22 0.90 1.15
Does not undermine
collective agreement 1.00 1.00 1.00
Undermines collective agreement 0.77 0.81 1.04
Does not create a climate of fear 1.00 1.00 1.00
Creates a climate of fear 1.04 0.87 1.28
(continued)
taking account of members’ views). In addition, older workers are more
likely to display a stronger identification with the traditional model. Finally,
in the case of unions that either work with the employer or oppose change in
work reorganization, members are less likely to identify with traditional
models of collective action. This result certainly highlights the challenge for
a union to find a winning strategy in the context of work reorganization. The
only viable strategy would appear to be proactive in terms of exerting pres-
sure for an autonomous agenda for change. It also seems likely that it is as
much the legitimacy of the position adopted by the union, legitimated
through democratic practices, as it is the position itself that drives member
disaffection.
DISCUSSION
The central question examined in this article concerns the impact of
increasing labor market heterogeneity on the link between unions and their
414 WORK AND OCCUPATIONS / November 2005
TABLE 2 (continued)
Multiplier Estimations
Weak Versus Weak Versus Medium Versus
Strong Medium Strong
Variables Identification Identification Identification
Union
Takes account of members’ views 1.00 1.00 1.00
Does not take account of
members’ views 1.82** 1.87** 0.94
Explains decisions 1.00 1.00 1.00
Does not explain decisions 1.71** 1.30 1.45*
Union stance on work
reorganization
Exerts pressure for change on
the employer 0.97 1.20 0.79
Does not exert pressure for
change on the employer 1.00 1.00 1.00
Opposed to change 1.13 0.86 1.36*
Not opposed to change 1.00 1.00 1.00
Works jointly with the employer 1.53* 1.06 1.57**
Does not work jointly with
the employer 1.00 1.00 1.00
N
874 869 909
Pseudo
R
20.27 0.16 0.08
*
p
< .05.**
p
< .005.
members. Before responding to this question, it is important to recall four
limits of this study. We did not seek to take account of all the factors that
might affect the link between the individual union member and his or her
union, notably family socialization, nor did the study tackle the question of
the nonunionized, only the unionized, albeit in an institutional context where,
because most union members come to unions because their jobs are union-
ized and not because they choose unionization, weak union identities are
more likely to be present within the union. Given the fairly homogeneous
racial and ethnic character of the population that we were working with, we
were unable to evaluate the impact of racial and ethnic differences. Finally,
our methodology (telephone interviews supplemented by a postal survey) is
inevitably fairly static in relation to the complexity of shifting value frames
about collective action. Despite these limits, this study does offer unique
insights into the construction of union identities both because of its detailed
focus on the dynamics of union representation and its ability to contextualize
the views of individual members through multilevel analysis. So, does
increasing social differentiation lead to a weakening of union identity?
First, our study points to massive support for the relevance of unions. Only
a small minority of the 1,457 union members does not see unions as neces-
sary and would prefer not to be unionized. The support for the traditional
union model of collective action is, however, altogether more equivocal.
These results lend credence to Richard Hyman’s (1994) contention that in the
context of current socioeconomic transformations, it is not so much a crisis of
unionism as a crisis of a particular form of unionism, and unions must be
recast in the light of these transformations.
Second, differences in social identity do not appear to exert a strong influ-
ence on the evaluation of the relevance of unionism, but they have tremen-
dous consequences for the evaluation of its modes of action. Only the respon-
dent’s gender and sectoral location are associated with the evaluation of the
relevance of unions. Women in public services are less likely to consider the
union as relevant. This finding merits closer scrutiny because men in this
same sector do not manifest this same degree of disaffection. Among the
potential explanations to be considered are a sense of exclusion from deci-
sion making and a failure to connect with key women’s issues in this sector.
The consequences of social differentiation for modes of union action are
altogether more complex. Among those less likely to express their support
for the traditional model of union action are men and women in the public
sector, younger union members, those in precarious employment, and those
with a college or technical education. One of the most contentious issues
appears to be the use of the strike weapon. It should be noted that the right to
strike is subject to particular institutional constraints in the Quebec public
Lévesque et al. / DEMOCRACY AS UNION REVITALIZATION 415
sector (notably the obligation to ensure essential services that is subject to an
administrative determination). Moreover, the nature of the relationship
between service provider and service beneficiary, which is typically
gendered emotional labor for the provision of care, is probably an important
path to the understanding of value frames in this sector. There is also the issue
of the territory of representation and the notion of solidarity during a strike.
Men and women in the public sector, the young, workers with precarious
jobs, and those with a college or postsecondary, technical education were all
more likely to support the idea of a narrower frame of reference for union rep-
resentation as well as more individual than collective rights. This is not to
suggest that it is the majority of such workers, simply that these particular
groups of workers are more likely to express disaffection with the traditional
union model. Thus, differences in social identity appear to constitute a real
challenge for the traditional frame of union values.
Third, our study lends little support to the importance of material context
as a source of disaffection for the need for unions. The material context of
respondents is not associated with the way in which respondents assess the
relevance of unions. It is, however, associated with their evaluation of its
modes of action. In essence, greater economic or material pressures tend to
translate into a reduced level of disaffection for traditional union values, thus
lending some support to the importance of instrumentality. In this sense,
increased pressures can be translated into support for traditional union val-
ues, just as such a defensive instrumentality might subside in more favorable
economic circumstances. This result is important in several ways. In a con-
text of substantial pressures on workload and job security, a context that is in
fact very characteristic of most contemporary workplaces, there is a reduced
likelihood of disaffection from traditional union values. However, it should
also be recalled that the very groups most likely to suffer the ill effects of
weakened internal labor market regulation, notably women, the young, and
those with precarious jobs are more likely to express disaffection. Our results
also point to an ideological space for progressive employment relations poli-
cies. Where such policies can actually translate into improved working con-
ditions or lesser pressures at work, there is a greater probability of a weaken-
ing of support for the traditional union model of collective action. The
complexity of these results highlights the importance of grounded research
on the micromobilization of collective union identities within restructuring
workplaces (see, for example, Wallace & Leicht, 2004).
Fourth, employer strategy influences the perception of the relevance of
unions but not the traditional model of collective action. Contrary to some
prevailing conceptions about the negative effects of cooperative employers
for unions, it is the hostile employer who creates a climate of fear that exerts a
416 WORK AND OCCUPATIONS / November 2005
downward effect on the evaluation of the relevance of unions. However, an
employer who seeks to undermine the legitimacy of joint regulation (i.e., the
collective agreement) is more likely to be associated with a heightened
evaluation of the relevance of unionism.
Fifth, union behavior and strategy appear to influence the evaluation of
both the relevance of unions and their modes of action. Where the union
member perceives the local union to be less democratic, the support for both
the relevance of unions and the traditional value frame are likely to be
weaker. The effect of the union position on work organization is more equiv-
ocal. Union support for change in work organization is associated with
decreased union relevance, but there is alsodecreased support for traditional
union values when the union either supports or opposes work reorganization.
This result points to the contentious character of workplace change and the
role of the union therein. According to these results and in terms of member-
ship identification with their modes of action, unions are likely to fare better
with a proactive agenda for change but the legitimacy of the union in a con-
text of work reorganization would appear to be as important as its actual
position.
Finally, what most influences union disaffection? Is it social differentia-
tion, material context or actor behavior? Of the three approaches under scru-
tiny, it would appear that actor behavior exerts the greatest influence. In terms
of the relevance of unions, when all variables are considered simultaneously,
the perception of the democratic character of the local union is the single
most important factor. There is much greater support for the social differenti-
ation thesis in the case of modes of union action or support for the traditional
model of collective action. Once again, with the exception of the influence of
declining material conditions at work, the sole possible countervailing force
would appear to be the democratic character of the local union.
The proliferation of social identities in the contemporary labor market, so
characteristic of many advanced industrial societies, clearly constitutes a
challenge to the traditional framework of collective values for union action.
Some categories of worker—the young, those with technical education,
women and men in the public sector, those in precarious jobs—do appear to
express new sets of values. These workers embrace unionism as an institution
but do not necessarily espouse its traditional frame for collective action. A
key question is how to give voice to these new identities. Part of the response
lies in the sensitivity of these new groups of workers to the legitimacy of their
local unions. Local democracy matters because it reinforces support for the
union as an institution and its framing as a set of values and actions. In so
doing, democracy is also a resource likely to enhance the power of the local
union as a social actor (see Lévesque & Murray, 2002).
Lévesque et al. / DEMOCRACY AS UNION REVITALIZATION 417
UNION STRATEGIES FOR REVITALIZATION
The implications of this study for trade union identity concern the capac-
ity of local unions to deal with new labor market identities. It is clear that a
positive perception of local union democracy reinforces commitment to
union values. At the same time, it is the practice of union democracy that will
ultimately contribute to the emergence of new union values.
Our findings certainly add some credence to analyses suggesting that
union actions and strategies are out of sync with contemporary exigencies,
that their modes of action are often hierarchical and bureaucratic, that their
systems of communication are inadequate, and that the nature and objectives
of their actions need to be rethought (Fantasia & Voss, 2004; Hyman, 1994).
Similarly,from a different perspective,Heckscher (1988) has suggested that
current union structures have a great deal of difficulty giving expression to
new forms of solidarity, to these new labor market identities synonymous
with contemporary social differentiation. Such a contention is strongly ech-
oed by many feminist scholars who are highly critical of union performance
in giving voice to gender questions (Briskin & McDermott, 1993; Pocock,
1997).
A key challenge for union revitalization therefore concerns unions’ insti-
tutional capacity to take account of new labor market identities through
enhanced possibilities to participate in the life of the union. It should be
emphasized that solidarity does not exist a priori but is rather the result of a
process of socialization, of repeated efforts to develop alliances and complic-
ities between workers, despite the differences that might separate them
(Leisink et al., 1996). This insight is also at the core of much of the new litera-
ture on deliberative democracy. For example, Fung and Olin Wright (2003)
emphasize that “deliberative processes can affect the understanding individ-
uals have both of their interests and of the optimal strategies for satisfying
those interests” (p. 41). Conceptions of interests do not simply preexist their
engagement with union organizations; they are complex, often vague. Indi-
viduals define their interests in interactions with other actors and these inter-
actions affect the understanding of those interests (Fung & Olin Wright,
2003; Mansbridge, 1992). This processual nature of solidarity highlights the
real importance of participation in the debates about union activities and
strategies. Our results also point to the need for detailed ethnographic work
looking at the construction and transformation of union value frames.
Our findings do not point to a rupture with regard to unionism as an insti-
tution but they do highlight the need to renew its modes of action. This
renewal constitutes a profound yet somewhat paradoxical challenge. First,
the development of new values and modes of action that give expression to
418 WORK AND OCCUPATIONS / November 2005
these new labor market identities must start from the implication of these
groups of workers in union debates. Involvement in these debates therefore
constitutes a minimum point of departure and the challenge of how to achieve
this minimum remains a substantial impediment. Second, and more hope-
fully for the union as an institution, the democratic character of these debates
and the very fact of participation are likely to reinforce identification with an
emerging set of collective values. In other words, through this continuing
debate about the terms and nature of collective action, not only will there
emerge new value frames but many of the older values might well be appro-
priated and given renewed expression. In our view, that is why democracy is
at the very heart of union renewal projects.
NOTE
1. Given the three values associated with the dependent variable, we used logistic regression
to identify those independent variables that differentiate our respondents according to their vi-
sion of unions as absolutely, somewhat, or not relevant. The independent variables were
dichotomized to estimate their explanatory capacity (odds ratio) and the groupings of the de-
pendent variable (union relevance) were opposed, yielding probability estimates for each inde-
pendent variable. Gender and sector were combined into five categories because of their high de-
gree of correlation: men in goods, women in private services, men in private services, women in
public services, and men in public services.
REFERENCES
Baccaro, L., Hamann, K., & Turner, L. (2003). The politics of labour movement revitalization:
The need for a revitalized perspective.European Journal of Industrial Relations,9, 119-133.
Barling, J., Fullagar, C., & Kelloway, E. K. (1992). The union & its members: A psychological
approach. New York: Oxford University Press.
Benford, R. D., & Snow, D. A. (2000). Framing processes and social movements: An overview
and assessment. Annual Review of Sociology,26, 611-639.
Boltanski, L., & Chiapello, E. (1999). Le nouvel esprit du capitalisme [The new spirit of capital-
ism]. Paris: Gallimard.
Briskin, L., & McDermott, P. (1993). Womenchallenging unions: Feminism, democracy and mil-
itancy. Toronto, Canada: University of Toronto Press.
Bronfenbrenner, K. S., Friedman,S., Hurd, R. W., Oswald, R. A., & Seeber,R. L. (Eds.). (1998).
Organizing to win: New research on union strategies. Ithaca, NY: ILR Press.
Bronfenbrenner, K. S., & Hickey, R. (2004). Changing to organize:A national assessment of un-
ion strategies. In R. Milkman & K. Voss(Eds.), Rebuilding labor: Organizing and organizers
in the new union movement (pp. 17-61). Ithaca, NY: ILR Press.
Carré, F., Ferber, M. A., Golden, L., & Herzenberg, S. A. (Eds.). (2000). Nonstandard work: The
nature and challenges of emerging employment arrangements. Ithaca, NY: ILR Press.
Lévesque et al. / DEMOCRACY AS UNION REVITALIZATION 419
Cornfield, D. B., & McCammon, H. J. (Eds.). (2003). Labor revitalization: Global perspectives
and new initiatives. Research in the Sociology of Work,11, 1-294.
Fairbrother, P., & Yates, C. A. B. (Eds.). (2003). Trade unions in renewal: A comparative study.
London: Continuum.
Fantasia, R., & Voss,K. (2004). Hard work: Remaking the American labor movement. Berkeley,
CA: University of California Press.
Fiorito, J., Gallagher, D. G., & Fukami, C. V. (1988). Satisfaction with union representation. In-
dustrial and Labor Relations Review,41, 294-307.
Freeman, R. B., & Rogers, J. (1999). What workers want. Ithaca, NY: ILR Press.
Frege, C. M., & Kelly, J. (Eds.). (2004). Varieties of unionism: Strategies for union revitalization
in a globalizing economy. Oxford, UK: Oxford University Press.
Frenkel, S., Korczynski, M., Shire, K. A., & Tam, M. (1999). On the front line: Organization of
work in the information economy. Ithaca, NY: ILR Press.
Fudge, J., & Vosko, L. (2001). Gender,segmentation and the standard employment relationships
in Canadian labour law,legislation and policy. Economic and Industrial Democracy,22,281-
310.
Fung, A., & Olin Wright, E. (Eds.). (2003). Deepening democracy: Institutional innovations in
empowered participatory governance. London: Verso.
Gallie, D. (1996). Trade union allegiance and decline in British urban labour markets. In D.
Gallie, R. Penn, & M. Rose (Eds.), Trade unionism in recession (pp. 140-174). Oxford, UK:
Oxford University Press.
Heckscher, C. (1988). The new unionism: Employee involvement in the changing corporation.
New York: Basic Books.
Hobsbawn, E. (1994). Age of extremes: The short twentieth century 1914-1991. London: Aba-
cus.
Hyman, R. (1994). Changing trade union identities and strategies. In R. Hyman & A. Ferner
(Eds.), New frontiers in European industrial relations (pp. 108-139). Oxford, UK: Basil
Blackwell.
Hyman, R. (2001). Understanding European trade unionism: Between market, class, & society.
London: Sage.
Klandermans, B. (1989). Union commitment: Replications and tests in the Dutch context. Jour-
nal of Applied Psychology,74, 869-875.
Leisink, P., Van Leemput, J., & Vilrokx, J. (Eds.). (1996). The challenges to trade unions in Eu-
rope: Innovation or adaptation. Cheltenham, UK: Edward Elgar.
Lévesque, C., & Murray, G. (2002). Local versus global: Activating local union power in the
global economy. Labor Studies Journal,27, 39-66.
Lipset, S. M., Meltz, N. M., Gomez, R., & Katchanovski, I. (2004). The paradox of American
unionism. Ithaca, NY: ILR Press.
Madsen, M. (1996). Trade union democracy and individualisation: The cases of Denmark and
Sweden. Industrial Relations Journal,27, 115-128.
Mansbridge, J. R. (1992). A deliberativetheory of interest representation. In M. P. Petracca (Ed.),
The politics of interests: Interest groups transformed (pp. 32-57). Boulder, CO: Westview.
McCallum, R. (2002). Trade union recognition and Australia’s neo-liberal voluntary bargaining
laws. Relations Industrielles/Industrial Relations,57, 225-251.
Muller-Jentsch, W. (1988). Industrial relations theory and trade union strategy. International
Journal of Comparative Labour Law and Industrial Relations,4, 177-190.
Newton, L. A., & McFarlane Shore, L. (1992). A model of union membership:Instrumentality,
commitment, and opposition. Academy of Management Review,17, 275-298.
420 WORK AND OCCUPATIONS / November 2005
Offe, C., & Wiesenthal, H. (1985). The two logics of collective action: Theoretical notes on so-
cial class and organizational force. In C. Offe (Ed.), Disorganized capitalism (pp. 170-220).
Cambridge, MA: MIT Press.
Peetz, D. (2002). Decollectivist strategies in Oceania. Relations Industrielles/Industrial Rela-
tions,57, 252-281.
Piore, M. (1995). Beyond individualism: How social demands of the new identity groups chal-
lenge American political and economic life. Cambridge, MA: Harvard University Press.
Pocock, B. (Ed.). (1997). Strife: Sex and politics in labour unions. New South Wales, Australia:
Allen and Unwin.
Sharpe, T. (2004). Union democracy and succesfull campaigns: The dynamics of staff authority
and worker participation in an organizingunion. In R. Milkman & K. Voss(Eds.), Rebuilding
labor: Organizing and organizers in the new union movement (pp. 62-87). Ithaca, NY: ILR
Press.
Tannock, S. (2001). Youth at work: The unionized fast-food and grocery workplace. Philadel-
phia: Temple University Press.
Touraine, A., Wieviorka, M., & Dubet, F. (1984). Le mouvement ouvrier [The Workers’ Move-
ment]. Paris: Fayard.
Turner, L., Katz, H. C., & Hurd, R. W. (Eds.). (2001).Rekindling the movement: Labor’s quest
for relevance in the 21st century. Ithaca, NY: ILR Press.
Valkenburg, B. (1996). Individualization and solidarity: The challenge of modernization. In P.
Leisink, J. Van Leemput, & J. Vilrokx(Eds.), The challenges to trade unions in Europe:Inno-
vation or adaptation? (pp. 89-104). Cheltenham, UK: Edward Elgar.
Waddington,J., & Whitson, C. (1997). Why do people join unions in a period of membership de-
cline? British Journal of Industrial Relations,35, 515-546.
Wallace, M., & Leicht, K. T. (2004). Culture wars in the workplace? Cultural antecedents of
workers’ job entitlement. Work and Occupations,31, 3-37.
Christian Lévesque is an associate professor of industrial relations at HEC
Montréal and codirector of the Interuniversity Research Centre on Globaliza-
tion and Work (CRIMT). His main research interests focus on work reorganiza-
tion, unionism, and new employment relationship systems in the global era.
During the past few years, he has done extensive fieldwork in Canada and
Mexico on labor relations and union renewal. He is currently conducting a re-
search project, with several CRIMT researchers, on the impact of international
union alliances on local and national union capacity building.
GregorMurray is a professor in the School of Industrial Relations at Montréal
University and director of the Interuniversity Research Centre on Globaliza-
tion and Work (CRIMT). This center is conducting an international research
program under the auspices of the Social Sciences and Humanities Research
Council of Canada’s Major Collaborative Research Initiatives on Rethinking
Institutions for Work and Employment and Employment in a Global Era
(www.crimt.org). Gregor’s main research interests concern unionism, work or-
ganization, globalization, and industrial relations theory, on which he has
published various books and articles, including The Social Regulation of the
Lévesque et al. / DEMOCRACY AS UNION REVITALIZATION 421
Global Firm, L’Organisation de la Production et du Travail: Vers un Nouveau
Modèle, Work and Employment Relations in the High-Performance Work-
place, and La Representation Syndicale.
Stéphane Le Queux is a senior lecturer in the Department of Management at
Griffith University (Australia) and a research associate in the Interuniversity
Research Centre on Globalization and Work (CRIMT), for which he is conduct-
ing research on new identities and workers’voice in comparative perspective.
After being involved in research on the Europeanization of industrial relations
for the European Trade Union Institute (ETUI) and then on youth attitudes to-
ward unions in Australia and internationally, his latest work has been princi-
pally focusing on the relationship between antiglobalization movements and
trade unions in Australia, Canada, and France and its potential for union re-
newal and the revitalization of labor politics.
422 WORK AND OCCUPATIONS / November 2005
