Full text of DOI:10.1111/1467-8543.00195

British Journal of Industrial Relations
39:2 June 2001 0007±1080 pp. 167±182
Exit,Voice and Loyalty Reactions to
Job Insecurity in Sweden: Do Unionized
and Non-unionized Employees Differ?
Magnus Sverke and Johnny Hellgren
Abstract
Although job insecurity has received growing recognition in connection with
the transformation of working life, little is known about how unionization
affects its consequences. Data from Swedish health care employees indicate
that job insecurity is related primarily to coping strategies in the exit and voice
domains. Union members were less inclined to make use of the exit and voice
options compared with their non-unionized co-workers, and more typically ex-
pressed loyalty to the organization. The collective support derived from union
membership may make individual voice expressions less important, a finding
that has to be replicated in countries with lower unionization rates.
1. Introduction
The past decades have witnessed a strong and intense pressure for organ-
izational change. This trend has been fuelled by factors such as the new
information technology, intensified global competition and a recessionary
climate. As a result, organizations in many countries have sought to improve
organizational effectiveness and to reduce labour costs through downsizing
or `right-sizing' (Hitt et al. 1994) of their activities. Clearly, these changes
have dramatic consequences for those who lose their jobs, but a growing
body of literature also bears witness to the negative effects on those who
remain in the organization (Latack and Dozier 1986). While a trend towards
individual flexibility, increasing employability requirements and intensified
demands to take responsibility for one's competence development all represent
characteristics of the new working life (Howard 1995), the experience of
employment uncertainty is perhaps the most dramatic consequence for
the so-called survivors of organizational changes involving layoffs. The
Magnus Sverke and Johnny Hellgren are in the Department of Psychology, Stockholm
University.
# Blackwell Publishers Ltd/London School of Economics 2001. Published by Blackwell Publishers Ltd,
108 Cowley Road, Oxford OX4 1JF, UK and 350 Main Street, Malden, MA 02148, USA.

168 British Journal of Industrial Relations
detrimental effects of such job insecurity for employee attitudes and well-
being are well documented (Ashford et al. 1989; Barling and Kelloway 1996;
Hartley et al. 1991).
Less is known, however, about the various alternatives employees may
use in coping with job insecurity, and about how the choice of alternatives
may differ between unionized and non-unionized employees. For instance,
do unionized and non-unionized employees make use of divergent strategies
to cope with job insecurity? In this paper we will depart from the exit, voice
and loyalty framework (Hirschman 1970) and examine the extent to which
employees utilize these alternative responses in their attempts to handle
uncertain employment situations. Although employees may exit from the
organization, use their voice (e.g. protest against organizational change) or
express loyalty to the organization in order to reduce future insecurity,
it is likely that union members and unaffiliated employees use different
means to cope with their experiences of employment uncertainty. Without
the collective support derived from union membership, it may be argued, an
employee will have greater difficulty in coping with job insecurity (Dekker
and Schaufeli 1995). On the other hand, those who stay outside the unions
may trust their own capacity to redress insecurity (Rosenblatt and Ruvio
1996). These issues will be empirically addressed using questionnaire data
from a sample of Swedish health care employees with a high degree of
unionization.
2. A typology of job insecurity and union membership
Job insecurity refers to a `powerlessness to maintain desired continuity in
a threatened job situation' (Greenhalgh and Rosenblatt 1984: 438). Because
it is a subjective experience based on interpretations of the immediate work
environment, job insecurity can be described as a discrepancy between pre-
ferred and experienced security in the employment situation (Hartley et al.
1991). Although the fear of losing one's job is likely to increase in
connection with corporate retrenchment, this suggests not only that the
same objective situation can evoke dissimilar experiences but also that
job insecurity can be an important factor in seemingly unthreatened job
situations (Rosenblatt and Ruvio 1996). We can expect, however, that
employees whose work situation is strongly affected by organizational
change and who see their co-workers losing their jobs are more likely to
experience high job insecurity, while those who are relatively unaffected by
the transformation process would experience lower levels of insecurity.
Even if innumerable layoffs have occurred in unionized work settings,
very few studies have addressed downsizing in relation to unionization
(Mellor 1992), and the issue of job insecurity has received scarce attention in
the industrial relations literature (Bender and Sloane 1999). However, there
is research to suggest that, on a general level, an individual emphasis on
job security represents an important determinant of union membership
# Blackwell Publishers Ltd/London School of Economics 2001.

Job Insecurity and Union Membership 169
(Crockett and Hall 1987). It appears, for example, that both members and
officers of labour organizations rank job security as perhaps their most
important bargaining issue (Dworkin et al. 1988). Furthermore, union
members seem to give job security a higher priority than non-members
(Brown Johnson et al. 1992). While this suggests that unionized workers
differ from their non-unionized counterparts in terms of the priority placed
on job security as well as the appraisal of the degree of insecurity in the
surrounding work situation, there is likely to be substantive variation also
among union members Ð even if they are employed at the same workplace.
Figure 1 presents a typology of job insecurity and union membership.
Combining high and low levels of insecurity with affiliation/non-affiliation
results in four different groups. `Insecure members' represent workers who
belong to a trade union and feel that the subjective probability of job loss
is relatively high. `Insecure non-members' have similar experiences of job
insecurity but are not affiliated to a union. The high level of job insecurity
shared by individuals in these groups is likely to result in intense efforts to
cope with the uncertainty, but it is possible that unionized and non-
unionized employees differ in terms of how they react to job insecurity
(Rosenblatt and Ruvio 1996). Not only may unionized employees receive
support from their union, but members and non-members may also differ in
terms of individual-level strategies to cope with insecurity. `Secure members'
and `secure non-members' share similar perceptions of low job uncertainty,
which suggests that they will be less inclined to be involved in coping efforts
than insecure employees. They do differ with respect to formal relationships
with labour organizations, however, something that could be manifested in
FIGURE 1
A Typology of Union Membership and Job Insecurity
Relationship to union
Not affiliated
Affiliated
Low
Secure
non-members
Secure
members
High
Insecure
non-members
Insecure
members
# Blackwell Publishers Ltd/London School of Economics 2001.

170 British Journal of Industrial Relations
their investments in strategies to cope in a relatively secure employment
situation.
3. Exit, voice and loyalty
In his seminal book Exit, Voice, and Loyalty: Responses to Decline in Firms,
Organizations, and States, Hirschman (1970) provides a framework for
understanding how various actors may respond to unsatisfactory conditions
in a variety of institutional settings. He also shows that the activities these
actors engage in represent a valuable means by which management finds out
about its failures and can learn for the future. While Hirschman's arguments
were developed to describe primarily how organizations respond to decline,
industrial relations research undoubtedly has benefited from this frame-
work; the result has been enhanced recognition and understanding of the
wide range of reactions individuals may display to unsatisfying employment
conditions (Farrell 1983; Withey and Cooper 1989). According to Hirsch-
man (1970), discontented employees can find a better job and quit (exit),
can work to improve the situation (voice), or can stay and support the
organization (loyalty).
Exit is a major alternative when firms, organizations and states are in
decline (Hirschman 1970). For instance, dissatisfied customers can switch to
another supplier, and dissatisfied voters can decide to support a different
political party. Similarly, employees who are dissatisfied with the degree of
job security in their existing positions can choose to exit from the job or the
organization (Hartley et al. 1991). Indeed, job insecurity has consistently
been found to be associated with withdrawal responses as manifested in
intensified turnover intentions (Arnold and Feldman 1982; Brockner 1988;
Burke and Nelson 1998; Davy et al. 1997; Dekker and Schaufeli 1995;
Hartley et al. 1991). In this respect, it has been argued (Greenhalgh and
Rosenblatt 1984; Hartley et al. 1991; Hirschman 1970) that management
may be more inclined to listen to the criticism of valuable employees, but
also that such qualified workers will more easily find a new job and thus will
be more likely to quit.
Like exit, voice is a reaction that may bring about change. Both exit and
voice are liable to make management rethink the dissatisfying conditions
and possibly rectify their origins, especially if these reactions are exhibited
by highly valuable employees. It is also conceivable that, in reacting to
unfavourable job situations, employees may attempt to change the source of
dissatisfaction prior to choosing the exit option (March and Simon 1958). In
contrast to the more economical nature of exit, however, voice is a political
response. Hirschman (1970: 30) described voice as `any attempt at all to
change rather than to escape from an objectionable state of affairs', some-
thing that is likely to be characteristic of individuals with substantial involve-
ment in the organization. Voice can be conceived of as interest articulation,
and it has been operationalized for instance as union membership (Freeman
# Blackwell Publishers Ltd/London School of Economics 2001.

Job Insecurity and Union Membership 171
and Medoff 1984), but it also may involve other actions and protests. In
Sweden, with an average rate of unionization of around 85 per cent
(Kjellberg 1998), union membership is part of everyday life and cannot in
itself be described as a manifestation of voice. Therefore, we decided to
investigate other forms of voice, such as expressions of discontentment with
a process of change and its goals.
While both exit and voice represent rather dramatic, active and overt
reactions to dissatisfaction with the employment situation, loyalty can
perhaps best be described as a more passive form. According to Hirschman
(1970), loyalty is the product of various factors that bind the individual to
the organization and thus make exit costly and voice troublesome. In this
vein, loyalty may be the reaction of employees with lower status or less
crucial skills in the firm in their attempts to redress job insecurity. Loyalty
can be manifested in improved performance or high levels of commitment to
the organization, for example.
However, both theoretical and empirical research on job insecurity calls
this argument into question. It has been found, for instance, that job
insecurity typically is associated with impaired organizational commitment
(Ashford et al. 1989; Brockner et al. 1992) or is unrelated to the construct
(Barling and Kelloway 1996). The issue of impaired loyalty as an inherent
consequence of job insecurity or enhanced loyalty as a manifestation of
attempts to redress uncertainty therefore merits further empirical scrutiny.
An important issue, which has hardly received attention in the literature,
concerns the role of unionization in employee reactions to job insecurity.
On a general level, it can be assumed not only that unions may reduce job
insecurity perceptions by providing protection (Greenhalgh and Rosenblatt
1984), but also that the collective support derived from union membership
will make members' reactions different from those of non-unionized
workers (Dekker and Schaufeli 1995). For instance, a study that included
the mere presence of union membership as an external coping resource
(Shaw et al. 1993) found positive relationships between union membership
and expressions of both loyalty (organizational commitment) and absence
of voice (positive attitudes towards the organizational change), but no inter-
action effects of job insecurity and union membership on these reactions.
Dekker and Schaufeli (1995) reported that social support derived from
union membership, co-workers and colleagues did not moderate the effects
of job insecurity on employee well-being. Most importantly, it has been
argued that unions lend stability to careers by binding members to the
employing organization (Bender and Sloane 1999; D'Amico 1984).
Given that social support is generally seen as an important factor in the
relationship between the stress induced by job insecurity and its outcomes
and that union membership is a potentially important source of support in
this context (e.g. Dekker and Schaufeli 1995; Hartley et al. 1991; Lim 1996),
one major objective of the present research is to examine whether unionized
and non-unionized employees who express differing levels of job insecurity
also differ in terms of how they cope with perceptions of employment
# Blackwell Publishers Ltd/London School of Economics 2001.

172 British Journal of Industrial Relations
uncertainty. Following Freeman and Medoff's (1984) reasoning, it can be
predicted that insecure non-members will be more inclined to leave the
organization, while insecure members tend to remain and instead express
their dissatisfaction with the insecure situation through various voice mech-
anisms. More specifically, our goal is to evaluate whether, and to what
extent, union membership has direct effects on the coping strategies and
moderates the effects of job insecurity.
4. Data
The empirical material for the study comes from a questionnaire survey
among all staff members of a Swedish emergency hospital undergoing
transformation. In 1994, the hospital was formed into a municipal corpor-
ation with the aim of providing better health care at a lower cost. In
connection with this, personnel and activities of an adjacent hospital that
was closed down were taken over, which meant a 35 per cent increase in
patients and a 10 per cent increase in work hours. In the following two
years, the staff was reduced from 1450 to 1200 employees by natural
attrition, outsourcing, reorganizations and layoffs. Starting in 1996, the
political leadership of the county council under which the hospital is
organized decided to reduce costs in the health care services by 17 per cent
in the three succeeding years. These cost reductions affected the ten or so
emergency hospitals in the area differently, and there was great freedom and
flexibility for the hospital to obtain the savings.
In consequence, two clinical departments were sold out and there were
closures and mergers of wards, although layoffs per se were relatively
infrequent. The changes took place in a period of high demand for health-
care personnel and a favourable situation for such employees on the labour
market. Great measures were taken by the management to minimize the
negative impact on the personnel while human resource management inter-
ventions were selected so as to meet organizational demands. The local
unions at the hospital were actively involved in the change process, and, in
general, the climate between management and unions was co-operative.
Sample
In 1998 questionnaires were mailed to the homes of all 1270 employees
(including temporary workers), accompanied by a letter describing the
general purpose of the research, explaining that participation was entirely
voluntary, and assuring recipients that their responses would be treated in
confidence. Approximately three weeks after the first mail-out, a reminder
postcard was sent to those who had not returned their questionnaires. A
total of 736 employees responded to the survey, for a response rate of 58 per
cent. The present study is based on 635 employees with complete data on
all study variables. The average age of the sample was 41 years (SD = 10),
# Blackwell Publishers Ltd/London School of Economics 2001.

Job Insecurity and Union Membership 173
with the youngest participant being 18 and the oldest 67. Females con-
stituted 79 per cent of the respondents, who had worked for the hospital for
an average of 10 years (SD = 9).
Questionnaire
Apart from union membership and demographics, which were assessed
using single questions, the study variables were measured using multiple
indicators and responses were given on Likert scales ranging from 1 (strongly
disagree) to 5 (strongly agree). Indices were constructed by averaging over
the relevant items after negatively phrased items had been reverse-coded.
(a) Union membership and job insecurity
Participants indicated to which union they belonged (90.4 per cent) or that
they were not union members (9.6 per cent). For the purpose of the present
study, a dummy variable was constructed such that unaffiliated employees
were coded 0 while membership in any of the unions operating at the
hospital was coded 1. The assessment of job insecurity was based on
Ashford et al.'s (1989) scale which reflects perceived threats to the total job,
and the ten questions were redrafted into statements (e.g., `It is likely that
I may be laid off permanently'). The scale demonstrated adequate internal
consistency (alpha = 0.83) and is similar in operationalization to other
established measures of job insecurity (Caplan et al. 1975; Hartley et al.
1991; Kuhnert and Vance 1992; Roskies and Louis-Guerin 1990).
The job insecurity measure was bifurcated at its mean value (M = 1.79),
and individuals reporting higher levels were coded as high in job insecurity
while those scoring below the mean were treated as low in job insecurity.
Combining this bifurcated measure with the dichotomous union member-
ship variable enabled us to operationalize the typology of union membership
and job insecurity: insecure members (n = 240), secure members (n = 334),
insecure non-members (n = 32) and secure non-members (n = 29).1
(b) Exit variables
We assessed two forms of exit Ð organizational withdrawal and job with-
drawal Ð each measured by three items. Each of these variables is based on
a well established conceptualization of turnover intention (Mobley et al.
1979) which has been operationalized to reflect the specific tendencies to
turnover from the organization and the job, respectively (Cohen 1998).
Sample items were `I am actively looking for jobs outside this hospital'
(organizational withdrawal) and `I am actively looking for another assign-
ment' (job withdrawal). The reliability estimate was satisfactory for both
organizational (alpha = 0.87) and job withdrawal (alpha = 0.80).
(c) Voice variables
In previous research, voice has been the most problematic aspect of the
Hirschman (1970) framework in terms of operationalization, which has been
# Blackwell Publishers Ltd/London School of Economics 2001.

174 British Journal of Industrial Relations
reflected in low reliability (see e.g. Withey and Cooper 1989). Our strategy
to avoid such problems was to operationalize individual voice based on an
attitudinal framework and, where possible, to rely on established measures.
Disengagement in the downsizing goal was assessed using Hollenbeck et al.'s
(1989) eight-item goal commitment scale (e.g., `Quite frankly, I don't care if
I achieve the goals of the hospital's cost savings or not'). The items were
slightly modified (i.e., `this goal' was throughout replaced by `the hospital's
cost savings') and the scale was reverse-coded in order to reflect disengage-
ment (alpha = 0.81). Protest against the downsizing process was measured
using six items specifically developed for the present study (e.g., `The
measures that have been taken to save money in the hospital are not
acceptable'). The scale was internally consistent (alpha = 0.81).
(d) Loyalty variables
Organizational commitment was assessed using a four-item measure
(Sjoberg and Sverke 1996) based on items from different commitment
scales (Allen and Meyer 1990; Cook and Wall 1980; Guest and Dewe 1991;
Mowday et al. 1979). The items reflect affective attachment to the organ-
ization (e.g. `I really feel as if this hospital's problems are my own') and the
scale demonstrated internal consistency (alpha = 0.72). A second form of
loyalty Ð work intensity Ð was operationalized using a five-item scale (e.g.
`When there's a job to be done, I devote all my energy to getting it done')
developed by Brown and Leigh (1996). The internal consistency for this
scale was 0.76.
5. Results
Table 1 reports descriptive statistics and intercorrelations for all study
variables. Union members typically were older than non-members, and
`insecure' employees comprised a larger proportion of women than their
more `secure' co-workers. There was no statistically significant relation-
ship between job insecurity and union membership, indicating that members
and non-members overall do not differ in perceptions of job insecurity.
Job insecurity was positively related to both exit variables and both voice
variables; and, while negatively related to organizational commitment, it was
unrelated to the other loyalty variable (work intensity). These bivariate
relationships thus indicate that job insecurity is related primarily to coping
strategies within the exit and voice dimensions. Union membership status
was found to relate negatively to both the exit and the voice attributes and
positively to the loyalty variables. This suggests that union members are less
prone to withdraw from the organization and the job, less inclined to protest
against the changes (voice), but tend to express more loyalty to the
organization than the non-affiliated employees.
To test for direct and interaction effects of job insecurity and union
membership on the three different coping dimensions (exit, voice and loyalty),
# Blackwell Publishers Ltd/London School of Economics 2001.

TABLE 1
Descriptive Statistics and Correlations for the Study Variables.
Variable
M
SD
Alpha
1
2
3
4
5
6
7
8
9
1. Job insecuritya
0.43
0.90
0.50
0.30
Ð
Ð
2. Union membershipb
70.06
Exit
3. Organizational withdrawal
4. Job withdrawal
2.10
2.56
1.21
1.19
0.87
0.80
0.21
0.24
70.13
70.10
0.72
Voice
5. Disengagement in the downsizing goal
6. Protest against the downsizing process
2.78
3.19
0.70
0.78
0.81
0.81
0.17
0.16
70.15
70.10
0.39
0.27
0.34
0.26
0.70
Loyalty
7. Organizational commitment
8. Work intensity
2.88
4.53
0.81
0.48
0.72
0.76
70.09
70.04
0.15
0.10
70.42
70.19
70.33
70.15
70.51
70.17
70.38
70.02
0.27
Covariates
9. Age
41.04
0.79
10.20
0.41
Ð
Ð
0.03
70.11
0.18
0.08
70.12
70.17
70.21
70.11
70.18
70.07
70.16
70.04
0.21
0.03
0.08
0.13
10. Sexc
70.01
N = 634; for correlations ꢀ 0.09, p50.05.
a
Low insecurity = 0, high insecurity = 1.
Non-members = 0, members = 1.
Male = 0, female = 1.
b
c

176 British Journal of Industrial Relations
we relied on a multivariate analysis of variance procedures (MANCOVA).
Age and sex were used as covariates to avoid biased conclusions on the
effects of job insecurity and union membership. The MANCOVA gener-
ated significant overall effects of job insecurity (F [6,623] = 2.65, p50.01)
and union membership (F [6,623] = 2.19, p50.01) on the coping variables,
but there was no significant multivariate interaction between job insecurity
and union membership (F [6,623] = 0.99, n.s.). While these results suggest
that, overall, both job insecurity and union membership are related to
the coping strategies chosen, they also indicate that union membership
does not moderate the effects of job insecurity. The covariates were related
to the set of dependent variables as reflected in significant multivariate
effects of both age (F [6,623] = 8.52, p50.001) and sex (F [6,623] = 3.88,
p50.01).
The results of the follow-up univariate tests are presented in Table 2 along
with mean values in the coping strategies for the four groups. Among the
exit variables, there was no significant effect of job insecurity on the inten-
tion to withdraw from the organization; on the other hand, job withdrawal
was significantly affected by job insecurity (F [1,628] = 8.23, p50.01). There
was a significant univariate effect of union membership on organizational
withdrawal (F [1,628] = 4.60, p50.05), but not on job withdrawal. While
the mean values in the exit variables were below 3 on the five-point scale
for all four groups, thus suggesting that exit is not a preferred alternative
for any of them, some important differences are apparent. Not only is
considering exit from the job more frequent in the groups characterized
by higher levels of job insecurity, but also, the results indicate that non-
unionized employees are more prone to withdraw from the organization
than their unionized counterparts.
In terms of the voice variables, job insecurity displayed significant univari-
ate effects on both disengagement in the downsizing goal (F [1,628] = 7.50,
p50.01) and protest against the downsizing process (F [1,628] = 9.47,
p50.01). Thus, voice reactions are more frequent among employees report-
ing high job insecurity. Union membership evidenced a significant effect on
disengagement in the downsizing goal (F [1,628] = 8.14, p50.01) but not on
protest against the downsizing process. Hence, compared with non-affiliated
employees, union members expressed lower mean values in one of the voice
variables.
The MANCOVA revealed no significant effects of job insecurity on any
of the loyalty variables. In contrast, there were significant univariate effects
of union membership on both organizational commitment (F [1,628] = 6.99,
p50.01) and work intensity (F [1,628] = 4.17, p50.05). While these results
suggest that groups distinguished by high and low insecurity do not differ
substantially with respect to the levels of organizational commitment or
work intensity reported, they also implicate that unionized employees are
more prone than non-unionized employees to make use of the loyalty option
in an effort to cope with job insecurity and a changing organizational
reality.
# Blackwell Publishers Ltd/London School of Economics 2001.

TABLE 2
Mean Values in Exit, Voice and Loyalty Reactions, and Tests for Direct and Interaction Effects of Union Membership and Job Insecurity
Insecure
members
Secure
members
Insecure
non-members
Secure
non-members
Effects (F)
Covariates (F)
Insecurity
Union
I6U
Age
Sex
Exit
Organizational withdrawal
Job withdrawal
2.37
2.89
1.85
2.29
2.44
2.90
2.46
2.63
2.63
8.23**
4.60*
1.27
2.99
1.12
8.56**
30.32***
13.57***
4.64*
Voice
Disengagement in the downsizing goal
Protest against the downsizing process
2.88
3.20
2.65
3.07
3.17
3.55
2.90
3.15
7.50**
9.47**
8.14**
2.49
0.05
0.67
17.26***
14.59***
1.41
0.21
Loyalty
Organizational commitment
Work intensity
2.83
4.53
2.96
4.55
2.49
4.36
2.72
4.44
3.01
0.53
6.99**
4.17*
0.18
0.26
22.92***
2.65
0.10
9.29**
* p50.05, ** p50.01,*** p50.001; d.f. = 1,628.

178 British Journal of Industrial Relations
6. Discussion and conclusions
The results of this study indicate that perceptions of job insecurity derived
from organizational transformation and retrenchment are related primarily
to coping strategies in the exit and voice domains, but not to loyalty re-
actions. In our data, employees who reported high levels of insecurity were
more inclined than less insecure workers to exit from the job but not from
the organization. This finding partly replicates previous studies that have
found support for a relationship between job insecurity and turnover
intentions (e.g. Brockner 1988; Davy et al. 1997). It also expands previous
research by suggesting that insecure employees do not necessarily consider
leaving the organization, but rather may look for a more secure position
within the organization. Clearly, this finding also argues for the mean-
ingfulness of a conceptual distinction between different forms of withdrawal
cognitions (cf. Cohen 1998).
Our results also expand previous research on the outcomes of job in-
security by suggesting that employees with high insecurity are more prone to
protest against downsizing implementations and to express disengagement
in their goals than their less insecure co-workers. The finding that the level
of job insecurity did not evoke dissimilar loyalty reactions is interesting.
While the theoretical framework derived from Hirschman (1970) suggests
that high insecurity would lead to enhanced loyalty reactions in order to
redress one's attractiveness in the organization (thereby possibly remedying
insecurity), empirical research typically has found job insecurity to be asso-
ciated with impaired loyalty to the organization. The results of our study
take a position in between and, taken together with other studies that have
not found empirical support for a relationship between insecurity and, for
instance, organizational commitment, clearly suggest that the effects of in-
security on various loyalty reactions merit further theoretical as well as
empirical investigation.
The most notable finding of the present study, however, concerns the
relationships between union membership and employee coping strategies.
There seem to be substantial differences between unionized and non-unionized
employees in terms of how they react to job insecurity. In contrast to what
might perhaps be expected, our results suggest that non-affiliated workers
tend primarily to engage in exit and voice coping, while those employees
who belong to a trade union are rather more prone to cope with job
insecurity by expressing their loyalty to the organization. Non-unionized
workers displayed higher mean values than union members in organiz-
ational withdrawal and disengagement in the downsizing goal. On the other
hand, affiliated employees reported higher levels of loyalty as operation-
alized by organizational commitment and work intensity.
These results are congruent with the Shaw et al. (1993) study, which found
union membership to be positively related to organizational commitment
(i.e. loyalty) and positive attitudes towards an ongoing organizational
restructuring (i.e. low levels of voice). They are also in accordance with
# Blackwell Publishers Ltd/London School of Economics 2001.

Job Insecurity and Union Membership 179
previous research which has suggested that unionized workers are more
inclined than their non-unionized counterparts to remain with the organ-
ization (Bender and Sloane 1999; D'Amico 1984). One potential explanation
of these results, derived from research on dual commitment to company and
union, is that a co-operative industrial relations climate Ð a characteristic
of the Swedish industrial relations system Ð will facilitate employee ex-
pressions of loyalty to both the union and the employer (cf. Angle and Perry
1986; Magenau et al. 1988). Thus, a most interesting finding of the present
study, which also receives support from the work of Freeman and Medoff
(1984) and Hirschman (1986), is that union membership not only may
reduce exit but also may lead to increased loyalty (and productivity). Such
beneficial consequences of union presence have obvious practical implica-
tions for management.
With respect to the finding that unionized employees, in contrast to the
non-members, were less inclined to use the voice option by expressing dis-
engagement in the downsizing goal, it could be that our operationalization
failed to capture the essence of voice. Given the difficulties in assessing voice
experienced in previous studies (e.g. Withey and Cooper 1989), more re-
search is needed to uncover how best to capture individual voice. A more
substantial explanation to our finding, however, is that affiliated employees
benefit from collective forms of voice expressed by their unions, and hence
feel less need to engage in individualized forms. Freeman and Medoff (1984)
argue that the presence of a union on the workplace (a voice function in
their framework) generally has a positive effect on organizational prod-
uctivity by putting `pressure on management to tighten job-production
standards and accountability in order to preserve profits in the face of
higher wages' (p. 14). In what may be termed the `organizational success of
unionism' argument, they maintain:
Economic theorists of all persuasions have increasingly recognized that unions'
ability to enforce labor agreements, particularly those with deferred claims, creates
the possibility for improved labor contracts and arrangements and higher
economic efficiency. (Freeman and Medoff 1984: 11)
It should be recalled, however, that, even though we found support for
positive effects of union membership, this variable did not moderate the
effects of job insecurity. While the absence of such a buffering effect of
union support has been observed also in a study that predicted employee
well-being (Dekker and Schaufeli 1995), it is possible that the non-
significance of the interaction between job insecurity and union membership
is a statistical artifact rather than a substantial finding. The small number of
individuals in the non-member groups indicates that the present study had
little power to identify a significant interaction effect Ð even if there were in
fact a substantial one.2
Nevertheless, whether as a moderator of job insecurity or as a direct
predictor of coping strategies, union membership appears to have some
crucial consequences. Regardless of their level of job insecurity, union
# Blackwell Publishers Ltd/London School of Economics 2001.

180 British Journal of Industrial Relations
members were less inclined to make use of the exit and voice options than
their non-unionized co-workers, and they more typically expressed loyalty
to the organization. Given that there exist hardly any studies on the
potential differences in terms of job insecurity reactions between unionized
and non-unionized employees, our results could be taken as preliminary
evidence for such positive effects. However, the high degree of unionization
in Sweden calls the generalizability of this issue into question. Clearly,
therefore, our results need replication in other industrial relations contexts
before any firm conclusions can be drawn regarding the positive effects of
union membership on how employees cope with job insecurity.
Final version accepted 20 October 2000.
Acknowledgements
The research reported here was supported by a grant from the Swedish
Council for Work Life Research. An earlier version of the paper was pre-
sented at the conference on `Plant Closures and Downsizing in Europe: How
Do European Workers and the Trade Union Movement React to Job In-
security and the Threat of Unemployment?' at Leuven, Belgium, January 1999.
Notes
1. It may be argued that it is the median value rather than the mean that should be
used as cutoff point. However, research indicates that mean and median splits
tend to produce similar results (Sverke and Sjoberg 1994), and all results reported
in this paper remain essentially the same using a median split (median = 1.60).
2. This reasoning applies to all findings of the present study. While the conservative
MANCOVA procedures minimize the risks of obtaining significance for true non-
relationships (type I error), the small proportion of non-unionized employees in
the sample, combined with the conservative statistical analyses, involve a risk of
not identifying true relationships (type II error).
References
Allen, N. J. and Meyer, J. P. (1990). `The measurement and antecedents of affective,
continuance and normative commitment to the organization'. Journal of Occu-
pational Psychology, 63: 1±18.
Angle, H. L. and Perry, J. L. (1986). `Dual commitment and labor±management
relationship climates'. Academy of Management Journal, 29: 31±50.
Arnold, H. J. and Feldman, D. C. (1982). `A multivariate analysis of the
determinants of job turnover'. Journal of Applied Psychology, 67: 350±60.
Ashford, S. J., Lee, C. and Bobko, P. (1989). `Content, causes, and consequences of
job insecurity: a theory-based measure and substantive test'. Academy of Manage-
ment Journal, 4: 803±29.
Barling, J. and Kelloway, K. E. (1996). `Job insecurity and health: the moderating
role of workplace control'. Stress Medicine, 12: 253±9.
# Blackwell Publishers Ltd/London School of Economics 2001.

Job Insecurity and Union Membership 181
Bender, K. A. and Sloane, P. J. (1999). `Trade union membership, tenure and the
level of job insecurity'. Applied Economics, 31: 123±35.
Brockner, J. (1988). `The effects of work layoffs on survivors: research, theory, and
practice. In B. M. Staw and L. L. Cummings (eds.), Research in Organizational
Behavior, Vol. 10. Greenwich, Conn.: JAI Press.
ÐÐ Tyler, T. R. and Cooper-Schneider, R. (1992). `The influence of prior commit-
ment to an institution on reactions to perceived unfairness: the higher they are, the
harder the fall'. Administrative Science Quarterly, 37: 241±61.
Brown, S. P. and Leigh, T. W. (1996). `A new look at psychological climate and its
relationships to job involvement, effort, and performance'. Journal of Applied
Psychology, 81: 358±68.
Brown Johnson, N., Bobko, P. and Hartenian, L. S. (1992). `Union influence on
local union leaders' perceptions of job insecurity: an empirical test'. British Journal
of Industrial Relations, 30: 45±60.
Burke, R. J. and Nelson, D. (1998). `Mergers and acquisitions, downsizing, and
privatization: a North American perspective'. In M. K. Gowing, J. D. Kraft and
J. C. Quick (eds.), The New Organizational Reality: Downsizing, Restructuring, and
Revitalization. Washington, DC: American Psychological Association, pp. 21±54.
Caplan, R. D., Cobb, S., French, J. R. P. Jr., Van Harrison, R. V. and Pinneau, S. R.
Jr. (1975). Job Demands and Worker Health: Main Effects and Occupational
Differences. Washington, DC: US Department of Health, Education, and Welfare.
Cohen, A. (1998). `An examination of the relationship between work commitment
and work outcomes among hospital nurses'. Scandinavian Journal of Management,
14: 1±17.
Cook, J. and Wall, T. D. (1980). `New work attitude measures of trust, organiz-
ational commitment and personal need non-fulfillment'. Journal of Occupational
Psychology, 53: 39±52.
Crockett, G. and Hall, K. (1987). `Salaried professionals and union membership: an
Australian perspective'. Journal of Industrial Relations, 29: 49±65.
D'Amico, R. (1984). `Industrial feudalism reconsidered'. Work and Occupations, 11:
407±37.
Davy, J. A., Kinicki, A. J. and Scheck, C. L. (1997). `A test of job security's direct
and mediated effects on withdrawal cognitions'. Journal of Organizational
Behavior, 18: 323±49.
Dekker, S. W. A. and Schaufeli, W. B. (1995). `The effects of job insecurity on
psychological health and withdrawal: a longitudinal study'. Australian Psychol-
ogist, 30: 57±63.
Dworkin, J. B., Feldman, S. P., Brown, J. M. and Hobson, C. J. (1988). `Workers'
preferences in concession bargaining'. Industrial Relations, 27: 7±20.
Farrell, D. (1983). `Exit, voice, loyalty, and neglect as responses to job dissatisfaction:
a multidimensional scaling study'. Academy of Management Journal, 26: 596±607.
Freeman, R. B. and Medoff, J. L. (1984). What Do Unions Do? New York: Basic
Books.
Greenhalgh, L. and Rosenblatt, Z. (1984). `Job insecurity: toward conceptual
clarity'. Academy of Management Review, 3: 438±48.
Guest, D. E. and Dewe, P. (1991). `Company or trade union: which wins workers
allegiance? A study of commitment in UK electronic industry'. British Journal of
Industrial Relations, 29: 75±96.
Hartley, J., Jacobson, B., Klandermans, and van Vuuren, T. (1991). Job Insecurity:
Coping with Jobs at Risk. London: Sage.
# Blackwell Publishers Ltd/London School of Economics 2001.

182 British Journal of Industrial Relations
Hitt, M. A., Keats, B. W., Harback, H. F. and Nixon, R. D. (1994). `Rightsizing-
building and maintaining strategic leadership: a long-term competitiveness'.
Organizational Dynamics, 23: 18±32.
Hirschman, A. O. (1970). Exit, Voice, and Loyalty: Responses to Decline in Firms,
Organizations, and States. Cambridge, Mass.: Harvard University Press.
ÐÐ (1986). Rival Views of Market Society. New York: Viking.
Hollenbeck, J. R., Williams, C. R. and Klein, H. J. (1989). `An empirical examin-
ation of the antecedents of commitment to difficult goals'. Journal of Applied
Psychology, 74: 18±24.
Howard, A. (1995). The Changing Nature of Work. San Francisco: Jossey-Bass.
Kjellberg, A. (1998). `Sweden: restoring the model?' In A. Ferner and R. Hyman
(eds.), Changing Industrial Relations in Europe. Blackwell, Oxford, pp. 74±117.
Kuhnert, K. V. and Vance, R. J. (1992). `Job insecurity and moderators of the
relation between job insecurity and employee adjustment'. In J. C. Quick, L. R.
Murphy and J. J. Hurrell (eds.), Stress and Well-being at Work: Assessments and
Interventions for Occupational Mental Health. Washington, DC: American
Psychological Association, pp. 48±63.
Latack, J. C. and Dozier, J. B. (1986). `After the axe falls: job loss on a career
transition'. Academy of Management Review, 11: 375±92.
Lim, V. K. G. (1996). `Job insecurity and its outcomes: moderating effects of work-
based and nonwork-based social support'. Human Relations, 49: 171±94.
Magenau, J. M., Martin, J. E. and Peterson, M. M. (1988). `Dual and unilateral
commitment among stewards and rank-and-file union members'. Academy of
Management Journal, 31: 359±76.
March, J. G. and Simon, H. A. (1958). Organizations. New York: John Wiley.
Mellor, S. (1992). `The influence of layoff severity on post-layoff union commitment
among survivors: the moderating effect of the perceived legitimacy of a layoff
account'. Personnel Psychology, 45: 579±600.
Mobley, W. H., Griffeth, R. W., Hand, H. H. and Meglino, B. M. (1979). `Review
and conceptual analysis of the employee turnover process'. Psychological Bulletin,
86: 493±522.
Mowday, R. T., Steers, R. M. and Porter, L. W. (1979). `The measurement of
organizational commitment'. Journal of Vocational Behavior, 14: 224±47.
Rosenblatt, Z. and Ruvio, A. (1996). `A test of a multidimensional model of job
insecurity: the case of Israeli teachers'. Journal of Organizational Behavior, 17:
587±605.
Roskies, E. and Louis-Guerin, C. (1990). `Job insecurity in managers: antecedents
and consequences'. Journal of Organizational Behavior, 11: 345±59.
Shaw, J. B., Fields, W. M., Thacker, J. W. and Fisher, C. D. (1993). `The availability
of personal and external coping resources: their impact on job stress and employee
attitudes during organizational restructuring'. Work and Stress, 7: 229±46.
Sjoberg, A. and Sverke, M. (1996). `Predicting turnover intention among nurses: the
role of work values'. In V. V. Baba (ed.), Work Values and Behaviour: Research
and Applications. Montreal: International Society for the Study of Work and
Organizational Values, pp. 213±23.
Sverke, M. and Sjoberg, A. (1994). `Dual commitment to company and union in
Sweden: an examination of predictors and taxonomic split methods'. Economic
and Industrial Democracy, 15: 531±64.
Withey, M. J. and Cooper, W. H. (1989). `Predicting exit, voice, loyalty, and neglect'.
Administrative Science Quarterly, 34: 521±39.
# Blackwell Publishers Ltd/London School of Economics 2001.