Content uploaded by Steven E. Abraham
Author content
All content in this area was uploaded by Steven E. Abraham on Feb 19, 2016
Content may be subject to copyright.
Employee Responsibilities and Rights Journal, Vol. 17, No. 4, December 2005 (
C
2005)
DOI: 10.1007/s10672-005-9049-0
The Impact of Union Membership on Intent to Leave:
Additional Evidence on the Voice Face of Unions
1
Steven E. Abraham,
2,4
Barry A. Friedman,
2
and Randall K. Thomas
3
The impact of union membership on employees’ intent to leave their jobs was examined to
test the effect of unions’ “voice” face. Regression analyses showed a significant, negative
relationship between union membership and employees’ intent to leave. In addition, the
data revealed significant interactions between union membership and job satisfaction and
between union membership and organizational commitment. Dissatisfied nonunion mem-
bers are much more likely to intend to leave their jobs than are union members. Similarly,
nonunion members with low organizational commitment are much more likely to intend to
leave their jobs than are union members. Both of these results support the conclusion that
union membership reduces employees’ intention to leave their jobs, and provides evidence
that the voice face of unions matter.
KEY WORDS: unions; satisfaction; commitment; intention to leave.
INTRODUCTION
In their seminal work What Do Unions Do? (1984), Freeman and Medoff point out that
unions in the US have two “faces,” a monopoly face and a collective voice face. Unions’
monopoly face is the one with which most people are familiar (Kaufman, 2002). Espoused
as far back as 1944 by Simons, the monopoly face theory argues that unions are similar to
monopolies, in that they distort the labor market outcomes from what they would be under
perfect competition (i.e., without unions). Through the monopoly face, unions negotiate
wages that often are above the equilibrium wage (Lewis, 1963), may force management into
using more employees than needed (Simler, 1962), negotiate restrictive work rules that tend
to limit management’s ability to operate the enterprise most efficiently (Addison, 1984)
and force management to adopt less efficient personnel practices (Kaufman & Kaufman,
1987).
1
An earlier version of this research was presented at the 2004 Association on Employee Practices and Principles
Conference, Ft. Lauderdale, Florida.
2
School of Business, SUNY-Oswego, Oswego, New York.
3
Harris Interactive, Rochester, New York.
4
To whom correspondence should be addressed at School of Business, SUNY-Oswego, Oswego, New York 13126;
e-mail: abraham@oswego.edu.
201
0892-7545/05/1200-0201/0
C
2005 Springer Science+Business Media, Inc.
202 Abraham, Friedman, and Thomas
The voice face is based on the exit-voice theory from Hirschman (1970). As applied
to the workplace, exit-voice theory posits that employees can address dissatisfaction with
the workplace in one of the two ways: by expressing their dissatisfaction (i.e., voice) or by
leaving the firm (i.e., exit). Since unions provide employees with more effective voice than
they have in non-union workplaces, there will be less turnover among union members than
among nonunion members. According to proponents of the voice face of unions, there will
be more efficient labor market outcomes in unionized firms.
This paper provides a test of unions’ voice face by investigating whether employees
who are members of unions have less intention to leave their firms than nonmembers.
Further, this study examines the relationship between union membership and intent to
leave across different levels of job satisfaction and organizational commitment. If union
members have less intention to leave their firms than nonunion employees, the voice
face of unions will be supported as well. The rest of this paper is organized as follows:
Section 2 contains a literature review, Section 3 discusses the hypotheses to be tested,
Section 4 describes the method, Section 5 provides and discusses the results and Section 6
concludes.
LITERATURE REVIEW
Much previous literature has investigated issues associated with the voice face of
unions. As stated by Addison and Belfield, the “...impact of the collective voice model on
research into the economic consequences of unions is hard to exaggerate” (2004, p. 563).
Recently, a number of papers addressing the issue appeared in Journal of Labor Research
(Bennett, 2004; Addison & Belfield, 2004; Hirsch, 2004). These papers summarize much
of the literature that has examined many of the propositions offered by Freeman and
Medoff (1984) and offer new evidence on the question of unions’ voice face as well.
For example, Hirsch (2004) examines union effects on productivity, growth, profits and
investment, relating his findings to the conclusions presented by Freeman and Medoff.
Similarly, Addison and Belfield (2004) summarize many studies that examine the effects of
unions on productivity, productivity growth, profitability and investments in tangible and
intangible capital. In addition, the authors also summarize much of the literature dealing
with unionism and turnover and add their own evidence on that issue.
Research addressing unions’ voice face contends that union voice manifests itself in
a number of different ways, such as union workers being more productive, union firms
having more wage equality, etc. Of greatest relevance to this paper, however, proponents
of unions having a voice face argue that unionized firms will experience less turnover than
nonunion firms.
Several of the early studies addressing the impact of unionization on turnover are
summarized in Chapter 6 of What Do Unions Do? (Freeman & Medoff, 1984), and all
of these studies point to the same general conclusion: holding wages constant, turnover is
reduced and employee tenure is increased among unionized workers. Depending on the
data set involved, the results show that the reduction in quits under unionization is between
31 and 65%, while the increase in tenure varies from 23 to 32%. In more detailed analyses,
the authors determine that the reduction in turnover under unions is more attributable to
unions’ voice face than to the monopoly face. For example, using data from the Panel Study
of Income Dynamics, 1971–1979, they report that a 20% increase in wages reduces quits
The Impact of Union Membership on Intent to Leave 203
by 8% (the monopoly face), while the voice face, controlling for wages, reduces quits by
at least 31%.
In a study of the trucking industry that casts doubt on the voice face argument, Delery
et al. (2000) reports that, while reduced turnover is associated with unionism, the majority
of that effect is due to unions’ monopoly face. In other words, the reason union members
are less likely to quit than nonunion members is because unionized employees receive
greater wages than their nonunion counterparts. This study casts doubt on whether unions
have a voice face. On the other hand, a study of the telecommunications industry by Batt
et al. (2002) shows that quits are between 4.9 and 6.5% less in union than nonunion
establishments.
Rees (1991) found that teachers in school districts with the strongest grievance proce-
dures had lower probabilities of quitting than teachers covered by weaker procedures. He
concluded that these results support the voice theory by contending that these procedures
should be considered as a proxy for the voice provided by unions. Groothuis (1994) looked
at the effects of union status and firm size on turnover and found that unionization lowered
both dismissal rates and quit rates in large firms but not in small firms. He concluded that
his results support the voice face of unions as well.
Even more recently, Iverson and Currivan (2003) tested the exit-voice hypothesis
by examining data on 674 unionized teachers from 405 schools. The dependent variable
was voluntary turnover and the independent variables of interest were job satisfaction
and union participation. The authors used activity of each teacher in the union to provide
a measure of union participation and employed event history analysis to investigate the
effects of union participation and job satisfaction on turnover. In addition, the authors
investigated whether union participation and job satisfaction interacted to predict turnover.
Iverson and Currivan’s results showed that both job satisfaction and union participation
had significant, negative effects on turnover. Further, the authors found an interaction
between union participation and job satisfaction. Specifically, union membership had a
significant, negative effect on turnover for employees who were both very satisfied and very
unsatisfied.
Research has shown support for unions having a voice face in other countries as well.
Lower quit rates and increased tenure due to unions have been found in Japan (Muramatsu,
1984; Osawa, 1989); Canada (Kupferschmidt & Swidensky, 1989), the United Kingdom
(Elias & Blanchflower, 1989; Addison & Belfield, 2004) and Australia (Kornfeld, 1990;
Miller & Mulvey, 1991).
In sum, a substantial body of literature has confirmed “The Two Faces of Unionism”
theory first espoused by Freeman and Medoff as far back as 1979 and popularized
in 1984. Unions provide collective voice to employees, and this collective voice will
be associated with less turnover and greater tenure in union firms than nonunion
firms.
HYPOTHESES
This paper provides a new test of the voice face of unions with one major difference.
Most previous investigations of the issue have examined the impact of unionism on the
probability that a union employee exits the firm (see, e.g., Freeman & Medoff, 1984,
p. 96; Bemmels, 1997). This paper, however investigates the impact of union membership
204 Abraham, Friedman, and Thomas
on employees’ intent to leave, defined here as a conscious and deliberate decision to
voluntarily leave their organization (Tett & Meyer, 1993).
The distinction between intention to leave and actual turnover in this study is not triv-
ial. As pointed out by Lee and Mitchell (1999) and Lee et al. (1999), the turnover process
(i.e., an employee’s decision to leave a job) is a complex one that may proceed through a
number of different psychological paths. Also, even if employees intend to leave their jobs,
other factors beyond their control may prevent them from doing so (e.g., the labor mar-
ket). For example, during economic downturns and high unemployment levels, employees
perceive fewer alternative employment opportunities. When economic conditions improve,
the quit rate tends to increase as employees see alternative and more attractive employment
opportunities (Carsten & Spector, 1987). By testing whether union membership moderates
the relationship between job satisfaction and employees’ intention to leave, we can get a
more accurate test of whether the collective voice provided by unions has an effect on the
workplace.
In addition to employees’ union membership, this paper includes two other de-
terminants of employees’ intent to leave their jobs: job satisfaction and organizational
commitment. Previous research has shown that there are negative relationships between
job satisfaction and employees’ intent to leave and between organizational commitment
and employees’ intent to leave (Tett & Meyer, 1993; Meyer et al., 2002). For exam-
ple, in a meta-analysis of 178 independent samples, Tett and Meyer (1993) concluded
that satisfaction and commitment contribute independently to the prediction of turnover
intentions, and Meyer et al. (2002) found that commitment correlated negatively with
turnover intentions. Hammer and Avgar (2005) reviewed the research pertaining to job
satisfaction, organizational commitment, turnover and unions and concluded that stud-
ies of organizational commitment have paid scant attention to the role of union status
(p. 257).
On the basis of the literature discussed above, the following hypotheses were tested:
H1: Union members will have less intention to leave their organization than will non-union
members.
This is the basic exit-voice hypothesis that was originally articulated by Freeman and
Medoff, and tested in the literature reviewed above.
H2: The relationship between job satisfaction and intention to leave will be stronger for
non-union members than for union members.
Locke (1976) defined job satisfaction as a pleasurable state resulting from the appraisal
of one’s job as attaining important outcomes or values, and meeting basic needs. Models
of employee intention to leave often include job satisfaction as a predictor of employee
turnover (Hammer & Avgar, 2005). The exit-voice hypothesis claims that the voice mech-
anism, unions provide for employees will diminish the negative relationship between job
satisfaction and intention to leave. Because they do not have a collective voice to address
their dissatisfaction, nonunion employees are more likely to want to leave their jobs when
they are dissatisfied with their jobs. A union membership and job satisfaction interaction
is predicted, where the job satisfaction–itention to leave relationship will be stronger for
nonunion employees.
The Impact of Union Membership on Intent to Leave 205
H3: The relationship between organizational commitment and intention to leave will be
stronger for non-union employees than for union members.
Mowday et al. (1979) defined organizational commitment as the extent that employees
identify with and wish to remain with their organization. Generally, employees who iden-
tify with their organizations have fewer intentions to leave as compared to employees who
are less committed to their organizations (Hammer & Avgar, 2005), but it is predicted that
union membership will influence this relationship. Similar to Hypothesis 2, it is predicted
that union employees will have less desire to leave their jobs even if they are less com-
mitted to their organization. A union membership–organizational commitment interaction
is predicted, where the relationship between organizational commitment and intention to
leave will be weaker for union members.
METHOD
Sample
The data were collected in June 2001 as part of a national Harris Poll
employee
satisfaction survey. Survey respondents were members of the Harris Poll Online multi-
million member panel who were 18 years of age or older, from the US. Data obtained
from Harris Poll Online respondents have been found to provide results equivalent to
those obtained in random digit dial telephone surveys (Krosnick et al., 2005; Thomas
et al., 2004). Potential respondents were randomly selected by strata (age, gender, and
region of country) and invited by e-mail to participate in a survey on attitudes to-
ward work. Of the 10,434 individuals that responded to the survey, 7,744 indicated
that they were not planning to retire in the next few years and were included in the
study. Employees close to retirement were excluded from the study, as their responses
would not be meaningful for the study. Out of the 7,744 respondents, 7,068 (91.3%)
held a full-time position while 676 (8.7%) held a part-time position in their companies.
Table I reports descriptive statistics for the sample. Males made up 53.4% of the sample,
14.8% were minority, and the average age was 39.94 years (SD = 10.10). Respondents
from organizations of fewer than 500 employees constituted 45.8% of the sample and
17.9% were from organizations of greater than 10,000. Respondents who reported hav-
ing annual incomes of between $25,000 and 75,000 were 63.2% of the sample and
59.6% reported 1–10 years with their company and 23.9% indicated 10 or more years.
The sample consisted of secretarial/clerical (14.2%), production/hourly (26.8%), pro-
fessional/technical (31.4%), lead worker/team leader (6.1%), supervisor (5.4%), man-
ager/senior manager (14.6%), and senior leadership (1.5%) employees. The respondents
were from the service (50.8%), manufacturing (9.5%), transportation/communication
(8.6%), publication administration/military (6.7%), finance (6.7%), and retail (5.9%)
industries.
Table I also presents the same descriptive statistics for the union and nonunion sub-
samples separately. These statistics are similar to prior research regarding differences be-
tween union and nonunion employees. For example, union members tended to be employed
by larger organizations than nonunion members (Z =−10.43, p<.001) and union mem-
bers tended to have lower-level jobs (Z =−11.07, p<001). Related to the issues being
206 Abraham, Friedman, and Thomas
Table I. Sample Descriptive Characteristics
Total Non-union Union
Variable Category N % N % N %
Education Less than high school 10 0.1 8 0.1 2 0.2
Completed some high school 76 1.0 61 0.9 15 1.5
High school graduate or equivalent 947 12.2 804 12.0 143 13.9
Completed some college 2847 36.8 2469 36.8 378 36.7
College graduate 2273 29.4 2050 30.5 223 21.7
Some graduate school 587 7.6 481 7.2 106 10.3
Completed graduate school 1004 13.0 842 12.5 162 15.7
Minority status Non-minority 6600 85.2 5732 85.4 868 84.4
Minority 1144 14.8 983 14.6 161 15.6
Gender Male 4137 53.4 3595 53.5 542 52.7
Female 3607 46.6 3120 46.5 487 47.3
Total household Less than 15,000 249 3.6 234 3.9 15 1.6
income ($) 15,000–24,999 690 10.0 634 10.6 56 6.0
25,000–34,999 1071 15.5 939 15.7 132 14.2
35,000–49,999 1510 21.9 1258 21.1 252 27.1
50,000–74,999 1780 25.8 1485 24.9 295 31.8
75,000–99,999 876 12.7 759 12.7 117 12.6
100,000–124,999 392 5.7 351 5.9 41 4.4
125,000–149,999 157 2.3 143 2.4 14 1.5
150,000–199,999 102 1.5 96 1.6 6 0.6
200,000–249,999 41 0.6 40 0.7 1 0.1
250,000 or more 34 0.5 34 0.6 0 0.0
Company size Less than 500 3549 45.8 3249 48.4 300 29.2
(employees) 500–999 843 10.9 697 10.4 146 14.2
1,000–2,999 945 12.2 775 11.5 170 16.5
3,000–9,999 1019 13.2 846 12.6 173 16.8
More than 10,000 1388 17.9 1148 17.1 240 23.3
Job category Secretarial/Clerical 1103 14.2 971 14.5 132 12.8
Production/hourly employee 2072 26.8 1642 24.5 430 41.8
Lead worker/Team leader 470 6.1 399 5.9 71 6.9
Professional/Technical 2435 31.4 2098 31.2 337 32.8
First-Level supervisor 420 5.4 386 5.7 34 3.3
Manager/Director 957 12.4 937 14.0 20 1.9
Senior manager 170 2.2 167 2.5 3 0.3
Senior leadership 117 1.5 115 1.7 2 0.2
Years employed Less than 6 months 522 6.7 489 7.3 33 3.2
by current 6 months to less than 1 year 756 9.8 703 10.5 53 5.2
employer 1 year to less than 2 years 1096 14.2 1026 15.3 70 6.8
2 years to less than 4 years 1479 19.1 1332 19.8 147 14.3
4 years to less than 6 years 1027 13.3 901 13.4 126 12.2
6 years to less than 8 years 600 7.7 517 7.7 83 8.1
8 years to less than 10 years 417 5.4 337 5.0 80 7.8
More than 10 years 1847 23.9 1410 21.0 437 42.5
Industry Services 3934 50.8 3409 50.8 525 51.0
Agricultural 46 0.6 42 0.6 4 0.4
Finance 522 6.7 519 7.7 3 0.3
Transportation/Communication 663 8.6 538 8.0 125 12.1
Construction 137 1.8 113 1.7 24 2.3
Manufacturing 735 9.5 647 9.6 88 8.6
Public administration/Military 522 6.7 374 5.6 148 14.4
Mining 13 0.2 12 0.2 1 0.1
Retail 459 5.9 424 6.3 35 3.4
Wholesale trade 74 1.0 71 1.1 3 0.3
Other industry 639 8.3 566 8.4 73 7.1
The Impact of Union Membership on Intent to Leave 207
tested in this paper, union members tended to have longer tenure than nonunion members
(Z =−16.30, p<.001). Forty-two percent of union members had been employed by their
current employer for over 10 years, while the percentage was only 21.0% for nonunion
members. There were no significant differences for industry between union and nonunion
members.
Measures
Control Variables
We measured gender, age, education, minority status, household income, years em-
ployed by current employer, company size (number of employees at company) and indus-
try. Previous research on unionism and turnover has found these variables to be related
to turnover. For example, turnover rates have been found to be related to firm size and
unionization and their interaction (Groothuis, 1994). It appears that unionization low-
ers the likelihood of an individual quitting or being dismissed at large firms, but that
layoffs may be more likely in especially large and small firms. Faber and Saks (1980)
found that higher wage earners were less likely to perceive union advantages than lower
wage earners, a finding replicated often. Faber and Saks (1980) also found that black
employees were more likely to vote for unionization, and that older employees were less
likely to vote for unionization. Business sectors may also differ with respect to labor
supply, outsourcing, and availability of alternative employment opportunities, and thus
influence employees’ intention to leave. As a result, industry was included as a control
variable.
Dummy codes were used for gender (female, 1; male, 0), minority status (minor-
ity, 1; non-minority, 0), and Job Category (secretarial/clerical, production or hourly, lead
worker/team leader, professional/technical, first-level supervisor, manager/director, and
senior leader). The authors also dummy coded industry using Standard Industry Codes
(SIC) as follows: services, agricultural, finance, transportation/communication, construc-
tion, manufacturing, public administration/military, mining, retail, and wholesale trade.
Household income, age, and education were treated as continuous variables.
Independent Variables
For the primary independent variable of interest, union membership, respondents were
asked “Do you belong to a union?” We adjusted union membership so that non-members
had a value of −1 and members had a value of 1 (setting either group to 0 would not prop-
erly capture variance due to category membership in an interaction term). Job satisfaction
was measured by asking respondents “Using a scale of 1 to 10 where ‘1’ means ‘Poor’
and ‘10’ means ‘Excellent,’ overall, how would you rate your satisfaction with your job?”
Organizational commitment was measured with the question “How committed are you
personally to your company’s success?” with five category fully anchored scale with the
response categories of “Not at all committed,” “Slightly committed,” “Somewhat commit-
ted,” “Committed,” and “Very committed.” To allow for proper treatment of the interaction
terms in the subsequent regression analysis, we range-adjusted the job satisfaction and
organizational commitment measures to range on a scale of 0–1.
208 Abraham, Friedman, and Thomas
Table II. Mean Employee Intention to Leave, Company Satisfaction, Organizational Commitment
for Union (N = 1029) and Non-Union (N = 6715) Employees
Non-union Union member
Variable Mean SD Mean SD Difference t
Intention to leave
a
4.07 3.29 2.87 2.81 1.20 11.10
∗∗∗
Job satisfaction
b
6.77 2.28 6.53 2.35 0.24 3.17
∗∗
Organizational commitment
c
3.89 1.07 3.79 1.08 0.10 2.68
∗∗
a
Intention to leave was measured using a 10 category scale where “1” meant “Very Unikely” and “10”
meant “Very Likely.”
b
Job satisfaction was measured using a 10 category scale where “1” was “Poor” and “10” was
“Excellent.”
c
Organizational commitment was measured using a 5 category scale ranging from “Notatallcom-
mitted”to“Very committed.”
∗∗
p<.01;
∗∗∗
p<.001.
Dependent Variable
The dependent variable in the regression analysis was employees intention to leave
their jobs. Respondents were asked, “Using a scale of 1 to 10 where ‘1’ means Very Unlikely
and ‘10’ means Ver y L ik el y, how likely are you to voluntarily leave your company within
the next year?”
Table II presents descriptive statistics for the dependent and independent variables of
primary interest. The mean intention to leave was 4.07 for nonunion members and 2.87 for
union members. The difference (t = 11.10) was highly significant (p<.001). Nonunion
members’ reported satisfaction was 6.77, while the reported satisfaction for union members
was 6.53. It is interesting to note that the difference in satisfaction between union members
and nonunion members was only .24 on a scale of 1–10. While this difference was significant
(t = 3.17, p<.002), it was rather small compared to the differences reported in many other
studies (Bryson et al., 2004; Bender & Sloan, 1998). These results are consistent with the
observation by Hammer and Avgar (2005), who conclude that there remains a small but
significant negative union effect on job satisfaction after job and workplace characteristics
were accounted for. Finally, nonunion members’ organizational commitment was a modest
amount more (3.89 versus 3.79) though significant (t = 2.68, p<.007).
Regression Analysis
A simultaneous regression analysis was utilized to test the hypotheses stated above.
Intention to leave, the dependent variable, was regressed on the control variables and the
independent variables of interest. Following procedures on the treatment of interactions
in Cohen et al. (2003), we dummy coded the categorical variable of union membership
and mean centered the continuous variables of job satisfaction and organizational commit-
ment to allow us to also enter interaction terms as predictors into the regression equation
without introducing nonessential multicollinearity. In addition, since approximately 12.5%
of respondents declined to answer, we inserted the mean value (on an 11-category scale)
for analytic purposes. The coefficients on each variable were examined in light of the
hypotheses stated above.
The Impact of Union Membership on Intent to Leave 209
Table III. Intention to Leave Regressed on Union Membership, Company Satisfaction,
Organizational Commitment, and Control Variables
Variables β t
Control variables
a
Gender −0.02 −2.67
∗∗
Age −0.09 −8.85
∗∗∗
Education 0.04 4.32
∗∗∗
Minority status 0.02 2.95
∗∗
Total household income 0.00 0.60
Number of years employed by current employer −0.14 −13.94
∗∗∗
Number of employees at company −0.06 −6.80
∗∗∗
Services −0.01 −0.62
Agricultural −0.01 −0.62
Finance −0.01 −0.37
Transportation/communication −0.02 −1.99
∗
Construction −0.01 −0.34
Manufacturing −0.03 −2.45
∗
Public administration/military −0.05 −4.17
∗∗∗
Mining −0.02 −2.89
∗∗
Retail −0.01 −0.21
Independent variables
Union membership −0.09 −9.22
∗∗∗
Satisfaction −0.41 −36.34
∗∗∗
Organizational commitment −0.13 −11.74
∗∗∗
Interactions
Job satisfaction × union membership 0.06 5.45
∗∗∗
Organizational commitment × union 0.03 2.49
∗
Membership
a
All job category dummy variables were non-significant (secretarial/clerical, production, team
leader, professional/technical, supervisor, manager, senior manager, and leader). To simplify
the table, job category dummy variables were omitted from the table. All job category
dummy variable β weights, t statistics, and significance levels may be obtained from the first
author.
∗
p<.05;
∗∗
p<.01;
∗∗∗
p<.001.
RESULTS AND DISCUSSION
Table III summarizes the results for the regression analysis. The overall model
was significant (F
(21,7722)
= 155.98, p<.001, adjusted R
2
= .296). For the demographic
control variables, factors associated with higher intention to leave were gender (males),
age (younger employees), those with more education, minorities, those with fewer years
on the job and those employed at companies with fewer employees. With regard to the
industry categories, employees involved in transportation/communication, manufacturing,
public administration, and mining were less likely to intend to leave their jobs.
The coefficients on all the variables of interest were negative and highly significant.
Union membership, job satisfaction, and organizational commitment were each negatively
related to intent to leave.
Both interaction terms (the union membership by job satisfaction interaction and the
union membership by organizational commitment interaction) were significantly related to
intention to leave as well. Relative to union members, non-union employees’ job satisfaction
is more clearly related to their expressed intention to leave, with less satisfaction being more
predictive of intention to leave. Union employees demonstrate a lower correspondence
210 Abraham, Friedman, and Thomas
Fig. 1. Employee intention to leave as a function of union membership and job satisfaction.
between satisfaction and intention to leave. The same results hold true with respect to
organizational commitment. There is a much stronger relationship between intent to leave
and organizational commitment for nonunion than for union.
The results just discussed support each of the three hypotheses stated above. In support
of the exit-voice hypothesis (Hypothesis 1), union membership was negatively related to
intention to leave after the control variables, job satisfaction, and organizational commit-
ment were taken into account. Hypotheses 2 and 3 were supported by the data as well.
Perhaps the support for these hypotheses is best illustrated by Figs. 1 and 2. These figures
illustrate separately for union members and nonunion members, the relationships between
job satisfaction and intent to leave and between organizational commitment and intent to
leave. As shown in Fig. 1, at low levels of job satisfaction, nonunion members are much
more likely to intend to leave their jobs than union members are. As job satisfaction in-
creases, the difference in intent to leave between union members and nonunion members
is much less pronounced. This clearly supports Hypothesis 2.
As shown on Fig. 2, at low levels of organizational commitment, nonunion members
are much more likely to intend to leave their jobs than are union members. As organiza-
tional commitment increases, the difference in intent to leave between union members and
nonunion members is much less pronounced. This clearly supports Hypothesis 3.
These results provide support for the conclusion that unions have a voice face, as
union members clearly have less intention to leave their organizations than do nonmembers.
Interestingly, the results reported here are quite similar to the results reported by Iverson
and Currivan (2003). In their study and ours, unionization (union membership in this paper,
union participation in that paper) and satisfaction were associated with lower turnover.
In addition, there was a negative relationship between unionization and turnover among
employees with low satisfaction and high satisfaction in both papers. Further, these results
add more support to the voice hypothesis because they show that the relationship between
The Impact of Union Membership on Intent to Leave 211
Fig. 2. Employee intention to leave as a function of union membership and organizational commitment.
organizational commitment and intent to leave is stronger for nonunion employees than for
union members.
As pointed out in other studies that have used tenure and quit rates to test the voice face
of unions, however, some might argue that these results do not show that unions have a voice
face. Their argument is that union members have longer tenure and lower quit rates because
union jobs have higher wages and better benefits than nonunion jobs, making unionized
jobs more attractive and giving unionized employees less desire to quit their jobs. While
the data included in this survey do not allow us to rule out conclusively the possibility that
it is union wage premia that gives union members less intention to leave their organizations
than do nonmembers, the results help to reduce that possibility. First, one of the control
variables included in the regression equation was “Total Household Income Before Taxes.”
As expected, the coefficient on that variable was negative and significant (see Table III),
meaning that employees who came from households with higher income had less desire
to leave their jobs. Total household income is largely comprised of wages. Of importance
to this paper is that including total household income in the regression model controls for
the impact of household income on employees’ desire to leave their jobs. While we did
not directly control for the union wage effect, the union wage premia as a source of union
members’ reduced desire to leave their jobs, we can be more confident in concluding that
these results support unions having a voice face.
Further, it must be acknowledged that these results do not necessarily explain why
employees leave their jobs and do not add a great deal to the discussion of the turnover
process overall. As noted above, the current research on turnover has determined that an
employee’s decision to leave a job is a complex one that may proceed through a number
of different psychological paths (Lee & Mitchell, 1994; Lee et al., 1997). Job satisfaction
and organizational commitment are but two of many factors that may contribute to that
212 Abraham, Friedman, and Thomas
decision. In as much as unionism operates to reduce employee turnover by influencing
employee response to job satisfaction, the results shown here may not indicate that unions
and the voice provided by unions do a great deal to influence the overall level turnover in
firms. Nevertheless, that was not the objective of this paper either. Rather, the objective
was to test whether unions provide employees with collective voice in the workplace that
will reduce employees’ intent to leave. Our results show that they do, on the basis of the
empirical support for all three hypotheses tested in this paper.
CONCLUSION
This paper provides further support for unions having a voice face. The reduced
turnover provided by the voice face of unions will benefit both parties to the employment
relationship. In a large sample of employees, union members had less intention to leave
their jobs, after a number of other determinants of the desire to leave were controlled. The
control variables used in the analyses helped to discount other potential explanations and
support unions’ voice as accounting for these results. This has important implications for
organizations.
As employees’ organizational tenure increases, their productivity is likely to improve,
as they acquire firm-specific human capital. In addition, it is costly for firms to have to
replace workers who quit because they have no voice to express their dissatisfaction. Finally,
research indicates that intention to leave is a precursor to further organizational dysfunction
(Koys, 2001; Chen et al., 1998) and can serve as a valuable diagnostic tool to understand
how organizational changes might impact the workforce.
REFERENCES
Addison, J. T. (1984). Trade unions and restrictive practices. In Rosa, J. (Ed.), The Economics of Trade Unions:
New Directions. Boston: Kluwer Nijhoff.
Addison, J. T., & Belfield, C. (2004). Union voice. Journal of Labor Research, 25(4), 583–596.
Batt, R., Colvin, J.S.A., & Keefe, J. (2002). Employee voice, human resource practices, and quit rates: evidence
from the telecommunications industry. Industrial and Labor Relations Review, 55(July 2002), 573–594.
Bemmels, B. (1997). Exit, voice, and loyalty in employment relationships. In Lewin, D., Mitchell, D. J. B., &
Zaidi, M. A. (Eds.), The Human Resource Management Handbook, Part II, 245–259. Greenwich, CT: JAI
Press.
Bender, K. A., & Sloane, P. J. (1998). Job satisfaction, trade unions, and exit-voice revisited. Industrial and Labor
Relations Review, 51(2), 222–241.
Bennett, J. T., et al. (2004). What do unions do? A twenty-five year perspective [Introduction to the symposium].
Journal of Labor Research, 25(3), 334–343.
Bryson, A., Cappellari, L., & Lucifora, C. (2004). Does union membership really reduce job satisfaction? British
Journal of Industrial Relations, 24(3), 439–457.
Carsten, J. M., & Specter, P. E. (1987). Unemployment, job satisfaction, and employee turnover: A Meta analytic
test of the Muchinsky model. Journal of Applied Psychology, 72, 374–381.
Chen, X. P., Chun, H., & Sego, D. J. (1998). The role of organizational citizenship behavior in turnover: Concep-
tualization and preliminary test of key hypotheses. Journal of Applied Psychology, 83, 922–931.
Cohen, J., Cohen, P., West, S. G., & Aiken, L. S. (2003). Applied Multiple Regression/Correlation for the
Behavioral Sciences. 3rd edn., Hillside, NJ: Erlbaum.
Delery, J. E., Gupta, N., Douglas Shaw, J., Jenkins, J. R., & Ganster, M. (2000). Unionization, compensation and
voice effects on quits and retention. Industrial Relations, 39(4), 625–645.
Elias P., & Blanchflower, D. G. (1989). Occupations, earnings, and work histories of young adults: Who gets the
good jobs? Research Paper no. 68. Department of Employment, London.
Faber, H. S., & Saks, D. H. (1980). Why workers want unions: The role of relative wages and job characteristics.
The Journal of Political Economy, 88(2), 349–369.
The Impact of Union Membership on Intent to Leave 213
Freeman, R., & Medoff, J. (1984). What do unions do? New York: Basic Books.
Groothuis, P. A. (1994). Turnover: The implication of establishment size and unionization, Quarterly Journal of
Business and Economics, 33, 41–53.
Hammer, T. H., & Aver, A. (2005). The impact of unions on job satisfaction, organizational commitment and
turnover, Journal of Labor Research, 26(2), 241–266.
Hirsch, B. T. (2004). What do unions do for economic performance? Journal of Labor Research, 25(3).
Hirshman, A. O. (1970). Exit, Voice and Loyalty. Cambridge, MA: Harvard University Press.
Iverson, R. D., & Currivan, D. B. (2003). Union participation, job satisfaction, and employee turnover: An event
history analysis of the exit-voice hypothesis. Industrial Relations, 42, 101–105.
Kaufman, B. E. (2002). Reflections on six decades in industrial relations: An interview with John Dunlop;
Interview. Industrial and Labor Relations Review, 55(2), 324–334.
Kaufman, R. S., & Kaufman, R. T. (1987). Union effects on productivity, personnel practices, and survival in the
automotive parts industry. Journal of Labor Research, 8(4), 333.
Kornfeld, R. (1990). Australia’strade unions wage, finge and job tenure effects. Mimeo Cambridge, Mass, National
Bureau of Economic Research.
Koys, D. (2001). The effects of employee satisfaction, organizational citizenship behavior, and turnover on
organizational effectiveness: A unit-level, longitudinal study. Personnel Psychology, 54, 101–114.
Krosnick, J., Nie, N., & Rivers, D. (2005). Web survey methodologies: A comparison of survey accuracy. Paper
presented at the 60th Annual Conference of the American Association for Public Opinion Research, Miami
Beach, Florida.
Kupferschmidt, M., & Swidensky, R. (1989). Longitudinal estimates of the union effect on wages, wage dispersion,
and pension fringe benefits. University of Guelph, Ontario, Canada. Processed.
Lewis, H. G. (1963). Unionism and Relative Wages in the United States. Chicago: University of Chicago Press.
Locke, E. A. (1976). The nature and cause of job satisfaction. In Dunnette, M. D. (Ed.), Handbook of Industrial
and Organizational Psychology, 1297–1349. Chicago: Rand McNally.
Lee, T. W., & Mitchell, T. R. (1994). An alternative approach: The unfolding model of voluntary employee
turnover. Academy of Management Journal, 19, 51–89.
Lee, T. W., Mitchell, T. R., Holton, B. C., McDaniel, L., & Hill, J. W. (1999). Theoretical development and
extension of the unfolding of voluntary turnover. Academy of Management Journal, 42, 450–462.
Mowday, R. T., Steers, R. M., & Porter, L. W. (1979). The measurement of organizational commitment. Journal
of Vocational Behavior, 14, 224–247.
Meyer, J. P., Stanley, D. J., Herscovitch, L., & Topolnytsky, L. (2002). Affective, continuance and normative
commitment to the organization: A meta-analysis of antecedents, correlates, and consequences. Journal of
Vocational Behavior, 61, 20–52.
Miller, P., & Mulvey, C. (1991). Australian evidence on the exit voice model of the labor market. Industrial and
Labor Relations Review, 45, 44–57.
Muramatsu, K. (1984). The effect of trade unions on productivity in Japanese manufacturing industries. In Aoki
A. (Ed.), The Economic Analysis of Japanese Firms, Elsevier, Amsterdam.
Osawa, M. (1989). The Service ecoinomy and industrial relations in small and medium-size firms in Japan. Japan
Labor Bulletin, 1, 1–10.
Rees, D. I. (1991). Grievance procedure strength and teacher quits. Industrial and Labor Relations Review, 45,
31–43.
Simler, N. J. (1962). The Economics of featherbedding. Industrial and Labor Relations Review, 16(1), 111–121.
Simons, H. (1944). Some reflections on syndicalism Journal of Political Economy, LII, 1–19, excerpted in Richard
A. Lester, Labor Readings on Major Issues, 200–209, New York: Random House.
Tett, R. P., & Meyer, J. P. (1993). Job satisfaction, organizational commitment, turnover intention, and turnover:
Path analysis based on meta-analytic findings. Personnel Psychology, 46, 259–293.
Thomas, R. K., Krane, D., & Taylor, H. (2004). On the convergent validity of attitude measurement in phone
and online surveys. Paper presented at the 59th Annual Conference of the American Association for Public
Opinion Research, Phoenix, Arizona.
