Journal of Retailing 89 (2, 2013) 207–218

Engaging Dissatisfied Retail Employees to Voice Promotive Ideas: The Role

of Continuance Commitment
Jeffrey P. Boichuk a,∗ , Bulent Menguc b,1
a Department of Marketing and Entrepreneurship, C.T. Bauer College of Business, University of Houston, 334 Melcher Hall, Houston, TX 77204, United States
b Department of Management, King’s College London, 150 Stamford St., London SE1 9NH, United Kingdom

Abstract
Retailers rely on employees’ promotive work-related ideas to spur service delivery innovations. Yet a well-established finding in the literature
is that employees refrain from sharing such ideas when they are dissatisfied, and a mountain of evidence suggests that job dissatisfaction is an
epidemic in the retail industry. The intuitive solution would be for supervisors to support these employees; by willfully listening to employees’
problems and providing help, supervisors could expect employees to voice their ideas. However, our results, from a field study and a controlled
experiment, suggest that support should only be provided if a dissatisfied retail employee is also committed to his or her organization out of
necessity. Otherwise, support ends up inducing levels of employee voice that are not significantly different than would be the case had the support
been withheld, yielding the support a misallocation of effort.
Published by Elsevier Inc on behalf of New York University.

Keywords: Job dissatisfaction; Retail employee voice; Continuance commitment; Internal marketing; Employee engagement

The academic literature is rife with studies that point to
the primary drivers of employee satisfaction (e.g., MacKenzie,
Podsakoff, and Ahearne 1998). Yet a recent industry survey,
conducted by Mercer LLC at the end of 2010, suggests that
employee satisfaction remains a major concern in business
today – an estimated 32 percent of U.S. employees are ready
to quit (Needleman 2011). Another survey, conducted glob-
ally, estimates that only 31 percent of employees are engaged
(BlessingWhite 2011). That is, despite our extensive knowledge
of what leads to job satisfaction, a multitude of deleterious
factors – including low pay and limited upward mobility –
remain prevalent in practice (Katzenbach and Santamaria 1999;
Osterman and Shulman 2011; Segal 2012).
This paper departs from the extant retailing literature in the
sense that it investigates employees’ reactions to job dissat-
isfaction, rather than drivers of job satisfaction. The central
research question addressed is: how can retailers stimulate
dissatisfied employees to react to their dissatisfaction in an
active and constructive manner? In particular, our purpose is to
∗
Corresponding author. Tel.: +1 713 743 4579.
E-mail addresses: jpboichuk@uh.edu (J.P. Boichuk),
bulent.menguc@kcl.ac.uk (B. Menguc).
1 Tel.: +44 020 7848 4531.
identify conditions under which retail employees who are dis-
satisfied engage in change-oriented voice behaviors, which we
define to be voluntary communication efforts directed at chal-
lenging the status quo of a work unit through the suggestion of
creative and promotive solutions (Van Dyne and LePine 1998).
We follow previous research and theorize that voice behaviors
are the result of a calculative cognition process, whereby retail
employees consider the costs and benefits of speaking up before
they decide whether to express their constructive ideas and opin-
ions (cf. Detert and Edmondson 2011; Liang, Farh, and Farh
2012). While a primary benefit or motivation behind employee
voice is the hope “to change, rather than to escape from, an
objectionable state of affairs” (Hirschman 1970, p. 30), poten-
tial costs associated with voice include the potential to initiate
interpersonal conflict, the fear of negative feedback, and the risk
of being perceived as a troublemaker (LePine and Van Dyne
1998).
Previous authors acknowledge that the voice literature
remains scant with respect to studies that examine interactive
effects in predicting voice behaviors. For instance, Tangirala
and Ramanujam (2008) state that “the relationships between
voice and some of its antecedents may be more nuanced than
previously assumed” (p. 1190). Others suggest that proximal
or transient influences are likely to impact employees’ voice
behaviors (Liang, Farh, and Farh 2012; Venkataramani and

0022-4359/$ – see front matter. Published by Elsevier Inc on behalf of New York University.
http://dx.doi.org/10.1016/j.jretai.2013.01.001
208 J.P. Boichuk, B. Menguc / Journal of Retailing 89 (2, 2013) 207–218
Tangirala 2010). Still others are proponents of studies that
“develop predictive models that consider the independent and
joint effects of theoretically relevant personal and situational
variables” (LePine and Van Dyne 1998, p. 855).
We respond to these calls for research by investigating how
continuance commitment (i.e., a retail employee’s calculative
attachment to his or her organization) and supervisor support
impact the job dissatisfaction – voice behaviors relationship. By
means of a field study, we find that these factors play a critical
role for retail employees who have low levels of job satisfaction.
In our sample, relatively unsatisfied employees contributed to a
voluntary service development program, by suggesting creative
solutions to problems and encouraging others to do the same,
more frequently if they felt stuck in their current organization
and worked for a supportive supervisor. By contrast, the rela-
tively unsatisfied employees who instead perceived alternative
employment options to be available elsewhere were less likely
to contribute even when supervisor support was provided.
We extend this key finding with a controlled experiment.
Chief among our goals here was to directly analyze a situation
where individuals were reacting to a disaffecting environment.
The results mirror those found in the field and suggest that
supervisors’ supportive efforts are not answered with greater
likelihoods of employee voice when dissatisfied retail employ-
ees have ample external employment options. Regardless of a
supervisor’s support, these employees are relatively unlikely to
share their constructive ideas. It follows that supervisors are
better off spending their limited time and resources to support
dissatisfied employees who are committed to their organization
out of necessity.
Theoretical and managerial implications of these findings
are covered in the discussion section of this paper. First, how-
ever, we provide a theoretical overview of exit-voice theory, the
framework we use to motivate the study of employee voice, and
develop our central hypothesis. A complete discussion of the
two studies we employ in this paper is couched between these
two sections.
Theoretical and hypotheses development
Exit-voice theory
Within the realm of internal marketing, exit-voice theory
refers to the four reactions employees may have to job dissatis-
faction (Farrell 1983; Hirschman 1970). Among them are exit (to
voluntarily depart from an organization), neglect (to withdraw
from work psychologically or behaviorally), loyalty (to remain
committed, hoping that change is on the horizon), and voice (to
advocate change with respect to some source of unrest). Impor-
tantly, the only reaction that constitutes an active, constructive
response is voice (Rusbult et al. 1988). Recognizing this distin-
guishing characteristic and the fundamental role that voice plays
in business relationships, marketing scholars have devoted con-
siderable attention to the study of voice in the inter-firm context
(e.g., Ping 1997) and in the consumer complaint behaviors lit-
erature (e.g., Mittal, Huppertz, and Khare 2008; Singh 1990).
However, this research stream has yet to transcend to the internal
marketing context in the marketing literature.
That being said, the management literature has made notable
progress developing and testing conceptual models centered
around voice when employees are the purveyors. In this liter-
ature, voice is conceptualized as an employee behavior that is
discretionary, communicative, and challenging in nature with
regards to some status quo (LePine and Van Dyne 1998; Van
Dyne and LePine 1998). Voice can be studied laterally (i.e.,
peer-to-peer; Edmondson 1999) or vertically (i.e., subordinate-
to-supervisor; Burris, Detert, and Chiaburu 2008; Detert and
Burris 2007; Gao, Janssen, and Shi 2011; Morrison, Wheeler-
Smith, and Kamdar 2011). And employee voice can take on
either a promotive tone, in which case challenging statements
are complemented with suggestions for improvement, or a pro-
hibitive tone, in which case problems are raised without an
accompanied plan for resolution (Liang, Farh, and Farh 2012).
We study voice in the vertical realm where voice is promo-
tive in nature. Examples of these types of voice behaviors may
include making helpful recommendations to one’s supervisor
concerning problems that customers face or getting involved in
service development planning by sharing creative solutions. On
these grounds, we delineate this paper’s scope and define retail
employee voice as:
Voluntary communication effort that is conveyed by retail
employees to their supervisors and directed at challenging the
status quo of a work unit through the suggestion of creative
and promotive solutions.
Furthermore, given that customer service delivery innova-
tions are an emerging research topic in the marketing literature
(Ostrom et al. 2010), we study voice in the realm of retail
employees’ change-oriented suggestions regarding the deliv-
ery of services to a retailer’s new or existing customers (e.g.,
Chen, Tsou, and Huang 2009; Ordanini and Parasuraman 2011;
Umashankar, Srinivasan, and Hindman 2011). In our first study,
we employ a unique dataset that captures retail employees’ vol-
untary contributions to a service development program that was
launched to facilitate voice behaviors. In our second study, we
present our experiment’s participants with a voice opportunity
that poses to improve the service provided to a retailer’s customer
base. Next, we develop our central hypothesis.
The job dissatisfaction – employee voice relationship
Employees are less likely to share their suggestions for
improvement when they are less satisfied with their job (e.g.,
Brehm 1966; LePine and Van Dyne 1998; Morrison, Wheeler-
Smith, and Kamdar 2011; Ng and Feldman 2011; Rusbult
et al. 1988). The psychological driver behind this effect can be
explained as follows: when employees are dissatisfied with the
salient aspects of their job, they are likely to enter into an eco-
nomic exchange with their employer (Blau 1964). Under such
circumstances, employees form relations with organizations on
a quid pro quo basis, exchanging their services for a salary on
a tit-for-tat basis. This situation makes it unlikely that voice,
which is an extra-role behavior (Van Dyne and LePine 1998),
 J.P. Boichuk, B. Menguc / Journal of Retailing 89 (2, 2013) 207–218
will be chosen in response to job dissatisfaction (see MacKenzie,
Podsakoff, and Ahearne 1998 for an account of this general
phenomenon).
Thus, a well-supported finding in the literature is that job
dissatisfaction is negatively related to voice. For retailers, this
means that bottom-up innovations are least likely to occur when
retail employees are disengaged. In order to understand how this
relationship can be assuaged, we study boundary conditions in
the form of continuance commitment and supervisor support.
The moderating roles of continuance commitment and
supervisor support
Of the three facets of employee commitment (i.e., affective,
continuance, and normative; Allen and Meyer 1990), con-
tinuance commitment is the most relevant to this study. As
mentioned previously, voice can be viewed as the summation of
a cognitive calculus process that weighs the benefits of speaking
up against the costs of such (cf. Detert and Edmondson 2011;
Liang, Farh, and Farh 2012). Continuance commitment captures
much of the cost portion of that mental calculation, which is why
researchers have considered it to be a critical condition for voice
(Zhou and George 2001).
By definition, continuance commitment refers to the pro-
portion of an employee’s commitment that stems from costs
associated with leaving an organization (Allen and Meyer
1990). Retail employees who are high in continuance commit-
ment are committed to stay with a given retailer, regardless
of their emotional detachment, because they have significant
organization-specific investments, too few options for alterna-
tive employment, or both. With fewer alternatives in the market,
continuancely committed employees should have less reason to
stay silent about their dissatisfaction. Under this circumstance,
the option to improve one’s satisfaction through changing jobs
is improbable, making the option of voice more attractive. As
a result, we expect less satisfied retail employees to engage in
voice behaviors more frequently when they are strongly com-
mitted to a given retailer for reasons related to extended tenure,
organization-specific skills, or a depressed economy (i.e., when
continuance commitment is high), and less voice when they can
move on to greener pastures (i.e., when continuance commit-
ment is low; Meyer and Allen 1984).
Furthermore, supervisor support is a factor that – when
directed at relatively unsatisfied or dissatisfied retail employ-
ees who are continuancely committed to their organization –
should influence employee’s engagement in voice behaviors.
Supervisor support can be defined as the degree to which retail
employees can rely on their supervisor to both help and resolve
issues, especially during times of difficulty (House 1981). As a
function of relational internal marketing, supervisor support is a
mechanism through which vertical relationships foster promo-
tive environments (Bell, Menguc, and Stefani 2004).
Based on the above, we expect continuancely committed
employees who have potential solutions to their problems to
reciprocate supervisors’ empathetic behaviors by engaging in
voice behaviors. By contrast, we expect subordinates who are
relatively unsatisfied with their job and have “one foot out
209
the door” due to low levels of continuance commitment to be
unreceptive to supervisors’ supportive efforts. Social learning
theorists highlight that a recipient’s attention must be aroused
for messages to be internalized and for behavioral change to
occur (e.g., the expression of voice; Ormrod 2003). Seeing as
retail employees who have many alternatives and low barriers
to exit should have relatively low attention levels, we contend
that these employees should fail to internalize supervisors’ sup-
portive efforts and, as a result, be unresponsive to supervisor
support. The above arguments lead to the following hypothesis:
Under the states of low job satisfaction or high job dissatis-
faction, supervisor support has a positive effect on employee
voice when continuance commitment is high and no effect
when it is low.
In the next two sections, we outline the methodology and
results of a field-based data collection project and of a controlled
experiment. In Study 1, we test our hypothesis with a multisource
dataset from a company in the travel and tourism sector. In Study
2, we stimulate job dissatisfaction in the minds of participants
and manipulate continuance commitment and supervisor sup-
port to provide additional support for their interactive effect on
employee voice.
Study 1
Sample and data collection procedure
The data were collected from a large retailer that operates
50 travel agencies across the English-speaking provinces of
Canada. Each agency’s manager received a set of packets that
contained consent forms, surveys, and postage-paid envelopes.
These managers were in charge of survey administration, and to
eliminate concerns of common method bias, they responded to
a survey that inquired about their travel agents’ voice behaviors.
This survey was focused on employees’ involvement in a service
development program the company had in place. The program
gave employees an outlet to contribute ideas for service improve-
ment, and all contributions were considered voluntary (i.e., not
formally required). Travel agents responded to a separate sur-
vey that included measures of the independent, moderating, and
control variables.
Through two waves of mailing, we received 384 usable sur-
veys from travel agents (a response rate of 71 percent). The
number of travel agent responses across travel agencies ranged
from 2 to 22 with an average of 7.68 responses per agency. A
majority of the travel agents were female (72 percent), their aver-
age age was 32.5 years old, and the average travel agent’s job
tenure was 10.2 years in length. A comparison of the sample
profile to the company records indicated that the sample was
representative of the population of retail employees.
Measures
All measurement items employed in our survey were rated
using five-point Likert scales, anchored by the dichotomous
statements “strongly disagree” (1) and “strongly agree” (5).
210 J.P. Boichuk, B. Menguc / Journal of Retailing 89 (2, 2013) 207–218

Table 1
Descriptive statistics and correlations for Study 1.
1 2 3 4 5 6 7 8 9 10 11 12 13 14

1. Gender
2. Age −.04
3. Tenure −.10* .54**
4. Education −.23** −.16** −.18**
5. Training −.03 .12* −.07 .05
6. Salary increase −.01 −.22** −.28** .02 .20**
7. Affective commitment .03 .10 .08 .04 .02 .04
8. Autonomy .08 .10* .07 −.03 .03 .04 .40**
9. Organizational support −.17** .19** .05 .20** .05 .05 .37** .24**
10. Coworker support .03 .02 −.01 −.01 −.03 −.03 .41** .25** .20**
11. Job satisfaction (r) .03 −.15** −.08 .04 .04 −.03 −.65** −.50** −.48** −.37**
12. Continuance commitment .12* −.16** .19** .07 −.15** −.02 .05 −.06 −.22** .20** .15**
13. Supervisor support −.24** .09 −.04 .10 .10* .15** .37** .21** .42** .42** −.33** −.01
14. Employee voice .11* .19** .07 .05* .13* .01 .40** .35** .29** .21** −.40** .07 .12*
Mean 1.81 32.49 10.21 1.95 5.51 −.07 4.51 4.05 4.19 4.11 1.70 3.17 4.11 4.20
SD 0.40 13.40 8.38 0.79 8.45 −.08 0.54 0.67 0.80 0.86 0.57 0.91 0.88 0.77

Notes. (r) = reverse-scored.

* p < .05 (two-tailed test).
** p < .01 (two-tailed test).

Agency managers responded to a survey that included a four-
item scale that measured travel agents’ voice behaviors, which
we adapted from Van Dyne and LePine (1998). In the travel agent
survey, job satisfaction was captured with a three-item scale cre-
ated by O’Reilly and Caldwell (1981). We tested our model with
the reverse-coded scores of this measure. Continuance com-
mitment was obtained with a four-item scale from Allen and
Meyer (1990). Supervisor support was operationalized using a
three-item scale adapted from House (1981).
Previous studies indicate that employee voice is influenced
by a whole set of variables, including demographic, attitudinal,
interpersonal relationship, and compensation-related factors.
Accordingly, to avoid model misspecification and to improve the
nomological validity of our model, we control for demographics,
affective commitment, autonomy, organizational support, and
coworker support. The demographic variables we control for
include: gender (1-male; 2-female), age (in years), tenure (i.e.,
length of employment with current organization), education (1-
high school, 2-bachelor’s degree, 3-graduate degree), training
(i.e., hours of training received last year), and salary increase
(i.e., salary increase over the past two years). We measured
affective commitment with a six-item scale taken from Allen and
Meyer (1990), autonomy with a three-item scale developed by
Spreitzer (1995), coworker support with three items from House
(1981), and organizational support with a four-item scale from
Eisenberger et al. (1986).
See Appendix A for the scales used in Study 1 and a discus-
sion of the psychometric properties of these scales. We report
the descriptive statistics and intercorrelations of these variables
in Table 1.
Analytic approach
The data we collected were nested (i.e., travel agents were
nested within travel agencies), making a random coefficients
modeling procedure in hierarchical linear modeling (HLM) a
likely candidate to test our hypotheses. HLM accounts for inter-
dependence in multilevel data by dichotomizing and estimating
within- and between-group variance components separately
(Raudenbush and Bryk 2002). We employed the maximum like-
lihood estimation option in HLM. This option provides deviance
statistics, which allow us to assess model improvement across
nested models: additional predictors are judged based on signif-
icant decreases in the deviance statistic (Raudenbush and Bryk
2002).
To ensure that HLM is the appropriate approach to employ,
we ran a null model with no predictors of voice. If a signifi-
cant degree of variance in voice behaviors falls between travel
agencies, ordinary least squares regression is not appropriate
and HLM is necessary (Hofmann 1997). The results indi-
cated significant between-travel agency variance (ICC1 = .30; χ2
[49] = 72.03, p < .001), suggesting that HLM analysis is appro-
priate.
In the following tests, we entered all variables, except for con-
tinuance commitment and job (dis)satisfaction, into the model
after centering them by their group mean (i.e., travel agency
mean) (Hofmann and Gavin 1998). The reason we chose this
specification is that continuance commitment and supervisor
support may influence how travel agents evaluate their jobs.
Similarly, travel agents’ continuance commitment may be influ-
enced by the support they receive from supervisors. Therefore,
unless these sources of endogeneity are controlled for, the linear
and moderated relationships in our model may not be estimated
accurately (see Luo, Rindfleisch, and Tse 2007 for details).
To mitigate this potential source of misspecification, we com-
puted the residual value of our reverse-coded job satisfaction
measure by regressing it against the two moderators in our
model and the residual value of continuance commitment by
regressing it against supervisor support. The residual values of
job (dis)satisfaction and continuance commitment that are not
 J.P. Boichuk, B. Menguc / Journal of Retailing 89 (2, 2013) 207–218
accounted for by these regressors were then used as indicators
in the model. In addition, we created two-way and three-way
interaction terms using these residual values (Luo, Rindfleisch,
and Tse 2007).
Hypothesis testing
We followed a hierarchical approach to test our hypothesis.
Table 2 reports the results for the linear and moderated effects
on voice.
The controls-only model (Model 1) resulted in a sig-
nificant decrease in model deviance over the null model
(deviance = 165.85; df = 10; p < .001), and explained 29 per-
cent of total variance in voice. Model 2 added all main effects
including the reverse-coded job satisfaction term. This model
resulted in a significant decrease in model deviance over
the controls-only model (deviance = 48.56; df = 3; p < .001),
explaining 35 percent of total variance in voice. Next, we
entered the two-way interaction effect of job (dis)satisfaction
and continuance commitment into the model. Model 3 resulted
in a significant decrease in model deviance over Model 2
(deviance = 25.65; df = 1; p < .001), and explained 42 per-
cent of total variance in voice. Model 4 included all control
variables, main effects, two-way interactions, and the three-
way interaction. Overall, Model 4 resulted in a significant
decrease in model deviance over Model 3 (deviance = 11.67;
df = 3; p < .01) and explained 49 percent of the total
variance.
Table 2 (Model 4) indicates that the main effect of job
(dis)satisfaction is negatively and significantly related to voice
behaviors (γ = −.19, p < .001). The main effects of the moder-
ating variables (i.e., continuance commitment and supervisor
support) were not significantly related to voice behaviors. The
two-way interaction of job (dis)satisfaction and continuance
commitment was positive and statistically significant (γ = .24,
p < .01).
We proposed in the hypothesis development section that
when continuance commitment and supervisor support are both
high, job dissatisfaction has the strongest, most positive rela-
tionship with travel agents’ voice behaviors. The three-way
interaction term is positive and significant (γ = .39, p < .01),
providing support for this hypothesis. We examined the nature
of this interaction effect by plotting it at high (one standard
deviation above the mean) and low (one standard deviation
below the mean) levels of the moderator variables (i.e., con-
tinuance commitment and supervisor support) (Aiken and West
1991). Simple slope analysis (Dawson and Richter 2006) also
indicated that the relationship between reverse-coded job sat-
isfaction and voice is significantly positive when continuance
commitment and supervisor support are both high (γ = .27,
p < .01), whereas the same relationship is either significantly
negative or not significantly different from zero in all other con-
ditions (high continuance commitment/low supervisor support:
γ = −.22, p < .05; low continuance commitment/high supervisor
support: γ = −.70, p < .001; low continuance commitment/low
supervisor support: γ = −.10, ns) (see Fig. 1).
211
Additional analysis
In this paper, we focus on the calculative process that con-
tributes to employees’ engagement in voice behaviors. However,
Table 2 indicates a strong main effect of affective commit-
ment on voice behaviors, and we recognize that the process
could be affectively charged as well. Therefore, it is theoreti-
cally relevant to show affective and continuance commitment’s
differential effects. We tested this possibility by conducting
an additional analysis to estimate the three-way interaction
of job dissatisfaction, affective commitment, and supervisor
support. Controlling for the effect of continuance commit-
ment and all other control variables, we found the following
results: job (dis)satisfaction × affective commitment (γ = .09,
ns), job (dis)satisfaction × supervisory support (γ = .05, ns),
affective commitment × supervisory support (γ = .23, p < .01),
job (dis)satisfaction × affective commitment × supervisory sup-
port (γ = .04, ns). The full model (i.e., control variables, main
effects, two-way interactions, and the three-way interaction)
explained 41 percent of the total variance in voice behaviors.
These nonsignificant findings give us greater confidence that, at
least in our context, the process driving employees’ decisions
was calculative in nature.
Discussion
The purpose of Study 1 was to identify circumstances when
employees with low levels of job satisfaction engage in voice
behaviors. We found these employees to be more likely to
contribute to the organization’s service development program
under the conditions of high continuance commitment and high
supervisor support. Evidently, supervisor support has differ-
ential effects with subordinates who have low levels of job
satisfaction, depending on their level of continuance com-
mitment. These employees engaged in voice behaviors more
frequently if they needed to continue working for their orga-
nization as a result of organization-specific investments or
external job market scarcity. By contrast, those who had lit-
tle stake in the organization or plentiful external options for
employment did not respond as such. This core finding sug-
gests that supervisors should be selective when they provide
support.
Limitations of Study 1 relate to the research context within
which the data were collected. Since the data were collected
cross-sectionally from a single organization within a single
industry, concerns related to causality and external validity can-
not be addressed. Further, since job satisfaction was measured as
opposed to job dissatisfaction, the results do not directly capture
retail employees’ reactions to a dissatisfying work environment.
Understanding these limitations, we designed a controlled exper-
iment that first stimulated job dissatisfaction and later captured
participants’ likelihood of sharing a promotive idea of theirs.
The experimental materials also remained inexplicit with respect
to the retail sector studied for generalizability purposes. In the
following section, we present the method and results of this
controlled experiment.
212 J.P. Boichuk, B. Menguc / Journal of Retailing 89 (2, 2013) 207–218

Table 2
HLM results for Study 1.
Model 1 Model 2 Model 3 Model 4

Dependent variable: voice behaviors

Constant 4.43*** 4.46*** 4.48*** 4.45***
Control variables
Gender 0.18** 0.13* 0.13* 0.14
Age −0.01 −0.01 −0.01 0.00
Tenure 0.01 0.01 0.01 0.00
Education −0.08** −0.05 −0.05 −0.05
Training −0.01 0.03 0.04 0.04
Salary increase −0.03 −0.05 −0.06 −0.06
Affective commitment 0.54*** 0.35*** 0.22*** 0.23**
Autonomy 0.08* 0.01 0.01 0.01
Organizational support −0.03 0.02 0.05 0.05
Coworker support 0.05 0.04 0.02 0.02
Main effect
Job satisfaction (r) (1) −0.20*** −0.19*** −0.19***
Moderators
Continuance commitment (2) 0.05 0.05 0.05
Supervisor support (3) 0.02 0.06 0.06
Two-way interactions
(1) × (2) 0.24*** 0.24**
(1) × (3) −0.03
(2) × (3) 0.03
Three-way interaction
(1) × (2) × (3) 0.39**
Deviance (df) 452.80 (13) 404.24 (16) 378.59 (17) 366.92 (20)
Decrease in deviance (df) 165.85 (10)*** 48.56 (3)*** 25.65 (1)*** 11.67 (3)*
Pseudo-R2(within group) 0.41 0.45 0.54 0.64
Pseudo-R2(between group) 0.01 0.12 0.13 0.13
Total R2 0.29 0.35 0.42 0.49

Notes. (r) reverse-scored; total R2 = R2(within group) × (1 − ICC1) + R2between-groups × ICC1 (Hirst et al. 2009).
* p < .05 (two-tailed test).
** p < .01 (two-tailed test).
*** p < .001 (two-tailed test).

Study 2 4.2 years of retail experience. Participation in the experiment

In Study 2, we examine a scenario where individuals make
likelihood decisions about expressing voice under a state of
job dissatisfaction (for a similar application, see Ganesan et al.
2010). With random assignment and a 2 (continuance com-
mitment: high vs. low) × 2 (supervisor support: high vs. low)
between-subjects factorial design, we aim to provide further
support for the notion that dissatisfied retail employees are
responsive to supervisor support when continuance commitment
is high (but not low).
We recruited 150 participants through MTurk, an online labor
system (cf. Goodman, Cryder, and Cheema 2012). They read
that the objective of the experiment is to understand how people
react to certain job conditions in the retail sector. Four pas-
sages followed, each of which described a separate aspect of
an employment situation (see Appendix B). Participants were
asked to imagine that they found themselves in this situation.
Post hoc checks show that they considered the experiment mate-
rial realistic: the mean score to the question “How realistic was
the situation described in the four passages?” was 5.87 out of
7. Approximately 41 percent of participants were female, their
average age was 33 years old, and the average participant had
was limited to residents of the United States.
The first passage stimulated job dissatisfaction. We prepared
this passage to include various facets that enter into employ-
ees’ job evaluations (Churchill, Ford, and Walker 1974), albeit
without mention of dissatisfaction with supervision due to its
later manipulation. The second passage manipulated continu-
ance commitment and the third passage manipulated supervisor
support. Both of these manipulations were generated to reflect
high and low levels of their respective scales in Study 1. The
fourth passage introduced a voice opportunity; all participants
were told about a novel idea of theirs regarding the sales con-
test schedule used by their organization. This situation was
used because it represents a common problem in practice (see
Steenburgh and Ahearne 2012). Manipulation checks can be
viewed in Appendix C.
Dependent measure
Participants indicated how they would react given the job
conditions specified in the four passages by assessing their like-
lihood of sharing their sales contest idea with their supervisor.
Responses fell on a scale from 0 = “Not at all likely to voice my
 J.P. Boichuk, B. Menguc / Journal of Retailing 89 (2, 2013) 207–218 213

Low Continuance Commitment High Continuance Commitment

6 6

5 5

Voice Behaviors
Voice Behaviors

4 4

3 3

2 2

1 1
High Job Satisfaction Low Job Satisfaction High Job Satisfaction Low Job Satisfaction

Low SS (1) 5.24 4.71 Low SS (3) 4.95 5.13

High SS (2) 4.79 4.39 High SS (4) 4.77 4.66
Notes: SS = supervisor support; the numbers in parentheses act as labels of the four slopes and are used in the slope difference tests below.

Slope difference tests:

Pair of slopes t-value for slope difference p-value for slope difference
(4) and (3) 3.17 0.00
(4) and (2) 6.17 0.00
(4) and (1) 2.71 0.01
(3) and (2) 3.52 0.00
(3) and (1) -1.08 0.20
(2) and (1) -3.53 0.00

Fig. 1. Study 1 results: three-way interaction effect involving continuance commitment and supervisor support when job satisfaction varies from high to low.

idea” to 100 = “Extremely likely to voice my idea.” We report
the experimental results in the following section (see also Fig. 2).
Results
In line with our central hypothesis, the results show that the
interactive effect of continuance commitment and supervisor
support is statistically significant and positive (F1, 149 = 4.36,
p < .05). Further, the simple effects of supervisor support are
indicative of participants being responsive to supervisor support
when continuance commitment is high, but not when con-
tinuance commitment is low. Supervisor support resulted in
significantly greater voice intentions when continuance com-
mitment was high (low supervisor support = 62.82 vs. high
supervisor support = 75.56, p < .05), but not when continuance
commitment was low (low supervisor support = 53.56 vs. high
supervisor support = 47.79, ns). Further, based on the results, the
90
85
80
75
Voice Intentions
main effect of continuance commitment was positive (β = 18.52,
p < .01), whereas the main effect of supervisor support was not
significantly different from zero (β = 3.49, ns).
Additional analysis
We also ran an additional experiment to test the contingency
role of affective commitment with 150 participants who were
again recruited through MTurk. This experiment was a 2 (affec-
tive commitment: high vs. low) × 2 (supervisor support: high vs.
low) between-subjects factorial design under the condition of
high job dissatisfaction. The manipulation we used for affective
commitment, along with the complete results, can be obtained
from the authors upon request. All other study materials were
identical to those used in Study 2. For brevity purposes, we
only report that the interaction between affective commitment
and supervisor support is not significantly different from zero
(F1, 149 = 1.40, ns). Again, affective commitment was not found
to play a contingency role in employees’ decisions to engage
in voice behaviors. This result lends further evidence to our
theorization that the process is calculative in nature.

70
65 Low Supervisor
Support Discussion
60
55 High Supervisor
50 Support
45
These results provide greater support for those found in Study
40 1 and extend the context studied to a situation where job dissat-
35
30
isfaction is high. Having manipulated continuance commitment
Low High and supervisor support independently, we can more confidently
Continuance Commitment Continuance Commitment
state that disaffected retail employees are more likely to be unre-
Fig. 2. Study 2 results: interaction effect of continuance commitment and super- ceptive to supervisor support if they perceive barriers to exit to
visor support when job dissatisfaction is high. be few. As social learning theorists would contend (cf. Ormrod
214 J.P. Boichuk, B. Menguc / Journal of Retailing 89 (2, 2013) 207–218
2003), recipients of support must be attentive for messages deliv-
ered to be internalized and for behavioral change to occur. Posing
continuance commitment as a calculative mechanism that deter-
mines the attentiveness of subordinates is new to the literature.
These results suggest that continuance commitment is an impor-
tant factor that retail managers should consider prior to offering
support to a given subordinate.
General discussion
Altogether, the results indicate that retailers have a control-
lable means (i.e., supervisor support) to solve a problem that
is otherwise uncontrollable (i.e., staffs of discontented retail
employees who have nowhere else to go may otherwise remain
silent). That is, an employee group that may reduce extra effort
(e.g., Meyer et al. 1989) can be influenced to deposit voice into
the organizational system if support is provided by supervisors.
Theoretical implications
This paper makes significant contributions to the internal
marketing and employee engagement literatures by adopting an
exit-voice framework and studying the topic of retail employee
dissatisfaction (Hirschman 1970). Previously, little was known
about the consequences of continuance commitment to retail
organizations. This knowledge gap existed because much of the
extant research in the marketing literature has focused on affec-
tive rather than continuance commitment. For instance, Payne
and Webber (2006) found that affective commitment plays a con-
tingency role in the relationship between job (dis)satisfaction
and service-oriented citizenship behaviors. Less understood
prior to this study was the contingency role that continuance
commitment plays. Therefore, we contribute to marketing scho-
lars’ understanding about retail employees who are dissatisfied,
yet committed to their organization out of necessity by identify-
ing that these employees are receptive to supervisors’ supportive
efforts.
Specifically, we establish that supervisors who are reliable,
helpful, and willing to listen to retail employees’ work-related
problems (House 1981) influence voice from retail employ-
ees who are dissatisfied, yet committed out of necessity. In
effect, under this condition, dissatisfaction leads to a construc-
tive employee reaction (i.e., voice), which runs contrary to the
destructive reactions that the exit-voice framework predicts (i.e.,
exit or neglect).
In addition, when continuance commitment was used in
previous research, its impact was predominantly hypothesized
to contribute to an economic exchange rather than a social
exchange (Chandrashekaran et al. 2000). In that regard, results
from this study run opposite to results of previous research in
the marketing field, but parallel to principles of exit-voice theory
(Hirschman 1970) and to applications of continuance com-
mitment in the innovation literature (Zhou and George 2001).
Therefore, the results of this paper bring new light to continuance
commitment in the marketing literature and help to understand
how voice can be fostered in the retail setting.
Managerial implications
This paper also poses practical lessons for retail organi-
zations. Several conditions, internal and external to retailers,
provide opportunities for them to reap the benefits of highly sup-
portive supervisors. For instance, internal to the organization, if
retail employees are highly committed to an organization due to
organization-specific investments, such as organization-specific
training, supervisor support should result in higher levels of retail
employee voice. External to the organization, if the economy is
in a downturn and employment is sparse, retailers can also expect
increases in voice from supervisor support.
Also, it is evident from our results that supervisor support
should not be provided to dissatisfied subordinates who have
low levels of continuance commitment. This lesson is crucial
for supervisors to learn, because both conditions must hold in
order for retailers to foster feedback mechanisms from the front-
line. For instance, given that supervisors have limited time and
social resources (Konovsky and Pugh 1994), this is an impor-
tant finding as it informs practitioners about how they can better
allocate their team building and development efforts.
It is then advised that retail managers continuously con-
duct a mental checklist to determine whether or not supportive
efforts should be provided to retail employees who are dissat-
isfied. Such a checklist should take into consideration at least
two factors. First, the specific characteristics of an employee
who is dissatisfied (e.g., low vs. high levels of investment in
an organization) should be considered to determine the like-
lihood that he or she will be receptive to support. Second,
retail managers should consider the economic conditions of their
industry in order to understand whether or not retail employees
are likely to have alternative options of employment. Support
should only be provided to a retail employee who is dissatis-
fied if, after conducting a mental checklist, the retail manager
deems the focal employee to be high in continuance com-
mitment. Please note that this prescription is independent of
a retail employee’s performance level. Conditions may arise
where the performance of a retail employee is incessantly low
and continuance commitment is high. In this scenario, supervi-
sors may be better off coaching the retail employee out of the
organization.
Future research directions
Grounded in exit-voice theory, the findings presented in this
paper advance the internal marketing and employee engagement
literatures, and open the door for future researchers to raise and
address other research questions related to managing a dissat-
isfied retail sales force. Particularly, the findings would benefit
from research that investigates additional boundary conditions
of the job dissatisfaction – retail employee voice relationship
(e.g., autonomy). We studied continuance commitment in this
paper due to our theorization that voice behaviors depend on
a calculative mental process, yet conditions may exist where
the process operates in an affective realm. Future research that
further investigates the differential effects of affective and con-
tinuance commitment would improve our understanding of the
 J.P. Boichuk, B. Menguc / Journal of Retailing 89 (2, 2013) 207–218
situational factors that place voice decisions in each of these
respective realms.
Acknowledgements
The first author gratefully acknowledges financial support
from the Social Sciences and Humanities Research Council by
way of the Joseph-Armand Bombardier Canada Graduate Schol-
arship Program. This support assisted the development of this
project as part of his master’s thesis at Brock University, Canada.
In addition, the authors would like to thank the University of
Houston Marketing Faculty, Seigyoung Auh, and Narongsak
Thongpapanl for their helpful comments on an earlier version
of this article, and three anonymous JR reviewers as well as the
Editor, Shankar Ganesan, for their insight and guidance during
the review process.
Appendix A. Measurement assessment for Study 1
Since the data we employ in Study 1 came from two sources
(i.e., the agency manager survey and the travel agent survey),
we estimated two measurement models (Anderson and Gerbing
1988). Through a confirmatory factor analysis (CFA), we found
all factor loadings of the measurement items to be above .40, all
normalized residuals to be less than 2.58, and all modification
indices to be less than 3.84 (Anderson and Gerbing 1988). Fur-
thermore, the models demonstrated good fit to the data: (agency
manager survey: χ(2) 2 = 4.02; GFI = .98, TLI = .99, CFI = .99,
RMSEA = .05; travel agent survey: χ(278)2 = 607.2, GFI = .91;
TLI = .94, CFI = .94, RMSEA = .06).
With overall model fit assured, measurement model fit indices
were applied to address concerns of convergent and discriminant
validity. To assess convergent validity, we examined the com-
posite reliability (CR) and average variance extracted (AVE) of
the constructs (see Table A.1). All constructs exceeded their
respective thresholds of acceptance for both CR (.70) and AVE
(.50); thus, support was provided for the notion of convergent
validity.
Similarly, based on three tests we conducted, discriminant
validity was supported. First, none of the 95 percent confi-
dence intervals for the correlations included one (Anderson
and Gerbing 1988). Second, AVE estimates for all respective
pairs of constructs in the model were greater than their match-
ing squared correlation (Fornell and Larcker 1981). Third, we
compared unconstrained models with constrained models (i.e.,
the correlation between a given pair of constructs was set to
1) (Anderson and Gerbing 1988). Chi-square difference tests
favored the unconstrained models for each pair of constructs,
further supporting the notion of discriminant validity.
Appendix B. Experimental manipulations for Study 2
Stimulating job dissatisfaction: You are a retail salesperson,
your current employer pays you very little (just about mini-
mum wage), and your chances of advancement are very limited.
In fact, there are no job openings posted on your company’s
215
internal job board, and pay increases are dismal. There is no
doubt that you are stuck in your current position at a low pay
grade for the foreseeable future. You remember a time when
assisting customers was rewarding, but now all you seem to hear
is, “I’m just looking, thanks.” Armed with product and price
information from the Internet, customers no longer need retail
salespeople’s assistance – unless they have something to com-
plain about, of course. Then you get their undivided attention
while they blame you for every mistake your coworkers have
ever made. You share nothing in common with your cowork-
ers either, and talking with them about the weather has become
a chore. Plus, they are very artificial, often talking about you
behind your back.
Continuance commitment manipulation [low commit-
ment in parentheses]: No one (everyone) seems to be hiring
and attractive retail positions generally require applicants to have
more (less) experience than you currently have. As such, it would
be very difficult (easy) for you to find another organization that
compares to your current one. If you were to decide to leave your
current organization, it would be very inconvenient (convenient)
to find a suitable replacement. You have no (ample) time to look
for alternative employment – too (not) much is going on in your
personal life. When you aren’t working, you have many (few)
family demands, leaving you with no (lots of) time to search
for alternative employment. Unfortunately (Fortunately), you
(don’t) need to keep working for your current organization.
Supervisor support manipulation [low support in paren-
theses]: Your supervisor is very (anything but) sympathetic. She
cares a great deal (very little) about the welfare of her staff, and
she is (nor is she) helpful when it comes to resolving work-
related issues. When you find yourself in a tough position, she
is one of the first (last) people you go to. Listening skills are
definitely (not) her strong suit – she often (seldom) listens to the
problems and concerns of her employees. Overall, the help she
gives you is second to none (lackluster at best).
Stimulating a voice opportunity: A certain aspect of your
job is particularly vexing. Sales contests are designed to meet
internal goals rather than to coincide with customers’ purchase
cycles. In slow months, a contest will no doubt be introduced.
During fast months, no such contest is ever designed. This
practice is penny wise and pound foolish, in your opinion. Con-
tests do nothing more than incent your company’s sales staff
to badger the few customers that enter your store during slow
months with pushy sales tactics, because targets require a high
conversion rate. And contests would move the needle consider-
ably if implemented during fast months because targets guard
against complacency. Your novel idea is that changing the sales
contest schedule entirely could improve customer satisfaction
and optimize sales for your organization. Although ideas such
as yours represent creative ways to solve current problems in
organizations, they also challenge the status quo. A primary
motivation behind sharing ideas with supervisors is often to
change, rather than escape from, unfortunate situations. Poten-
tial costs associated with sharing ideas include (a) the potential to
initiate interpersonal conflict, (b) the fear of negative feedback,
and (c) the risk of being perceived as a troublemaker. Since shar-
ing creative solutions to work-related problems is typically not
216 J.P. Boichuk, B. Menguc / Journal of Retailing 89 (2, 2013) 207–218

Table A.1
Confirmatory factor analysis results for Study 1.
Constructs and items Factor loading t-Value

Travel agency manager survey

Voice behaviors (α = .87;
CR = .88; AVE = .64)
This travel agent. . .
. . .makes helpful recommendations to others in this branch concerning problems that customers face. .82a –
. . .encourages others in this branch to get involved with service development planning. .79 16.95
. . .gets involved in service development planning by sharing creative solutions. .83 17.28
. . .speaks up in this branch with ideas about creative ways to deliver service to customers. .76 15.58

Travel agent survey

Job satisfaction (r) (α = .75; CR = .76; AVE = .52)
Overall, I am satisfied with my job (r). .64 11.32
I am satisfied with the important aspects of my job (r). .80 22.10
I would prefer another, more ideal job. .71a –
Supervisor support (α = .90; CR = .92; AVE = .79)
My supervisor can be relied upon when things get difficult in my job. .98a –
My supervisor is willing to listen to my job-related problems. .86 26.50
My supervisor is helpful to me in getting my job done. .82 24.04
Continuance commitment (α = .83; CR = .84; AVE = .57)
Right now, staying with this organization is a matter of necessity as much as desire. .65a –
It would be very hard for me to leave my organization right now, even if I wanted to. .92 14.20
Too much of my life would be disrupted if I decided I wanted to leave my organization right now. .87 14.22
I feel that I have too few options to consider leaving this organization. .51 9.18
Affective commitment (α = .87; CR = .88; AVE = .56)
I really care about the fate of this organization. .70 14.70
I am willing to put in a great deal of effort, beyond what normally is expected, in order to help this .72 14.64
organization be successful.
My organization really inspires me to put forth my best effort in customer service. .71 14.50
I would accept almost any type of job assignment in order to keep working for this organization. .76 16.90
I am extremely glad that I chose this organization over others I was considering at the time I joined. .76 16.47
Overall, I am very committed to this organization. .82a –
Autonomy (α = .78; CR = .78; AVE = .54)
I have significant autonomy in determining how I do my job. .73 11.36
I can decide on my own how to go about doing my job. .68 11.65
I have considerable opportunity for independence and freedom in how I do my job. .79a –
Organizational support (α = .88; CR = .89; AVE = .67)
Excellent customer service is encouraged at this organization. .88a –
This organization recognizes those who deliver excellent customer service. .84 7.22
Our ability to offer excellent service is respected by this organization’s top managers. .69 6.89
The organizational reward system encourages excellence in customer service. .86 7.04
Coworker support (α = .90; CR = .92; AVE = .78)
I find the feedback that I receive form my co-workers on my service to customers very useful. .93 26.26
My co-workers provide me with valuable information about how to improve my customer service. .87 23.52
The feedback I receive from my coworkers helps me improve my customer service. .88a –
a Item was fixed to 1 to set the scale of the construct; (r) = reverse-scored item; α = Cronbach’s alpha; CR = composite reliability; AVE = average variance extracted.

part of a retail salesperson’s formal job requirements, commu- Manipulation checks for job dissatisfaction (adapted from
nicating these ideas to supervisors is considered voluntary. A Churchill, Ford, and Walker 1974 and O’Reilly and Caldwell
decision you face is whether or not to share your idea with your 1981, respectively)
supervisor, and that decision is entirely up to you. Please rate your level of dissatisfaction according to the
following aspects of the job (1 = “Not at all dissatisfied” to
7 = “Extremely dissatisfied”):
Appendix C. Measures and manipulation checks for
Study 2
1. Pay
Items were rated on seven-point Likert scales from 2. Advancement opportunity
1 = “strongly disagree” to 7 = “strongly agree,” unless otherwise 3. Customers
stated. 4. Coworkers
 J.P. Boichuk, B. Menguc / Journal of Retailing 89 (2, 2013) 207–218
Please indicate your level of agreement with the following
statements.
1. Overall, I would be dissatisfied with this job.
2. I would be dissatisfied with the important aspects of this job.
3. I would prefer another, more ideal job.
Manipulation check for continuance commitment
(adapted from Allen and Meyer 1990)
Please indicate your level of agreement with the following
statements.
1. I feel that I would have too few options to consider leaving
this organization.
2. It would be very hard for me to leave this organization, even
if I wanted to.
3. Staying with this organization is a matter of necessity as much
as desire.
4. Too much of my life would be disrupted if I decided I wanted
to leave this organization.
Manipulation check for supervisor support (adapted from
House 1981)
Please indicate your level of agreement with the following
statements.
1. My supervisor can be relied upon when things get difficult
in my job.
2. My supervisor is willing to listen to my job-related problems.
3. My supervisor is helpful to me in getting my job done.
Manipulation check for the voice opportunity (original
scale)
Please indicate your level of agreement with the following
statements.
1. My idea represents a creative solution to a problem that my
organization faces.
2. My idea challenges the current practices implemented by my
organization.
3. There are benefits associated with sharing my idea with my
supervisor.
4. There are costs associated with sharing my idea with my
supervisor.
5. Sharing my idea with my supervisor is under my control.
Voice intentions (adapted from Ganesan et al. 2010)
How likely is it that you would share your sales contest idea
with your supervisor? (0 = “Not at all likely to voice my idea”
to 100 = “Extremely likely to voice my idea”)
Appendix D. Manipulation check results for Study 2
The continuance commitment and supervisor support manip-
ulations were successful. As expected, participants in the high
continuance commitment conditions rated external options to
217
be less available than those in the low continuance commit-
ment conditions (5.61 vs. 2.01, t = 18.47, p < .001). Similarly,
participants in the high supervisor support conditions rated their
supervisor to be more helpful than those in the low supervisor
support conditions (6.17 vs. 1.68, t = 29.12, p < .001).
Study 2 also included a passage that intended to stimulate job
dissatisfaction and a passage that introduced a voice opportunity.
These passages worked as expected. Suggesting that participants
judged the job described in passage 1 as being dissatisfying, job
dissatisfaction was rated as being greater than the scale mid-
point on both the Churchill, Ford, and Walker (1974) scale (5.62,
t = 12.11, p < .001) and the O’Reilly and Caldwell (1981) scale
(6.51, t = 41.03, p < .001). Suggesting that participants judged
the sales contest idea as being a chance to suggest a creative
solution to an organizational problem, the mean score for the
voice opportunity scale was greater than the scale midpoint
(5.71, t = 28.69, p < .001).
As an additional check, we examined the correlation matrix
among the manipulation checks of the independent factors
included in the experiment (i.e., job dissatisfaction, continuance
commitment, and supervisor support). We find no statistically
significant pairwise correlations in this matrix, suggesting that
our manipulations were not affected by each other.
References
Aiken, Leona S. and Stephen G. West (1991), Multiple Regression: Testing and
Interpreting Interactions, Newbury Park, CA: Sage.
Allen, Natalie J. and John P. Meyer (1990), “The Measurement and Antecedents
of Affective, Continuance and Normative Commitment to the Organization,”
Journal of Occupational Psychology, 63 (1), 1–18.
Anderson, James C. and David W. Gerbing (1988), “Structural Equation Mod-
eling in Practice: A Review and Recommended Two-Step Approach,”
Psychological Bulletin, 103 (3), 411–23.
Bell, Simon J., Bulent Menguc and Sara L. Stefani (2004), “When Customers
Disappoint: A Model of Relational Internal Marketing and Customer Com-
plaints,” Journal of the Academy of Marketing Science, 32 (2), 112–26.
Blau, Peter M. (1964), Exchange and Power in Social Life, New York: Wiley.
BlessingWhite (2011), “Employee Engagement Report,” (retrieved on
12.28.12), [from http://www.blessingwhite.com/eee report.asp].
Brehm, Jack W. (1966), A Theory of Psychological Reactance, New York:
Academic.
Burris, Ethan R., James R. Detert and Dan S. Chiaburu (2008), “Quitting Before
Leaving: The Mediating Effects of Psychological Attachment and Detach-
ment on Voice,” Journal of Applied Psychology, 93 (4), 912–22.
Chandrashekaran, Murali, Kevin McNeilly, Frederick A. Russ and Detelina
Marinova (2000), “From Uncertain Intentions to Actual Behavior: A Thresh-
old Model of Whether and When Salespeople Quit,” Journal of Marketing
Research, 37 (4), 463–79.
Chen, Ja-Shen, Hung Tai Tsou and Astrid Ya-Hui Huang (2009), “Service Deliv-
ery Innovation: Antecedents and Impact on Firm Performance,” Journal of
Service Research, 12 (1), 36–55.
Churchill, Gilbert A. Jr., Neil M. Ford and Orville C. Walker Jr. (1974), “Mea-
suring the Job Satisfaction of Industrial Salesmen,” Journal of Marketing
Research, 11 (3), 254–60.
Dawson, Jeremy F. and Andreas W. Richter (2006), “Probing Three-Way
Interactions in Moderated Multiple Regression: Development and Appli-
cation of a Slope Difference Test,” Journal of Applied Psychology, 91 (4),
917–26.
Detert, James R. and Ethan R. Burris (2007), “Leadership Behavior and
Employee Voice: Is the Door Really Open?,” Academy of Management
Journal, 50 (4), 869–84.
218 J.P. Boichuk, B. Menguc / Journal of Retailing 89 (2, 2013) 207–218
Detert, James R. and Amy C. Edmondson (2011), “Implicit Voice Theories:
Taken-for-Granted Rules of Self-Censorship at Work,” Academy of Man-
agement Journal, 54 (3), 461–88.
Edmondson, Amy C. (1999), “Psychological Safety and Learning Behavior in
Work Teams,” Administrative Science Quarterly, 44, 350–83.
Eisenberger, Robert, Robin Huntington, Steven Hutchison and Debora Sowa
(1986), “Perceived Organizational Support,” Journal of Applied Psychology,
71 (3), 500–7.
Farrell, Dan (1983), “Exit, Voice, Loyalty, and Neglect as Responses to Job Dis-
satisfaction: A Multidimensional Scaling Study,” Academy of Management
Journal, 26 (4), 596–607.
Fornell, Claes and David F. Larcker (1981), “Evaluating Structural Equation
Models with Unobservable Variables and Measurement Error,” Journal of
Marketing Research, 18 (1), 39–50.
Ganesan, Shankar, Steven P. Brown, Babu John Mariadoss and Hillbun Ho
(2010), “Buffering and Amplifying Effects of Relationship Commitment
in Business-to-Business Relationships,” Journal of Marketing Research, 47
(2), 361–73.
Gao, Liping, Onne Janssen and Kan Shi (2011), “Leader Trust and Employee
Voice: The Moderating Role of Empowering Leader Behaviors,” The Lead-
ership Quarterly, 22 (4), 787–98.
Goodman, Joseph K., Cynthia E. Cryder and Amar Cheema (2012), “Data
Collection in a Flat World: The Strengths and Weaknesses of Mechani-
cal Turk Samples,” Journal of Behavioral Decision Making, [online first at
http://dx.doi.org/10.1002/bdm.1753].
Hirschman, Albert O. (1970), Exit, Voice, and Loyalty: Responses to Decline
in Firms, Organizations, and States, Cambridge, MA: Harvard University
Press.
Hirst, Giles, Daan van Knippenberg and Jing Zhou (2009), “A Cross-Level
Perspective on Employee Creativity: Goal Orientation, Team Learning
Behavior, and Individual Creativity,” Academy of Management Journal, 52
(2), 280–93.
Hofmann, David A. (1997), “An Overview of the Logic and Ratio-
nale of Hierarchical Linear Models,” Journal of Management, 23 (6),
723–44.
Hofmann, David A. and Mark B. Gavin (1998), “Centering Decisions in Hierar-
chical Linear Models: Implications for Research in Organizations,” Journal
of Management, 24 (5), 623–41.
House, James S. (1981), Work Stress and Social Support, Reading, MA: Addison
Wesley.
Katzenbach, Jon R. and Jason A. Santamaria (1999), “Firing Up the Front Line,”
Harvard Business Review, 77 (3), 107–17.
Konovsky, Mary A. and S.D. Pugh (1994), “Citizenship Behavior and Social
Exchange,” Academy of Management Journal, 37 (3), 656–69.
LePine, Jeffrey A. and Linn Van Dyne (1998), “Predicting Voice Behavior in
Work Groups,” Journal of Applied Psychology, 83 (6), 853–68.
Liang, Jian, Crystal I. Farh and Jiing-Lih Farh (2012), “Psychological
Antecedents of Promotive and Prohibitive Voice: A Two-Wave Examina-
tion,” Academy of Management Journal, 55 (1), 71–91.
Luo, Xueming, Aric Rindfleisch and David K. Tse (2007), “Working with Rivals:
The Impact of Competitor Alliances on Financial Performance,” Journal of
Marketing Research, 44 (1), 73–83.
MacKenzie, Scott B., Philip M. Podsakoff and Michael Ahearne (1998), “Some
Possible Antecedents and Consequences of In-Role and Extra-Role Sales-
person Performance,” Journal of Marketing, 62 (3), 87–98.
Meyer, John P. and Natalie J. Allen (1984), “Testing the “Side-Bet Theory” of
Organizational Commitment: Some Methodological Considerations,” Jour-
nal of Applied Psychology, 69 (3), 372–8.
Meyer, John P., Sampo V. Paunonen, Ian R. Gellatly, Richard D. Goffin and
Douglas N. Jackson (1989), “Organizational Commitment and Job Perfor-
mance: It’s the Nature of the Commitment That Counts,” Journal of Applied
Psychology, 74 (1), 152–6.
Mittal, Vikas, John W. Huppertz and Adwait Khare (2008), “Customer Com-
plaining: The Role of Tie Strength and Information Control,” Journal of
Retailing, 84 (2), 195–204.
Morrison, Elizabeth Wolfe, Sara L. Wheeler-Smith and Dishan Kamdar (2011),
“Speaking Up in Groups: A Cross-Level Study of Group Voice Climate and
Voice,” Journal of Applied Psychology, 96 (1), 183–91.
Needleman, Sarah E. (2011), “One-Third of U.S. Workers Ready to Quit,”
(retrieved on 09.08.11), [from http://blogs.wsj.com/in-charge/2011/06/20/
one-third-of-u-s-workers-ready-to-quit/].
Ng, Thomas W.H. and Daniel C. Feldman (2011), “Employee Voice Behavior: A
Meta-Analytic Test of the Conservation of Resources Framework,” Journal
of Organizational Behavior, 33 (2), 216–34.
O’Reilly, Charles A. and David F. Caldwell (1981), “The Commitment and Job
Tenure of New Employees: Some Evidence of Post-Decisional Justification,”
Administrative Science Quarterly, 26 (4), 597–616.
Ordanini, Andrea and A. Parasuraman (2011), “Service Innovation Viewed
Through a Service-Dominant Logic Lens: A Conceptual Framework and
Empirical Analysis,” Journal of Service Research, 14 (1), 3–23.
Ormrod, Jeanne Ellis (2003), Human Learning, 4th ed. Upper Saddle River, NJ:
Pearson Prentice Hall.
Osterman, Paul and Beth Shulman (2011), Good Jobs America: Making Work
Better for Everyone, New York: Russell Sage.
Ostrom, Amy L., Mary Jo Bitner, Stephen W. Brown, Kevin A. Burkhard,
Michael Goul, Vicki Smith-Daniels, Haluk Demirkan and Elliot Rabinovich
(2010), “Moving Forward and Making a Difference: Research Priorities for
the Science of Service,” Journal of Service Research, 13 (1), 4–36.
Payne, Stephanie C. and Sheila Simsarian Webber (2006), “Effects of Service
Provider Attitudes and Employment Status on Citizenship Behaviors and
Customers’ Attitudes and Loyalty Behavior,” Journal of Applied Psychology,
91 (2), 365–78.
Ping, Robert A. Jr. (1997), “Voice in Business-to-Business Relationships: Cost-
of-Exit and Demographic Antecedents,” Journal of Retailing, 73 (2), 261–81.
Raudenbush, Stephen W. and Anthony S. Bryk (2002), Hierarchical Linear
Models: Applications and Data Analysis Methods, London: Sage Publica-
tions.
Rusbult, Caryl E., Dan Farrell, Glen Rogers and Arch G. Mainous (1988),
“Impact of Exchange Variables on Exit, Voice, Loyalty, and Neglect: An
Integrative Model of Responses to Declining Job Satisfaction,” Academy of
Management Journal, 31 (3), 599–627.
Segal, David (2012), “Apple’s Retail Army, Long on Loyalty but Short
on Pay,” (retrieved on 06.23.12), [from http://www.nytimes.com/
2012/06/24/business/apple-store-workers-loyal-but-short-on-pay.html?
pagewanted=all& r=0].
Singh, Jagdip (1990), “Voice, Exit, and Negative Word-of-Mouth Behaviors: An
Investigation Across Three Service Categories,” Journal of the Academy of
Marketing Science, 18 (1), 1–15.
Spreitzer, Grechen M. (1995), “Psychological Empowerment in the Workplace:
Dimensions, Measurement, and Validation,” Academy of Management Jour-
nal, 38 (5), 1442–65.
Steenburgh, Thomas and Michael Ahearne (2012), “Motivating Salespeople:
What Really Works,” Harvard Business Review, 90 (7/8), 71–5.
Tangirala, Subrahmaniam and Rangaraj Ramanujam (2008), “Exploring Non-
linearity in Employee Voice: The Effects of Personal Control and
Organizational Identification,” Academy of Management Journal, 51 (6),
1189–203.
Umashankar, Nita, Raji Srinivasan and Dustin Hindman (2011), “Develop-
ing Customer Service Innovations for Service Employees: The Effects of
NSD Characteristics on Internal Innovation Magnitude,” Journal of Service
Research, 14 (2), 164–79.
Van Dyne, Linn and Jeffrey A. LePine (1998), “Helping and Voice Extra-Role
Behaviors: Evidence of Construct and Predictive Validity,” Academy of Man-
agement Journal, 41 (1), 108–19.
Venkataramani, Vijaya and Subrahmaniam Tangirala (2010), “When and Why
Do Central Employees Speak Up? An Examination of Mediating and Mod-
erating Variables,” Journal of Applied Psychology, 95 (3), 582–91.
Zhou, Jing and Jennifer M. George (2001), “When Job Dissatisfaction Leads to
Creativity: Encouraging the Expression of Voice,” Academy of Management
Journal, 44 (4), 682–96.