Content uploaded by Minseo Kim
Author content
All content in this area was uploaded by Minseo Kim on Jan 09, 2018
Content may be subject to copyright.
Content uploaded by Minseo Kim
Author content
All content in this area was uploaded by Minseo Kim on Jan 09, 2018
Content may be subject to copyright.
Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=rijh20
Download by: [Central Michigan University] Date: 08 January 2018, At: 06:42
The International Journal of Human Resource
Management
ISSN: 0958-5192 (Print) 1466-4399 (Online) Journal homepage: http://www.tandfonline.com/loi/rijh20
Empowering leadership: leading people to
be present through affective organizational
commitment?
Minseo Kim & Terry A. Beehr
To cite this article: Minseo Kim & Terry A. Beehr (2018): Empowering leadership: leading people
to be present through affective organizational commitment?, The International Journal of Human
Resource Management, DOI: 10.1080/09585192.2018.1424017
To link to this article: https://doi.org/10.1080/09585192.2018.1424017
Published online: 07 Jan 2018.
Submit your article to this journal
View related articles
View Crossmark data
THE INTERNATIONAL JOURNAL OF HUMAN RESOURCE MANAGEMENT, 2018
https://doi.org/10.1080/09585192.2018.1424017
Empowering leadership: leading people to be present
through aective organizational commitment?*
MinseoKim and Terry A.Beehr
Department of Psychology, Central Michigan University, Mount Pleasant, MI, USA
ABSTRACT
This study tested a mediation model in which empowering
leadership was negatively related to three withdrawal
behaviors: lateness, absenteeism, and turnover intention,
with aective organizational commitment as a mediator. With
294 full-time US employees, results from structural equation
modeling indicated that empowering leader behaviors at
one time were positively related to estimates of aective
organizational commitment at a second time, which in turn
was negatively related to absenteeism and turnover intention
at a nal time. Additionally, no signicant direct eect was
found between empowering leadership and withdrawal
behaviors, further supporting the mediation model. However,
neither empowering leadership nor aective commitment
inuenced followers’ lateness. Empowering leadership, which
provides employees with autonomy and developmental
support, may have a favorable eect on employees’ decisions
to attend and stay in the organization, as well as their aective
reaction to the organization in the form of psychological
commitment. This study extended prior research models by
examining a full range of withdrawal behaviors in relation
to empowering leadership and showed that commitment
may explain why empowering leader behaviors can aect
employees’ retention decisions.
Considerable attention has been given to employees’ withdrawal behaviors that
reect work-related attitudes and behavioral responses, because employee with-
drawal (e.g. lateness, absenteeism, and turnover intention) is linked to high
organizational costs (Berry, Lelchook, & Clark, 2012; Hancock, Allen, Bosco,
McDaniel, & Pierce, 2013). Given that employee withdrawal is an ongoing issue
and can cause a substantial nancial burden for the organization, it is important
to determine potential antecedents that are closely associated with withdrawal
behaviors. Lateness, absenteeism, and turnover are conceptually similar behaviors
in that employees can engage in them by being out of the organization. ey have
© 2018 Informa UK Limited, trading as Taylor & Francis Group
KEYWORDS
Empowering leadership;
affective organizational
commitment; lateness;
absenteeism; turnover
intention
CONTACT M inseo Kim kim4m@cmich.edu
*Some of these data were presented at the August 2016 Academy of Management meeting in Anaheim CA.
Downloaded by [Central Michigan University] at 06:42 08 January 2018
2 M. KIM AND T. A. BEEHR
long been considered forms of employee withdrawal (e.g. Beehr & Gupta, 1978;
Hanisch & Hulin, 1990, 1991), but they dier in terms of timing (lateness at the
beginning of the workday and absenteeism all of the work day) and permanence
(with turnover the employee never comes back to work), and their measures do
not necessarily correlate so highly with each other that they are redundant (r=.20
to .57 in the present study). Hanisch and Hulin argued they might be forms of a
larger construct of withdrawal that would also include retirement. ey found
however, that the three forms of withdrawal were not part of a single latent factor,
because lateness and absenteeism were more likely to form their own latent factor
without turnover intention. ey reported that retirement intention might form a
latent factor with turnover, but subsequent research showed that retirement was
quite dierent from the others (Adams & Beehr, 1998). Overall, organizational
research has been interested in employee withdrawal – ways that employees are
not present in organizations; most studies have not attempted to combine these
three withdrawal behaviors into a single latent construct but have instead inves-
tigated them separately.
A body of research has addressed this topic and suggested several predictors
of withdrawal behaviors. ese include job satisfaction, leadership styles, and
characteristics of the work environment such as stress and autonomy (Ferris, 1985;
Frooman, Mendelson, & Kevin Murphy, 2012; Grieth, Hom, & Gaertner, 2000;
Waldman, Carter, & Hom, 2015). e present study contributes to the withdrawal
literature by focusing on a specic leadership style – empowering leadership – in
relation to a broad array of withdrawal criteria: lateness, absenteeism, and turn-
over intentions.
Amundsen and Martinsen (2014) dene empowering leadership as motivat-
ing an employee intrinsically by sharing power and by providing support for the
employee’s development. ere are many ways for appointed leaders to attempt to
inuence their followers, and empowering leadership can be seen as a somewhat
unique way, because at its core, the followers are leading themselves, a kind of
downward power transfer (Amundsen & Martinsen, 2014); some theorists clas-
sify empowering leadership as a form of subordinates’ self- or shared-leadership
(e.g. Pearce & Sims, 2002; Pearce et al., 2003). e seeming paradox of a leader
inuencing people to lead themselves is solved by empirical research identifying
the behaviors comprising empowering leadership. For example, it is encouraging
(rather than directing or forcing) subordinates’ opportunity thinking, self-reward,
self-leadership, participation in goal setting, and teamwork (Pearce et al., 2003).
Because the leadership domain is well-established and contains many leader-
ship types or constructs, when empowering leadership began to be discovered,
promoted and measured, it was necessary to show its relationship with and dif-
ferences from other leadership types. Empowering leadership is somewhat more
closely related to consideration than to initiating structure for example, and to
participative decision-making and showing concern, among several managerial
behaviors (Arnold, Arad, Rhoades, & Drasgow, 2000), and it is positively related
Downloaded by [Central Michigan University] at 06:42 08 January 2018
THE INTERNATIONAL JOURNAL OF HUMAN RESOURCE MANAGEMENT 3
to both transformational leadership and leader-member exchange (e.g. Amundsen
& Martinsen, 2014). Important for its discriminant validity, however, the strength
of its relationship with other types of leadership is about the same as other lead-
ership types have with each other (Arnold et al., 2000), factor analyses show it
is empirically distinguishable from other types of leadership (e.g. Pearce et al.,
2003; Tekleab, Sims, Yun, Tesluk, & Cox, 2008), and it predicts variance in crite-
rion variables even aer controlling for other leadership styles (e.g. Amundsen &
Martinsen, 2014; Arnold et al., 2000; Pearce & Sims, 2002; Tekleab et al., 2008).
e autonomy provided by empowering leaders helps employees experience
more responsibility and expand their skills, because these experiences can theoret-
ically lead to aective commitment (Johnson, Chang, & Yang, 2010). Empowering
leader styles are oen advocated, because they are expected to lead to employee
behaviors that benet the organization, but relatively little research has investi-
gated the links between empowering leadership and employee withdrawal. In
order to elaborate on research ndings showing that some other leadership styles
had a signicant relationship with employee withdrawal processes and actual
turnover (Ferris, 1985; Frooman et al., 2012; Waldman et al., 2015), the present
study examines the relationship between empowering leadership and a wide array
of subordinates’ withdrawal behaviors. Furthermore, if empowering leadership is
indeed theoretically related to less employee withdrawal, we propose and test an
explanatory mechanism for why this relationship occurs, i.e. the development of
increased organizational commitment. us, the present study contributes specif-
ically to the empowering leadership literature by determining the extent to which
this newer form of leader behavior can favorably inuence employees’ retention
decisions and multiple forms of withdrawal.
If empowering leadership can aect employees’ withdrawal, then that eect
should show across as variety of withdrawal types. A meta-analysis examined
the relationship between employees’ and groups’ feelings of empowerment and
one indicator of employee withdrawal, turnover intentions (Seibert, Wang, &
Courtright, 2011), but the present study focused on leaders’ behaviors that spe-
cically are empowering as a predictor of a wide range of withdrawal indicators.
Consistent with our own review of the literature, a review by Sharma and Kirkman
(2015) reported that one previous study examined the relationship between
empowering leadership and turnover intentions (Chen, Sharma, Edinger, Shapiro,
& Farh, 2011), but little or no research has examined the eects of empowering
leader behaviors at the individual level on a full range of withdrawal behaviors.
at is, one study examined the relationship between empowering leadership and
turnover intentions, but (1) it was a cross-level study of leadership at the team level
and turnover intentions at the individual level, and (2) it only addressed turnover
intentions but not other withdrawal behaviors (i.e. lateness and absenteeism in the
present study). e concept of employee withdrawal is broader than just turno-
ver, however, and we sought to determine whether empowering leadership could
predict it. To extend prior research models, the present study takes into account
Downloaded by [Central Michigan University] at 06:42 08 January 2018
4 M. KIM AND T. A. BEEHR
three forms of work withdrawal indicators, namely lateness, absenteeism, and
turnover intention as consequences of empowering leadership and tests a model
illustrating how empowering leaders may aect withdrawal decisions of followers.
In order to learn more about the psychological process resulting in employee
withdrawal, the present study proposes the mediator, aective organizational com-
mitment that is a psychological tie binding an employee with the organization.
As employees’ commitment plays an inuential role in predicting behavioral out-
comes (Meyer, Stanley, Herscovitch, & Topolnytsky, 2002; Shore & Wayne, 1993),
it could be a potential mediator that may be inuenced by leader empowerment
styles and may help explain how empowering leadership behaviors (ELBs) can
help to diminish followers’ withdrawal behaviors (Figure 1). e present study
contributes to the literature on empowering leadership by examining its ability to
predict a wide range of individuals’ voluntary withdrawal in the forms of lateness,
absenteeism, and turnover intentions via employee commitment.
Model development and hypotheses
Employee empowerment is a positive state that can lead to favorable employee
reactions including performance (e.g. Rapp, Gilson, Mathieu, & Ruddy, 2016),
and the very notion of ELBs stems from the idea that there are specic things
that leaders can do to make subordinates feel empowered. Manz and Sims (1987)
argued that the role of leaders in empowered work environments is dierent
from the role of traditional leaders in terms of ‘the shi in source of control from
the leader to the follower’ and ‘leading others to lead themselves’ (p. 119). ey
emphasized the development of followers’ self-management or self-leadership
(Manz & Sims, 1987, 1991). Arnold et al. (2000) described ELBs as behaviors
that facilitate team members’ performance by raising their level of autonomy,
encouraging subordinates to express opinions and ideas, promoting teamwork,
Figure 1.Hypothesized model.
Downloaded by [Central Michigan University] at 06:42 08 January 2018
THE INTERNATIONAL JOURNAL OF HUMAN RESOURCE MANAGEMENT 5
information sharing, and collaborative decision-making. e above two deni-
tions have been operationalized in research at the team level especially, focusing
on the role of leaders in self-managing work teams.
However, at the individual level, the eect of ELBs on outcomes needs to be
developed more fully, because leaders may treat and empower individual subor-
dinates dierently. In fact, LMX theory argues that leaders usually do not treat
all subordinates the same, suggesting that all leadership behaviors should be
examined at the individual level (Graen & Uhl-Bien, 1995). Researchers have
suggested that empowering leadership specically can be seen as a dyadic relation-
ship between a supervisor and an individual follower (Ahearne, Mathieu, & Rapp,
2005; Zhang & Bartol, 2010). Supporting this view, a few studies of empowering
leadership practices were conducted at the individual subordinate level (Keller &
Dansereau, 1995), because supervisors dierentiated among subordinates in the
amounts of latitude they allowed (Ahearne et al., 2005). Additionally, a study by
Zhang and Bartol (2010) supported the analysis of empowering leadership at an
individual level, showing that there was no signicant dierence between with-
in-group variance and between- group variance. Results of these studies imply
that focusing on the eect of ELBs at the individual-level (as in the present study)
is appropriate. It is expected that ELBs could aect individual subordinates to a
greater extent or at least to the same degree as teams and as a result, may lead to
dierent levels of psychological responses and behavioral outcomes compared to
those of team-level studies.
More recently, at the individual level, Amundsen and Martinsen (2014) con-
ceptualized the key elements of empowering leadership as follows: Empowering
leadership consists of inuencing a subordinate in three ways: (1) power sharing
and (2) motivation support, which are two forms of autonomy; and (3) develop-
ment support. ese are intended ‘to promote their experience of self-reliance,
motivation, and capability to work autonomously within the boundaries of overall
organizational goals and strategies’ (p. 489). e process of power sharing and
motivation support reects the leader behaviors of delegating, coordinating, infor-
mation sharing, encouraging initiative, encouraging goal focus, ecacy support,
and inspiring. e process of development support represents the behaviors of
modeling and guiding. Consistent with Amundsen and Martinsen (2014), the
present study examined ELBs as these three behaviors directed at individuals.
Previous research provided support for the inuence of empowerment on some
favorable attitudes and behaviors at the individual level including job satisfaction
(e.g. Amundsen & Martinsen, 2015), work engagement (e.g. Tuckey, Bakker, &
Dollard, 2012), career success (e.g. Kim & Beehr, 2017a), in-role and extra-role
behaviors (e.g. Humborstad, Nerstad, & Dysvik, 2014; Raub & Robert, 2010), and
lowered depression (e.g. Kim & Beehr, 2017b). erefore, if a leader successfully
empowers subordinates, both employees and employers may benet.
Taken together, several studies suggested that empowering leadership was asso-
ciated with positive work outcomes, and thus empowering leadership can be an
Downloaded by [Central Michigan University] at 06:42 08 January 2018
6 M. KIM AND T. A. BEEHR
eective leadership style for both employees and organizations. Considering the
conceptualization of empowering leadership, empowering leader behaviors are
likely to be associated with positive work outcomes because employees feel some
form of intrinsic motivation, a positive feeling about and attraction to the work
and the workplace. e present study tests the resulting positive feeling in the
form of aective organizational commitment and attraction in the form of (less)
employee withdrawal from the organization, as modeled in Figure 1.
Relationship between empowering leadership and aective organizational
commitment
As noted earlier, empowering leadership emphasizes followers’ autonomy, partic-
ipation, and development through the encouragement of self-direction, and thus
empowering leaders are willing to share power with followers and support them.
Accordingly, empowering leadership will elicit positive psychological reactions
and attitudes from followers. e present study focuses on followers’ aective
commitment, an ‘emotional attachment to, identication with, and involvement in
the organization’ as a result of positive work experiences (Meyer & Allen, 1991, p.
67). Aective commitment reects a deep emotional relationship of the employee
with the organization (Meyer & Allen, 1991) as opposed to remaining because
of feelings of obligation or working specically for tangible reasons (e.g. nan-
cial return). Because empowering leaders behave in ways that foster motivation
and ecacy among followers, as well as promote their involvement in the work
processes, followers may feel more condent and have positive experiences and
emotions about their work. Aective commitment especially develops through
work experiences such as interactions with supervisors and work group (Meyer
& Allen, 1997). us, aective commitment (a kind of positive feeling) is more
likely to represent the inuence of empowering leader behaviors than the two
other forms of commitment.
e employee development behaviors of empowering leadership include mod-
eling, guiding, and coaching to give followers opportunities to increase their
job-related skill, ability, and learning and mastery experiences, thereby enhancing
personal ecacy or perceived personal competence. is feeling may be related to
raising aective organizational commitment according to psychological contract
theory. In that theory, employees tend to consider promises from their leader to
be promises from their organization (Rousseau, 1998), and in turn, reciprocity
theory (e.g. Blau, 1964; Cropanzano & Mitchell, 2005) predicts that employees
will show more loyalty and commitment to the organization. In a similar vein,
organizational support theory maintains that employees view treatment, eval-
uation, and support received from leaders as indicative of organizational treat-
ment, evaluation, and support (Eisenberger et al., 2010). ese theories suggest
that employees perceive their leaders as representing the organization. Previous
research argued that employees tend to become committed to their organization to
Downloaded by [Central Michigan University] at 06:42 08 January 2018
THE INTERNATIONAL JOURNAL OF HUMAN RESOURCE MANAGEMENT 7
the extent that their organization or leader provides for growth and achievement
needs (Mathieu & Zajac, 1990). us, the more employees receive developmen-
tal support from empowering leaders, the more they will likely have a favorable
exchange relationship with the leader or organization, developing positive mood
and feelings of obligation, willingness to emotionally bond with their organization.
In addition to empowering leadership in the form of development support,
empowering leadership facilitates followers’ self-leadership with the behaviors
such as the delegation of decision-making, information sharing, encouraging
personal initiative, self-dened goal focus, and inspiring, which is represented by
the autonomy support process facets of empowering leadership. When employees
are encouraged to autonomously self- manage their work and to be accountable
for work outcomes in their organization, they are more likely to foster intrinsic
motivation and favorable attitudes. Eects of such motivation and attitudes are
informed by goal setting theory (Erez & Arad, 1986) and self-determination the-
ory (Deci & Ryan, 1985). A participative goal-setting process increases employees’
experiences of intrinsic motivation and a sense of self-control, providing them
with opportunities to satisfy their higher psychological needs for autonomy and
personal growth. Self-determination theory also maintains that employees are
intrinsically motivated when their needs of autonomy, competence, relatedness
are met. Empowering leadership can be seen as an autonomy-supporting form
of leadership, thereby satisfying the psychological needs and promoting intrinsic
motivation of employees, which consequently results in positive attitudes such
as aective commitment. Supporting this assumption, in a study by Chen et al.
(2011), ELBs including expressing condence in the team’s ability to accomplish
its task successfully, allowing the team to self-manage its work and make deci-
sions on its own led team members to feel more emotionally involved in work
processes, thereby feeling aectively committed to their organization. It was also
found that followers feel more deeply committed to the organization when they
are given individualized support and sucient opportunity to voice opinions on
their work (Den Hartog & De Hoogh, 2009). Given the norm of reciprocity and
social exchange theory (Blau, 1964; Cropanzano & Mitchell, 2005), individual
employees are likely to reciprocate benecial empowering leader behaviors by
demonstrating higher aective commitment when they perceive that leaders take
care of them by providing the needed support, consulting about important deci-
sions, giving more autonomy, and removing unnecessary bureaucratic constraints.
Hypothesis 1: Empowering leadership is positively related to aective commitment.
Relationship between aective commitment and withdrawal behaviors
Dierent types of organizational withdrawal including lateness, absenteeism, and
turnover intention may be signs of employees’ negative behaviors toward the job
and the organization, representing a distancing from work physically and psycho-
logically. Employee lateness refers to arriving late at work, whereas absenteeism
Downloaded by [Central Michigan University] at 06:42 08 January 2018
8 M. KIM AND T. A. BEEHR
is missing at least a full day of work (Koslowsky, Sagie, Krausz, & Singer, 1997).
Turnover intention can be dened as thinking of leaving, desirability of leaving,
and likelihood of leaving the workplace or one’s current job (Hanisch & Hulin,
1990, 1991). Previous studies do not seem to provide a single model that explains
the relationship between the dierent withdrawal behaviors. Some researchers
argued that no relationship exists (Ross, 1988), while others suggested positive
relationships or a spillover model (Iverson & Deery, 2001). ese inconsistent nd-
ings imply that it may be necessary to seek out the dierences between withdrawal
behaviors through predictors such as empowering leadership and job attitudes.
e present study, therefore, examines whether each of three forms of withdrawal
behaviors can be inuenced by the perception of leaders’ empowering behaviors
and followers’ aective commitment.
Employees will be likely motivated to care about their attendance or withdrawal
when they experience the specic positive emotion of aective commitment to
their organization. Such employees will be less inclined to be late, absent, and
leave to their organization than those who experience negative commitment. In
addition, employees who are emotionally committed to the organization may
want to stay and maintain membership in their organization, which makes them
exert eorts for the organization and avoid undesirable behaviors. Supporting
this assumption, aective commitment was found to have a negative relationship
with lateness frequency and lateness duration (Foust, Elicker, & Levy, 2006), as
well as with absenteeism and a positive relationship with intent to remain (e.g.
El Akremi, Colaianni, Portoghese, Galletta, & Battistelli, 2014; Jaros, Jermier,
Koehler, & Sincich, 1993; Somers, 1995; anacoody, Newman, & Fuchs, 2014).
Further, meta-analyses showed that organizational commitment was negatively
related to withdrawal behaviors including lateness, absenteeism, turnover inten-
tion, and turnover (Grieth et al., 2000; Harrison, Newman, & Roth, 2006; Tett
& Meyer, 1993).
ese ndings suggested that staying with an organization was strongly asso-
ciated with feelings of loyalty, aection, and belongingness, which are all com-
ponents of aective commitment. In other words, employees’ level of aective
commitment inuences formation of a tendency to leave from or stay with the
organization, and thus an individual who is aectively committed with the organ-
ization will be less likely to think about quitting and other withdrawal processes
such as lateness and absenteeism (Jaros et al., 1993). erefore, it is assumed that
the emotional bonds to the organization stemming from high levels of aective
commitment result in diminishing lateness, absenteeism, and turnover intention.
Hypothesis 2: Aective commitment is negatively related to lateness.
Hypothesis 3: Aective commitment is negatively related to absenteeism.
Hypothesis 4: Aective commitment is negatively related to turnover intention.
Downloaded by [Central Michigan University] at 06:42 08 January 2018
THE INTERNATIONAL JOURNAL OF HUMAN RESOURCE MANAGEMENT 9
Aective commitment as a mediator
If supervisors’ empowering leadership style can enhance employees’ aective
commitment, empowering leadership may have a favorable eect on employees’
decisions to attend and stay in the organization. Because leaders can be considered
as representatives of the organization (Rhoades & Eisenberger, 2002; Simons &
Roberson, 2003), their actions can result in employee reactions to the whole work-
place, represented by aective organizational commitment. Employees’ presence
at work (lack of withdrawal) can therefore be a reaction to the leader that is trans
-
ferred by the eect of the leader behaviors on the employee’s commitment. Taken
together, therefore, the model in Figure 1 and the set of hypotheses 1 through 4
propose that aective commitment can help to explain the relationship between
empowering leadership and withdrawal behaviors. us, the nal hypothesis is
the most important and proposes a mediation eect.
Hypothesis 5a: Empowering leadership predicts lateness by means of aective
commitment.
Hypothesis 5b: Empowering leadership predicts absenteeism by means of aective
commitment.
Hypothesis 5c: Empowering leadership predicts turnover intention by means of aec-
tive commitment.
e present study examines employee commitment as a mediator explaining
why empowering leadership can result in a full range of employee withdrawal
behaviors: lateness, absenteeism, and turnover intention. By testing the model
developed for Figure 1, the present study contributes by reporting the potential
eects of empowering leadership on a wide range of employee withdrawal behav-
iors at the individual level. Importantly, an individual aective variable, aective
commitment, is proposed as the theoretical explanation for why this relationship
may exist.
Methods
Participants and procedure
Participants were recruited through Amazon’s Mechanical Turk, an online service
for connecting researchers and respondents. Eligible participants were full-time
employees working in the United States and holding at least a 95% approval rating
from previous Mturk assignments, consistent with existing recommendations (e.g.
Mason & Suri, 2012). Data were collected at two time points three weeks apart
in order to reduce the eects of common method bias (Podsako, MacKenzie, &
Podsako, 2012). To further reduce response bias eects, social desirability was
used as a control variable in case there was socially desirable responding about
employees’ own withdrawal from the organization. Initially, 599 participants com-
pleted the rst survey, but 517 responses remained aer dropping 82 responses
due to low eort responding (e.g. above 80% of the same answers or quick answers
Downloaded by [Central Michigan University] at 06:42 08 January 2018
10 M. KIM AND T. A. BEEHR
– completing the survey more than twice as fast as the average time) on this
survey. Of these, 311 employees completed the second survey. Aer removing
unmatched codes from Time 1 and Time 2 of the survey, 294 matched responses
were available from all time points to be used for analyses. Of this nal sample,
57.3% were female, 83.0% were white, 65.7% had at least a bachelor’s degree, and
61.8% were between the ages of 18 and 39. e participants were employed in a
variety of industries (e.g. education, healthcare, and nance).
Measures
e rst survey (Time 1) measured the predictor, empowering leadership, and
demographics. e second survey (time 2) measured the three criteria, withdrawal
behaviors, and social desirability. e mediator, aective commitment, was meas-
ured at both times to use their mean as an estimate of the score at the midpoint
(designated as Time 1 ½ in Figures 1 and 2) between Time 1 and Time 2. e
mediator should have its eects between Time 1 and Time 2, but instituting an
additional data collection is likely to further reduce sample size, and therefore we
used a dierent method to estimate mediator scores between Time 1 and Time
2. If a variable changes its values from Time 1 and Time 2, its value between
these two time periods is highly likely to be between its values at those two time
points. Below, we report statistical evidence that scores on the mediator indeed
were dierent between Times 1 and 2, thereby suggesting that the mean between
the two measurements captures people’s scores at some point between Times 1
and 2. is method was also explained and used successfully in a recent study on
leadership (Kim & Beehr, 2017c).
Figure 2.Structural equation modeling with standardized estimatesa.
Note: aPaths in structural model analysis are significant at p<.01. Paths with dashed lines are nonsignificant.
Downloaded by [Central Michigan University] at 06:42 08 January 2018
THE INTERNATIONAL JOURNAL OF HUMAN RESOURCE MANAGEMENT 11
Empowering leadership was measured using the 18-item Empowering Leadership
Scale (ELS) (Amundsen & Martinsen, 2014). Amundsen and Martinsen described
empowerment conceptually as consisting of three dimensions, power sharing,
development support, and motivational support, but aer factor analyses of their
scale, they concluded that there were two main factors, autonomy support and
developmental support; power sharing and motivational support were combined
into one factor and labeled as autonomy support. In order to have more than two
manifest indicators of empowering leadership, we used the development support
facet as one indicator and the two conceptually separate parts of the autonomy
support facet, labeled as Amundsen and Martinsen labeled them, motivational
support, and the other power sharing. We also compared a 2-factor model with the
3-factor model. e result of a conrmatory factor analysis produced the same t
indices: the two factors model consisting of autonomy support and developmen-
tal support, χ2(134, N=294) = 731.42, p<.01, CFI=.96, IFI=.96, NNFI=.96,
RMSEA=.12; the three factor model loading the power sharing, motivational
support, and developmental support separately, χ
2
(132, N=294) = 653.92, p<.01,
CFI=.96, IFI=.96, NNFI=.96, RMSEA=.12.
Four items (α= .88) represent power sharing, six items (α= .93) represent
development support, and eight items (α=.91) represent motivational support.
An example item is ‘My leader encourages me to take initiative’ rated on a 7-point
Likert scale from 1 (strongly disagree) to 7 (strongly agree). e three subscales
were highly correlated each other, ranging from r=.73 to r=.81. Reliability for
the full scale was .96.
Aective commitment was assessed with eight items (Time 1 α=.90; Time 2
α=.92) from Allen and Meyer’s (1990) scale. An example item is ‘is organiza-
tion has a great deal of personal meaning for me,’ answered on a 7-point Likert
scale from 1 (strongly disagree) to 7 (strongly agree). e mean absolute dierence
in aective commitment from T1 to T2 was .58 with a standard deviation of .65,
indicating that many employees’ scores had changed (increasing for some employ-
ees and decreasing for others) during this three-week interval. A one-sample t-test
showed that the average change was signicant, t(293) = 15.38, p<.01. erefore,
the mean of T1 and T2 was used to estimate the score between T1 and T2.
Lateness was measured with the three items used in Hanisch and Hulin (1990,
1991). An example item is ‘How oen are you late for work or scheduled assign-
ments?’ (α=.71), rated on a 7-point Likert scale from 1 (never) to 7 (every day).
Absenteeism was also measured with three items from Hanisch and Hulin (1990,
1991) on the same 7-point scale as lateness. An example item is ‘How oen do
you think about being absent from work when you are scheduled to be there?’
(α=.72).
Turnover intention was measured using the ve items (α=.94) from Walsh,
Ashford, and Hill (1985). An example item is ‘I am thinking about quitting my
job’ rated on a 5-point Likert scale from 1 (disagree) to 5 (agree).
Downloaded by [Central Michigan University] at 06:42 08 January 2018
12 M. KIM AND T. A. BEEHR
Social desirability was used as a control variable, because it is desirable for
employees to report their faithful attendance at work. It was measured with the
ve-item (α = .75) Socially Desirable Response Set (SDRS) (Hays, Hayashi, &
Stewart, 1989). An example item is ‘No matter who I’m talking to, I am always
a good listener.’ Respondents were asked to indicate the degree to which each
statement is true or false on a 5-point Likert scale from 1 (denitely true) to 5
(denitely false). Eects of the model on the criteria were thus controlled for
social desirability; importantly, eects were also controlled for common method
variance, because social desirability was measured with the same method, similar
to the concept of a marker variable analysis (Podsako et al., 2012).
Results
Means, standard deviations, reliabilities, and correlations of all the variables are
presented in Table 1. Empowering leadership was positively related to the medi-
ator, aective commitment (r=.60, p<.01); it was also negatively related to two
of the criteria (empowering leadership to absenteeism, r=−.21, p<.01, and to
turnover intention, r=−.43, p<.01). e mediator was negatively related to the
same two criteria (aective commitment to absenteeism, r=−.30, p<.01 and to
turnover intention r=−.66, p<.01).
Because empowering leadership and aective commitment were highly cor-
related (r=.60, p<.01), a conrmatory factor analysis (CFA) using LISREL 8.8
was run to test the t of a two-factor model for these measures. e two-factor
model produced an excellent t, χ2(8, N=294) = 16.13, p<.05; RMSEA=.06;
SRMR=.03; GFI=.98; NNFI=.99; CFI=1.00, which was better than a one-factor
model, χ2(9, N=294) = 453.47, p<.01; RMSEA=.41; SRMR=.14; GFI=.66;
NNFI=.63; CFI=.78; this provided evidence of discriminant validity of empow-
ering leadership and aective commitment.
We also ran CFAs focused on the outcome variable set (withdrawal behav-
iors). A three-factor and one-factor model were compared; the three-factor model
showed a better t on all indices, χ2 (41, N=294) = 199.47, p<.01, CFI=.95,
IFI=.95, NNFI=.93, RMSEA=.11 than the one-factor model, χ2 (44, N=294)
= 935.07, p<.01, CFI=.77, IFI=.77, NNFI=.71, RMSEA=.26. Because Hanisch
Table 1.Means, standard deviations, reliabilities, and correlations of all variables.
Notes: N=294. Reliabilities are in italics on the diagonal.
**p<.01; *p<.05.
Variables Mean SD 1 2 3 4 5 6
1. Empowering leadership (T1) 4.89 1.14 .96
2. Affective commitment (mean
of T1 + T2)
4.53 1.29 .60** .90/.92
3. Lateness (T2) 2.05 1.10 −.09 −.11 .71
4. Absenteeism (T2) 2.30 1.20 −.21** −.30** .57** .72
5. Turnover intention (T2) 2.32 1.13 −.43** −.66** .20** .38** .94
6. Social desirability (T2) 3.59 .75 .10 .12* −.21** −.24** −.11 .75
Downloaded by [Central Michigan University] at 06:42 08 January 2018
THE INTERNATIONAL JOURNAL OF HUMAN RESOURCE MANAGEMENT 13
and Hulin (1990, 1991) had reported two factors (turnover versus absenteeism
and lateness), we also tested a two-factor model, but it t the data more poorly
than the three-factor model on all t indices, χ2 (43, N=294) = 244.90, p<.01,
CFI=.93, IFI= .93, NNFI = .91, RMSEA=.13. erefore, the three facets of
withdrawal behaviors used in the present study can be empirically distinct, and
we examined them as separate outcomes.
Hypotheses and model testing
Structural equation modeling (SEM) was used to test the hypotheses and ver-
ify the indirect eect of empowering leadership on withdrawal behaviors. For
full structural equation modeling with latent variables, empowering leadership’s
three subscales were used as indicators of a latent variable. As recommended by
Matsunaga (2008), the item-parceling method was used for the unidimensional
construct of aective commitment; its eight items were randomly parceled to
form three indicators of a latent variable (two consisting of three items each and
one of two items). For the withdrawal behaviors, each item served as an indicator
of the lateness (three items), absenteeism (three items), and turnover intention
(ve items) constructs.
All the model t indices are shown in Table 2. According to the recommenda-
tions of Kline (2015), multiple indices to assess the t of the model were used: A
satisfactory model t can be inferred when the χ
2
/df ratio is below 3.00 and values
for the comparative t index (CFI), incremental t index (IFI), and nonnormed
t index (NNFI) should not be smaller than .90. In addition to these t indexes,
the root mean square error of approximation (RMSEA) measured model lack of
t. For the RMSEA, values of about .05 or less indicate a close t, values between
.05 and .08 indicate a reasonable t, and values between .08 and .10 indicate
marginal t (Browne & Cudeck, 1992). e values obtained for the CFI, IFI,
NNFI, and RMSEA from the CFA results were .96, .96, .95, and .08, respectively.
Additionally, the value for the χ2/df ratio was 2.74. ese results suggested that
the measurement model had a satisfactory t to the data. e hypothesized model
also t the data moderately well, χ2(202, N=294) = 663.81, p<.01, CFI=.94,
IFI=.94, NNFI=.93, RMSEA=.09. e value of RMSEA was somewhat high,
but still in an acceptable range for marginal t (.08–.10; Browne & Cudeck, 1992).
Table 2.Summary of model fit indices.
Notes: N=294. χ2-values for the structural models are significant at p<.01.
Model test χ2df CFI IFI NNFI RMSEA ∆χ²∆ df p
Measurement model 530.93 194 .96 .96 .95 .08
Hypothesized structural model 663.81 202 .94 .94 .93 .09
Alternative model: Direct paths
from empowering leadership to
each of the three outcomes
660.74 199 .94 .94 .93 .09 3.07 3 .38
Downloaded by [Central Michigan University] at 06:42 08 January 2018
14 M. KIM AND T. A. BEEHR
Additionally, we compared the original model using the average of aective
commitment (Time ½) with the two alternative models: (1) the predictor and
mediator measured at the same time (2) the mediator and criterion variables meas-
ured at the same time. We found that the t statistics for these two new models
were identical at two decimal places; model 1, χ2(202, N=294) = 699.71, p<.01,
CFI=.93, IFI=.93, NNFI=.92, RMSEA=.09, and model 2, χ2 (202, N=294)
= 686.22, p<.01, CFI=.93, IFI=.93, NNFI=.92, RMSEA=.09. In addition,
neither model ts the data better than the original model. Studying the results
of each model, we think the best explanation of any dierences in the models is
more psychometric than theoretical. at is, the T1-T1-T2 model nds stronger
parameter coecients for the le side of the model, and the T1-T2-T2 model
nds stronger parameter coecients for the right side of the model. erefore, the
results are entirely in line with a common-method explanation: When variables
were measured at the same time, they related more strongly to each other than
when they were measured at dierent points in time. us, we tested hypotheses
with the original form of the mediator (T1½).
Standardized parameters are in Figure 2. Empowering leadership was posi-
tively related to aective commitment (β=.66, p<.01), supporting Hypothesis
1, the rst link of the model. Hypothesis 2, that aective commitment would be
negatively related to lateness, was not supported because the path coecient for
aective commitment to lateness (β=−.13, ns) was not signicant. However, the
results supported Hypothesis 3, that aective commitment would be negatively
related to absenteeism (β=−.17, p<.01), as well as Hypothesis 4, that aective
commitment would be negatively related to turnover intention (β=−.69, p<.01).
Social desirability was included as a control variable, because it might espe-
cially aect the self-reports of withdrawal behaviors. Social desirability showed
signicant relationships with lateness (r=−.21, p<.01) and absenteeism (r=−.24,
p<.01), but not with turnover intention (r=−.11, ns). Although the path coef-
cients from social desirability to lateness (β=−.31, p<.01) and to absenteeism
(β=−.35, p<.01) were signicant, the conclusions in the present study were still
the same aer controlling the eect of social desirability. In summary, the SEM
results suggested that empowering leadership had indirect eects on absenteeism
and turnover intention via aective commitment but not on lateness, and these
eects remained even aer controlling for social desirability response sets.
In addition to testing each hypothesis, an alternative model was tested, in which
empowering leadership had three direct paths to the criteria: lateness, absentee-
ism, and turnover intention. e alternative model also provided an adequate t
to the data, χ2 (199, N=294) = 660.74, p<.01, CFI=.94, IFI=.94, NNFI=.93,
RMSEA=.09. However, it did not improve the χ
2
t signicantly, ∆χ
2
(3, N=294)
= 3.07, p=.38, and none of the t statistics of the original model was changed.
Furthermore, direct eects between empowering leadership and the three types
of withdrawal behaviors were all nonsignicant. erefore, given no noticeable
Downloaded by [Central Michigan University] at 06:42 08 January 2018
THE INTERNATIONAL JOURNAL OF HUMAN RESOURCE MANAGEMENT 15
improvement in t, the originally hypothesized model was judged as the practi-
cally better model because it was more parsimonious.
Mediation and bootstrapping
To further test the mediation eects in the model, bootstrapping analyses were
calculated using the PROCESS macro, a computational tool for mediation analysis
(Hayes, 2013). Table 3 presents the direct eects and bootstrapped estimates for
the indirect eects with 95% condence intervals. Following the recommendations
of Preacher and Kelley (2011), k2 (Kappa-squared) is also reported as a media-
tion eect size. e k2 is not sensitive to sample size, because the k2 is the ratio of
the indirect eect to the maximum possible size of the indirect eect given the
constraints of the data (Hayes, 2013). To determine the criteria for describing the
magnitude of eect sizes, Cohen’s guidelines dening small (.01), medium (.09),
and large (.25) eect sizes were considered (Preacher & Kelley, 2011).
Aective commitment signicantly mediated the relationships between
empowering leadership and two withdrawal behaviors – absenteeism and turn-
over intention – because their condence intervals did not include a zero, sup-
porting mediation eects for these two criteria. As seen in Table 3, the direct
eects from empowering leadership to absenteeism (c’=−.05, p = .47) and to
turnover intention (c’=−.06, p=.24) were not signicant. e indirect eects of
empowering leadership through aective commitment to absenteeism (ab=−.17,
CI (95%): LL=−.26, UL=−.09, k2=.13) and turnover intention (ab=-.37, CI
(95%): LL=−.46, UL=−.28, k2=.34) were signicant. e indirect eect size of
empowering leadership to absenteeism via aective commitment can be inter-
preted as medium strength, whereas the eect size of k2 for aective commitment
for the relationship between empowering leadership and turnover intention can
be interpreted as a large eect. Additionally, the completely standardized indirect
coecients were −.16 and −.37 respectively, indicating that followers’ absenteeism
decreases by .16 and turnover intention by .37 for every one standard deviation
increase in empowering leader behaviors indirectly via aective commitment.
Table 3.Results of bootstrapping tests for estimating indirect effects with 95% confidence inter-
vals.
Notes: N=294. β=c’ (direct effect). ab=unstandardized indirect effect. SE=bootstrap standard error. ab, SE, and CI
95% were obtained from 10,000 bootstrap samples. abcs=completely standardized indirect effect. k2=indirect
effect/ maximum possible mediation effect.
Predictor Mediator Outcome
Direct
eects Indirect eects
β (p) ab SE CI 95% abcs k2
Empowering
leadership
Affective
Commit-
ment
Lateness −.04(.57) −.05 .04 −.13, .02 −.05 .04
Absenteeism −.05(.47) −.17 .04 −.26, −.09 −.16 .13
Turnover
Intention
−.06(.24) −.37 .04 −.46, −.28 −.37 .34
Downloaded by [Central Michigan University] at 06:42 08 January 2018
16 M. KIM AND T. A. BEEHR
Overall, both the model t statistics, the alternative model t testing, and boot-
strapping results all support the proposition that aective commitment mediates
the association of empowering leadership with two forms of followers’ with-
drawal: absenteeism and turnover intention (but not lateness). us, mediation
Hypothesis 5b and 5c was supported.
Additional analyses
We tried two more alternative analyses with the model. First, we substituted
the three facets of empowering leadership for the overall empowering leader-
ship variable. e new model had good t, χ2 (338, N=294) = 893.75, p<.01,
CFI= .96, IFI=.96, NNFI=.95, RMSEA=.07, with the new paths for power
sharing (β=−.04, ns), motivational support (β=.50, p<.01), and development
support (β=.22, p<.01). e lack of a signicant path from power sharing to
aective commitment is due to its multicollinearity, especially with motivational
support (r=.90). As noted in the Method section, power sharing and motiva-
tional support have sometimes been considered a single facet of empowerment
(Amundsen & Martinsen, 2014).
Finally, we also considered the control variable, social desirability, as a control
for aective commitment as well as for the withdrawal behaviors. Because aective
commitment is a report about the self (as are the withdrawal behaviors), social
desirability might aect it. Note however, social desirability was not signicantly
correlated with aective commitment (Table 1), and controlling it did not change
the t indices or path coecients in the model.
Discussion
e present study examined the eect of empowering leadership on followers’
aective commitment and three forms of employee withdrawal, namely lateness,
absenteeism, and turnover intention. e results suggest that empowering lead-
ership may aect forms of employee attendance, but primarily because it leads
to the experience of commitment: Commitment is an explanatory variable that
shows why empowering leadership can result in less employee withdrawal.
Overall, the present research contributes to knowledge about the possible eects
of empowering leadership on employee withdrawal. It shows that (1) empower-
ing leadership can inuence a wide variety of withdrawal behaviors (lateness,
absenteeism, and turnover intentions), (2) it does this at the individual level,
and (3) aective organizational commitment explains the eects of empowering
leadership. Regarding empowering leadership and withdrawal, the concept of
withdrawal is broad, ranging from very short-term (e.g. being late by an hour
or less) to medium-term (e.g. being absent for a week or less) to permanent (e.g.
turning over). All of these employee responses have in common the fact that the
employee is missing from the organization, but they are distinct too, as our CFAs
Downloaded by [Central Michigan University] at 06:42 08 January 2018
THE INTERNATIONAL JOURNAL OF HUMAN RESOURCE MANAGEMENT 17
showed. erefore, it is important to know the degree to which empowering
leadership can aect each type of withdrawal.
Furthermore, we have evidence that the eect of empowering leadership on
these dierent withdrawal behaviors is transmitted through empowering lead-
ership engendering organizational commitment in subordinates. Empowering
leaders contribute to creating an environment for followers to develop a sense of
aective commitment to the organization by letting them actually make decisions
or by oering opportunities to voice their opinions in performing their work,
thereby leading to an increased sense of responsibility. Subsequently, employees
with higher levels of aective commitment showed lower levels of absenteeism and
turnover intention. Furthermore, there were no signicant direct links between
empowering leadership and withdrawal behaviors, suggesting that aective
commitment was an important psychological state that is necessary in order to
translate the positive eect of empowering leader behaviors to decreased with-
drawal of followers. erefore, withdrawal behaviors may be decreased if leaders
can strengthen the followers’ emotional bond to the organization by oering
the needed support, recognition, guidance, and chances to participate in work
processes.
Unexpectedly, neither empowering leadership nor the mediator, aective com-
mitment, had a signicant relationship with lateness. e fact that lateness acted
dierently from the other two forms of employee withdrawal provides some sup-
port for the independence model of withdrawal (Koslowsky et al., 1997), even
though lateness did correlate positively with absenteeism and turnover intention.
One possible reason for the independence is that employees might be habitually
late, no matter how much they are motivated and committed to their job and
organization. at is, if arriving a few minutes aer start time for work would
not be considered inappropriate behavior or there is no big penalty for slightly
being late to work, employees are likely to engage in lateness behavior (or not)
regardless of their leaders’ behaviors, organizational commitment, or withdrawal
intentions. Hence, lateness may be well predicted by individual characteristics
including personality and disposition as suggested in some previous studies (Blau,
1994; Bowling, Burns, & Beehr, 2010; Koslowsky, 2000; Richard & Slane, 1990), or
by penalties other than leadership styles. Lateness is dierent from absenteeism
and turnover in important ways. For one, the late employee is actually at work
during the day (aer the ocial starting time), but employees who are absent
or who turnover are not present during each day. Organizational commitment
(or leadership style) may not have a strong eect on lateness if the committed
employees still show up for work and believe they are contributing strongly to
the organization with their work performance.
Additionally, lateness is addressed as only one of three withdrawal behav-
iors, with other types of withdrawal being absenteeism and turnover intention.
However, lateness is a relatively milder form of withdrawal and thus might not well
represent stronger withdrawal signs such as more complete avoidance of work.
Downloaded by [Central Michigan University] at 06:42 08 January 2018
18 M. KIM AND T. A. BEEHR
For these reasons, leadership style and aective commitment may fail to predict
employees’ lateness behavior.
Employees considering withdrawal from their organization are likely to reect
low motivation as well as low commitment to the organization, thereby decreasing
their eort at work, which may also negatively aect their coworkers’ morale and
motivation (Koslowsky, 2000; Tett & Meyer, 1993). is may make it more dicult
for an organization to function eectively. Moreover, organizations should pay
attention to employees’ withdrawal, because absenteeism is a cost to the organ-
ization, and intent to turnover is one of the strongest proximal predictors for
actual turnover that is costly for the organization (Grieth et al., 2000). Given the
ndings that empowering leadership and aective commitment oer benets to
organizations with regard to employee retention and attendance in the present
study, providing leaders with empowering leadership training and development
programs would have desirable outcomes in terms of decreasing employee with-
drawal. Previous research has demonstrated the viability of training management
students to use empowering leadership styles (e.g. Sumpter, Gibson, & Porah,
2016), and thus such training seems feasible.
Research and practical applications
Although aective commitment was shown to be an important mediator in the
present study, future research would be encouraged to investigate other poten-
tial intervening variables (e.g. organization-based self-esteem and psychological
contract) linking empowering leadership to employees’ desirable behaviors. Trust
is inherent in empowering leadership (Srivastava, Bartol, & Locke, 2006), that
is, leaders are more likely to empower subordinates who they trust to use that
empowerment wisely. Future research could examine trust as a mediating vari-
able of empowering leadership with positive employee attitudes and behaviors;
indeed, employees’ feelings of being trusted (and oen reciprocating that trust)
have been shown related to a variety of outcomes such as organizational commit-
ment, citizenship behaviors, and team performance (De Jong, Dirks, & Gillespie,
2016; Dirks & Ferrin, 2001). Trust is oen reciprocal, and one party, supervisor
or subordinate, shows explicit trust in the other, the other party has reason to
develop more trust in the relationship also. Future research should investigate
trust, as both a cause of leader empowerment and a result of it.
Other forms of withdrawal also could be examined in future studies using
the basic model in Figure 1, such as psychological withdrawal, withdrawal from
coworkers or others in the workplace (e.g. not responding to communications),
or taking long work breaks. Lastly, in addition to withdrawal outcomes, there are
many other outcomes that may be inuenced by empowering leader behaviors
such as followers’ job strain, happiness, or career success. us, future research
can further our knowledge by examining additional outcome domains. Research
is also needed on the antecedents or potential causes of leader empowering
Downloaded by [Central Michigan University] at 06:42 08 January 2018
THE INTERNATIONAL JOURNAL OF HUMAN RESOURCE MANAGEMENT 19
behaviors that may be found in the organizational environment. One facilitator
of leaders’ empowering behavior is organizational climate or culture. Leaders
may be encouraged to empower subordinates if they experience the organiza-
tion as a whole to consist of striving, capability, and high internal motivation.
at kind of organization is ripe for leadership styles that include sharing power
to enhance employees’ already natural motivation tendencies, and to encourage
employee development, all characteristics of empowering leadership (Amundsen
& Martinsen, 2014; Zhang & Bartol, 2010). Organization-level research could
examine this possibility by studying culture as a promotive factor for empowering
leadership.
e study also has applied implications for organizations. In order to retain
core employees, organizations should encourage leaders or managers to engage in
empowering leadership by developing reward systems or training programs based
on a number of empowering behaviors that enhance autonomy and development
support among followers. Organizations can also employ the sets of empowering
behaviors measured in the present study to use as a tool when selecting its manag-
ers for rewards such as pay increases or advancement. Leader behaviors operation-
alized as empowering in the present study were development support by providing
continuous learning and development opportunities through leaders’ guidance
and role modeling, as well as autonomy/motivation support by expressing con-
dence in subordinates, providing opportunities for subordinates to participate in
decision-making along with sharing information, and encouraging initiative and
goal focus. e content of the scales used to measure ELBs could be guidelines for
using these behaviors eectively in developing empowering leadership programs.
By encouraging and showing supervisors or leaders ways to enhance employee
involvement, initiative, and self-development (empowering leadership), organi-
zations can indirectly decrease employees’ absenteeism and the tendency to leave
their organization. Considering the positive consequences to organizations of
having aectively committed employees, as well as the lack of a direct eect of
empowering leadership on withdrawal behaviors, making followers feel emo-
tionally attached to the organization (aective commitment) through the ELBs
of power sharing, motivational support, and development support may be a nec-
essary strategy to actually increase the employee’s desire to be in the organization.
Limitations
Although the present study provided new insights to the relationships among
empowering leadership, aective commitment, and employee withdrawal, some
limitations need to be addressed. Even though the variables were measured at dif-
ferent times in the temporal order matching their place in the proposed model, the
ndings do not ensure strong causal relations among the study variables. Future
experimental studies manipulating empowering leadership in randomly assigned
training groups of leaders would provide stronger causal inferences regarding
Downloaded by [Central Michigan University] at 06:42 08 January 2018
20 M. KIM AND T. A. BEEHR
its direct causal eects on employees’ commitment and indirect eects on their
withdrawal. If future studies use the multiple-time-point method of the present
study rather than experimental methods, they could address another issue regard-
ing causality: It is possible that there are unmeasured variables that inuence
multiple variables in the model, and they are at least partial causes of some of the
paths being signicant. One example variable that could cause both empowering
leadership and the outcomes is organizational culture. A culture that is respectful
of employees might lead to many organizational practices and conditions that
employees react well to, including empowering leadership. However, in addition
to empowering leadership, the culture itself could directly make employees more
committed to the organization and less inclined to withdraw from it. is could
cause any of the pairs of variables linked by paths in the model to be related to
each other. Future research on the present model could measure and control not
only organizational culture but also other variables that could theoretically cause
multiple variables in the model.
e present study adopted self-reported measures of all withdrawal behaviors,
because self-report data have been used successfully in previous studies (Hanisch
& Hulin, 1990, 1991; Koslowsky & Dishon-Berkovits, 2001), suggesting this is a
viable way for obtaining good withdrawal measures (Johns, 1994). Moreover, care
was taken to reduce the eects of common method variance. First, the data were
collected at multiple time points, separating the predictor from the outcomes
(which can help to reduce common method eects; Podsako et al., 2012). Second,
we controlled for social desirability, a substantive variable that can aect self-re-
ports. ird, because social desirability was measured with the same method as
the rest of the variables, controlling for it also controlled for method variance, the
way a marker variable does (Podsako et al., 2012). We note that the correlations
between the study’s variables varied and were only moderately strong, suggesting
no strong and consistent eect of the self-report method across variables. us,
there are multiple empirical reasons to believe the issue of common method bias
eects was not a serious problem. Moreover, Spector’s (2006) review showed that
problems caused by common method variance are oen somewhat overstated and
seldom serious enough to invalidate research ndings based on well-designed
measures.
Commitment to the organization is theoretically a reason why people would
be willing to attend rather than withdraw from the workplace, because it is a
psychological tie to the organization, and empirical research had generally con-
rmed this link (Grieth et al., 2000; Harrison et al., 2006; Tett & Meyer, 1993). A
larger question was whether and to what degree commitment would result from
empowering leadership, and as a mediator, would transfer its eect to employee
withdrawal. Although we found the proposed mediator eects for commitment,
other mediator variables could also be proposed. An intuitive mediator would
be psychological empowerment, for example. We recommend that future studies
include it as well as other variables of interest. In addition, we note that although
Downloaded by [Central Michigan University] at 06:42 08 January 2018
THE INTERNATIONAL JOURNAL OF HUMAN RESOURCE MANAGEMENT 21
empowerment is expected to be a positive state, it could have negative eects,
such as feelings of being abandoned by the leader (e.g. Wong & Giessner, 2017),
and some employees may directly resist empowerment (e.g. Maynard, Mathieu,
Marsh, & Ruddy, 2007).
Finally, even though we controlled social desirability, there still could be some
problems with it in the data. ere are other variables that are very similar to social
desirability that might aect relationships between variables, such as impression
management or self-presentation. Any one social desirability measure is unlikely
to control for all similar variables, and therefore some inuence of social desira-
bility-like variables remains.
Conclusion
e present study examined the relationship of empowering leader behaviors
with followers’ withdrawal and tested the mediating role of aective commit-
ment in these relationships. Empowering leadership has motivational inuences
on followers by promoting their sense of control, condence on their jobs, and
feelings of being empowered, conferring greater authority and autonomy to them.
e more employees have opportunities to get involved in decisions and work
activities as well as to engage in self-development, the more they will have loyalty
and aection for the organization (Chen et al., 2011; Den Hartog & De Hoogh,
2009). Subsequently, aectively committed employees develop positive attitudes
and behaviors toward the organization, and they thereby engage less in with-
drawal processes, especially absenteeism and turnover intention. Together, the
present study extended our prior understanding of the eects of positive forms
of leadership on employee behaviors by highlighting that empowering leadership
is also an important driver in shaping positive attitude (aective commitment)
and behavioral intentions of followers.
Disclosure statement
No potential conict of interest was reported by the authors.
ORCID
Minseo Kim http://orcid.org/0000-0002-7522-184X
References
Adams, G. A., & Beehr, T. A. (1998). Turnover and retirement: A comparison of their similarities
and dierences. Personnel Psychology, 51, 643–665.
Ahearne, M., Mathieu, J., & Rapp, A. (2005). To empower or not to empower your sales
force? An empirical examination of the inuence of leadership empowerment behavior on
customer satisfaction and performance. Journal of Applied Psychology, 90, 945–955.
Downloaded by [Central Michigan University] at 06:42 08 January 2018
22 M. KIM AND T. A. BEEHR
Allen, N. J., & Meyer, J. P. (1990). e measurement and antecedents of aective, continuance
and normative commitment to the organization. Journal of Occupational Psychology, 63,
1–18.
Amundsen, S., & Martinsen, Ø. L. (2014). Empowering leadership: Construct clarication,
conceptualization, and validation of a new scale. e Leadership Quarterly, 25, 487–511.
Amundsen, S., & Martinsen, Ø. L. (2015). Linking empowering leadership to job satisfaction,
work eort, and creativity: e role of self-leadership and psychological empowerment.
Journal of Leadership & Organization Studies, 22, 304–323.
Arnold, J. A., Arad, S., Rhoades, J. A., & Drasgow, F. (2000). e empowering leadership
questionnaire: e construction and validation of a new scale for measuring leader behaviors.
Journal of Organizational Behavior, 21, 249–269.
Beehr, T. A., & Gupta, N. (1978). A note on the structure of employee withdrawal. Organizational
Behavior and Human Performance, 21, 73–79.
Berry, C. M., Lelchook, A. M., & Clark, M. A. (2012). A meta-analysis of the interrelationships
between employee lateness, absenteeism, and turnover: Implications for models of withdrawal
behavior. Journal of Organizational Behavior, 33, 678–699.
Blau, P. M. (1964). Exchange and power in social life. New York, NY: Wiley.
Blau, G. (1994). Developing and testing a taxonomy of lateness behavior. Journal of Applied
Psychology, 79, 959–970.
Bowling, N. A., Burns, G. N., & Beehr, T. A. (2010). Productive and counterproductive
attendance behavior: An examination of early and later arrival to and departure from work.
Human Performance, 23, 305–322.
Browne, M. W., & Cudeck, R. (1992). Alternative ways of assessing model t. Sociological
Methods & Research, 21, 230–258.
Chen, G., Sharma, P. N., Edinger, S. K., Shapiro, D. L., & Farh, J. (2011). Motivating and
demotivating forces in teams: Cross-level inuences of empowering leadership and
relationship conict. Journal of Applied Psychology, 96, 541–557.
Cropanzano, R., & Mitchell, M. S. (2005). Social exchange theory: An interdisciplinary review.
Journal of Management, 31, 874–900.
De Jong, B. A., Dirks, K. T., & Gillespie, N. (2016). Trust and team performance: A meta-analysis
of main eects, moderators, and covariates. Journal of Applied Psychology, 101, 1134–1150.
Deci, E., & Ryan, R. (1985). Intrinsic motivation and self-determination theory. New York, NY:
Plenum Press.
Den Hartog, D. N., & De Hoogh, A. H. (2009). Empowering behavior and leader fairness
and integrity: Studying perceptions of ethical leader behavior from a levels-of-analysis
perspective. European Journal of Work and Organizational Psychology, 18, 199–230.
Dirks, K. T., & Ferrin, D. L. (2001). e role of trust in organizational settings. Organization
Science, 12, 450–467.
Eisenberger, R., Karagonlar, G., Stinglhamber, F., Neves, P., Becker, T. E., Gonzalez-Morales,
M. G., & Steiger-Mueller, M. (2010). Leader–member exchange and aective organizational
commitment: e contribution of supervisor’s organizational embodiment. Journal of
Applied Psychology, 95, 1085–1103.
El Akremi, A., Colaianni, G., Portoghese, I., Galletta, M., & Battistelli, A. (2014). How
organizational support impacts aective commitment and turnover among Italian nurses:
A multilevel mediation model. e International Journal of Human Resource Management,
25, 1185–1207.
Erez, M., & Arad, R. (1986). Participative goal-setting: Social, motivational, and cognitive
factors. Journal of Applied Psychology, 71, 591–597.
Ferris, G. R. (1985). Role of leadership in the employee withdrawal process: A constructive
replication. Journal of Applied Psychology, 70, 777–781.
Downloaded by [Central Michigan University] at 06:42 08 January 2018
THE INTERNATIONAL JOURNAL OF HUMAN RESOURCE MANAGEMENT 23
Foust, M. S., Elicker, J. D., & Levy, P. E. (2006). Development and validation of a measure of
an individual’s lateness attitude. Journal of Vocational Behavior, 69, 119–133.
Frooman, J., Mendelson, M. B., & Kevin Murphy, J. (2012). Transformational and passive
avoidant leadership as determinants of absenteeism. Leadership & Organization Development
Journal, 33, 447–463.
Graen, G. B., & Uhl-Bien, M. (1995). Relationship-based approach to leadership: Development
of leader-member exchange (LMX) theory of leadership over 25years: Applying a multi-
level multi-domain perspective. e Leadership Quarterly, 6, 219–247.
Grieth, R. W., Hom, P. W., & Gaertner, S. (2000). A meta-analysis of antecedents and correlates
of employee turnover: Update, moderator tests, and research implications for the next
millennium. Journal of Management, 26, 463–488.
Hancock, J. I., Allen, D. G., Bosco, F. A., McDaniel, K. R., & Pierce, C. A. (2013). Meta-analytic
review of employee turnover as a predictor of rm performance. Journal of Management,
39, 573–603.
Hanisch, K. A., & Hulin, C. L. (1990). Job attitudes and organizational withdrawal: An
examination of retirement and other voluntary withdrawal behaviors. Journal of Vocational
Behavior, 37, 60–78.
Hanisch, K. A., & Hulin, C. L. (1991). General attitudes and organizational withdrawal: An
evaluation of a causal model. Journal of Vocational Behavior, 39, 110–128.
Harrison, D. A., Newman, D. A., & Roth, P. L. (2006). How important are job attitudes? Meta-
analytic comparisons of integrative behavioral outcomes and time sequences. Academy of
Management Journal, 49, 305–325.
Hayes, A. F. (2013). Introduction to mediation, moderation, and conditional process analysis: A
regression-based approach. New York, NY: e Guilford Press.
Hays, R. D., Hayashi, T., & Stewart, A. L. (1989). A ve-item measure of socially desirable
response set. Educational and Psychological Measurement, 49, 629–636.
Humborstad, S. I. W., Nerstad, C. G., & Dysvik, A. (2014). Empowering leadership, employee
goal orientations and work performance: A competing hypothesis approach. Personnel
Review, 43, 246–271.
Iverson, R. D., & Deery, S. J. (2001). Understanding the “personological” basis of employee
withdrawal: e inuence of aective disposition on employee tardiness, early departure,
and absenteeism. Journal of Applied Psychology, 86, 856–866.
Jaros, S. J., Jermier, J. M., Koehler, J. W., & Sincich, T. (1993). Eects of continuance, aective,
and moral commitment on the withdrawal process: an evaluation of eight structural equation
models. Academy of Management Journal, 36, 951–995.
Johns, G. (1994). How oen were you absent? A review of the use of self-reported absence
data. Journal of Applied Psychology, 79, 574–591.
Johnson, R. E., Chang, C.-H. (D),, & Yang, L-Q. (2010). Commitment and motivation at work:
e relevance of employee identity and regulatory focus. Academy of Management Review,
35, 226–245.
Keller, T., & Dansereau, F. (1995). Leadership and empowerment: A social exchange perspective.
Human Relations, 48, 127–146.
Kim, M., & Beehr, T. A. (2017a). Directing our own careers, but getting help from empowering
leaders. Career Development International, 22, 300–317.
Kim, M., & Beehr, T. A. (2017b). Can empowering leaders aect subordinates’ well-being and
careers because they encourage subordinates’ job craing behaviors? Journal of Leadership
& Organizational Studies, Advance online publication. doi:10.1177/1548051817727702
Kim, M., & Beehr, T. A. (2017c). Self-ecacy and psychological ownership mediate the eects
of empowering leadership on both good and bad employee behaviors. Journal of Leadership
& Organizational Studies, 24, 466–478.
Downloaded by [Central Michigan University] at 06:42 08 January 2018
24 M. KIM AND T. A. BEEHR
Kline, R. B. (2015). Principles and practice of structural equation modeling (4th ed.). New York,
NY: e Guilford Press.
Koslowsky, M. (2000). A new perspective on employee lateness. Applied Psychology, 49, 390–
407.
Koslowsky, M., & Dishon-Berkovits, M. (2001). Self-report measures of employee lateness:
Conceptual and methodological issues. European Journal of Work and Organizational
Psychology, 10, 145–159.
Koslowsky, M., Sagie, A., Krausz, M., & Singer, A. (1997). Correlates of employee lateness:
Some theoretical considerations. Journal of Applied Psychology, 82, 79–88.
Manz, C. C., & Sims, H. P., Jr (1987). Leading workers to lead themselves: e external
leadership of self- managing work teams. Administrative Science Quarterly, 32, 106–128.
Manz, C. C., & Sims, H. P. (1991). SuperLeadership: Beyond the myth of heroic leadership.
Organizational Dynamics, 19, 18–35.
Mason, W., & Suri, S. (2012). Conducting behavioral research on Amazon’s mechanical turk.
Behavior Research Methods, 44, 1–23.
Mathieu, J. E., & Zajac, D. M. (1990). A review and meta-analysis of the antecedents, correlates,
and consequences of organizational commitment. Psychological Bulletin, 108, 171–194.
Matsunaga, M. (2008). Item parceling in structural equation modeling: A primer. Communication
Methods and Measures, 2, 260–293.
Maynard, M. T., Mathieu, J. E., Marsh, W. M., & Ruddy, T. M. (2007). A multilevel investigation
of the inuences of employees’ resistance to empowerment. Human Performance, 20, 147–
171.
Meyer, J. P., & Allen, N. J. (1991). A three-component conceptualization of organizational
commitment. Human Resource Management Review, 1, 61–89.
Meyer, J. P., & Allen, N. J. (1997). Commitment in the workplace: eory, research, and application.
ousand Oaks, CA: Sage Publications.
Meyer, J. P., Stanley, D. J., Herscovitch, L., & Topolnytsky, L. (2002). Aective, continuance,
and normative commitment to the organization: A meta-analysis of antecedents, correlates,
and consequences. Journal of Vocational Behavior, 61, 20–52.
Pearce, C. L., & Sims, H. P., Jr (2002). Vertical versus shared leadership as predictors of
the eectiveness of change management teams: An examination of aversive, directive,
transactional, transformational, and empowering leader behaviors. Group Dynamics: eory,
Research, and Practice, 6, 172–197.
Pearce, C. L., Sims, H. P., Jr, Cox, J. F., Ball, G., Schnell, E., Smith, K. A., & Trevino, L. (2003).
Transactors, transformers and beyond: A multi-method development of a theoretical
typology of leadership. Journal of Management Development, 22, 273–307.
Podsako, P. M., MacKenzie, S. P., & Podsako, N. P. (2012). Sources of method bias in social
science research and recommendations on how to control it. Annual Review of Psychology,
63, 539–569.
Preacher, K. J., & Kelley, K. (2011). Eect size measures for mediation models: Quantitative
strategies for communicating indirect eects. Psychological Methods, 16, 93–115.
Rapp, T. L., Gilson, L. L., Mathieu, J. E., & Ruddy, T. (2016). Leading empowered teams: An
examination of the role of external team leaders and team coaches. e Leadership Quarterly,
27, 109–123.
Raub, S., & Robert, C. (2010). Dierential eects of empowering leadership on in-role and
extra-role employee behaviors: Exploring the role of psychological empowerment and power
values. Human Relations, 63, 1743–1770.
Rhoades, L., & Eisenberger, R. (2002). Perceived organizational support: A review of the
literature. Journal of Applied Psychology, 87, 698–714.
Downloaded by [Central Michigan University] at 06:42 08 January 2018
THE INTERNATIONAL JOURNAL OF HUMAN RESOURCE MANAGEMENT 25
Richard, D., & Slane, S. (1990). Punctuality as a personality characteristic: Issues of measurement.
e Journal of Psychology, 124, 397–402.
Ross, J. G. (1988). Relations among Lateness, absence, and turnover: Is there a progression of
withdrawal? Human Relations, 41, 517–531.
Rousseau, D. M. (1998). e ‘problem’ of the psychological contract considered. Journal of
Organizational Behavior, 19, 665–671.
Seibert, S. E., Wang, G., & Courtright, S. H. (2011). Antecedents and consequences of
psychological and team empowerment in organizations: A meta-analytic review. Journal of
Applied Psychology, 96, 981–1003.
Sharma, P. N., & Kirkman, B. L. (2015). Leveraging leaders: A literature review and future
lines of inquiry for empowering leadership research. Group & Organization Management,
40, 193–237.
Shore, L. M., & Wayne, S. J. (1993). Commitment and employee behavior: Comparison of
aective commitment and continuance commitment with perceived organizational support.
Journal of Applied Psychology, 78, 774–780.
Simons, T., & Roberson, Q. (2003). Why managers should care about fairness: e eects of
aggregate justice perceptions on organizational outcomes. Journal of Applied Psychology,
88, 432–443.
Somers, M. J. (1995). Organizational commitment, turnover and absenteeism: An examination
of direct and interaction eects. Journal of Organizational Behavior, 16, 49–58.
Spector, P. E. (2006). Method variance in organizational research truth or urban legend?
Organizational Research Methods, 9, 221–232.
Srivastava, A., Bartol, K. M., & Locke, E. A. (2006). Empowering leadership in management
teams: Eects on knowledge sharing, ecacy, and performance. Academy of Management
Journal, 49, 1239–1251.
Sumpter, D. M., Gibson, C. B., & Porah, C. (2016). Act expediently, with autonomy: Vicarious
learning, empowered behaviors, and performance. Journal of Business and Psychology, 1–15.
Tekleab, A. G., Sims, H. P., Jr, Yun, S., Tesluk, P. E., & Cox, J. (2008). Are we on the same
page? Eects of self-awareness of empowering and transformational leadership. Journal of
Leadership & Organizational Studies, 14, 185–201.
Tett, R. P., & Meyer, J. P. (1993). Job satisfaction, organizational commitment, turnover
intention, and turnover: Path analyses based on meta-analytic ndings. Personnel Psychology,
46, 259–293.
anacoody, P. R., Newman, A., & Fuchs, S. (2014). Aective commitment and turnover
intentions among healthcare professionals: e role of emotional exhaustion and
disengagement. e International Journal of Human Resource Management, 25, 1841–1857.
Tuckey, M. R., Bakker, A. B., & Dollard, M. F. (2012). Empowering leaders optimize working
conditions for engagement: A multilevel study. Journal of Occupational Health Psychology,
17, 15–27.
Waldman, D. A., Carter, M. Z., & Hom, P. W. (2015). A multilevel investigation of leadership
and turnover behavior. Journal of Management, 41, 1724–1744.
Walsh, J. P., Ashford, S. J., & Hill, T. E. (1985). Feedback obstruction: e inuence of the
information environment on employee turnover intentions. Human Relations, 38, 23–46.
Wong, S. I., & Giessner, S. R. (2017). e thin line between empowering and laissez-faire
leadership: An expectancy-match perspective. Journal of Management, Accepted (Online
First). Obtained early online, October 17, 2017.doi:10.1177/0149206315574597
Zhang, X., & Bartol, K. M. (2010). Linking empowering leadership and employee creativity:
e inuence of psychological empowerment, intrinsic motivation, and creative process
engagement. Academy of Management Journal, 53, 107–128.
Downloaded by [Central Michigan University] at 06:42 08 January 2018
