Sie sind auf Seite 1von 15
LEAQUA-01041; No of Pages 15 The Leadership Quarterly xxx (2015) xxx – xxx Contents lists

LEAQUA-01041; No of Pages 15

Pages 15 The Leadership Quarterly xxx (2015) xxx – xxx Contents lists available at ScienceDirect The

Contents lists available at ScienceDirect

The Leadership Quarterly

journal homepage: www.elsevier.com/locate/leaqua

Quarterly journal homepage: www.elsevier.com/locate/leaqua Empowering leadership and managers' career perceptions:

Empowering leadership and managers' career perceptions:

Examining effects at both the individual and the team level

Torsten Biemann a, , Eric Kearney b,1 , Kathrin Marggraf c,2

a Department of Management, University of Mannheim, 68161 Mannheim, Germany

b Department of Business Administration, University of Potsdam, 14482 Potsdam, Germany

c HR Strategy & Organizational Effectiveness, Henkel AG & Co. KGaA, Henkelstraße 67, 40589 Düsseldorf, Germany

article info

Article history:

Received 11 November 2013 Received in revised form 11 January 2015 Accepted 12 March 2015 Available online xxxx

Handling editor: Shelley Dionne

Keywords:

Empowering leadership

Career self-efcacy

Career satisfaction

Multilevel analysis

abstract

In a multilevel model of leadership behavior, we investigated whether and how empowering leadership affects individuals' career perceptions. We developed a conceptual model that links empowering leadership at the individual level and at the group level (mean as well as dispersion) to individuals' career self-efcacy and career satisfaction. To test our model, we used questionnaire data from a multilevel data set of 2493 employees in leadership positions nested in 704 teams from a large German corporation. Hierarchical linear regression analyses showed that empowering leadership at the individual level was positively related to career self-ef cacy, which in turn mediated the relationship between empowering leadership and career satisfaction. Empowering leadership at the group level was positively related to career self-efcacy when it was conceptualized as leadership differentiation (i.e., the standard deviation of empowering leadership ratings), but not when it was conceptualized as leadership climate (i.e., mean empowering leadership ratings). Career self-efcacy in turn mediated the relationship between empowering leadership differentiation and career satisfaction. Finally, we found a negative relationship between empowering leadership differentiation and career satisfaction.

© 2015 Elsevier Inc. All rights reserved.

Introduction

There is an ongoing trend for organizations to become leaner and more cost-efcient. As a consequence of establishing atter hierarchies, employees' responsibilities at lower hierarchical levels expand (Argyris, 1998; Forrester, 2000). Moreover, it is increasing- ly the respective individuals themselves, rather than their organizations, who are responsible for their careers (Sullivan, 1999). Long- term career planning by organizations has become more difcult and has partly been replaced by employees' own career manage- ment, as described in boundaryless ( Arthur & Rousseau, 1996 ) or protean career concepts ( Hall & Moss, 1998 ). Empowered employees with individual career plans may create problems for organizations insofar as employees' career planning must not necessarily include a continuous career in the current organization. But especially because of an increasing reliance on complex knowledge work and rapid technological advancements, the retention of managers and other high-quality employees is vital for the success of today's organizations ( Grant, 1996; Holtom, Mitchell, Lee, & Eberly, 2008 ). Promising development opportunities and career perspectives for employees have been shown to decrease turnover (Kraimer, Seibert, Wayne, Liden, & Bravo, 2011) and thus offer organizations the means to enhance career satisfaction and retain valued employees. Leadership plays an important role

Corresponding author. Tel.: +49 621 181 1502. E-mail addresses: biemann@bwl.uni-mannheim.de (T. Biemann), kearney@uni-potsdam.de (E. Kearney), marggraf.kathrin@gmail.com (K. Marggraf).

1 Tel.: +49 331 977 3593.

2 Tel.: +49 171 9591057.

1048-9843/© 2015 Elsevier Inc. All rights reserved.

2 T. Biemann et al. / The Leadership Quarterly xxx (2015) xxx – xxx in

2

T. Biemann et al. / The Leadership Quarterly xxx (2015) xxxxxx

in facilitating access to career development opportunities. Gaining knowledge on the impact that supervisors have through their lead- ership behaviors on the career self-efcacy and career satisfaction of key employees can thus help organizations to establish practices to increase performance and prevent or at least decrease turnover that is motivated by a perceived lack of career opportunities. In this regard, scholars and practitioners alike have in recent years shown great interest in the concepts of empowerment and empowering leadership. Leaders may lead others to lead themselves(Manz & Sims, 1987: 119) and foster employee empowerment by exhibiting empowering leadership behaviors that shift responsibility and authority from the leader to the subordinates (Amundsen

& Martinsen, 2014; Zhang & Bartol, 2010). The construct of psychological empowerment is typically dened by the four cognitions

impact, competence, autonomy, and meaningfulness (Spreitzer, 1995). Empowerment has been shown to foster employee motivation as well as attitudinal and performance outcomes at both the individual (e.g., Zhang & Bartol, 2010) and the team level of analysis (e.g., Kirkman & Rosen, 1999). However, most studies on empowerment have examined the effects of psychological empowerment itself, rather than the effects of empowering leadership ( Seibert, Silver, & Randolph, 2004; Seibert, Wang, & Courtright, 2011;

Spreitzer, 2008). Scholars have identied different antecedents of psychological empowerment such as organizational structure, organizational culture, task characteristics, and work design (Maynard, Gilson, & Mathieu, 2012; Maynard, Mathieu, Gilson, O'Boyle

& Cigularov, 2013; Seibert et al., 2011). Nevertheless, in recent years there has been a surge of interest in the effects of empowering

leadership behaviors (e.g., Ahearne, Mathieu, & Rapp, 2005; Chen et al., 2011; Lorinkova et al., 2013; Martin et al., 2013). A strong interest in this leadership style appears to be justi ed given that such leadership behaviors are in line with the trend to grant employees greater discretion at work to foster motivation and unlock the potential of an increasingly better educated and more skilled workforce (Srivastava, Bartol, & Locke, 2006). We argue that the concept of empowering leadership lends itself particularly well to explore the link between leader behaviors and employees' career perceptions. However, the association between empowering leadership and various outcomes appears to be rather complex. Although on average empowering leadership has positive effects on satisfaction and performance, the strength of these relationships seems to strongly depend on the context (e.g., Stewart, 2006). Some studies have been unable to conrm positive effects of empowering lead- ership (e.g., Mathieu, Gilson, & Ruddy, 2006), and others have yielded surprising ndings regarding which employees benet more and which employees benet less from empowering leadership behaviors (e.g., Ahearne et al., 2005). Current knowledge of the effects of empowering leadership in organizations is still incomplete and many important issues remain unresolved (Lorinkova et al., 2013). We contend that different ways through which empowering leadership behaviors might impact individual outcomes must be distin- guished to explain these mixed ndings. This is in line with Kirkman and Rosen (1999), who reasoned that it is important to conduct multi-level studies to determine optimal levels of empowerment at the individual and the group level. To date, however, most research has examined the effects of empowering leadership at either the individual or the group level, but not at both levels simul- taneously. A notable exception is a study by Chen, Kirkman, Kanfer, Allen, and Rosen (2007), who investigated the effects of leadermember exchange (LMX) and empowering leadership on individual and team performance. With respect to empowering leadership, however, this study actually examined two different constructs leadership climate and psychological empowerment. But to disen- tangle effects of a predictor at different levels of analysis, the most straightforward approach is to study the same construct at multiple levels simultaneously (Firebaugh, 1978). We develop a multilevel model of empowering leadership behavior that distinguishes among three distinct ways in which empowering leadership impacts subordinates' career perceptions, each of which offers unique theoretically and managerially mean- ingful implications that complement those derived from the respective other two aspects: rst, a leader can directly exhibit empowering behavior toward an individual follower; second, a leader can empower the follower's whole team; and third, a leader

can differentiate among followers and enact empowering behavior in different degrees in relation to different followers in a team. The rst of these aspects pertains to the individual, whereas the second and third aspects pertain to the group level of analysis. We use arguments from LMX theory (Dansereau, Graen, & Haga, 1975; Erdogan & Liden, 2002; Graen, 1976; Graen & Uhl-Bien, 1995; Liden, Sparrowe, & Wayne, 1997) to explore the underlying mechanisms for the three ways in which empowering leadership affects followers. LMX focuses on dyadic relationships between leaders and followers and on how the quality of these relationships affects outcomes such as satisfaction and performance (Dulebohn, Bommer, Liden, Brouer, & Ferris, 2012; Liden et al., 1997). LMX theory as- sumes that leaders develop different types of relationships with different followers (Ilies, Nahrgang, & Morgeson, 2007). We adopt this underlying logic to examine the relationship between empowering leadership enacted by superiors on the one hand and the career self-efcacy and career satisfaction of subordinate managers on the other. We thus assume that leaders differ from one another both with respect to the mean levels of empowering leadership behaviors they enact toward their respective group of followers and concerning the extent to which they exhibit leadership differentiation that is, the degree to which they differentially empower followers. With the exception of LMX, leadership approaches typically assume that leaders generally behave similarly toward different sub- ordinates. Thus, virtually all studies of empowering leadership conducted at the group level have conceptualized and operationalized empowering leadership as leadership climate (i.e., mean levels of empowering leadership ratings provided by followers; e.g., Chen et al., 2011; Lorinkova et al., 2013; Srivastava et al., 2006). However, we argue that it would be unrealistic to assume that empowering leadership behaviors are always and necessarily enacted similarly toward all followers. If leaders wish to enhance the career self- efcacy and career satisfaction of focal individuals (e.g., highly valued employees), it is important to ask not only about the extent to which empowering behaviors are conducive to this end, but also whether the leader should empower the members of his or her group similarly or differentially. Whereas the importance of examining the question of whether equal or differential treatment of group members is more conducive to individual-level outcomes has received attention in the LMX literature (e.g., Liao, Liu, & Loi, 2010), it has been all but ignored in the empowering leadership literature. For the LMX literature, Gooty and Yammarino (in press) have shown that the simultaneous investigation of LMX at different levels of analysis, including the examination of LMX

T. Biemann et al. / The Leadership Quarterly xxx (2015) xxx – xxx 3 differentiation

T. Biemann et al. / The Leadership Quarterly xxx (2015) xxxxxx

3

differentiation at the group level, enriches understanding of the effects of LMX on individual-level outcomes. In a similar vein, we pro- pose that the effects of empowering leadership on individual-level career perceptions can only be fully understood by examining three different aspects of empowering leadership at two different levels of analysis. Aside from examining (at the individual level) the effects of empowering leadership as perceived by an individual follower, we simultaneously examine (at the group level) the ef- fects of both empowering leadership climate and empowering leadership differentiation and argue that these different aspects of empowering leadership in part have differential effects on career perceptions. In sum, our goal is to delineate and empirically test the different ways in which leadership is associated with individual follower career self-efcacy and career satisfaction. We focus on empowering leadership because these behaviors are intended and have been shown in previous research (e.g., Kirkman & Rosen, 1999; Zhang & Bartol, 2010) to enhance followers' psychological empowerment, which we argue is particularly relevant to how employees experience their careers. Moreover, we draw on LMX theory to delineate different processes at the individual and the group level through which these leader behaviors might impact subordinates' career perceptions.

Literature review and hypotheses

Lorinkova et al. (2013) have dened empowering leadership as sharing power with subordinates and raising their level of auton- omy and responsibility(p. 573). Researchers have identied positive effects of empowering leadership on different performance out- comes at the individual level (e.g., Zhang & Bartol, 2010), the team level (e.g., Srivastava et al., 2006), and the organizational level of analysis (e.g., Carmeli, Schaubroeck, & Tishler, 2011). At the team level, for example, two meta-analyses have shown that, on average, empowering leadership accounts for at least as much variance in team performance as do transformational, motivational, and consid- erate leadership, respectively (Burke, Stagl, Klein, Goodwin, Salas, & Halpin, 2006; Stewart, 2006). At the individual level, the media- tors whereby empowering leadership engenders its effects on performance outcomes include psychological empowerment and intrinsic motivation (Zhang & Bartol, 2010), as well as self-efcacy and adaptability (Ahearne et al., 2005). Arguably, empowering leadership constitutes a new leadership stylethat is in important ways conceptually distinct from related and more established constructs such as transformational leadership (Arnold, Arad, Rhoades, & Drasgow, 2000). Although transfor- mational leadership often fosters psychological empowerment of followers through coaching and mentoring, it strongly relies on the leader's charisma and vision and may also foster follower dependence on the leader (Kark, Shamir, & Chen, 2003). In a similar vein, Bass and Riggio (2006, p. 1112) have noted that transformational leaders can be either directive or participative. By contrast, empowering leadership aims to foster follower autonomy and independence from the leader ( Burke et al., 2006 ), as the goal of empowering leadership behaviors is to strengthen followers' self-leadership skills and proactive behaviors ( Martin et al., 2013; Tuckey, Bakker, & Dollard, 2012). Individuals' careers unfold in and outside of organizations (Hall, 2002), which makes career plan- ning and perceptions a highly individual process that entails some degree of independence from the leader. Hence, we consider empowering leadership as the most suitable approach to capture leader behaviors that are intended to foster subordinates' career development. Arnold et al. (2000) describe ve dimensions of empowering leadership behaviors that can easily be linked to subordinate managers' career perceptions. Empowering leaders set high performance standards for themselves and work hard. They lead by example and may therefore serve as role models for successful careers that are based on effort and merit rather than luck, political maneuvering, or organizational arbitrariness. Subordinates may therefore learn from observing the work and work-related attitudes of their supervisors, and these observations provide them with information on how to become successful in their own career. Empowering leadership is furthermore characterized by coaching activities, which pertain to a direct interaction between supervisor

Multilevel contextual model of leadership behavior

Group level SD Empowering leadership (L2) Mean H3 + H5 + H6 - Empowering Career
Group level
SD
Empowering
leadership (L2)
Mean
H3 +
H5 +
H6 -
Empowering
Career
Career self-efficacy
leadership (L1)
Satisfaction
H1 +
H2/H4/H7 + (mediation)

Individuallevel

Fig. 1. Multilevel contextual model of leadership behavior.

4 T. Biemann et al. / The Leadership Quarterly xxx (2015) xxx – xxx and

4

T. Biemann et al. / The Leadership Quarterly xxx (2015) xxxxxx

and employee in that supervisors help subordinates to improve their performance and become more condent about their work ( Arnold et al., 2000 ). Moreover, showing concern provides subordinates with opportunities to discuss personal issues, including those that are related to career progression. The informing dimension of empowering leadership emphasizes the practical knowledge of aspects of development opportunities that subordinates might gain from their leaders. Finally, participative decision making affords subordinates an opportunity to voice ideas and instigate discussions (Arnold et al., 2000). Although most prior studies that have investigated empowering leadership at the group level have focused on leadership climate, there has recently been growing interest in the effects of differentiated leadership on individual and group outcomes. With few exceptions (e.g., Wu, Tsui & Kinicki, 2010, who have examined differentiation with respect to transformational leadership), such studies have been conducted within the paradigm of LMX. Heeding the call to examine LMX relationships in conjunction with the focal individual's perception of the relationships his or her leader maintains with other subordinates (e.g., Liden, Erdogan, Wayne, & Sparrowe, 2006), several researchers have investigated the effects of LMX differentiation (i.e., the extent to which a leader estab- lishes differential relationships with different members of a unit; Gooty & Yammarino, in press; Liao et al., 2010). Findings in this regard are mixed, especially when it comes to the association between LMX differentiation and individual (rather than team) outcomes (e.g., Henderson, Liden, Glibkowski & Chaudhry, 2009; Henderson, Wayne, Shore, Bommer, & Tetrick, 2008; Liao et al., 2010). In the following, we primarily draw on theory and research from the LMX literature to develop a multi-level model that spec- ies three ways in which empowering leadership affects subordinates' career perceptions (see Fig. 1). We rst develop individual- level hypotheses that pertain to the links between subordinates' ratings of the extent to which his or her leader exhibits empowering leadership and the respective subordinates' career perceptions. Subsequently, at the group level, we examine the effects of empowering leadership climate (i.e., the mean level of empowering leadership evaluations provided by group members). And nally, we examine relationships between leadership differentiation (i.e., divergence among the empowering leadership ratings provided by followers) and subordinates' career self-efcacy and career satisfaction.

Individual level effects

Empowering leadership and career self-efcacy Self-efcacy captures a person's belief that he or she can attain high levels of performance (Bandura, 1997). Career self-efcacy pertains to the extent to which a person is condent that he or she will manage career-related challenges and experience successes such as regular pay raises and promotions. Although to the best of our knowledge this link has never been studied directly, there are many arguments for how empowering leadership behaviors can promote career self-ef cacy. Empowering leaders serve as role models, as they lead by example. Observing an empowering leader may positively affect the career self-ef cacy of subordinates through vicarious learning (Conger & Kanungo, 1988). In addition, the feedback and support received from supervisors directly re- veals to subordinates how their work can be improved, which should likewise foster their career efcacy beliefs. Similarly, coaching may also serve as verbal persuasion with regard to the career goals of subordinates and make them feel more condent about future career development options. Furthermore, by expressing their invited opinions, subordinates may elicit positive feedback and thus enable further efcacy-enhancing experiences. These opportunities prepare subordinates for the next steps in their career progression that entail greater responsibilities. Perhaps most importantly, empowering leadership signals to subordinates that their leader trusts them, has condence in their abilities, and is willing to support them and provide them with necessary resources. This explanation for a positive link between empowering leadership and subordinates' career self-efcacy draws on LMX theory (Liden et al., 1997) and social exchange theory ( Cropanzano & Mitchell, 2005 ). Leader actions have been shown to be the strongest predictors of the quality of leader member relationships (Dulebohn et al., 2012). Empowering leader behaviors are precisely the kinds of behaviors that are likely to engender high-quality relationships. These behaviors include sharing power and responsibility, granting followers latitude with respect to how they do their work and discretion in decision-making, and communicating high expectations and trust in followers ( Arnold et al., 2000; Kirkman & Rosen, 1999; Lorinkova et al., 2013). We therefore argue that empowering leadership is conducive to moving leader member interactions beyond mere econom- ic exchanges (i.e., a formally agreed on balanced reciprocation of tangible assets) and toward the development of a relationship that is marked by commitment, trust, mutual helping, and loyalty ( Dulebohn et al., 2012; Erdogan & Liden, 2002 ), which also entails an affective attachment between leaders and followers ( Liden & Maslyn, 1998; Uhl-Bien & Maslyn, 2003 ). Empowering leadership behaviors are likely to result in a leader member relationship marked by feelings of mutual obligation and reciproc- ity ( Liden et al., 1997 ). We contend that followers who are part of such high-quality dyadic relationships will believe that they can count on their leaders to help and support them in their career progression. Aside from direct leader support, empowering leadership also gives followers the freedoms to explore different options concerning how best to further their careers. By contrast, the absence of empowering leadership is likely to be perceived as a lack of trust and con dence of the leader in the follower. Support and positive evaluations from the leade r are likely to be a major determinant of career self-ef cacy because direct superiors have strong in uence over career-related outcomes such as salary and promotions. A subordinate who is unsure of his or her leader's support and is being granted little autonomy is therefore less likely to exhibit high levels of career self- ef cacy. Hence, we argue that the relationship between empowering leadership and followers' career self-ef cacy is positive, primarily because empowering leadership engenders the high-quality leader member relationships marked by mutual com- mitment that are likely to boost a follower's belief that he or she can overcome any possible career-related dif culties and remain on a successful career trajectory.

T. Biemann et al. / The Leadership Quarterly xxx (2015) xxx – xxx 5 Hypothesis

T. Biemann et al. / The Leadership Quarterly xxx (2015) xxxxxx

5

Hypothesis 1. There is a positive relationship between empowering leadership and career self-efcacy at the individual level.

Career self-efcacy as a mediator Previous research has established that employees who are satised with their career progression and prospects have lower turn- over rates, perform better, and experience a better work-life balance (Finegold & Mohrman, 2001; Kraimer et al., 2011). Organizations that succeed in attaining high levels of career satisfaction among their employees may thus benet from lower turnover and higher- performing employees. We argue that empowering leadership increases subordinates' career satisfaction and can thus help to retain managers and other highly valued employees that constitute vital assets to their organization. Furthermore, we posit that career self- efcacy mediates this relationship. Career satisfaction (or subjective career success) is dened as an individual's reactions to unfolding career experiences (Hughes, 1937). It is contrasted with objective indicators of career success, most importantly salary, hierarchical level, and number of subordi- nates (Heslin, 2005). Meta-analytic results suggest a strong link between perceived supervisor support and career satisfaction (Ng, Eby, Sørensen, & Feldman, 2005). However, much less is known about the process that might translate supervisor support into a higher degree of career satisfaction. Some studies indicate that career self-efcacy may play a vital role in this relationship. For exam- ple, Day and Allen (2004) examined the impact of mentoring on career success and suggested career self-efcacy as a potential me- diator of this relationship, but their data offered only marginal support for this hypothesis. We argued above that empowering leadership is likely to affect career self-efcacy. To establish a mediation hypothesis, the link between career self-efcacy and career satisfaction needs further exploration. In this regard, longitudinal studies have found an impact of self-efcacy on job satisfaction (Saks, 1995) and perceived career success (Higgins, Dobrow, & Chandler, 2008). Moreover, studies of core self-evaluations indicate a strong relationship between self-efcacy and job satisfaction (Judge & Bono, 2001). Heslin (2005) pointed out that subjective career success entails reactions to both actual and anticipated career-related achievements. High career self-efcacy primarily pertains to the latter. If individuals believe that they are capable of effectively navigating their own careers, evaluations of anticipated career devel- opments will tend to be more positive. Despite empirical support for the relationship between supervisor support and career success, and between career self-efcacy and career success, previous research has not analyzed the mediating effects of career self-efcacy on the relationship between lead- ership behaviors and career satisfaction. LMX theory posits that a high-quality dyadic relationship between leader and follower fos- ters, among other variables, general satisfaction on the part of the follower ( Dulebohn et al., 2012 ). This is due to the leader establishing interaction patterns that are more social in nature (rather than being based mainly on economic and formal exchanges) and enhance the follower's perception of justice and sense of psychological empowerment (Dulebohn et al., 2012). Extending our above arguments, we contend that empowering leadership fosters career satisfaction primarily because it facilitates a leadermember relationship that is marked by mutual commitment, trust, and support. Embedded in such a high-quality relationship, the follower is likely to believe that he or she can meet career-related goals and meet career-related challenges, in no small measure because he or she has condence in continued leader support. These career self-efcacy beliefs, in turn, are likely to engender positive appraisals of and satisfaction with one's career development. With an empowering leader, a subordinate will feel better informed about career op- portunities and better prepared to develop the necessary skills for the next career step, which will result in higher career self-efcacy and, in turn, higher levels of satisfaction with regard to anticipated career success.

Hypothesis 2. Career self-efcacy partially mediates the positive relationship between empowering leadership and career satisfac- tion at the individual level.

Group level effects

Consensus effects of empowering leadership This study is based on the premise that empowering leadership enacted toward the team as a whole has effects that go beyond the effects generated in dyadic leadermember relations. We argue that individuals' career perceptions are not only inuenced by the perceived level of empowering leadership in dyadic leadermember interactions, but also by the leaders' empowering behavior to- ward other group members or the group as a whole. Leaders differ in the level of empowerment that they provide when leading groups (Chen et al., 2007), with some leaders providing more and others less empowerment (e.g., Mathieu et al., 2006). We refer to this average level of leader's empowerment as empowering leadership climate. It has been shown to foster group ef cacy (e.g., Srivastava et al., 2006), and we argue that empowering leadership climate is also positively related to individual-level career self-efcacy beliefs, which in turn are positively associated with career satisfaction. The empowering leadership climate constitutes an important aspect of the context within which career perceptions develop. In the case of a strong empowering leadership climate, individuals are more likely to perceive that, independent of and in addition to their dyadic relationship with the leader, the leader con- sistently establishes and maintains the conditions that are conducive to career development and success. Thus, if the leader enacts high levels of empowering leadership, individual followers can be more condent that contextual conditions are and will continue to be in place that will help them achieve their career goals, thus fostering individual career self-ef cacy and, in turn, career satisfaction. From the perspective of LMX theory (Liden et al., 1997), high levels of empowering leadership signal to the team members that the leader is willing to engage in high-quality relationships marked by mutual commitment and trust with each and every one of his or her followers. Consequently, followers will believe that such a leader will support each individual team member's career

6 T. Biemann et al. / The Leadership Quarterly xxx (2015) xxx – xxx development.

6

T. Biemann et al. / The Leadership Quarterly xxx (2015) xxxxxx

development. If, by contrast, empowering leadership climate is weak, individual subordinates will perceive contextual conditions to be less conducive to helping them attain their career-related objectives. Even if their dyadic relationship with the leader is currently marked by high levels of empowerment, individuals may view low levels of empowering leadership climate as revealing that high empowerment is not the norm and cannot be taken for granted. They realize that the leader is not generally striving for high- quality relationships with each follower. Under such conditions, followers may be less likely to expect that contextual conditions will necessarily remain supportive of their career aspirations, thus diminishing their career self-efcacy and, in turn, their career satisfaction. In Hypotheses 1 and 2, we posited links between dyadic leadermember relations and career perceptions. Here, we extend the frame of reference from one dyadic relationship to the leader's relationships with all members of a group.

Hypothesis 3. There is a positive relationship between empowering leadership climate and individual career self-efcacy.

Hypothesis 4. Individual career self-efcacy partially mediates the positive relationship between empowering leadership climate and individual career satisfaction.

Effects of empowering leadership differentiation As has frequently been argued in the recent leadership literature, variance among team member's evaluations of leader behaviors may constitute relevant information rather than being mere error variance ( Chen et al., 2011; Gooty & Yammarino, in press; Henderson et al., 2009; Liao et al., 2010). Dispersion in leader style evaluations of group members reveals the level of consensus among the team members (Feinberg, Ostroff, & Burke, 2005). High variance in group evaluations signies low consensus, whereas low variance indicates high consensus. Recent studies that have examined dispersion in leadership evaluations have con rmed that consensus matters, but yielded mixed results (e.g., Boies & Howell, 2006; Cole & Bedeian, 2007; Cole, Bedeian, & Bruch, 2011; Feinberg et al., 2005). Some researchers found evidence for positive effects of leadership differentiation on group-level outcomes. For example, Liden et al. (2006) identied a positive relationship between high LMX differentiation and performance when LMX quality was low. Even less clear, however, is the relationship between LMX differentiation and individual-level outcomes (Gooty & Yammarino, in press). We propose that leadership differentiation can simultaneously have two opposing effects on subordinates' career perceptions. On the one hand, it signals the possibility that some individuals can hope to establish a particularly high-quality relationship with the leader. On the other hand, it increases uncertainty and violates principles of equality and consistency, thereby reducing individuals' satisfaction. We argue that leadership differentiation is thus positively associated with career self-efcacy and negatively related to career satisfaction. In the following, we will rst describe the hypothesized positive effect and then elaborate on the posited negative effect. Members of a group can inuence the perceived realities of one another such that consensus with regard to leadership style may be seen as the quality of the shared social environment (Cole & Bedeian, 2007) or a shared reality (Gooty & Yammarino, in press). Heterogeneity in leader behavior assessments may re ect that a leader treats his or her followers differentially (e.g., Nishii & Mayer, 2009 ). Such a leader differentiates among employees to a greater extent than a leader with consistent assessments. This possibility might be particularly germane with respect to empowering leadership, given that different followers might require either objectively or in the perception of the leader different levels of autonomy and responsibility. We argued above that career self-efcacy comprises individuals' beliefs that they are capable of navigating their own careers. A more differentiated treatment of employees implies higher competition among employees and also higher incentives to attempt to become a better supported employee in the group ( Wayne & Ferris, 1990 ). Subordinates in a group with a leader who exhibits more differentiated empowering behaviors might perceive this as an opportunity to proactively engage in leadermember interac- tions with the goal of securing for themselves more empowering behaviors, preferential treatment, and better access to resources from the leader. In short, empowering leadership differentiation may serve as a cue to group members that those who try have a chance of establishing for themselves higher quality relationships with the leader who does not treat all followers similarly (cf. Erdogan & Liden, 2006; Henderson et al., 2009). To this end, subordinates could ask for additional coaching, career-relevant infor- mation, challenging tasks, or greater responsibilities. We argue that this would increase subordinates career self-efcacy because it fosters feelings of inuence and control over their individual work situation and their skill and career development. By contrast, in a group with a leader who empowers all subordinates equally (i.e., no leadership differentiation), subordinates will be less likely to perceive that they themselves can direct their careers, since variance in individual pro-activity and initiative apparently does not lead to variance in the degree to which leaders empower different followers. Subordinates may thus believe that there is little that they themselves can do to elicit more empowering behaviors from their leaders and forge preferential relationships with them, which we argue will undermine their self-efcacy.

Hypothesis 5. There is a positive relationship between empowering leadership differentiation and individuals' career self-efcacy.

At the same time, studies on leadership differentiation have argued that a lower consensus implies a higher probability of conicts and competition among the group members (Boies & Howell, 2006; Cole et al., 2011). For example, Sherony and Green (2002) have found that when leaders establish differential relationships with different members of a group, this also affects relationships among the group members themselves. Moreover, Harris, Li, and Kirkman (2014) found that high quality LMX relationships have a weaker effect on OCB and turnover intentions when LMX differentiation is high. And Hooper and Martin (2008) found that greater LMX

T. Biemann et al. / The Leadership Quarterly xxx (2015) xxx – xxx 7 differentiation

T. Biemann et al. / The Leadership Quarterly xxx (2015) xxxxxx

7

differentiation negatively inuences job satisfaction and well-being of group members. All of these ndings suggest that empowering leadership differentiation may not only engender the advantageous effects posited above, but also adverse consequences. Leadership differentiation reduces predictability and stability of future leadermember relations because it enables some subordi- nates to develop closer relationships with the leader, whereas other group members may build relationships with the leader that are characterized by greater distance and/or lower quality. In a group with high leadership differentiation, subordinates might therefore compete for the leader's resources with the goal of establishing themselves among those that are shown preferential treatment by the leader. This higher competition and lower stability may adversely affect subordinates' satisfaction with their careers. In a similar vein, Gooty and Yammarino (in press) have recently argued that LMX differentiation indicates a lack of a shared reality with respect to lead- er fairness, inequality in resource allocation, and uncertainty in regard to future leader actions. These same authors contend that mem- bers in the same status groups (e.g., subordinate managers reporting to the same leader) generally desire equal, rather than differential, treatment (also see Fiske, 1992). Leaders that differentially empower their followers violate these needs and expectations of their followers, which we argue is likely to diminish follower career satisfaction.

Hypothesis 6. There is a negative relationship between empowering leadership differentiation and individual's career satisfaction.

Hypotheses 2 and 4 propose that career self-ef cacy mediates the relationship between empowering leadership (at both the individual and the group level) and career satisfaction (at the individual level). We also argued above that higher career self- efcacy results in higher career satisfaction, because subordinates will anticipate a higher future career success if they believe that they can successfully manage their careers. Moreover, Hypothesis 5 proposes a link between empowering leadership differentiation (at the group level) and career self-efcacy. Hence, in combining these assumptions we posit:

Hypothesis 7. Individual career self-efcacy partially mediates the relationship between empowering leadership differentiation and career satisfaction.

Methods

Sample and data collection

We used a multilevel data set consisting of questionnaire and personnel data from 2493 employees in leadership positions in 704 groups of a large pharmaceutical company with headquarters in Germany. The survey was conducted online in May 2010 with a response rate of about 75%. Survey invitations were sent out from the account of the HR Board member. The population included all managers above the lowest executive level (entry level) and below the highest hierarchical level of this company. These managers were based in Germany, the U.S., the U.K., or France. Participation in the survey was voluntary. We matched personnel data of the managers to the survey answers. Respondents agreed to this procedure before completing the survey. Only 28 (0.7%) of the respon- dents rejected this matching procedure. The personnel records contained additional information on demographics (age, gender, ten- ure) and job characteristics (hierarchical level, compensation, organizational unit, contractual status). Information on the direct supervisor was included in the personnel records, which was then used to identify groups and their leaders. The average age in the sample was 48.47 years (SD = 6.73), which is relatively high because of this study's focus on managers in leadership positions (i.e., above entry level). 71.2% of the survey answers came from German executives, 24.8% from the U.S., 2.4% from France, and 1.6% from the U.K.

Measures

The survey was conducted in 3 different languages (German, English, and French). Translation and back-translation resulted in no signicant inconsistencies among the versions. We pre-tested the questionnaire by sending it to 100 randomly drawn executives in the rm. Based on the answers of 58 respondents in the pretesting phase, we tested for reliability of the scales and slightly adapted the scales by modifying the wording of some items. Empowering leadership, career self-efcacy and career satisfaction were all measured with individual survey responses. Items were answered on a 5-point-Likert scale ranging from strongly disagree(1) to strongly agree(5).

Empowering leadership The original scale of empowering leadership by Arnold et al. (2000) consisted of 38 items and comprised 5 different factors. In our

survey, we adapted the shorter scale used by Srivastava et al. (2006), which measures empowering leadership on a single dimension. Our nal scale consisted of 10 items with a Cronbach's alpha of 0.92. A sample item from the scale was: My direct supervisor encour- ages team members to express ideas/suggestions.For measuring empowering leadership at the individual level, we calculated the mean score of the ten items for each respondent. At the group level, we computed the group means of the individual mean scores. We justied aggregating empowering leadership to a higher level on the basis of ICC and r WG values: ICC(1) was 0.13 ( σ within =

= 0.075; F ratio = 1.55, p b .001), ICC(2) was 0.35, and mean r WG(J) uniform was 0.85 for a uniform distribution with

0.498, σ between

a random variance of σ 2 = 2. The distribution of empowering leadership in the sample was slightly skewed (skew = 0.62). There- fore, we calculated r WG(J) values for a slightly skewed distribution, which resulted in a mean r WG(J) measure-speci c value of 0.72, assum-

ing a variance of 1.34 for the underlying null distribution (Biemann, Cole, & Voelpel, 2012). ICC and r WG values indicated sufcient

2

2

8 T. Biemann et al. / The Leadership Quarterly xxx (2015) xxx – xxx within-group

8

T. Biemann et al. / The Leadership Quarterly xxx (2015) xxxxxx

within-group reliability and within-group agreement. Thus, we also used group values of empowering leadership in testing our hypotheses. We further calculated the standard deviation of individuals' empowering leadership mean scale values for each group, following the procedure described in Biemann and Kearney (2010) to account for varying group sizes.

Career self-efcacy The scale for career self-efcacy was based on the conceptualization of general self-efcacy by Chen, Gully, and Eden (2001). We reformulated the eight items with respect to the career context (see Appendix A). A sample item was: When facing difcult assign- ments in my career, I am certain that I will accomplish them.Cronbach's alpha for this scale was 0.76. Hypotheses 3 and 5 suggested higher-level predictors of career self-efcacy. ICC(1) values of 0.13 (σ within = 0.163, σ between = 0.024; F ratio = 1.52, p b .001) indi- cated signicant between-group variance, which justies adding higher-level predictors in a multilevel model (Biemann et al., 2012).

2

2

Career satisfaction We used Greenhaus, Parasuraman, and Wormley's (1990) scale to measure career satisfaction. It is the most widely used scale to measure career satisfaction (Heslin, 2005). A sample item was: I am satised with the progress I have made toward meeting my goals for the development of new skills.The 5 item measure yielded a Cronbach's alpha of 0.84 and the ICC(1) value was 0.09 (σ within =

2

2

0.501, σ between

= 0.049; F ratio = 1.35, p b .001).

Control variables At the group level, we added the division within the company (six dummies), location (Germany, U.S., U.K., and France), group size, and age of the supervisor as control variables to the statistical model. We included these group level controls to account for possible within-company effects on career outcomes. For example, previous research has revealed mean differences in measures of general self-efcacy across countries, although the underlying construct is congurally equivalent (Scholz, Doña, Sud, & Schwarzer, 2002). At the individual level, we added demographic data (age, gender, marital status, number of children below the age of 14) and a variety of career-related control variables (previous years in part-time; current employment status; organizational tenure; number of previous companies, functions, countries, promotions, and subgroups). We used these variables to control for previous career experiences that might inuence individuals' career perceptions. We further included annual income and hierarchical level (four levels) to control for effects of objective career success on career self-efcacy and career satisfaction. Control variables were gathered from personnel records, with the exception of the number of children, which was assessed in the questionnaire.

Analysis

The data had a hierarchical data structure with two levels of analysis (individuals in groups). Therefore, we applied a multilevel model that reected this non-independence in the data and allowed entering variables at different levels of analysis (Bliese, 2000). In ordinary least squares regressions, centering issues are relatively straightforward in that researchers may decide whether or not variables should be mean-centered (i.e., subtracting a variable's overall mean from each case of that variable; Aiken & West, 1991; Echambadi & Hess, 2007). Centering in multilevel models requires more attention, because lower level variables can be centered at the grand mean (CGM) or the deviation from a group's mean can be calculated (centering within clusters, CWC; Enders & Toghi, 2007). In the latter option, mean group values are calculated and the deviation of each case from the case's group mean is computed. CWC removes all between-cluster variation and, thus, group-mean centered variables are uncorrelated with higher-level variables. Statistically, there is no correct choice between CGM and CWC and both yield correct results (Kreft, De Leeuw, & Aiken, 1995). How- ever, both centering options generate parameter estimates that differ in value and also in meaning (Enders & Toghi, 2007). Centering decisions should therefore be based on the underlying research questions. For tests of the direct effect (Hypothesis 1) and the contex- tual effect (i.e. predictor at both individual and team level simultaneously; Hypothesis 3), we entered empowering leadership at the individual level CGM, because this allowed for a straightforward test of a contextual effect (Enders & Toghi, 2007). For the mediation hypotheses, we followed the recommendation of Zhang, Zyphur, and Preacher (2009). That is, empowering leadership for the medi- ating processes posited in Hypotheses 2, 4, and 7 was entered CWC in the model to disentangle lower level and higher level variance. There was missing data in our variables, which varied between zero (e.g., gender) and 3.6% (marital status). We applied multiple imputation (Rubin, 1987; Sinharay, Stern, & Russell, 2001) to avoid a loss of information and statistical power that is inherent in the common procedure of listwise deletion (Graham, 2009; Roth, 1994; Schafer & Graham, 2002). Since the nested data structure must be accounted for when imputing data, we applied an imputation algorithm described by Schafer (2001) and Schafer and Yucel (2002) that is specically designed for clustered data. We conducted imputations with the PAN extension package using the R language for statistical computing (R Development Core Team, 2008; the code is available on request). Following Sinharay et al. (2001), we cre- ated m = 5 imputations. Datasets were stored and analyses were performed on each of the ve datasets. Subsequently, we combined the results following the rules suggested by Rubin (1987).

Results

Table 1 displays the means, standard deviations, and zero-order correlations among the study variables. We conducted hypotheses tests with the software HLM (Raudenbush & Bryk, 2002). For tests of direct relationships, we estimated multilevel models with career self-efcacy and career satisfaction as dependent variables (see Table 2). The data supported Hypothesis 1, which posited a positive relationship between empowering leadership at the individual level and career self-efcacy (gamma = 0.071, p b .001). However,

9T.

Biemann et al. / The Leadership Quarterly xxx (2015) xxxxxx

Please cite this article as: Biemann, T., et al., Empowering leadership and managers' career perceptions: Examining effects at both the individual and the team level, The Leadership Quarterly (2015), http://dx.doi.org/10.1016/j.leaqua.2015.03.003

Table 1 Means, standard deviations, and correlations.

 

M

SD

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

1. Career satisfaction 2. Career self-ef cacy

3.52

0.74

0.84

3.98

0.43

.21

0.76

3. Empowering leadership 3.63

0.76

.36

.14

0.92

4. Gender (1 = female)

0.17

0.38

.05

.10

.01

5.

Age

48.47

6.73

.03

.13

.04

.13

6.

Organizational tenure

17.23

8.80

.03

.20

.04

.17

.63

7. Annual income 8. No. of companies 9. No. of functions 10. No. of countries

.10

.14

.01

.09

.20

.04

1.55

2.02

.00

.17

.02

.10

.02

.31

.06

1.87

1.01

.07

.03

.05

.09

.06

.12

.09

.07

1.01

1.20

.03

.05

.03

.06

.02

.01

.16

.00

.11

11. No. of subgroups

1.44

0.78

.01

.03

.02

.08

.01

.13

.04

.08

.22

.00

12.

Years in part-time

0.40

2.42

.01

.04

.05

.14

.02

.03

.11

.04

.00

.04

.01

13. Employment status 14. Married (1 = yes)

0.99

0.05

.01

.04

.01

.27

.09

.05

.26

.03

.04

.04

.00

.35

0.94

0.27

.02

.01

.03

.22

.02

.08

.04

.04

.03

.00

.05

.01

.00

15.

Children b 14

0.75

1.01

.03

.08

.05

.11

.46

.35

.07

.02

.04

.02

.00

.00

.06

.13

16. Age of supervisor 17. Emp. lead. group mean 18. Emp. lead. SD

49.68

6.01

.05

.03

.03

.04

.15

.12

.03

.01

.02

.00

.03

.01

.02

.03

.05

3.63

0.47

.23

.09

.61

.02

.01

.01

.04

.02

.05

.02

.00

.05

.02

.02

.00

.05

0.69

0.40

.12

.07

.19

.02

.01

.07

.01

.08

.01

.01

.02

.00

.02

.06

.03

.02

.31

19. Group size

4.45

2.35

.05

.03

.02

.00

.04

.04

.17

.04

.05

.01

.04

.03

.01

.00

.03

.12

.03

.06

.03 .01 .00 − .03 .12 .03 − .06 Notes . Cronbach's alphas in the diagonal

Notes. Cronbach's alphas in the diagonal (in italics); for rows 1 to 15 (individual level variables; N = 2493): p b .05 for r |.04|, p b .01 for r |.06|, and p b .001 for r |.07|; for rows 16 to 19 (group level variables; N = 704):

p b .05 for r |.08|, p b .01 for r |.10|, and p b .001 for r |.13|; correlation coefcients do not account for the nested data structure.

10 Table 2 Results from multilevel models. T. Biemann et al. / The Leadership Quarterly

10

Table 2 Results from multilevel models.

T. Biemann et al. / The Leadership Quarterly xxx (2015) xxxxxx

 

DV: career self-ef cacy

 

DV: career satisfaction

 

Model 1

Model 2

 

Model 3

Model 4

(Controls only)

(Controls only)

Individual level Annual income a

0.000

0.000

0.003⁎⁎

0.002⁎⁎

(0.001)

(0.001)

 

(0.001)

(0.001)

No. of promotions

0.027

0.024

0.097

⁎⁎⁎

0.074⁎⁎⁎

(0.012)

(0.012)

 

(0.022)

(0.020)

Hierarchical level dummies Gender (1 = female)

Yes

Yes

Yes

Yes

0.079 ⁎⁎

0.082

⁎⁎⁎

 

0.068

0.062

 

(0.024)

(0.024)

 

(0.043)

(0.040)

Age

0.007 ⁎⁎⁎

0.007

⁎⁎⁎

0.014⁎⁎⁎

0.011⁎⁎⁎

(0.002)

(0.002)

(0.003)

(0.003)

Organ. tenure

0.004 ⁎⁎

0.004 ⁎⁎

 

0.010

⁎⁎⁎

0.011⁎⁎⁎

(0.001)

(0.001)

 

(0.002)

(0.002)

No. of companies

0.013 ⁎⁎

0.012

⁎⁎

0.011

0.006

(0.004)

(0.005)

 

(0.008)

(0.008)

No. of functions

0.018

0.020

0.063⁎⁎⁎

0.058⁎⁎⁎

(0.008)

(0.008)

 

(0.015)

(0.014)

No. of countries

0.013

0.014

0.031

0.027

(0.007)

(0.007)

 

(0.012)

(0.012)

No. of subgroups

0.004

0.002

0.009

0.008

(0.011)

(0.011)

 

(0.020)

(0.019)

Years in part-time

0.001

0.000

0.002

0.002

(0.003)

(0.004)

 

(0.006)

(0.006)

Employment status

0.317

0.320

0.001

0.048

(0.150)

(0.192)

 

(0.343)

(0.317)

Married (1 = yes)

0.041

0.042

0.080

0.036

(0.037)

(0.033)

 

(0.062)

(0.057)

Children b 14

0.008

0.004

0.033

0.040

(0.009)

(0.009)

 

(0.017)

(0.016)

Empowering leadership (CGM)

0.071 ⁎⁎⁎

 

0.329⁎⁎⁎

 

(0.013)

 

(0.022)

Career self-ef cacy (CWC)

––

 

0.216⁎⁎⁎

 

(0.039)

Group level Age of supervisor

0.002

0.003

0.003

0.004

(0.002)

(0.001)

 

(0.003)

(0.003)

Group size

0.001

0.001

0.002

0.001

(0.004)

(0.004)

 

(0.007)

(0.007)

Division dummies Country dummies Empowering leadership (group mean)

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

0.000

0.045

 

(0.023)

 

(0.040)

Empowering leadership (SD)

0.073 ⁎⁎⁎

 

0.082

 

(0.022)

 

(0.038)

Career self-ef cacy (group mean)

––

 

0.328⁎⁎⁎

 

(0.066)

2

σ

within

(L1 variance)

0.154

0.152

 

0.478

0.414

2

(L2 variance)

0.003

0.002

0.022 ⁎⁎

0.017⁎⁎

σ

between

df (level 1, level 2)

(2466, 693)

(2463, 691)

 

(2466, 693)

(2461, 690)

Notes . Reported are average gamma coef cients based on ve data sets with imputed missing values (see text for details); standard errors in parentheses; CGM = centered at the grand mean; CWC = centered within clusters; df = degrees of freedom.

a Shown are values multiplied by 10 3 .

p b .05. p b .01. ⁎⁎⁎ p b 0.001.

⁎⁎

the data did not con rm Hypothesis 3 , which proposed a positive relationship between the group mean of empowering leader- ship and career self-ef cacy (gamma = 0.000, p N .05). With respect to the standard deviation of empowering leadership, we found support for Hypothesis 5 leadership differentiation was signi cantly positively related to career self-ef cacy (gamma = 0.073, p b .001) and Hypothesis 6 leadership differentiation was signi cantly negatively related to career satis- faction (gamma = 0.082, p b .05). Table 3 exhibits the results from the tests of our mediation hypotheses. In conducting these tests, we followed the procedure described in Zhang et al. (2009). We had posited that career self-efcacy mediates the respective paths to career satisfaction from empowering leadership at the individual level (Hypothesis 2), from the group mean of empowering leadership (Hypothesis 4), and from the groups' standard deviation of empowering leadership (Hypothesis 7). In step 1, we estimated the relationship between

T. Biemann et al. / The Leadership Quarterly xxx (2015) xxx – xxx Table 3

T. Biemann et al. / The Leadership Quarterly xxx (2015) xxxxxx

Table 3

Mediation analyses.

11

 

Step 1

Step 2

Step 3

Sobel Z

X Y

X M

X M Y

γ (s.e.)

γ (s.e.)

γ (s.e.)

Individual level Empowering leadership (CWC) a

0.344 ⁎⁎⁎*

0.078⁎⁎⁎

0.333

⁎⁎⁎

3.63

⁎⁎⁎

(0.022)

(0.015)

(0.025)

(Hypothesis 2)

Career self-ef cacy (mediator) (CWC)

0.213 ⁎⁎⁎

(0.042)

Group level Empowering leadership (group mean)

0.308 ⁎⁎⁎

0.000

0.045

0.01

(0.033)

(0.023)

(0.040)

(Hypothesis 4)

Empowering leadership (SD)

0.057

0.073⁎⁎⁎

0.082

2.78

⁎⁎

(0.038)

(0.022)

(0.038)

(Hypothesis 7)

Career self-ef cacy (group mean)

0.328 ⁎⁎⁎

(0.066)

a Empowering leadership was entered CGM for the mediation effects at group level to focus on contextual effects.

p b .05. p b .01. ⁎⁎⁎ p b 0.001.

⁎⁎

empowering leadership and career satisfaction without including career self-efcacy as a mediator in the model. It is noteworthy that in this model, the group mean of empowering leadership was signicantly positive (gamma = 0.308, p b .001), because empowering leadership was entered CWC at the individual level. CWC removes all between-group variance and yields individual-level scores that are unrelated to higher-level variables (Enders & Toghi, 2007). This is in line with the recommendations of Zhang et al. (2009) for cases in which a mediation effect of a lower-level variable is of interest, and not a contextual effect, as in Hypothesis 1. The gamma coefcient for empowering leadership at the individual level (gamma = 0.344, p b .001) was similar to the gamma coefcient of the group mean effect, which indicates that the slopes for the relationships between empowering leadership and career satisfaction are similar at both levels of analysis. Thus, the signicant gamma coefcient of the group mean of empowering leadership may not be interpreted as a contextual effect. Overall, the results (summarized in Table 3) support Hypothesis 2 (career self-efcacy partially mediated the relationship between empowering leadership at the individual level and career satisfaction) and Hypothesis 7 (career self-efcacy partially mediated the relationship between empowering leadership differentiation and career satisfaction). By contrast, Hypothesis 4 was not supported, because there was no signi cant relationship between empowering leadership at the group level and career self-ef cacy (gamma = 0.000, p N .05). Concerning Hypothesis 7, we obtained an inconsistent mediation model, because the effect of the standard deviation on the mediator was positive, but negative on the outcome (MacKinnon, Fairchild, & Fritz, 2007). Thus, the direction of the direct and indirect effects of the independent variable on the dependent variable differed. Although there was a positive relationship between the standard deviation of empowering leadership (X) and career self-efcacy (M), the relationship between leadership differentiation and career satisfaction (Y) was negative. As a result, the direct effect of the standard deviation on the outcome (X Y) is not signicant if the mediator is not included in the regression model (0.057, p N .05).

Discussion

Career perceptions of highly valued employees are important for organizations in that they can help decrease turnover and improve performance (e.g., Kraimer et al., 2011). It is therefore important for scholars and managers alike to understand what role empowering leadership behaviors can play in fostering followers' career self-efcacy and career satisfaction. In the present study, we explored this role of empowering leadership. In doing so, we attempted to address what we perceive to be a major weakness in the extant literature. Prior research on empowering leadership has typically studied the effects of empowering leadership at either the individual or the group-level analysis and, when examining the group-level, has assumed that empowering leaders generally behave similarly toward their different subordinates. We argued that a broad understanding of the link between empowering leadership and follower career perceptions is only possible when one examines the partially differential effects of three different conceptualizations and operationalizations of this construct. The results of our study provide strong evidence for the assumption that empowering leadership in uences subordinate managers' career self-efcacy and career satisfaction at the individual level. Although subordinates' perceptions within leadersubordinate dyads appear to be affected by empowering leadership, we were unable to conrm a leadership climate effect (i.e., empowering leadership operationalized as the mean of group member leader rat- ings). In line with our hypotheses, however, we did nd that leadership differentiation (i.e., empowering leadership operationalized as the standard deviation of group member leader ratings) was positively related to career self-efcacy and negatively related to career satisfaction. The latter effect was mediated by career self-efcacy in an inconsistent mediation. That is, although the respective relationships between differentiated empowering leadership and career self-efcacy and between career self-efcacy and career satisfaction were positive, the direct relationship between differentiated empowering leadership and career satisfaction was negative.

12 Theoretical implications T. Biemann et al. / The Leadership Quarterly xxx (2015) xxx –

12

Theoretical implications

T. Biemann et al. / The Leadership Quarterly xxx (2015) xxxxxx

Within the LMX literature, Gooty and Yammarino (in press) have shown that the simultaneous investigation of LMX at different levels of analysis, including the examination of LMX differentiation at the group level, enriches understanding of the effects of LMX on individual-level outcomes. In the present study, we adopted a similar logic and argued that not only the LMX literature, but also the literatures on other leadership constructs in our case, empowering leadership can be enriched by this approach. We relied on the logic underlying LMX theory (e.g., Liden et al., 1997) to develop hypotheses on how empowering leadership at the individual level of analysis and at the group level (leadership climate and leadership differentiation) may affect followers' career perceptions. The ben- ets of our approach are underscored by our ndings that these three different aspects of empowering leadership were differentially related to individual career perceptions. A major implication of our study concerns the distinction between individual level and group level effects of empowering leader- ship. In our study, we employed a model with the aim to disentangle effects at different levels of analysis. While we found signicant links between empowering leadership and career perceptions at the individual level (Hypotheses 1 and 2), hypotheses regarding mean group level effects were not supported (Hypotheses 3 and 4). This nding is in line with previous LMX research that highlighted the importance of levels of analysis when developing and testing hypotheses (Cogliser & Schriesheim, 2000; Schriesheim, Castro, & Yammarino, 2000). Our results indicate that subordinates' career perceptions are formed in dyadic relationships between leaders and subordinates. Thus, although we found some stability in leader's behaviors across leader member dyads, as exempli ed by high within-group agreement and reliability for empowering leadership in our data, there is no indication of a leadership climate effect that adds to the dyadic effect. For career perceptions, this could be explained by the individual-focused nature of career devel- opment. Only leadership behavior directed at oneself impacts career perceptions, whereas observing the group leader's empowering behavior toward other group members has no meaningful impact on one's own career perceptions. For more group-focused outcomes such as group efcacy or group cohesion, a stronger empowering leadership climate effect can be expected (e.g., Srivastava et al., 2006). However, previous leadership research primarily found effects at either the individual or the group level, but not both levels simultaneously. Our results show that there is a direct effect of the group mean of empowering leadership on career perceptions, but that this effect disappears when empowering leadership is added at the individual level (see Tables 1 and 3). This should be con- sidered in future leadership research when leadership style enters the statistical model only as an aggregated group level construct, but theoretical assumptions rest upon leadership climate effects. Moreover, the results of this study imply that empowering subordinates differentially is a double-edged sword. On the one hand, differentiated empowering leadership is likely to create an atmosphere of competition in a group, because the leader's resource allocation may not be xed, but instead may be contingent on subordinates' behavior. Our results show that this higher uncertainty that is likely to accompany differentiation with regard to future career support manifests in lower career satisfaction among subordi- nates. Previous research has shown similar effects of leadership differentiation for example, more withdrawal behavior and a negative impact on relationships among group members and work attitudes ( Erdogan & Bauer, 2010; Hooper & Martin, 2008; Nishii & Mayer, 2009; Schyns, 2006; Sherony & Green, 2002). On the other hand, differentiated empowering leadership increases career self-efcacy. Through observations and communication with their reference group, individuals experience that not all subor- dinates in their group are treated equally. This makes them believe that they have options to actively seek support and development opportunities, as their actions can in uence the trust, support, and amount of resources that they will receive from their leader ( Erdogan & Liden, 2002 ). While prior research on the effects of leadership differentiation typically emphasized either positive (e.g., Erdogan & Liden, 2006; Henderson et al., 2009 ) or negative effects (e.g., Gooty & Yammarino, in press; Hooper & Martin, 2008), our study suggests that empowering leadership differentiation simultaneously has both positive and negative effects on indi- vidual career perceptions. Combining both the positive effect of differentiated leadership on career self-efcacy and the negative effect on career satisfaction, we nd an inconsistent mediation in our data. It appears that leaders who establish leadermember relation- ships of different qualities directly decrease career satisfaction, but indirectly increase followers' satisfaction through the positive pathway via career self-efcacy, which in turn is positively associated with career satisfaction.

Managerial implications

Although organizations cannot guarantee lifetime employment, they must make sure to retain a capable and exible workforce, especially in critical positions within the organization. In our study, we argued that managers' career perceptions play an important role in this regard. We reasoned that an empowering leadership style can help to keep employees satised and achieve high employee loyalty. Our results offer several insights for executives and organizations. First, leadership differentiation is a double-edged sword. On the one hand, it increases individuals' beliefs that they can actively manage their careers, which in turn increases their satisfaction. On the other hand, leadership differentiation might foster feelings of competition, instability, and uncertainty in leadermember relation- ships, which in turn diminish satisfaction with one's own career. To buffer these unwanted side effects of leadership differentiation, leaders might set clear development goals and outline potential career pathways for subordinates within the organization. They could also communicate transparent criteria that determine the extent to which they will enact empowering leadership behaviors toward followers. These actions might help to reduce career insecurities and reduce individuals' turnover intentions. Second, our results indicate that leaders must establish individual relationships with each follower. At least with regard to career perceptions, empowering leadership is a dyadic concept. Leaders should not rely on merely establishing a general positive climate within the group, but must work on each leadermember relationship, which then positively impacts individual group members. Thus, leaders would be well advised to exhibit empowering leadership behaviors that focus on each member's individual needs.

T. Biemann et al. / The Leadership Quarterly xxx (2015) xxx – xxx Limitations, future

T. Biemann et al. / The Leadership Quarterly xxx (2015) xxxxxx

Limitations, future directions, and conclusion

13

We acknowledge several limitations of our study. First, we did not measure employee performance or turnover as objective work outcomes. Although career satisfaction and self-ef cacy have been shown to be related to relevant work outcomes ( Finegold & Mohrman, 2001; Kraimer et al., 2011), the lack of such outcome data did not allow us to replicate these ndings in our study. Further- more, in Hypotheses 1 and 2, we were interested in relationships between individuals' career perceptions and leadership styles and therefore obtained these measures from the same source that is, the individuals themselves. While this approach bears the risk of method bias that can affect the studied relationships in undesired ways (Podsakoff, MacKenzie, & Podsakoff, 2012), the substantial within-group agreement for the relevant constructs makes us believe that common method variance is not the main reason for the direction and size of our results. Within-group agreement indicates coherence of individuals' assessments in a group, thereby indicat- ing a common truesource of variance (e.g., leadership behavior) as the basis of their rating. For the same reason, Hypotheses 3 to 7 are not likely to be affected by this type of method bias, as they relate group values to individual-level outcomes. A second limitation of our study is the sample that consisted of employees in leadership position from only one German organi- zation. The organizational tenure of individuals in our sample was relatively high (M = 17.23 years), which underscores the impor- tance of career opportunities within the organization. However, further research is needed that examines the relevance of organizational and cultural differences as potential moderators of the studied relationships. Third, we only analyzed empowering leadership, but did not incorporate other leadership approaches such as transformational and transactional leadership (Bass, 1985). We do believe, however, that these approaches would also benet from studies that examine leadership effects at both the individual and the group level and simultaneously investigate the impact of leadership climate as well as leadership differentiation. Moreover, it is not only leadership behavior that affects employees' career perceptions (Ng et al., 2005). A promising area for future research might therefore be an analysis of the impact (and joint effect with leadership) of other organizational measures such as talent pools or career planning on career perceptions and work outcomes of highly qualied employees. Finally, our study showed that empowering leadership differentiation is positively related to career self-efcacy and negatively related to career satisfaction. The question therefore arises of whether, on balance, the positive or the negative effects are likely to prevail. A promising avenue for future research would thus be to explore the relative inuence of these countervailing effects of empowering leadership differentiation on outcomes such as performance and turnover. In conclusion, our study identi es the value that lies in examining the differential effects of different conceptualizations of empowering leadership at the individual and the team level of analysis. Our ndings reveal that this approach paints a richer and more complex picture and enables a broader understanding of the association between empowering leadership behaviors and indi- vidual followers' career perceptions.

Appendix A

Career self-efcacy (based on Chen et al., 2001 and adapted to the career context) Answer scale: (1 = strongly disagree, 2 = disagree, 3 = neither agree nor disagree, 4 = agree, 5 = strongly agree)

I will be able to achieve most of the career goals that I have set for myself.

When facing difcult assignments in my career, I am certain that I will accomplish them.

Regarding my career, I think that I can obtain outcomes that are important to me.

I believe I can succeed at almost any career endeavor to which I set my mind.

I will be able to successfully overcome many challenges in my career.

I am condent that I can perform effectively on many different assignments in my career.

Compared to other people, I can do most tasks in my job very well.

Even when job demands are tough, I can perform quite well.

References

Ahearne, M., Mathieu, J., & Rapp, A. (2005). To empower or not to empower your sales force? An empirical examination of the influence of leadership empowerment behavior on customer satisfaction and performance. Journal of Applied Psychology, 90(5), 945. Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and interpreting interactions. Sage. Amundsen, S., & Martinsen, O. L. (2014). Empowering leadership: Construct clarification, conceptualization, and validation of a new scale. The Leadership Quarterly, 25(3), 487511. Argyris, C. (1998). Empowerment: The emperor's new clothes. Harvard Business Review, 76(3), 98105. Arnold, J. A., Arad, S., Rhoades, J. A., & Drasgow, F. (2000). The empowering leadership questionnaire: The construction and validation of a new scale for measuring leader behaviors. Journal of Organizational Behavior, 21(3), 249269. Arthur, M. B., & Rousseau, D. M. (Eds.). (1996). The boundaryless career. A new employment principle for a new organizational era. New York: Oxford University Press. Bandura, A. (1997). Self-efficacy: The exercise of control. W.H. Freeman. Bass, B. M. (1985). Leadership and performance beyond expectations. New York: Free Press. Bass, B. M., & Riggio, R. E. (2006). Transformational leadership. Mahwah, NJ: L. Biemann, T., Cole, M. S., & Voelpel, S. (2012). Within-group agreement: On the use (and misuse) of rWG and rWG(J) in leadership research and some best practice guidelines. The Leadership Quarterly, 23, 6680. Biemann, T., & Kearney, E. (2010). Size does matter: How varying group sizes in a sample affect the most common measures of group diversity. Organizational Research Methods, 13(3), 582599. Bliese, P. D. (2000). Within-group agreement, non-independence, and reliability: Implications for data aggregation and analysis. In K. J. Klein, & S. W. Kozlowski (Eds.), Multilevel theory, research, and methods in organizations (pp. 349381). Jossey-Bass.

14 T. Biemann et al. / The Leadership Quarterly xxx (2015) xxx – xxx Boies,

14

T. Biemann et al. / The Leadership Quarterly xxx (2015) xxxxxx

Boies, K., & Howell, J. M. (2006). Leadermember exchange in teams: An examination of the interaction between relationship differentiation and mean LMX in explaining team-level outcomes. The Leadership Quarterly, 17(3), 246257. Burke, C. S., Stagl, K. C., Klein, C., Goodwin, G. F., Salas, E., & Halpin, S. M. (2006). What type of leadership behaviors are functional in teams? A meta-analysis. The Leadership Quarterly, 17(3), 288307. Carmeli, A., Schaubroeck, J., & Tishler, A. (2011). How CEO empowering leadership shapes top management team processes: Implications for firm performance. The Leadership Quarterly, 22(2), 399411. http://dx.doi.org/10.1016/j.leaqua.2011.02.013. Chen, G., Gully, S. M., & Eden, D. (2001). Validation of a new general self-efficacy scale. Organizational Research Methods, 4(1), 6283. Chen, G., Kirkman, B. L., Kanfer, R., Allen, D., & Rosen, B. (2007). A multilevel study of leadership, empowerment, and performance in teams. Journal of Applied Psychology, 92(2), 331. Chen, G., Sharma, P. N., Edinger, S. K., Shapiro, D. L., & Farh, J. L. (2011). Motivating and demotivating forces in teams: Cross-level influences of empowering leadership and relationship conflict. Journal of Applied Psychology, 96(3), 541. Cogliser, C. C., & Schriesheim, C. A. (2000). Exploring work unit context and leadermember exchange: A multilevel perspective. Journal of Organizational Behavior, 21(5), 487511. Cole, M. S., & Bedeian, A. G. (2007). Leadership consensus as a cross-level contextual moderator of the emotional exhaustionwork commitment relationship. The Leadership Quarterly, 18(5), 447462. Cole, M. S., Bedeian, A. G., & Bruch, H. (2011). Linking leader behavior and leadership consensus to team performance: Integrating direct consensus and dispersion models of group composition. The Leadership Quarterly. Conger, J. A., & Kanungo, R. N. (1988). The empowerment process: Integrating theory and practice. Academy of Management Review, 471482. Cropanzano, R., & Mitchell, M. S. (2005). Social exchange theory: An interdisciplinary review. Journal of Management, 31(6), 874900. Dansereau, F., Graen, G., & Haga, W. J. (1975). A vertical dyad approach to leadership within formal organizations. Organizational Behavior and Human Performance, 13,

Enders, C. K., & Tofighi, D. (2007). Centering predictor variables in cross-sectional multilevel models: A new look at an old issue. Psychological Methods, 12(2), 121138. Erdogan, B., & Liden, R. C. (2006). Collectivism as a moderator of responses to organizational justice: Implications for leader-member exchange and ingratiation. Erdogan, B., & Bauer, T. N. (2010). Differentiated leadermember exchanges: The buffering role of justice climate. Journal of Applied Psychology, 95(6), 1104. Erdogan, B., & Liden, R. C. (2002). Social exchanges in the workplace: A review of recent developments and future research directions in leadermember exchange theory. In L. L. Neider, & C. A. Schriesheim (Eds.), Leadership (pp. 65114). Greenwich, CT: Information Age Publishing. Feinberg, B. J., Ostroff, C., & Burke, W. W. (2005). The role of within-group agreement in understanding transformational leadership. Journal of Occupational and Organizational Psychology, 78(3), 471488. Finegold, D. L., & Mohrman, S. A. (2001). What do employees really want? The perception vs. the reality. Korn/Ferry International. Firebaugh, G. (1978). A rule for inferring individual-level relationships from aggregate data. American Sociological Review, 557572. Fiske, A. P. (1992). The four elementary forms of sociality: Framework for a unified theory of social relations. Psychological Review, 99(4), 689. Forrester, R. (2000). Empowerment: Rejuvenating a potent idea. Academy of Management Executive(19932005), 6780. Gooty, J., & Yammarino, F. J. (2015). The leadermember exchange relationship: A multisource, cross-level investigation. Journal of Management (in press). Graen, G. B. (1976). Role making process within complex organizations. In M. D. Dunnette (Ed.), Handbook of industrial organizational psychology (pp. 12011245). Chicago: Rand-McNally. Graen, G. B., & Uhl-Bien, M. (1995). Relationship-based approach to leadership: Development of leadermember exchange (LMX) theory of leadership over 25 years:

Applying a multi-level multi-domain perspective. The Leadership Quarterly, 6(2), 219247. Graham, J. W. (2009). Missing data analysis: Making it work in the real world. Annual Review of Psychology, 60, 549576. Grant, R. M. (1996). Toward a knowledge-based theory of the firm. Strategic Management Journal, 17, 109122. Greenhaus, J. H., Parasuraman, S., & Wormley, W. M. (1990). Effects of race on organizational experiences, job-performance evaluations, and career outcomes. Academy of Management Journal, 33(1), 6486. Hall, D. T. (2002). Careers in and out of organizations. Vol. 107, Sage. Hall, D. T., & Moss, J. E. (1998). The new protean career contract: Helping organizations and employees adapt. Organizational Dynamics, 26(3), 2237. Harris, T. B., Li, N., & Kirkman, B. L. (2014). Leadermember exchange (LMX) in context: How LMX differentiation and LMX relational separation attenuate LMX's influence on OCB and turnover intention. The Leadership Quarterly, 25, 314382. Henderson, D. J., Liden, R. C., Glibkowski, B. C., & Chaudhry, A. (2009). LMX differentiation: A multilevel review and examination of its antecedents and outcomes. The Leadership Quarterly, 20(4), 517534. Henderson, D. J., Wayne, S. J., Shore, L. M., Bommer, W. H., & Tetrick, L. E. (2008). Leadermember exchange, differentiation, and psychological contract fulfillment: A multilevel examination. Journal of Applied Psychology, 93(6), 1208. Heslin, P. A. (2005). Conceptualizing and evaluating career success. Journal of Organizational Behavior, 26, 113136. Higgins, M. C., Dobrow, S. R., & Chandler, D. (2008). Never quite good enough: The paradox of sticky developmental relationships for elite university graduates. Journal of Vocational Behavior, 72(2), 207224. Holtom, B. C., Mitchell, T. R., Lee, T. W., & Eberly, M. B. (2008). Turnover and retention research: A glance at the past, a closer review of the present, and a venture into the future. Academy of Management Annals, 2(1), 231274. Hooper, D. T., & Martin, R. (2008). Beyond personal leadermember exchange (LMX) quality: The effects of perceived LMX variability on employee reactions. The Leadership Quarterly, 19(1), 2030. Hughes, E. C. (1937). Institutional office and the person. The American Journal of Sociology, 43(3), 404413. Ilies, R., Nahrgang, J. D., & Morgeson, F. P. (2007). Leadermember exchange and citizenship behaviors: A meta-analysis. Journal of Applied Psychology, 92, 269277. Judge, T. A., & Bono, J. E. (2001). Relationship of core self-evaluations traitsself-esteem, generalized self-efficacy, locus of control, and emotional stabilitywith job satisfaction and job performance: A meta-analysis. Journal of Applied Psychology, 86(1), 80. Kark, R., Shamir, B., & Chen, G. (2003). The two faces of transformational leadership: Empowerment and dependency. Journal of Applied Psychology, 88(2), 246. Kirkman, B. L., & Rosen, B. (1999). Beyond self-management: Antecedents and consequences of team empowerment. Academy of Management Journal, 5874. Kraimer, M. L., Seibert, S. E., Wayne, S. J., Liden, R. C., & Bravo, J. (2011). Antecedents and outcomes of organizational support for development: The critical role of career opportunities. Journal of Applied Psychology, 96(3), 485. Kreft, I. G. G., De Leeuw, J., & Aiken, L. S. (1995). The effect of different forms of centering in hierarchical linear models. Multivariate Behavioral Research, 30(1), 121. Liao, H., Liu, D., & Loi, R. (2010). Looking at both sides of the social exchange coin: A social cognitive perspective on the joint effects of relationship quality and differentiation on creativity. Academy of Management Journal (AMJ), 53(5), 10901109. Liden, R. C., Erdogan, B., Wayne, S. J., & Sparrowe, R. T. (2006). Leadermember exchange, differentiation, and task interdependence: Implications for individual and group performance. Journal of Organizational Behavior, 27(6), 723746. Liden, R. C., & Maslyn, J. M. (1998). Multidimensionality of leadermember exchange: An empirical assessment through scale development. Journal of Management, 24(1), 4372. Liden, R. C., Sparrowe, R. T., & Wayne, S. J. (1997). Leadermember exchange theory: The past and potential for the future. Research in Personnel and Human Resources Management, 15, 47119. Lorinkova, N. M., Pearsall, M. J., & Sims, H. P., Jr. (2013). Examining the differential longitudinal performance of directive versus empowering leadership in teams. Academy of Management Journal, 56, 573596.

T. Biemann et al. / The Leadership Quarterly xxx (2015) xxx – xxx 15 MacKinnon,

T. Biemann et al. / The Leadership Quarterly xxx (2015) xxxxxx

15

Nishii, L. H., & Mayer, D. M. (2009). Do inclusive leaders help to reduce turnover in diverse groups? The moderating role of leadermember in the diversity to turnover relationship. Journal of Applied Psychology, 94, 14121426. Podsakoff, P. M., MacKenzie, S. B., & Podsakoff, N. P. (2012). Sources of method bias in social science research and recommendations on how to control it. Annual Review of Psychology, 63, 539569. R Development Core Team, R (2008). R: A language and environment for statistical computing. ISBN, 3(10), Vienna Austria: R Foundation for Statistical Computing. Raudenbush, S. W., & Bryk, A. S. (2002). Hierarchical linear models: Applications and data analysis methods. Sage Publications. Roth, P. L. (1994). Missing data: A conceptual review for applied psychologists. Personnel Psychology, 47, 537560. Rubin, D. B. (1987). Multiple imputation for nonresponse in surveys. Vol. 519, Wiley Online Library. Saks, A. M. (1995). Longitudinal field investigation of the moderating and mediating effects of self-efficacy on the relationship between training and newcomer adjustment. Journal of Applied Psychology, 80(2), 211. Schafer, J. L. (2001). Multiple imputation with PAN. Schafer, J. L., & Graham, J. W. (2002). Missing data: Our view on the state of the art. Psychological Methods, 7(2), 147177. Schafer, J. L., & Yucel, R. M. (2002). Computational strategies for multivariate linear mixed-effects models with missing values. Journal of Computational and Graphical Statistics, 11(2), 437457. Scholz, U., Doña, B. G., Sud, S., & Schwarzer, R. (2002). Is general self-efficacy a universal construct? Psychometric findings from 25 countries. European Journal of Psychological Assessment, 18(3), 242. Schriesheim, C. A., Castro, S. L., & Yammarino, F. J. (2000). Investigating contingencies: An examination of the impact of span of supervision and upward controllingness on leadermember exchange using traditional and multivariate within- and between-entities analysis. Journal of Applied Psychology, 85(5), 659. Schyns, B. (2006). Are group consensus in leadermember exchange (LMX) and shared work values related to organizational outcomes? Small Group Research, 37(1),

Seibert, S. E., Silver, S. R., & Randolph, W. A. (2004). Taking empowerment to the next level: A multiple-level model of empowerment, performance, and satisfaction. Academy of Management Journal, 332349. 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(5), 981. Sherony, K. M., & Green, S. G. (2002). Coworker exchange: Relationships between coworkers, leader member exchange, and work attitudes. Journal of Applied Psychology, 87(3), 542. Sinharay, S., Stern, H. S., & Russell, D. (2001). The use of multiple imputation for the analysis of missing data. Psychological Methods, 6(4), 317. Spreitzer, G. M. (1995). Psychological empowerment in the workplace: Dimensions, measurement, and validation. Academy of Management Journal, 14421465. Spreitzer, G. M. (2008). Taking stock: A review of more than twenty years of research on empowerment at work. Handbook of organizational behavior, 5473. Srivastava, A., Bartol, K. M., & Locke, E. A. (2006). Empowering leadership in management teams: Effects on knowledge sharing, efficacy, and performance. Academy of Management Journal, 49(6), 12391251. Stewart, G. L. (2006). A meta-analytic review of relationships between team design features and team performance. Journal of Management, 32(1), 2955. Sullivan, S. E. (1999). The changing nature of careers: A review and research agenda. Journal of Management, 25(3), 457484. 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(1), 1527. Uhl-Bien, M., & Maslyn, J. M. (2003). Reciprocity in managersubordinate relationships: Components, configurations, and outcomes. Journal of Management