The Leadership Quarterly 21 (2010) 530542

Contents lists available at ScienceDirect

The Leadership Quarterly

j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / l e a q u a

Empowering leadership: An examination of mediating mechanisms within a hierarchical structure

Robert P. Vecchio a, Joseph E. Justin b, Craig L. Pearce b,
a b

University of Notre Dame, United States Claremont Graduate University, United States

a r t i c l e

i n f o

a b s t r a c t
Drawing from recent theory and research on empowerment and resistance, data on leader behaviors and follower responses were collected from superiorsubordinate dyads in 179 public high schools. Structural equation modeling revealed that empowering leadership was associated with higher employee performance and satisfaction, as well as reduced dysfunctional resistance. Also, employee dysfunctional resistance partially mediated the relationship of empowering leadership with (a) employee performance and (b) employee satisfaction. These results are interpreted as supportive of a perspective that endorses the utility of empowering leadership at the dyadic level within a hierarchical power structure. 2010 Elsevier Inc. All rights reserved.

Keywords: Leadership Empowerment Empowering leadership Performance Satisfaction Resistance

Recent decades have witnessed the rise of employee empowerment (Lawler, Mohrman, & Benson, 2001; Spreitzer, in press). This movement is based on the notion that employees who are given greater opportunities for self-direction will manifest superior outcomes, such as higher levels of job performance and job satisfaction. Along with efforts to infuse empowerment through the fundamental redesign of job attributes (i.e., by altering contextual features so as to provide for more self-pacing and independent decision making, Conger & Kanungo, 1988; Kirkman & Rosen, 1997; Spreitzer, 1996, in press; Stewart, 2006; Thomas & Velthouse, 1990), it has been argued that the redesign of leadersubordinate power-relations can also achieve workplace gains (Arnold, Arad, Rhoades, & Drasgow, 2000; Bennis & Townsend, 1997; Kozlowski & Bell, 2003; Marks, Mathieu, & Zaccaro, 2001). Within the eld of leadership, this movement is also partially manifested in such notions as SuperLeadership where superiors are encouraged to lead others to lead themselves (Manz & Sims, 1990, 1995, 2001), and shared leadership where superiors are encouraged to deliberately share or distribute responsibility among members of a workgroup (Pearce & Conger, 2003; Carson, Tesluk, & Marrone, 2007; Kozlowski & Bell, 2003). Pearce and Conger (2003, p. 1) dene shared leadership as an interactive inuence process among a set of individuals that reects a broad distribution of inuence among the group members. As such, shared leadership is arguably a matter of degree and can also be manifest in settings that are inherently hierarchical in nature. Whether shared leadership is demonstrably related to superior workplace outcomes is, however, still very much open to question, as (e.g.) there may be settings wherein shared leadership is not the optimal approach (Locke, 2003). As observed by Locke (p. 273 276), successful organizational entities typically retain some elements of hierarchical control. Related research on empowering leadership (Arnold et al., 2000; Ensley, Hmieleski, & Pearce, 2006; Pearce et al., 2003; Srivastava, Bartol, & Locke, 2006) has deliberately focused on teams and team environments. Yet, the content of scale items developed explicitly to measure empowering leadership and shared leadership are also relevant to leadersubordinate relations in more traditional hierarchical work settings (Arnold et al., 2000, pp. 268269). And, as observed by Arnold et al. (2000, p. 351), considerable conceptual overlap exists among various scales that have been developed to assess aspects of leader behavior that relate to aspects of empowerment.

Corresponding author. Present address: Institute for Innovative Leadership, University of Nebraska, Lincoln, NE 68588 United States. Tel.: +1 402 472 0291. E-mail address: craig.l.pearce@gmail.com (C.L. Pearce). 1048-9843/$ see front matter 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.leaqua.2010.03.014

R.P. Vecchio et al. / The Leadership Quarterly 21 (2010) 530542

531

1. Empowering leadership and employee outcomes As stated by Liu, Lepak, Takeuchi, and Sims (2003), empowering leadership is a style of leadership that targets employees to develop self-control and to act on their own. Empowering leadership can be viewed, therefore, as essentially an approach that offers prescriptions to leaders for arranging the distribution and exercise of power. The historical and theoretical underpinnings of this approach are manifold. For example, one can identify notions of power sharing in behavioral self-management theory (Thorenson & Mahoney, 1974; Mahoney & Arnkoff, 1978), social cognitive theory (Bandura, 1986, 1989), situational leadership theory (Hersey & Blanchard, 1969), distributed versus focused leadership (Gibb, 1954), leadermember exchange theory (Graen, 1976; Graen & Graen, 2006), the normative participation models of VroomYettonJago (Vroom & Yetton, 1973; Vroom & Jago, 1995), Likert's systems of participative management (Likert, 1961, 1967), and cognitive behavior modication research (Meichenbaum, 1977). A further theoretical vein that relates to empowering employees can be identied in the substitutes for leadership notions of Kerr and Jermier (1978), who argued that employees who subscribe to professional standards and values should have a reduced need for a supervisory gure (as such employees are essentially self-managed). More recent statements of shared leadership (cf. Avolio, Jung, Murry, & Sivasubramaniam, 1996; Carson et al., 2007; Pearce & Conger, 2003; Pearce & Sims, 2000; Pearce & Sims, 2002; Pearce, Yoo, & Alavi, 2004) and empowering leadership (Arnold et al., 2000; Ahearne, Mathieu, & Rapp, 2005; Srivastava et al., 2006) posit the value of fostering employee self-directedness. As noted by Spreitzer and Doneson (2008), these two research streams essentially complement one another. In accord with the denitions offered by Carson et al. (2007, p. 1218) and Srivastava et al. (2006, p. 1240), we presently dene empowering leadership as behaviors that share power with subordinates. The sharing of power such that self-directedness is enhanced should reasonably be expected to generate a higher level of subordinate performance. Also, greater self-directness resulting from empowerment should be associated with superior subordinate attitudinal response (i.e., higher job satisfaction). While there has been substantial prior research on various aspects of power sharing (Cotton, Vollrath, Froggatt, Lengnick-Hall, & Jennings, 1988; Ledford & Lawler, 1994; Spreitzer, 1995, 1996; Vandenberg, Richardson, & Eastman, 1999), empirical demonstrations of the unique value of utilizing leader behaviors to foster empowerment within traditional work settings has been largely the focus of more contemporary research. Often, studies of workgroups or student groups (e.g., Avolio et al., 1996; Pearce & Sims, 2002; Srivastava et al., 2006) have reported the superiority of empowerment (either as a direct or indirect effect). However, short of creating genuine team structures (where positive process gains can be shown, Kauffeld, 2006), the notion that superiors can work toward sharing power through the encouragement of self-direction with specic employees has not been as often the focus of more recent empirical research. In a pioneering study in this area, Manz and Sims (1987) provided some of the rst evidence on specic leader behaviors that were empowering in nature. However, the items in their measure of leader behavior (p. 127) asked respondents to describe the behavior of their leader toward their entire group (i.e., in accordance with an average-leadership-style approach, cf. Dansereau, Graen, & Haga, 1975; Dienesch & Liden, 1986) rather than toward each individual. Further important early research on work teams by Susan Cohen and her associates (Cohen & Ledford, 1994; Cohen, Ledford, & Spreitzer, 1996; Cohen & Bailey, 1997) has laid the foundation for examining empowering leadership within self-managed teams, as well as within more traditional workgroups. Also, good conceptual arguments for an individual-level dynamic have been offered (Conger & Kanungo, 1988; Kanter, 1989; Kirkman & Rosen, 1997, 1999; Spreitzer, 1995, 1996). The presently suggested alternative approach, that focuses on the impact of developing leadersubordinate relations at the dyadic level (rather than on the impact of introducing a team structure) is of some practical importance as many work settings continue the tradition of relying on a hierarchical power structure and are not likely to move in the relatively radical direction of creating a genuine self-managed team structure (i.e., where team leadership is rotated or elected, and job cross-training is encouraged). For example, it is important to recall that (a) most private-sector employees in North America and Great Britain work in small-business settings (Headd, 2000; National Federation of Independent Business, 2007; Royal Bank of Canada, 2005; Federation of Small Businesses, 2006) where the owner/operators are likely to be reluctant to adopt more radical forms of power sharing, and (b) employees in governmental positions are not likely to be offered opportunities to be a part of a team culture because of inherent bureaucratic controls. Therefore, the study of power sharing within leadersubordinate dyads located within a traditional hierarchical structure is of considerable practical interest and importance. More specically, the present study sought to examine empirically whether leader efforts directed toward employee empowerment within a traditional hierarchical structure are linked with individual employee performance and satisfaction. Hypothesis 1. Empowering leadership will exhibit a positive relationship with employee (a) performance and (b) job satisfaction. The present study also examined the role of employee resistance to a leader in conjunction with empowering leadership. The impact of empowerment on forms of employee resistance, be they dysfunctional or constructive in nature, represents a gap in the domain of leadership research. Employee resistance may be dened as a set of responses to a supervisor's inuence attempts that includes degrees of both cooperativeness, i.e., constructive resistance, and opposition, i.e., dysfunctional resistance (cf. Tepper, Duffy, & Shaw, 2001, p. 975). While it is initially anticipated that empowering leadership would facilitate performance and satisfaction, as well as impact both forms of employee resistance, it is not altogether a certainty that empowering leadership will have uniformly positive effects across a range of outcomes. Also, it is important to note that prior studies of employee resistance have generally focused on negative features of leaders, where resistance is viewed as a likely response to aversive action by the leader (Ashforth, 1994; Pearce & Giacalone, 2003). While the study of employee resistance to aversive leadership is of importance, it does not in itself provide especially useful insights on how to manage employee resistance in a pro-active manner (beyond merely suggesting that a leader should be less oppressive). An empowering style of leadership, where the leader seeks to develop the employee, has not been studied in connection with employee resistance. As a result, it is also one of our intended contributions to determine whether an empowering style of

532

R.P. Vecchio et al. / The Leadership Quarterly 21 (2010) 530542

leadership can help manage employee resistance. Therefore, the present study also offers an original investigation of the potential interplay of two contemporary streams of research: empowering leadership and employee resistance. Hypothesis 2. Empowering leadership will be (a) positively related to functional employee resistance and (b) inversely related to dysfunctional employee resistance. 2. Employee resistance as a mediating mechanism Leadership in a hierarchical system involves, arguably, downward inuence attempts that are intended to drive performance. While employee responses to these attempts at social inuence should ideally reect greater commitment or at least compliance, they sometimes reect resistance (Barry & Watson, 1996; Yukl, Fu, & McDonald, 2003). As noted above, Tepper et al. (2001) have dened two subdimensions of employee resistance: functional (or constructive) resistance that reects efforts to open a dialogue with the supervisor (by requesting clarication and negotiation), and dysfunctional resistance that reects deance (by acting as if one did not hear the request or acting as if one has forgotten the request). Dysfunctional employee resistance is particularly challenging for supervisors as it has the potential to disrupt work ow, create more work for the more cooperative subordinates, and takes time, energy, and attention from other issues (see Falbe & Yukl, 1992). Constructive employee resistance is presumed to be a preferred employee response as it offers the chance to maintain and improve working relations, while keeping the focus on work accomplishment. As noted earlier, empowering leadership should have particular relevance to employee resistance in that supervisors who foster greater subordinate self-direction should see less dysfunctional, and more constructive, responses from their subordinates. Supervisors who encourage self-management should have employees who are more open to dialogue and negotiation as this is the basis of social interaction for such leaders when dealing with their followers. Similarly, such leaders should have less dysfunctional resistance as the increased empowerment of employees does not relieve the employees of their responsibility to be positive contributors. As a further consequence, lower dysfunctional employee resistance and higher constructive employee resistance should be positively related to employee performance and satisfaction. Although a number of correlates of employee resistance have been identied (Tepper et al., 2001), the dependent variable of employee performance has not received substantial empirical attention within this emerging literature (Tepper et al., 2006). The predicted linkages between forms of employee resistance and such employee outcomes as performance and satisfaction, however, may only reect a portion of a larger social dynamic, one that is driven initially by power sharing. Stated another way, empowering leader behavior will likely impact employee resistance (modifying both dysfunctional and constructive responses). Resistance, in turn, will be manifested in an employee's work contributions and should, subsequently, be reected in differences in supervisor ratings or individual employee performance. We should, therefore, expect to nd that the relationship of leadership and employee performance is mediated by the nature of employee resistance. Similarly, resistance on the part of the employee should be manifested in an employee's attitudes toward the job, as indexed by self-reports of satisfaction. Moreover, prior mediational research on empowering leadership has reported partial (rather than full) mediation (Kirkman & Rosen, 1999; Srivastava et al., 2006). Hence, we may reasonably anticipate that evidence of mediation will be partial, rather than full (i.e., empowering leadership will manifest direct associations with performance and satisfaction, yet have a mediated impact on these outcomes via forms of employee resistance). Hypothesis 3. The relationship between empowering leadership and employee performance will be partially mediated by (a) functional employee resistance and (b) dysfunctional employee resistance. Hypothesis 4. The relationship between empowering leadership and employee satisfaction will be partially mediated by (a) functional employee resistance and (b) dysfunctional employee resistance. 3. Performance and satisfaction Finally, there is good evidence that employee performance and satisfaction are linked, such that job performance more typically drives job satisfaction (rather than vice versa). Porter and Lawler's (1968) model of employee behavior was perhaps the rst clear statement of how performance may inuence employee satisfaction reactions. In their model, they make a strong case for how it is likely that performance (via rewards) can more easily drive satisfaction responses, than employee satisfaction can drive performance. Although a mechanism for performance being a function of satisfaction is still feasible, this alternative causal connection between these constructs is more circuitous, and less likely to be evidenced, than as originally proposed. Given prior evidence in support of this proposed dynamic (Greene, 1973; Kopelman, 1979; Iaffaldano & Muchinsky, 1985; Petty, McGee, & Cavender, 1984; Judge, Thoreson, Bono, & Patton, 2001), the present study offers the further hypothesis that employee job satisfaction will be a function of employee job performance. Hypothesis 5. Employee satisfaction will be a consequence of employee performance. 4. Method 4.1. Sample and procedure Leadership within public high schools was selected for study, as such locations meet the criteria of offering opportunities for leaders to empower employees as well as to be abusive (Blas & Blas, 2002). As argued by Liu et al. (2003), certain settings

R.P. Vecchio et al. / The Leadership Quarterly 21 (2010) 530542

533

(especially knowledge-based and professional in nature) may be more appropriate than others for revealing the true positive potential of empowerment via leader behavior. In addition, a recent Academy of Management Journal Forum (2005) called for greater research on public schools, with particular emphasis on the potential value of decentralizing authority within a traditionally hierarchical power structure. Moreover, public high schools are representative of the types of organizations that are neither capable of nor willing to (for a variety of reasons) transform themselves into a radical form of self-managed team structure, but that could potentially benet from greater power sharing within their traditional structure. In this regard, public high schools are representative of many forms of work settings wherein power sharing is likely to be dyadic, but limited, in nature. While it is recognized that teachers do perform some tasks in groups, their role is still fairly individualistic in nature, and the potential of principals as leaders to provide a positive impact on teachers is of considerable interest in the education literature (Barnett, McCormick, & Conners, 2001; Eden, 1998; Leithwood, 1994; Leithwood & Jantzi, 1997; Moye, Henkin, & Egley, 2005). The specic method employed to conduct this research consisted of mailing questionnaires directly to high school principals and to all department heads (a.k.a. lead) teachers at those schools whose principal returned a questionnaire.1 Starting randomly, every 10th high school of the California high school system was selected for surveying (California Department of Education, 2001), recycling through the list of schools until a minimum sample of 200 principals was attained. Of 491 questionnaire surveys that were ultimately mailed to principals, 223 (45.4%) were returned. A total of 1060 condential questionnaires were mailed to the department head (lead) teachers for those principals who had condentially responded (the average high school normally has ve lead positions). Of these, a total of 342 teachers responded (32.3%). As the present analyses were at the dyadic level of interaction between a principal and a head/lead teacher, a sample of 179 principalteacher dyads from 179 distinct schools was randomly selected from dyads where complete data were available (via a sampling procedure based on a table of random numbers, Lindquist, 1940).2 The average ages for the principals and teachers in these dyads were 52.18 and 48.17 years, respectively, with a near equivalence of men and women among the teachers (51.4% male) and two-thirds of the principals being male (67.8% male). 4.2. Measures The principal's questionnaire asked for ratings of the performance of each departmental head (or lead) teacher within the ve mandated instructional areas of English, History, Mathematics, Physical Education, and Science. Although the instructional areas of the lead teachers differed, it was believed (after interviewing a sample of eight school system administrators, principals, and teachers) that the overall job content and performance was sufciently similar that a common metric could be used to assess individual teacher performance. The specic items that were completed by the principal asked for ratings of each head/lead teacher on three issues that measured overall employee performance, adapted from a scale developed by MacKenzie, Podsakoff, and Fetter (1991) (This person is one of my best teachers, All things considered, this teacher is outstanding, and All things considered, this teacher performs his/her job the way I like to see it performed, response options, 1 = Strongly Disagree, 2 = Disagree, 3 = Neither Disagree nor Agree, 4 = Agree, 5 = Strongly Agree; Cronbach alpha = 0.93). For each principal who responded to the survey, a separate personally-addressed questionnaire was subsequently mailed directly to the department head/lead respondent. The questionnaire focused on the leadership style of the principal and the work reactions of the teacher. A ten-item scale, using modied items described by Manz and Sims (1987, 1990, 1995) and previously employed in empirical research on empowering leadership by Pearce and Sims (2002), provided a measure of the construct of empowering leadership from an individualized (i.e., dyadic, Dansereau, 1995; Mumford, Dansereau, & Yammarino, 2000) perspective (see Appendix A). These items were selected because they represent such critical and relevant elements as (a) encouragement of independent employee behavior (sample item: Encourages me to nd solutions to my problems without his/ her direct input), (b) the fostering of opportunistic thinking (sample item: Urges me to think of problems as opportunities rather than obstacles), and (c) the promotion of cooperative action (sample item: Urges me to work as a team with the other teachers who work at the school). Items for the leadership measure employed a 5-point response scale (1 = Strongly Disagree, 2 = Disagree, 3 = Neither Disagree nor Agree, 4 = Agree, 5 = Strongly Agree; Cronbach alpha = 0.90).3 Questions pertaining to employee resistance were taken from published works by Tepper et al. (1998, 2001). Tepper, Eisenbach, Kirby, and Porter (1998) reported supportive evidence of the construct validity of their measures of resistance (e.g., high internal reliability, a two-factor conrmatory factor structure). In the present study, the teachers were asked to report how frequently they did not comply with what their principal had requested. This technique of self-reporting has been used successfully in prior studies of incidences of resistance (Tepper et al., 2001). As mentioned earlier, employee resistance was
1 Lead teachers act as liaisons between the principals and the rank-and-le teachers, and typically oversee a broad range of assignments related to the educational mission (e.g., curriculum development and delivery, and communication with students, parents, business partners, and other teachers). Commonly, release time and a modest stipend are provided for handling these added responsibilities. 2 One leadersubordinate dyad was also selected from the set of respondents at each institution in order to avoid potential problems associated with nonindependence of descriptions of a common leader. This sampling approach was employed because nested data structures can generate parameter estimates that are incorrect (Bliese, 2000; Bliese & Hanges, 2004; Hofmann, 1997; Raudenbush & Bryk, 2002). 3 As the items in the empowering leadership scale incorporated an individualized format (where each respondent described how the leader acted toward him/ her as an individual), it would not be expected that these responses would necessarily reect a high degree of within-school agreement. To explore this issue, average deviation (AD) values were calculated for the entire sample of respondents. Because other agreement indices, e.g. rwg, may not be optimal for evaluating inter-rater agreement with very small numbers per group and as directly interpretable, AD values provide unique advantages (see Burke & Dunlap, 2002; Burke, Finkelstein, & Dusig, 1999; Dunlap, Burke, & Smith-Crowe, 2003; Smith-Crowe & Burke, 2003). The resultant scale AD values, which are given in terms of the original item metric, were 0.57 (for AD-means) and 0.48 (for AD-medians). These values suggest an acceptable, albeit not extremely high, level of within-school agreement in the descriptions of each principal.

534

R.P. Vecchio et al. / The Leadership Quarterly 21 (2010) 530542

Table 1 Descriptive statistics and intercorrelations among variables. Variable 1. 2. 3. 4. 5. Empowering leadership Performance Job satisfaction Dysfunctional resistance Functional resistance Mean 37.94 12.59 13.44 8.71 12.25 SD 6.18 2.78 2.06 3.34 3.82 1. (0.90) 0.33** 0.28** 0.40** 0.08 2. (0.93) 0.26** 0.27** 0.03 3. 4. 5.

(0.78) 0.37** 0.04

(0.83) 0.09

(0.77)

Note: Internal reliabilities are listed on the primary diagonal. **p b 0.01.

measured on two separate scales, a 6-item scale for dysfunctional resistance (sample items: I ignore the person and do what I want to do anyway, I act like I don't know about it; Cronbach alpha = 0.83) and a 4-item scale for functional (or constructive) resistance (sample items: I explain that it should be done a different way, and I present logical reasons for doing the task differently or at a different time; Cronbach alpha = 0.77). Response options included: 1 = I cannot remember using this tactic, 2 = I very seldom use this tactic, 3 = I occasionally use this tactic, 4 = I use this tactic moderately often, and 5 = I use this tactic very often. All teachers were also asked to respond to a three-item measure of job satisfaction developed by Cammann, Fichman, Jenkins, and Klesh (1983) (sample item: All in all, I am satised with my job). The response options for this measure were: 1 = Very Inaccurate, 2 = Moderately Inaccurate, 3 = Neither Inaccurate nor Accurate, 4 = Moderately Accurate, 5 = Very Accurate (Cronbach alpha = 0.78). The factor structure of the empowering leadership scale was examined to determine whether a single factor underlies the ten items (as, e.g., Ahearne et al., 2005, used a multi-dimensional measure of empowering leadership, but found that a single underlying dimension was a more appropriate representation of the construct). For the present data, a principal axis factoring analysis (using SPSS, version 14.0) yielded a single major factor, accounting for 47.43% of the variance. Parallel analysis (Horn, 1965; Hayton, Allen, & Scarpello, 2004) and a scree plot of the eigenvalues identied only a single factor. Therefore, a single composite score was created for the construct of empowering leadership by summing over all items. Because of Tepper et al.'s (2001) contention that resistance is essentially two-dimensional, a conrmatory factor analysis was conducted to test the proposed two-factor view of resistance. Specically, the results for a single-factor model (in which all ten resistance items loaded on a single, common factor) were contrasted with the results for a two-factor model (in which the six designated dysfunctional resistance items loaded on one factor and the four functional resistance items loaded on a second factor). The contrasting results (obtained via the AMOS package, Arbuckle and Wothke, 1999; Bryne, 2001) revealed a superior t for the two-factor model versus the single-factor model (Single-factor Model: 2 = 385.88, p b 0.01, normed t index = 0.89, relative t index = 0.83, TuckerLewis index = 0.85, comparative t index = 0.90 , root mean square error of approximation = 0.24; Two-

Fig. 1. Initial hypothesized model relating empowering leadership, employee resistance, performance, and satisfaction (nine paths).

R.P. Vecchio et al. / The Leadership Quarterly 21 (2010) 530542

535

factor Model: 2 = 72.69, p b 0.01, normed t index = 0.98, relative t index = 0.97, TuckerLewis index = 0.98, comparative t index = 0.99, rmsea = 0.08; 2 = 313.19, p b 0.01). Also, the lambda parameters for the two-factor model were all signicant (p b 0.01) and ranged from 0.40 to 0.87. 5. Results Table 1 presents the descriptive statistics, intercorrelations, and internal reliabilities (Cronbach, 1951) for the variables of interest. Signicant correlations were identied for empowering leadership with leader ratings of employee performance (r = 0.33, p b 0.01), job satisfaction (r = 0.28, p b 0.01), and dysfunctional resistance (r = 0.40, p b 0.01). However, functional resistance was not related to empowering leadership (r = 0.08, n.s.). Also, ratings of employee performance were signicantly correlated with reports of job satisfaction (r = 0.26, p b 0.01). Structural equation modeling was used to test the model that is implicit in the ve hypotheses. Specically, nine pathways, or links, were identied that connected: (1) empowering leadership to satisfaction; (2) empowering leadership to performance; (3) empowering leadership to functional resistance; (4) empowering leadership to dysfunctional resistance; (5) functional resistance to satisfaction; (6) functional resistance to performance; (7) dysfunctional resistance to satisfaction; (8) dysfunctional resistance to performance; and (9) performance to satisfaction. These nine paths, which represent the full set of hypotheses, are portrayed in Fig. 1. Fig. 1, therefore, may be labeled the initial hypothesized model. Structural equation modeling is divided into two parts: the estimation of a measurement model and the estimation of a structural model. The measurement model examines the relationships between the measured variables and the latent variables, while the structural model examines the relationships among the latent variables. In the present study, the items of the respective scales were taken as indicators of each latent variable. Following a successful demonstration that the items load on their respective latent variable, the parameters of the pathways are tested for statistical signicance, and the overall structural model is evaluated

Table 2 Unstandardized and standardized parameter estimates, and signicance levels for model in Fig. 1 (N = 179). Pathways Unstandardized Empowering satisfaction Empowering functional Empowering dysfunctional Empowering performance Functional satisfaction Functional performance Dysfunctional satisfaction Dysfunctional performance Performance satisfaction Construct Empowering leadership Items Emp. 1 Emp. 2 Emp. 3 Emp. 4 Emp. 5 Emp. 6 Emp. 7 Emp. 8 Emp. 9 Emp. 10 Dys. 1 Dys. 2 Dys. 3 Dys. 4 Dys. 5 Dys. 6 Funct. 1 Funct. 2 Funct. 3 Funct. 4 Perf. 1 Perf. 2 Perf. 3 Sat. 1 Sat. 2 Sat. 3 1.000 0.489 1.008 1.183 1.194 0.638 0.952 1.162 0.615 1.354 1.000 0.455 0.937 1.197 0.456 0.584 1.000 1.887 0.107 0.584 1.000 0.727 0.842 1.000 1.039 1.517 0.163 0.238 0.264 0.252 0.202 0.228 0.277 0.193 0.293 0.141 0.164 0.211 0.145 0.151 1.006 0.600 0.151 0.099 0.087 0.239 0.303 0.075 0.060 0.164 0.196 0.025 0.016 0.103 0.166 0.109 SE 0.028 0.043 0.049 0.057 0.022 0.030 0.035 0.054 0.040 Structural model estimates Standardized 0.073 0.060 0.162 0.189 0.025 0.015 0.102 0.162 0.111 Measurement model estimates 2.998 4.239 4.484 4.734 3.159 4.176 4.191 3.186 4.616 3.216 5.722 5.666 3.142 3.873 1.875 0.178 3.873 7.333 9.672 4.343 5.010 0.003 0.001 0.001 0.001 0.002 0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.002 0.001 0.061 0.086 0.001 0.001 0.001 0.001 0.001 Critical ratio 2.641 1.417 3.349 3.466 1.160 0.532 2.980 3.073 2.696 p 0.008 0.156 0.001 0.001 0.246 0.595 0.003 0.002 0.007

Dysfunctional resistance

Functional resistance

Performance

Satisfaction

Note: 2 = 280.04, df = 223, p = 0.006; RMSEA = 0.038.

536

R.P. Vecchio et al. / The Leadership Quarterly 21 (2010) 530542

Fig. 2. Trimmed model with standardized parameter estimates: six paths (**denotes p b 0.01).

for degree of t. Finally, modications in the model may be made based on the available evidence of parameter signicance and degree of model t. The results of the test of the initial hypothesized model for the sample of 179 respondents are given in Table 2. As all of the items were associated with their respective latent variables, the parameter estimates for the pathways were then examined. Of the nine hypothesized links, six were statistically signicant. All of the three non-signicant pathways pertained to the functional employee resistance variable as either an antecedent or outcome. Because the latent variable of functional resistance did not demonstrate signicant associations in the test of the initial model, a trimmed model was then created and tested (see Fig. 2). In the trimmed model, six pathways were tested (along with the measurement model for the relevant scales items, the results of which are not displayed here but are available upon request author). The resulting parameter estimates are given in Table 3. For all six pathways, the parameter estimates were statistically signicant. Moreover, the model's t was found to be somewhat improved (as indicated by a signicant reduction in chi-square; change in chi-square of = 280.04 172.60 = 107.44, exceeding the critical chi-square of 103.01 for degrees of freedom for the difference = 223 142 = 81). Although there was a signicant reduction in chi-square, it is important to note that the trimmed model, like the initial model, also had a signicant overall chi-square (suggesting poorness-of-t, and that the initial model could possibly be improved, Bentler, 1990; Browne & Cudeck, 1993). The root mean square error of approximation for both the initial and trimmed model were acceptable (RMSEA initial = 0.038, lower bound = 0.022, upper bound = 0.051; RMSEA trimmed = 0.040, lower bound = 0.020, upper bound = 0.056). The normed t index, the incremental t index, and the TuckerLewis t index for both models similarly suggested acceptable t (NFI, IFI, and TLI, initial = 0.980, 0.996, and 0.993; trimmed = 0.985, 0.997, and 0.997). In sum, the SEM analysis indicated support for Hypothesis 1a (empowering leadership was associated with satisfaction); Hypothesis 1b (empowering leadership was associated with performance); Hypothesis 2b (empowering leadership was inversely associated with dysfunctional resistance); Hypothesis 3b (dysfunctional resistance was

Table 3 Unstandardized and standardized parameter estimates and signicance levels for trimmed model in Fig. 2 (N = 179). Pathways Unstandardized Empowering satisfaction Empowering dysfunctional Empowering performance Dysfunctional satisfaction Dysfunctional performance Performance satisfaction 0.100 0.220 0.255 0.084 0.134 0.098 SE 0.032 0.050 0.055 0.030 0.047 0.038 Structural model estimates Standardized 0.098 0.215 0.243 0.085 0.131 0.100 Critical ratio 3.120 4.374 4.632 2.794 2.821 2.573 p 0.002 0.001 0.001 0.005 0.005 0.010

Note: 2 = 172.60, df = 142, p = .012; RMSEA = .040.

R.P. Vecchio et al. / The Leadership Quarterly 21 (2010) 530542 Table 4 Unstandardized and standardized parameter estimates and signicance levels for model in Fig. 1 with hold-out sample (N = 110). Pathways Unstandardized Empowering satisfaction Empowering functional Empowering dysfunctional Empowering performance Functional satisfaction Functional performance Dysfunctional satisfaction Dysfunctional performance Performance satisfaction 0.230 0.021 0.373 0.214 0.020 0.007 0.100 0.019 0.090 SE 0.061 0.052 0.083 0.077 0.049 0.019 0.048 0.046 0.045 Structural model estimates Standardized 0.219 0.021 0.349 0.209 0.019 0.007 0.101 0.020 0.087 Critical ratio 3.786 0.407 4.495 2.785 0.408 0.375 2.093 0.419 2.005

537

p 0.001 0.684 0.001 0.005 0.683 0.708 0.036 0.675 0.045

Note: 2 = 274.30, df = 223, p = 0.011; RMSEA = 0.046.

inversely associated with performance); Hypothesis 4b (dysfunctional resistance was inversely associated with job satisfaction); and Hypothesis 5 (job performance was associated with job satisfaction). Fig. 2 portrays the pathways of the trimmed model with the obtained standardized parameters estimates. As a further check on the reliability of the present ndings, a hold-out sample was tested with the same structural equation procedures. Specically, 110 teacher-principal dyads were selected from among those not included in the initial analyses. These 110 dyads were again selected so as to be unique members of their respective institutions (i.e., each teacher was associated with a single principal). The same initial hypothesized model, consisting of nine pathways, was again tested with a SEM approach (i.e., Fig. 1). Table 4 summarizes the results from the structural model estimation (the measurement model item estimates were again all signicant and associated with their respective scales). The pattern of signicant parameter estimates essentially replicated that found for the original sample, with the notable exception that the pathway from dysfunctional resistance to performance was no longer signicant. The overall model had reasonable t, as indicated by a RMSEA of 0.046, with a lower bound of 0.023 and an upper bound of 0.063. Plus, the NFI, IFI, and TLI were 0.968, 0.994, and 0.990, respectively. The overall chi-square was 274.28, with df = 223, p = 0.011. To test whether the results of the earlier trimmed model could be replicated, the hold-out sample was tested for t with the sixpathway model presented in Fig. 3. The results for a test of the trimmed model are given in Table 5. For this analysis, ve of the six pathways were found to be signicant (with the pathway from dysfunctional resistance to performance being non-signicant in the hold-out sample, see Fig. 3 for a portrayal of the pathways and the obtained standardized estimates). The overall t of this trimmed model for the hold-out sample, however, was relatively superior. For example, the overall chi-square was no longer statistically signicant (chi-square = 167.99, df = 141, p = 0.060, indicating a good t) and the RMSEA was .042 (lower

Fig. 3. Trimmed model with standardized parameter estimates for hold-out sample: six paths (*denotes p b 0.05, **denotes p b 0.01).

538

R.P. Vecchio et al. / The Leadership Quarterly 21 (2010) 530542

Table 5 Unstandardized and standardized parameter estimates and signicance levels for trimmed model in Fig. 2 with hold-out sample (N = 110). Pathways Unstandardized Empowering satisfaction Empowering dysfunctional Empowering performance Dysfunctional satisfaction Dysfunctional performance Performance satisfaction 0.234 0.379 0.215 0.106 0.025 0.089 SE 0.062 0.084 0.078 0.050 0.046 0.045 Structural model estimates Standardized 0.233 0.355 0.211 0.108 0.026 0.223 Critical ratio 3.787 4.538 2.773 2.141 0.531 1.993 p 0.001 0.001 0.006 0.032 0.596 0.046

Note: 2 = 167.99, df = 141, p = 0.060; RMSEA = 0.042.

bound = 0.000, upper bound = 0.064). Also, the NFI, IFI, and TLI were .977, .996, and .993, respectively. Further, the reduction in chi-square from the initial model to the trimmed model was signicant (274.28 167.99 = 106.29, exceeding the critical chisquare value of 103.01 for df for the difference = 223 142 = 81). Hence, we can conclude that ve of the six pathways identied as signicant in the test of the initial hypothesized model could be replicated with a hold-out sample. 6. Discussion Locke et al. (2001) have argued that management is presently operating in an age of groupism, where the contemporary focus on leadership and motivation endorses the value of group processes (with a concomitant reduced emphasis on the importance of the individual). Similarly, Goncalo and Staw (2006) identied a widespread enthusiasm for collectivist management techniques (versus individualistic techniques). Despite the popularity of this emerging perspective, many work settings have not adopted a full-blown team structure, but instead have tried to incorporate power sharing in a manner that seeks to motivate employees to strive for higher levels of accomplishment within a hierarchical system. Interestingly, Locke (2003) has suggested that leaders should not totally abdicate their responsibility for providing vision and strategic direction, but instead should develop a two-way social inuence dynamic with their subordinates while also promoting greater subordinate-tosubordinate sharing of job knowledge, power, resources, and information (Ilgen, 1995; Mohrman, Cohen, & Mohrman, 1999).4 In essence, Locke's integrative model reects a hybrid view of leadership that incorporates both top-down and bottom-up social inuence processes, as well as lateral (team) inuence. While Locke acknowledges the potential value of a pure lateral inuence model, this completely at scheme may not truly represent leadership, per se, as it is a system composed of genuine equals. Therefore, Locke's integrative model, further explicated in a theoretical letter exchange in Leadership Quarterly with Pearce and Conger (Pearce, Conger, & Locke, 2007), more closely represents the circumstances of many modern workplaces, wherein a nominal head is seeking to incorporate principles of shared leadership while still maintaining control within an essentially hierarchical structure. The results of the present study are, as a consequence, particularly relevant and informative for these types of work settings as they suggest important benets of engaging in empowering leadership. Evaluated in their totality, the present ndings are supportive of the suggested value of empowering leadership in that leaders who encouraged greater independence from authority and greater lateral cooperation had subordinates who reported higher levels of satisfaction. This is not a trivial nding as other research on empowering leadership (using Arnold et al.'s Empowering Leadership Questionnaire) has suggested that the association of empowering leadership and employee satisfaction may actually be inverse in some settings (Robert, Probst, Martocchio, Drasgow, & Lawler, 2000). A further challenge, therefore, lies in the specication of the situational attributes that invert this association. But of perhaps equal or greater importance is the present nding that the performance of employees who experienced empowering leadership was also higher. Furthermore, employees who experienced empowering leadership also reported engaging in less dysfunctional resistance. However, constructive (or functional) resistance was not found to be associated with empowering leadership. Perhaps it is too simplistic to equate constructive employee resistance with a positive, or supportive, demeanor. In the eyes of some leaders, constructive resistance is still basically resistance. Also, for some leaders, constructive employee resistance may be viewed as a cover for a deeper desire to not be cooperative (for either selsh reasons on the part of the employee or because of a masked desire to undermine the leader). For some leaders, resistance of all types is seen as a potential source of problems, relative to the possibly preferred response of simple compliance. Further evidence that leaders may view all forms of resistance in negative terms is available in the present nding that leader ratings of employee performance were inversely associated with subordinate reports of dysfunctional resistance. The near-zero association of employee performance with constructive resistance suggests that some leaders may not be sure how to read their subordinates behaviors (i.e., they may be suspicious as to the underlying motives of functional resistance). This raises an interesting question for future research: Does a leader's interpretation of a subordinate's possible disloyalty and lack of motivation inuence whether the leader regards a constructive resistance response in positive or negative terms (i.e., perceived disloyalty or perceived low motivation by a subordinate may moderate the association of employee constructive resistance and rated
4 Intriguingly, recent research by Goncalo and Staw (2006) also suggests further limits to a collectivist approach in that creativity has been found to be greater in individualistically-oriented groups.

R.P. Vecchio et al. / The Leadership Quarterly 21 (2010) 530542

539

performance). From a practical standpoint, the present ndings suggest that subordinates need to be cautious when using constructive resistance, so as to avoid having their actions interpreted as being driven by disloyalty to the leader or lack of motivation to perform. Further, the present ndings, in conjunction with the ndings in the literature on task-based conict, suggest that the relationship between constructive forms of disagreement and performance is highly complex (cf. De Dreu & Weingart, 2003). The results of the present analyses also offer some interesting insights in that the advocates of greater levels of empowering leadership have sought to specify the mechanisms by which power sharing might impact employees. The present results are of special interest because they also highlight dysfunctional social dynamics (a side of the social equation that is typically neglected by research that focuses on the hoped-for positive dynamics of empowering leadership prompting social reciprocity and selfefcacy motives). From the present analysis, it appears that both performance and satisfaction may be impacted by empowering leadership through the potential linkage of reduced dysfunctional resistance. Satisfaction may also be inuenced by empowering leadership through the pathway of increased performance. Empowering leadership may help to overcome dysfunctional resistance because it places greater responsibility on the employee and raises an employee's sense of self-worth through offering greater personal and professional challenges. 6.1. Limitations and further questions Because this study was essentially cross-sectional in nature and SEM-based results do not justify claims of causal direction, one cannot easily invoke causal dynamics (Podsakoff, MacKenzie, Lee, & Podsakoff, 2003; Podsakoff & Organ, 1986; Spector, 1987, 2006). However, the independence of the sources of the key variables (i.e., the obtainment of employee performance ratings from superiors, the obtainment of descriptions of leader behavior from subordinates, and the self-reporting by employees of their resistance behaviors) is a positive feature of this study's design. Also, it seems unlikely that employee reports of their leader's behavior drove leader evaluations of employee performance (although it is, of course, plausible that prior performance evaluations could inuence employee reports of their leader's behavior).5 Similarly, the identied linkage between empowering leadership and dysfunctional employee resistance offers a perspective for further insights. While the demonstration of mediation of these relationships is valuable, we do not have a complete understanding of how empowering leadership might be employed to overcome resistance, or what precise leader behaviors are more likely to induce resistance. Does social reciprocity underlie the relationship between empowering leadership and lower dysfunctional resistance, while an attempt to re-establish one's sense of social status underlies the relationship between employee performance and greater dysfunctional resistance? Perhaps multiple perspectives for assessing employee resistance, as well as replications in the for-prot sector, may further illuminate our understanding of employee response to leader behaviors. Certainly, questions remain as to the generalizability of the observed effects of empowering leadership for other organizational settings. In addition, it would be valuable to see future research on empowering leadership that spanned a fuller range of employee responses (from greater commitment, which is presumed the target of empowering leadership, to mere compliance, to forms of resistance). A further topic for future research pertains to determining whether empowering leadership might be especially valuable in reducing rivalry among peers. Competition among coworkers in the workplace is often lamented as placing a brake on unit performance and limiting unit morale. Empowering leadership, through the leader's active emphasis on mutual support among unit members, may help to boost member contributions by spotlighting such forms of subtle resistance as the withholding of both information and effort. Therefore, empowering leadership may be shown to have its benecial impact through discouraging and eliminating certain types of employee behavior that are tied to competitiveness among coworkers. One benecial side-effect of this social dynamic is that we should expect to nd less employee envy (i.e., coworker resentment) in workgroups that have leaders who engage in empowering leadership (Vecchio, 1995, 2007). Employee envy should be reduced as a result of a leader's emphasis on coworker collaboration and cooperation. By having subordinates focus on superordinate goals and mutual support, leaders should nd less in-ghting among their subordinates. It is important to note that the present study employed a measure of empowering leadership that was derived from the Manz and Sims (1995) and Pearce and Sims (2002) stream of leadership research. Future research may be better informed by utilizing a wider range of alternative measures of empowering leadership. For example, the Arnold et al. (2000) Empowering Leadership Questionnaire (ELQ) offers a multi-dimensional measure of the empowering leadership construct. The factorial structure of the ELQ could offer additional insights on the role of leader empowerment. Research that seeks to contrast and integrate alternative indices of empowering leadership would be a welcome addition to this literature. 6.2. Summary A common theme contained in recent efforts to connect empowerment and leadership is that key leader behaviors can enhance an employee's job satisfaction (Arnold et al., 2000; Aryee & Chen, 2006; Carless, 2004; Koberg, Boss, Senjem, & Goodman, 1999; Kirkman & Rosen, 1999) and job performance (Kirkman & Rosen, 1999; Koberg et al., 1999; Spreitzer, 1995). However, the
5 It should be noted that although a pilot sample of administrators, principals, and teachers endorsed the performance scale items as relevant to the performance appraisal task (i.e., the job descriptions for the lead teachers did not differ so greatly that a common metric could not be used), the present performance results may be limited by additional uncontrolled variance. As a consequence, it should have been more difcult to identify signicant associations (as the present statistical tests were, therefore, relatively conservative or less likely to detect effects).

540

R.P. Vecchio et al. / The Leadership Quarterly 21 (2010) 530542

mediating mechanisms by which empowering leadership may be linked to employee attitudes and performance are not well specied. The present study, while building on the original work of Tepper et al. (2001) on employee resistance, incorporated the dependent variables of employee performance and satisfaction, and examined empowering leadership in an attempt to more fully explain the possible connections among leader behavior, employee resistance, and employee performance and satisfaction. Of special note, empowering leadership was found to be associated (inversely) with dysfunctional (but not associated with functional) employee resistance. Moreover, it was found that the linkage between both performance and satisfaction with empowering leadership involves the mediation of employee dysfunctional resistance. As noted by Stewart (2006), empowering leadership has been identied as a factor that may help to improve workgroup dynamics. Because we expect that leaders may vary the extent to which they empower individual subordinates, the present results further show that empowering leadership can also be examined at the dyadic level within a traditional, hierarchical structure. Appendix A Empowering leadership Independent action 1. 2. 3. 4. Encourages me to nd solutions to my problems without his/her direct input. Urges me to assume responsibilities on my own. Advises me to solve problems when they pop up without always getting a stamp of approval. Encourages me to search for solutions without supervision. Opportunistic thinking 5. Urges me to think of problems as opportunities rather than obstacles. 6. Advises me to look for the opportunities in the problems I face. 7. Encourages me to view unsuccessful performance as a chance to learn. Cooperative action 8. Urges me to work as a team with the other teachers who work at the school. 9. Encourages me to work together with other teachers who work at the school. 10. Advises me to coordinate my efforts with the other teachers who work at the school. References
Academy of Management Journal Public Affairs Forum (2005). Making public schools work: Management reform is the key. Academy of Management Journal, 48, 929940. Ahearne, J. M., Mathieu, J. E., & Rapp, A. (2005). To empower or not to empower your sales force? An empirical examination of inuence of leadership empowerment behavior on customer satisfaction and performance. Journal of Applied Psychology, 90, 945955. Arbuckle, J. L., & Wothke, W. (1999). AMOS 4.0 user's guide. Chicago, IL: SPSS Inc. Arnold, J., Arad, S., Rhoades, J., & Drasgow, F. (2000). The empowering leadership questionnaire: The construction and validation of a new scale for measuring leader behaviors. Journal of Organizational Behavior, 21, 249269. Aryee, S., & Chen, Z. X. (2006). Leader-member exchange in a Chinese context: Antecedents, the mediating role of psychological empowerment and outcomes. Journal of Business Research, 59, 793801. Ashforth, B. (1994). Petty tyranny in organizations. Human Relations, 47, 755777. Avolio, B. J., Jung, D. I., Murry, W., & Sivasubramaniam, N. (1996). Building highly developed teams: Focusing on shared leadership processes, efcacy, trust, and performance. Advances in Interdisciplinary Studies of Work Teams, 3, 173209. Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory. Englewood Cliffs, NJ: Prentice-Hall. Bandura, A. (1989). Human agency in social cognitive theory. American Psychologist, 44, 11751184. Barnett, K., McCormick, J., & Conners, R. (2001). Transformational leadership in schoolsPanacea, placebo or problem? Journal of Educational Administration, 39, 2446. Barry, B., & Watson, M. R. (1996). Communication aspects of dyadic social inuence in organizations: A review and integration of conceptual and empirical developments. In B. R. Burleson & A. W. Kunkel (Eds.), Communication yearbook, 19 (pp. 269317). Thousand Oaks, California: Sage. Bennis, W. G., & Townsend, R. (1997). Reinventing leadership: Strategies to empower the organization. New York, NY: Morrow/Avon. Bentler, P. (1990). Comparative t indexes in structural models. Psychological Bulletin, 107, 238246. Blas, J., & Blas, J. (2002). The dark side of leadership: Teacher perspectives of principal mistreatment. Educational Administration Quarterly, 38, 671727. Bliese, P. D. (2000). Multilevel random coefcient modeling in organizational research. In F. Drasgow & N. Schmitt (Eds.), Measuring and analyzing behavior in organizations: Advances in measurement and data analysis (pp. 401445). San Francisco, CA: Jossey-Bass. Bliese, P. D., & Hanges, P. J. (2004). Being too liberal and too conservative: The perils of treating grouped data as though they were independent. Organizational Research Methods, 7, 400417. Browne, M. W., & Cudeck, R. (1993). Alternative ways of assessing model t. In K. A. Bollen & J. S. Long (Eds.), Testing structural equation models (pp. 136162). Beverly Hills, CA: Sage. Bryne, B. M. (2001). Structural equation modeling with AMOS: Basic concepts, applications, and programming. Mahwah, NJ: Lawrence Erlbaum. Burke, M. J., & Dunlap, W. P. (2002). Estimating interrater agreement with the Average Deviation Index: A user's guide. Organizational Research Methods, 5, 159172. Burke, M. J., Finkelstein, L. M., & Dusig, M. S. (1999). On average deviation indices for estimating inter-rater agreement. Organizational Research Methods, 2, 4968. California Department of Education (2001). California public school directory 2001. Sacramento: CDE Press. Cammann, C., Fichman, M., Jenkins, D., & Klesh, J. (1983). Assessing the attitudes and perceptions of organization members. In S. E. Seashore, E. E. Lawler, P. Mirvis, & C. Cammann (Eds.), Assessing organizational change (pp. 71138). New York, NY: Wiley. Carless, S. A. (2004). Does psychological empowerment mediate the relationship between psychological climate and job satisfaction? Journal of Business and Psychology, 18, 405425.

R.P. Vecchio et al. / The Leadership Quarterly 21 (2010) 530542

541

Carson, J. B., Tesluk, P. E., & Marrone, J. A. (2007). Shared leadership in teams: An investigation of antecedent conditions and performance. Academy of Management Journal, 50, 12171234. Cohen, S. G., & Bailey, D. E. (1997). What makes teams work: Group effectiveness research from the shop oor to the executive suite. Journal of Management, 23, 239290. Cohen, S. G., & Ledford, G. E. (1994). The effectiveness of self-managing teams: A quasiexperiment. Human Relations, 47, 1343. Cohen, S. G., Ledford, G. E., & Spreitzer, G. M. (1996). A predictive model of self-managing work team effectiveness. Human Relations, 49, 643676. Conger, J. A., & Kanungo, R. N. (1988). The empowerment process: Integrating theory and practice. Academy of Management Review, 13, 471482. Cotton, J., Vollrath, D. A., Froggatt, K. I., Lengnick-Hall, M. L., & Jennings, K. R. (1988). Employee participation: Diverse forms and different outcomes. Academy of Management Review, 13, 822. Cronbach, L. J. (1951). Coefcient alpha and the internal structure of tests. Psychometrika, 16, 297334. Dansereau, F. (1995). A dyadic approach to leadership: Creating and nurturing this approach under re. Leadership Quarterly, 6, 479490. Dansereau, F., Graen, G. B., & Haga, W. J. (1975). A vertical dyad linkage approach to leadership within formal organizations: A longitudinal investigation of the role-making process. Organizational Behavior and Human Performance, 13, 4678. De Dreu, C. K. W., & Weingart, L. R. (2003). Task versus relationship conict, team performance, and team member satisfaction: A meta-analysis. Journal of Applied Psychology, 88, 741749. Dienesch, R. M., & Liden, R. C. (1986). Leader-member exchange model of leadership: A critique and further development. Academy of Management Review, 11, 618634. Dunlap, W. P., Burke, M. J., & Smith-Crowe, K. (2003). Accurate tests of statistical signicance for Rwg and average deviation interrater agreement indexes. Journal of Applied Psychology, 88, 356362. Eden, D. (1998). The paradox of school leadership. Journal of Educational Administration, 36, 249261. Ensley, M. D., Hmieleski, K. M., & Pearce, C. L. (2006). The importance of vertical and shared leadership within new venture top management teams: Implications for the performance of startups. Leadership Quarterly, 17, 217231. Falbe, C. M., & Yukl, G. (1992). Consequences for managers of using single inuence tactics and combinations of tactics. Academy of Management Journal, 35, 638653. Federation of Small Businesses (2006). Small Business Stats. Available online http://www.fsb.org.uk/data/default.asp?id=64&loc=policy Gibb, C. A. (1954). Leadership. In G. Lindzey (Ed.), Handbook of social psychology (pp. 877917). Reading, MA: Addeson-Wesley. Goncalo, S. A., & Staw, B. M. (2006). Individualismcollectivism and group creativity. Organizational Behavior and Human Decision Processes, 100, 96109. Graen, G. B. (1976). Role making processes within complex organizations. In M. D. Dunnette (Ed.), Handbook of industrial and organizational psychology (pp. 12011245). Chicago, IL: Rand-McNally. Graen, G. B., & Graen, J. A. (2006). Sharing network leadership. Greenwich, CT: Information Age Publishing. Greene, C. N. (1973). Causal connections among managers merit pay, job-satisfaction, and performance. Journal of Applied Psychology, 58, 95100. Hayton, J. C., Allen, D. G., & Scarpello, V. (2004). Factor retention decisions in exploratory factor analysis: A tutorial on parallel analysis. Organizational Research Methods, 7, 191205. Headd, B. (2000). The characteristics of small-business employees. Monthly Labor Review, 123, 1318. Hersey, P., & Blanchard, K. H. (1969). Life cycle theory of leadership. Training & Development Journal, 23, 2634. Hofmann, D. A. (1997). An overview of the logic and rationale of hierarchical linear models: Implications for research in organizations. Journal of Management, 24, 623641. Horn, J. L. (1965). A rationale and test for the number of factors in factor analysis. Psychometrika, 32, 179185. Iaffaldano, M. T., & Muchinsky, P. M. (1985). Job-satisfaction and job-performance: A meta-analysis. Psychological Bulletin, 97, 251273. Ilgen, D. (1995). Teams embedded in organizations: Some implications. American Psychologist, 54, 129139. Judge, T. A., Thoreson, C. J., Bono, J. E., & Patton, G. K. (2001). The job satisfaction-job performance relationship: A qualitative and quantitative review. Psychological Bulletin, 127, 376407. Kanter, R. M. (1989). The new managerial work. Harvard Business Review, 67, 8592. Kauffeld, S. (2006). Self-directed work groups and team competence. Journal of Occupational and Organizational Psychology, 79, 121. Kerr, S., & Jermier, J. M. (1978). Substitutes for leadership: Their meaning and measurement. Organizational Behavior and Human Performance, 22, 375403. Kirkman, B. L., & Rosen, B. (1997). A model of work team empowerment. In R. W. Woodman & W. A. Pasmore (Eds.), Research in organizational change and development (pp. 131167). Greenwich, CT: JAI Press. Kirkman, B. L., & Rosen, B. (1999). Beyond self-management: Antecedents and consequences of team empowerment. Academy of Management Journal, 42, 5874. Koberg, C. S., Boss, W., Senjem, J. C., & Goodman, E. A. (1999). Antecedents and outcomes of empowerment: Empirical evidence from the health care industry. Group and Organization Management, 34, 7191. Kopelman, R. E. (1979). A causal-correlational test of the Porter and Lawler Framework. Human Relations, 32, 545556. Kozlowski, S. W. J., & Bell, B. S. (2003). Work groups and teams in organizations. In W. C. Borman & D. R. Ilgen (Eds.), Comprehensive handbook of psychology: Industrial and organizational psychology (pp. 333375). New York, NY: Wiley. Lawler, E. E., Mohrman, S. A., & Benson, G. (2001). Organizing for high performance: Employment involvement, TQM, reengineering, and knowledge management in the Fortune 1000. San Francisco: Jossey-Bass. Ledford, G. E., & Lawler, E. E. (1994). Research on employee participation: Beating a dead horse. Academy of Management Review, 19, 633636. Leithwood, K. (1994). Leadership for restructuring. Educational Administration Quarterly, 30, 498518. Leithwood, K., & Jantzi, D. (1997). Explaining variation in teachers perceptions of principals leadership: A replication. Journal of Educational Administration, 35, 312331. Likert, R. (1961). New patterns of management. New York, NY: McGraw-Hill. Likert, R. (1967). The human organization: Its management and value. New York, NY: McGraw-Hill. Lindquist, E. F. (1940). Statistical analysis in educational research. Boston, MA: Houghton-Mifin. Liu, W., Lepak, D. P., Takeuchi, R., & Sims, H. P. (2003). Matching leadership styles with employment modes: Strategic HRM perspective. Human Management Resource Review, 13, 127152. Locke, E. A. (2003). Leadership: Starting at the top. In C. L. Pearce & J. A. Conger (Eds.), Sharing leadership: Reframing the hows and whys of leadership (pp. 271284). Thousand Oaks, CA: Sage. Locke, E. A., Tirauer, D., Roberson, Q., Goldman, B., Latham, M., & Welden, E. (2001). The importance of the individual in an age of groupism. In M. Turner (Ed.), Groups at work: Theory and research. Mahwah, NJ: Lawrence Erlbaum. MacKenzie, S. B., Podsakoff, P. M., & Fetter, R. (1991). Organizational citizenship behavior and objective productivity as determinants of managerial evaluations of salespersons performance. Organizational Behavior and Human Decision Processes, 50, 128. Mahoney, M., & Arnkoff, D. (1978). Cognitive and self-control therapies. In S. Gareld & A. Bergin (Eds.), Handbook of psychotherapy and behavioral change. New York, NY: John Wiley & Sons. Manz, C. C., & Sims, H. P. (1987). Leading workers to lead themselves: The external leadership of self-managing work teams. Administrative Science Quarterly, 32, 106128. Manz, C. C., & Sims, H. P. (1990). Superleadership. New York, NY: Berkeley Books. Manz, C. C., & Sims, H. P. (1995). Superleadership: Beyond the myth of heroic leadership. In J. T. Wren (Ed.), The leader's companion (pp. 212221). New York, NY: Free Press. Manz, C. C., & Sims, H. P. (2001). The new superleadership: Leading others to lead themselves. San Francisco, CA: Berrett-Koehler. Marks, M. A., Mathieu, J. E., & Zaccaro, S. J. (2001). A temporally based framework and taxonomy of team processes. Academy of Management Review, 26, 356376. Meichenbaum, D. (1977). Cognitive-behavior modication: An integrative approach. New York, NY: Plenum Press. Mohrman, S., Cohen, S., & Mohrman, A. (1999). Designing team-based organizations: New forms of knowledge work. San Francisco, CA: Jossey-Bass.

542

R.P. Vecchio et al. / The Leadership Quarterly 21 (2010) 530542

Moye, M. J., Henkin, A. B., & Egley, R. J. (2005). Teacherprincipal relationships: Exploring linkages between empowerment and interpersonal trust. Journal of Educational Administration, 43, 260277. Mumford, M. D., Dansereau, F., & Yammarino, F. J. (2000). Followers, motivations, and levels of analysis on the case of individualized leadership. Leadership Quarterly, 11, 313340. National Federation of Independent Business (2007). Small business facts. Available online www.nb.com/object/smallBusinessFacts Pearce, C. L., & Conger, J. A. (2003). Sharing leadership: the hows and whys of leadership. Thousand Oaks, CA: Sage. Pearce, C. L., Conger, J. A., & Locke, E. A. (2007). Shared leadership theory. Leadership Quarterly, 18, 281288. Pearce, C. L., & Giacalone, R. A. (2003). Teams behaving badly: Counterproductive behavior at the team level of analysis. Journal of Applied Social Psychology, 33, 5875. Pearce, C. L., & Sims, H. P. (2000). Shared leadership: Toward a multi-level theory of leadership. In M. M. Beyerlein, D. A. Johnson, & S. T. Beyerlein (Eds.), Advances in interdisciplinary studies of work teams (pp. 115139). Amsterdam: JAI. Pearce, C. L., & Sims, H. P. (2002). Vertical versus shared leadership as predictors of the effectiveness of change management teams: An examination of aversive, directive, transactional, transformational, and empowering leader behaviors. Group Dynamics: Theory, Research, and Practice, 6, 172197. Pearce, C. L., Sims, H. P., Cox, J. F., Ball, G., Schnell, E., Smith, K. A., et al. (2003). Transactors, transformers and beyond: A multi-method development of a theoretical typology of leadership. Journal of Management Development, 22(4), 273307. Pearce, C. L., Yoo, Y., & Alavi, M. (2004). Leadership, social work and virtual teams: The relative inuence of vertical vs. shared leadership in the nonprot sector. In R. E. Riggio & S. Smith-Orr (Eds.), Improving leadership in nonprot organizations (pp. 180203). San Francisco: Jossey Bass. Petty, M. M., McGee, G. W., & Cavender, J. W. (1984). A meta-analysis of the relationships between individual job-satisfaction and individual-performance. Academy of Management Review, 9, 712721. Podsakoff, P., & Organ, D. W. (1986). Self reports in organization research: Problems and prospects. Journal of Management, 12, 531544. Podsakoff, P. M., MacKenzie, S. B., Lee, J. Y., & Podsakoff, N. P. (2003). Common method biases in behavioral research: A critical review of the literature and recommended remedies. Journal of Applied Psychology, 88, 879903. Porter, L. W., & Lawler, E. E. (1968). Managerial attitudes and performance. Homewood, IL: Irwin-Dorsey Press. Raudenbush, S. W., & Bryk, A. S. (2002). Hierarchical linear models: Applications and data analysis methods, 2nd ed. Thousand Oaks, CA: Sage. Robert, C., Probst, J. M., Martocchio, J. J., Drasgow, F., & Lawler, J. J. (2000). Empowerment and continuous improvement in the United States, Mexico, Poland, and India: Predicting t on the basis of the dimensions of power distance and individualism. Journal of Applied Psychology, 85, 643658. Royal Bank of Canada (2005). Current and future small business owners provide insights. Available online www.rbc.com/newsroom/20051003smallbus.html Smith-Crowe, K., & Burke, M. J. (2003). Interpreting the statistical signicance of observed AD interrater agreement values: Correction to Burke and Dunlap (2002). Organizational Research Methods, 6, 129131. Spector, G. (1987). Method variance as an artifact in self-reported affect and perceptions at work: Myth or signicant problem? Journal of Applied Psychology, 72, 438443. Spector, G. (2006). Method variance in organizational research: Myth or urban legend? Organizational Research Methods, 9, 221232. Spreitzer, G. (1995). Psychological empowerment in the workplace: Dimensions, measurement, and validation. Academy of Management Journal, 38, 14421465. Spreitzer, G. (1996). Social structural characteristics of psychological empowerment. Academy of Management Journal, 39, 483504. Spreitzer, G. (2008). Taking stock: A review of more than twenty years of research on empowerment. In C. Cooper, & J. Barling (Eds.), Handbook of Organizational Behavior (pp. 5472). Thousand Oaks, CA: Sage. Spreitzer, G., & Doneson, D. (2008). Musings on the past and future of employee empowerment. In T. G. Cummings (Ed.), Handbook of organizational development (pp. 311324). Thousand Oaks, CA: Sage. Srivastava, A., Bartol, K. M., & Locke, E. A. (2006). Empowering leadership in management teams: Effects on knowledge sharing, efcacy, and performance. Academy of Management Journal, 49, 12391251. Stewart, G. L. (2006). A meta-analytic review of relationships between team design features and team performance. Journal of Management, 32, 2954. Tepper, B. J., Duffy, M. K., & Shaw, J. D. (2001). Personality moderators of the relationship between abusive supervision and subordinates' resistance. Journal of Applied Psychology, 86, 974983. Tepper, B. J., Eisenbach, R. J., Kirby, S. L., & Porter, P. W. (1998). Test of a justice-based model of subordinates resistance to downward inuence attempts. Group and Organization Management, 23, 144160. Tepper, B. J., Uhl-Bien, M., Kohut, G. F., Rogelberg, S. G., Lockhart, D. E., & Ensley, M. D. (2006). Subordinates resistance and managers evaluations of subordinates performance. Journal of Management, 32, 185209. Thomas, K. W., & Velthouse, B. A. (1990). Cognitive elements of empowerment: An interpretive model of intrinsic task motivation. Academy of Management Review, 15, 666681. Thorenson, E. E., & Mahoney, M. J. (1974). Behavioral self-control. Austin, TX: Holt, Rinehart, & Winston. Vandenberg, R. J., Richardson, H. A., & Eastman, L. J. (1999). The impact of high involvement work processes upon organizational effectiveness: A 2nd-order latent variable approach. Group and Organizational Management, 24, 300339. Vecchio, R. P. (1995). It's not easy being green: Jealousy and envy in the workplace. In G. R. Ferris (Ed.), Research in personnel and human resources management (pp. 201244). Greenwich, CT: JAI Press. Vecchio, R. P. (2007). Cinderella and Salieri in the workplace: The envied and the envier. In S. W. Gilliland, D. W. Steiner, & D. P. Skarlicki (Eds.), Research in social issues in management (pp. 109134). Greenwich, CT: Information Age. Vroom, V. H., & Jago, A. G. (1995). Situation effects and levels of analysis in the study of leader participation. Leadership Quarterly, 6, 169181. Vroom, V. H., & Yetton, P. W. (1973). Leadership and decision making. Pittsburgh, PA: University of Pittsburgh Press. Yukl, G., Fu, P. P., & McDonald, R. (2003). Cross-cultural differences in perceived effectiveness of inuence tactics for initiating or resisting change. Applied Psychology: An International Review, 52, 6882.