Contingent reward transactional leadership, work attitudes, and organizational citizenship behavior: The role of procedural justice climate perceptions and strength

The Leadership Quarterly 19 (2008) 251–265

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

Contingent reward transactional leadership, work attitudes, and organizational citizenship behavior: The role of procedural justice climate perceptions and strength Fred O. Walumbwa a,⁎, Cindy Wu b,1, Bani Orwa c a b c

The Arizona State University, 4701 West Thunderbird Road, Phoenix, AZ 85069-7100, USA Baylor University, One Bear Place #98006, Waco, TX 76798, USA Kenyatta University, P.O. Box 43844-00100, Nairobi, Kenya

a r t i c l e

i n f o

Keywords: Contingent reward Leadership Organizational justice Climate Satisfaction Commitment Organizational citizenship behavior Level of analysis

a b s t r a c t Using a sample of 212 bank employees, we developed a model in which procedural justice climate perceptions and strength mediated the relationships between contingent reward leader behavior and follower satisfaction with supervisor, organizational commitment, and rated organizational citizenship behavior, controlling for perceived supervisor support. Results from the HLM analysis showed that procedural justice climate perceptions and strength completely mediated the relationships between contingent reward leader behavior and followers' satisfaction with supervisor and levels of organizational commitment, and partially mediated the relationship between contingent reward leader behavior and supervisor rated organizational citizenship behavior. Implications for research and practice of our findings are discussed. © 2008 Elsevier Inc. All rights reserved.

Contingent reward transactional leader behavior has been a focus of considerable organizational research over the last two decades (see Avolio, Bass, Walumbwa, & Zhu, 2004 for a review). Over these years, the majority of empirical and meta-analytic studies have demonstrated that contingent reward leader behavior has a positive effect on followers' attitudes and behaviors (Atwater, Camobreco, Dionne, Avolio, & Lau, 1997; Bass, Avolio, Jung, & Berson, 2003; Howell & Hall-Merenda, 1999; Lowe, Kroeck, & Sivasubramaniam, 1996; Podsakoff, MacKenzie, Paine, & Bachrach, 2000; Walumbwa, Wu, & Ojode, 2004), although negative associations have also been reported (Howell & Avolio, 1993; Yammarino & Bass, 1990). Two recent meta-analytic reviews reached consistent and positive conclusions regarding the significant unique effect of leader contingent reward behavior on a variety of important employee attitudes, perceptions, and behaviors (Judge & Piccolo, 2004; Podsakoff, Bommer, Podsakoff, & MacKenzie, 2006). Although previous studies have established a direct link between contingent reward leader behavior and several relevant work attitudes and behaviors, “surprisingly little attention has been given to articulating the reasons why leader reward behavior influences employee criterion variables” (Podsakoff et al., 2006, p. 115). Secondly, although there have been calls for research that uses a multilevel framework when investigating leadership (Graen & Uhl-Bien, 1995; Judge, Piccolo, & Ilies, 2004), relatively few studies have examined the effect of contingent reward leader behavior as a group-level variable on important work attitudes and behaviors. As multilevel researchers have repeatedly noted, a construct studied at the individual level of analysis is rarely equivalent to the same construct conceptualized at higher levels of analysis (Bliese, 2000; Morgeson & Hoffman, 1999; Yammarino, Dionne, Chun, & Dansereau, 2005). Therefore, the effect of contingent reward leader behavior at the individual level cannot be assumed to translate directly to the unit level of analysis; research must be conducted to test these relationships.

⁎ Corresponding author. Tel.: +1 602 543 6240; fax: +1 602 543 6221. E-mail addresses: [email protected] (F.O. Walumbwa), [email protected] (C. Wu), [email protected] (B. Orwa). 1 Tel.: +1 254 710 7672; fax: +1 254 710 1093. 1048-9843/$ – see front matter © 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.leaqua.2008.03.004

252

F.O. Walumbwa et al. / The Leadership Quarterly 19 (2008) 251–265

This study is an attempt to address this gap in the literature by developing and testing a framework in which group-level procedural justice climate perceptions (i.e., shared perceptions of procedural justice among individuals within a unit) and strength (within-unit variability in employees' perceptions of justice climate or the dispersion of employees' justice climate rating) mediate the relationship between group-level contingent reward leader behavior and individual followers' satisfaction with supervisor, organizational commitment, and rated organizational citizenship behavior (OCB), using justice judgment model (Leventhal, 1980) as a theoretical framework. Procedural justice climate—a group-level cognition about how a work group or unit as a whole is treated, has received increasing attention in recent years (Colquitt, Zapata-Phelan, & Roberson, 2005; Ehrhart, 2004; Liao & Rupp, 2005; Mossholder, Bennett, & Martin, 1998; Naumann & Bennett, 2000; Rupp & Cropanzano, 2002). As explained by Rupp, Bashshur, & Liao (2007), interest in general justice climate has been spurred by a number of events, including an increase in the use of team-based work systems within organizations, increase in research in multilevel issues as a method for understanding complex organizational phenomena, and the contemporary theoretical models of workplace justice that shows justice concerns are not always selfinterested, but that people also care about the treatment of others, and that third-party justice effects are far more common than was once thought. Although research has found positive relationships between justice climate and several relevant organizational work outcomes, there is still need to understand potential antecedents, such as leader behaviors (Cropanzano, Byrne, Bobocel, & Rupp, 2001; Ehrhart, 2004). Schneider, Gunnarson, & Niles-Jolly (1994) argued that climate perceptions are based on leaders' behavior and the actions they reward, because leaders are the “climate engineers” (Naumann & Bennett, 2000, p. 883). In particular, researchers have often suggested that justice climate and strength can arise from mere exposure to the same policies or leaders (Naumann & Bennett, 2000; Schneider & Reichers, 1983; Schneider, Salvaggio, & Subirats, 2003). However, as De Cremer, van Knippenberg, van Knippenberg, Mullenders, & Stinglhamber (2005) noted, “hardly any empirical research has looked at which well defined leadership behaviors may act in tandem with procedural fairness” (p. 3) or serve to create strong justice climate perceptions. Colquitt, Noe, & Jackson (2002) called for research that examines whether “certain management practices or reward systems may serve to create strong and favorable justice climates” (p. 104). The second goal of this study, therefore, is to address this gap in the literature by integrating leadership and organizational justice literatures, examining whether contingent reward leader behavior, as a group-level variable, serve as a potential antecedent to group-level procedural justice climate perceptions and strength. Specifically, we argue that because both contingent reward leader behavior and justice are theoretically rooted in expectancy and social exchange theories (Blau, 1964; Gouldner, 1960), justice climate perceptions and strength may act as powerful mediators of the influence of contingent reward leader behavior on individual follower work-related attitudes and behaviors such as satisfaction with supervisor, organizational commitment, and rated OCB. In sum, this study offers a unique integration of two of the visible literatures in organizational behavior: leadership and organizational justice (Pillai, Schriesheim, & Williams, 1999; Scandura, 1999) by theorizing and testing a model in which grouplevel procedural justice climate perceptions and strength mediate the effect of group-level contingent reward leader behavior on follower satisfaction with supervisor, organizational commitment, and rated OCB. In the same vein, we also tested whether grouplevel contingent reward leader behavior serves as an antecedent to procedural justice climate perceptions and strength. Below, we present extant research and theory that underpins the relationships among the variables investigated and offer specific hypotheses. The model of the relationships proposed in the present study is summarized in Fig. 1. 1. Theory and hypotheses 1.1. Contingent reward transactional leadership and level of analysis Contingent reward transactional (CRT) leader behavior refers to leader behaviors emphasizing clarifying role and task requirements, and providing followers with material or psychological rewards contingent on the fulfillment of contractual obligations (Bass, 1998). Such leadership behavior is focused on clarifying effort-reward relationships and involves exchanges that take place between a leader and his or her followers. That is, a leader discusses with followers what is required and clarifies how these outcomes are to be achieved and the reward they will receive in exchange for their satisfactory effort and performance. In general, CRT leaders provide tangible or intangible support and resources to followers in exchange for their efforts and performance, define rules regarding work duties, maintain standards, and determine the consequences of goal attainment. It is this focus on clarifying roles and expectations that distinguishes CRT leader behavior from transformational leadership. For example, while transformational leadership behaviors result in followers identifying with the needs of the leader, CRT leaders give followers something they want in exchange for something the leader wants. Bass (1998) argued that transformational and transactional are distinct leadership concepts and that “transformational leadership does not substitute for transactional leadership” (p. 21). There are several plausible levels of analysis at which leadership phenomena can operate (Schriesheim, Castro, Zhou, & DeChurch, 2006). Yammarino et al. (2005) note, “relatively few studies in any of the areas of leadership research have addressed levels-of-analysis issues appropriately in theory, measurement, data analysis, and inference drawing” (p. 10). In the present study, consistent with recent theorizing and arguments that leaders often engage in behaviors which are not directed toward specific individuals but toward the group as a whole (Hogg, 2001; Judge et al., 2004; van Knippenberg, van Knippenberg, De Cremer, & Hogg, 2004; Vecchio, 1982), we conceptualized CRT leader behavior as a group-level construct. We believe it is suitable to examine CRT leader behavior as a group-level variable because, whereas characteristics behavior patterns such as leader contingent reward

F.O. Walumbwa et al. / The Leadership Quarterly 19 (2008) 251–265

253

Fig. 1. Summary of study hypotheses. Note. The dashed headed arrows represent controls.

behavior may differ across the direct reports of a particular supervisor, all subordinates within a given unit are exposed to the same supervisor behavior. Empirical findings also support that leader behavior influences individuals both independently and as members of groups. For example, in a study of nature and antecedents of OCB, Smith, Organ, & Near (1983) found that not only between-group differences in leader supportiveness was positively related to individual OCB but this between-group differences were greater than withingroup difference. Katerberg & Hom (1981) examined between-units leadership effects by correlating the averaged Leadership Behavior Description Questionnaire's (LBDQ) Consideration and Initiating Structure scale scores in each unit with individual-level criteria. They found that both between-group and within-group variations significantly predicted subordinate reactions. Other researchers have also conceptualized leadership as representing a group influence (Cohen, Chang, & Ledford, 1997). House & Aditya (1997) explained group-level leadership as specific leader behaviors distributed throughout the entire work unit, in which different members of the work unit enact those behaviors simultaneously. House & Aditya (1997) concluded, “the exercise of leader behaviors can be shared by members of work units, as well as conducted by formal work unit managers” (p. 459). Therefore, we treat CRT leader behavior as a group-level phenomenon, which entails leadership as behaviors directed toward the group rather than the individual members. 1.2. CRT leader behavior as an antecedent of procedural justice climate perceptions and strength Chan's (1998) typology of composition models, particularly additive, direct consensus, and dispersion models, are most relevant to understanding the distinction between justice climate perceptions and strength. Additive composition models suggest that the meaning of a unit-level construct is an average of the lower level perceptions regardless of the variance among such perceptions. The direct consensus composition models use “within-group agreement of scores to index consensus at the lower level and to justify aggregation of lower level scores to represent scores at the higher level” (Chan, 1998, p. 237). Following Chan's additive and direct consensus models, we view procedural justice climate as shared and within-unit agreement of procedural justice among individuals within a unit. Conversely, in dispersion composition model, “within-group variance (or some derivative) is used as the operationalization of the purported group-level construct” (Chan, 1998, p. 139), which is statistically captured by the individual differences within a unit. Therefore, procedural justice climate strength in the present study entails the degree to which group members agree on the favorability of procedural justice within the unit, and is captured by the variance of the procedural justice climate perceptions of unit members. Lindell & Brandt (2000) suggested that climate quality (perceptions) and consensus (strength) arise from external and internal structural factors, such as a leader, constituting ambient stimuli available to all group members. Because procedural justice perceptions can be fostered by allowing employee input in decision making or by adhering to several rules of fair treatment (Colquitt et al., 2002), we expect that leader behavior directed at the group could be an antecedent to the forming and strengthening of justice procedural justice climate. For example, according to justice judgment model (Leventhal, 1980), decision making in organizations is a complex and a multifaceted process. The theory suggests that there are several criteria for judging

254

F.O. Walumbwa et al. / The Leadership Quarterly 19 (2008) 251–265

procedural justice including decision accuracy, consistency, setting of ground rules, creation of change mechanisms, and bias suppression. We argue that CRT leaders are likely to create and strengthen the fairness perceptions shared among the unit members because of their clear exchange rules that facilitate consistency and accuracy, and discourage bias in the effort-reward and performance-reward relationships (Vroom, 1964). Such leaders clarify work requirements in order for the followers to meet the agreed-upon performance expectations and administer rewards and recognition contingent upon followers successfully carrying out their roles and assignments (Bass, 1998). We further expect these leader behaviors to foster procedural justice climate perceptions and strength because the effort-performance and performance-reward relationships are made clear and applied consistently. Moreover, because CRT leaders stress consistency and clarify work requirements in exchange of reward, the reward decisions are likely to be perceived as accurate, and therefore the perceived biases in the reward decision and process are likely to be minimized. Indeed, Bass (1998) notes, “consistent honoring of transactional agreements builds trust, dependability, and perception of consistency” (p. 11). When a leader consistently demonstrates CRT leader behavior over time, a pattern of behavior emerges. Hofmann & Jones (2005) argued that leaders whose behaviors emphasize contingent rewards and reinforce the importance of meeting performance expectations create a shared norm that encourages efficient, systematic, and organized behavioral regularities. CRT leaders are more likely to enhance and create strong perceptions of procedural fairness, because followers under CRT leaders are exposed to the norm of fairness as a result of the leaders' consistency in honoring and clarifying the exchange rules (e.g., clarifying work requirements). This is consistent with social identity theory of leadership (Hogg, 2001), which suggests that key group attributes such as leadership influence group members' cognitive states, such as collective identity (in our case perceived unit procedural justice perceptions), which in turn is posited to reinforce unit members' attitudes and behaviors. Thus, given the norm of consistency resulting from shared perceptions of CRT leader behavior toward the unit and because CRT leaders tend to fulfill and promote the rules of fair treatment as defined by Leventhal (1980), we expect CRT leader behavior to foster strong and favorable justice climates (Colquitt et al., 2002). Similarly, social exchange theory (Blau, 1964) also suggests that if followers perceive their leaders are consistent and fair in their exchange policies, the more apt are followers to conclude there is procedural justice climate in their units. Under these circumstances, we expect there would be low variance in followers' shared perception of procedural justice climate, and therefore high justice climate strength. Finally, although we are aware of no study that has directly linked group-level CRT leader behavior to justice climate perceptions and strength, some studies have linked leadership behaviors to justice climate or team climate in general. For example, Ehrhart (2004), in studying grocery store chains, found that procedural justice climate partially mediated the relationship between servant leadership and unit-level OCB. Similarly, Korsgaard, Schweiger, & Sapienza (1995) in an experiment involving intact management teams found that leader consideration of team members had a positive impact on team members' shared perceptions of procedural fairness. Thus, on the basis of theory and past research, we suggest that group-level CRT leader behavior would be positively related to procedural justice climate perceptions and strength. Hypothesis 1. Contingent reward transactional leader behavior is positively associated with procedural justice climate perceptions. Hypothesis 2. Contingent reward transactional leader behavior is positively associated with procedural justice climate strength. 1.3. Consequences of procedural justice climate and justice climate strength Konovsky (2000, p. 504) noted, “almost wholly absent from consideration has been the influence of higher-order factors on procedural justice including culture and organization or group-level procedural justice antecedents and consequences” (see Mossholder et al., 1998 for an exception). Since then, interest in procedural justice or fairness at higher levels has received a great deal of research attention (Colquitt, 2004; Colquitt et al., 2002, 2005; Ehrhart, 2004; Liao & Rupp, 2005; Naumann & Bennett, 2000; Roberson, 2006; Roberson & Colquitt, 2005; Simons & Roberson, 2003). Much of this interest, in addition to other factors, seems also to be based on the assumption that fair procedures lead to favorable outcomes for both individual employees and the unit or the organization as a whole (Johnson, Selenta, & Lord, 2006; Judge & Colquitt, 2004; Konovsky, 2000). For example, Mossholder et al. (1998) demonstrated that individuals belonging to units with higher aggregate perceptions of procedural justice reported greater job satisfaction than would have been expected on the basis of their individual justice perceptions alone. Naumann & Bennett (2000) found that procedural justice climate was significantly associated with helping behaviors even when individual justice perceptions are controlled. Liao & Rupp (2005) found that procedural justice climate contributes to individual satisfaction with supervisor (when procedural justice climate is supervisor-focused) and OCB (when procedural justice is organizationfocused) above and beyond the individual procedural justice perceptions. Roberson (2006) used conversational data from people working in teams that had experienced different levels of procedural treatment and outcome favorability. She found that team member justice perceptions interacted with outcome favorability to predict justice climate strength. She also found team sensemaking positively related to justice climate strength. Evidently, although the emerging evidence seems to provide support for the contention that procedural justice climate is correlated with a variety of organizational-relevant outcomes, more work still needs to be done. Particularly, there has been little work linking justice climate strength to relevant organizational outcomes or its antecedents. We thus sought to extend research by examining the relationships between procedural justice climate perceptions and strength and satisfaction with supervisor, organizational commitment, and rated OCB drawing on exchange theory and perceived organizational support theory.

F.O. Walumbwa et al. / The Leadership Quarterly 19 (2008) 251–265

255

Exchange theory and perceived organizational support (POS) theory have one thing in common—they are both characterized or based on the principle of reciprocity or reciprocal exchanges (Blau, 1964; Gouldner, 1960). Therefore, we view them as interrelated and complimentary to our explanations. Exchange theory posits that individuals will “give back” commensurately what they perceive to have received (or fail to have received) from the other party in the relationship. POS, on the other hand, refers to employees “global beliefs concerning the extent to which organization values their contributions and cares about their well-being” (Eisenberger, Huntington, Hutchison, & Sowa, 1986, p. 501). Dansereau, Alutto, & Yammarino (1984) presented a general model of exchange theory discussing the role of fairness perceptions in the development of leader–follower relationships. Their perspective suggests that justice climate perceptions and strength may have far-reaching consequences in leader–follower relationships, thereby affecting follower attitudes and behaviors. For example, according to the group-value model (Tyler, Degoey, & Smith, 1996), procedural justice is important because it informs people about their social connections to groups and those who lead them. Particularly, group members will decide to take or not take pride in their group membership and leadership based on fair or unfair procedures applied in their group (Lind & Earley, 1992). Therefore, the greater and stronger the procedural justice climate, the more apt are employees to conclude that the supervisor is treating them well by contributing to the fairness procedures. Perceived organizational support theory (Eisenberger et al., 1986) also supports the view that perceptions of fair treatment by the organization contributes to employee perceptions of organizational motives. Lind & Tyler (1988) suggested that fair treatment is indicative of being valued by an authority, which reinforces employees' feelings of pride, respect, and self-esteem. Such feelings have been reported to guide individual attitudes and behaviors such as satisfaction, commitment to remain with the group, and engage in assertive extra activities such as volunteering to help others within the group (Judge & Ilies, 2004; O'Reilly & Chatman, 1986). It would then follow that employees who perceive fair procedures in their units are likely to be more satisfied with supervisor, remain committed to the organization and display citizenship related behaviors. Although OCB is defined as “individual behavior that is discretionary, not directly or explicitly recognized by the formal reward system, and that in the aggregate promotes the effective functioning of the organization” (Organ, 1988, p. 4), previous research has reported a positive relationship between CRT leader behavior and OCB (e.g., Podsakoff et al., 2000). Podsakoff et al. (2000) reasoned that this finding may be because employees have a broader conception of performance that includes OCB and therefore believe that their leader administers rewards contingent upon both task performance and OCB. Furthermore, based on social exchange theory, employees form economic or social exchange relationships with their employers (Blau, 1964), and quality social exchange relationships, which could be facilitated by CRT leader behavior because of its focus on fair exchanges, often result in higher employee satisfaction, commitment, and more citizenship behaviors (Liao & Rupp, 2005). In addition, the feeling of being treated fairly as a group serves to promote positive moods, which have been suggested to relate to the likelihood of perceiving extra-role tasks as in-role (Bachrach & Jex, 2000). Judge & Ilies (2004) argued that when employees are in positive moods, they typically are more creative, more motivated to perform at a high level, and are more helpful toward their co-workers. Prior research also demonstrates that employees tend to respond more positively (e.g., higher job satisfaction, commitment, and OCB) to the degree that the procedures associated with the outcomes are viewed as fair (Liao & Rupp, 2005; Lind & Tyler, 1988; Thibaut & Walker, 1975; Mossholder et al., 1998; Naumann & Bennett, 2000). Summarizing these arguments, we expect justice climate perceptions and strength to be positively associated with satisfaction with supervisor, organizational commitment, and engagement in extra-role behaviors. Accordingly, we propose and test the following hypotheses: Hypothesis 3. Procedural justice climate is positively associated with (a) satisfaction with supervisor, (b) organizational commitment, and (c) rated organizational citizenship behavior. Hypothesis 4. Procedural justice climate strength is positively associated with (a) satisfaction with supervisor, (b) organizational commitment, and (c) rated organizational citizenship behavior. Hypothesis 5. Procedural justice climate mediates the relationship of CRT leader behavior with (a) satisfaction with supervisor, (b) organizational commitment, and (c) rated organizational citizenship behavior. Hypothesis 6. Procedural justice climate strength mediates the relationship of CRT leader behavior (a) satisfaction with supervisor, (b) organizational commitment, and (c) rated organizational citizenship behavior. 2. Method 2.1. Participants and procedures Data for the present study were collected from supervisors and their direct reports in six banks representing 42 work units in the Midwest region of the United States over a period of 1 year as part of a larger project on leadership development. Respondents mostly performed administrative/professional and clerical duties in these organizations. The questionnaires for direct reports were distributed to 281 employees at three different points in time separated by approximately 12 weeks each. At Time 1, direct reports completed personal information (e.g., age, gender, organizational tenure), and evaluated their immediate supervisor's leadership behaviors. At Time 2, respondents completed a measure of procedural justice climate. Finally, at Time 3, respondents were asked to evaluate their level of satisfaction with their immediate supervisor and level of commitment to the organization. We asked respondents to provide their names so that we could match data collected

256

F.O. Walumbwa et al. / The Leadership Quarterly 19 (2008) 251–265

at later times and from their supervisors. However, letters from the investigators and bank presidents assured participants that their responses would remain confidential and would be used strictly for research purposes and that all identifying information would be removed and destroyed after data was coded. Completed questionnaires were sent directly to the first author. Approximately a week after all data from employees were coded, an e-mail was sent to supervisors asking them to rate their direct reports on OCB. Supervisors were identified with the help of each bank's HR department and were all familiar with the direct reports they evaluated. Each supervisor was provided with a list of employees in his or her unit to evaluate. Listwise deletion of individuals with missing information and those that could not be matched resulted in a final usable sample of 212 participants for this study. Fortytwo supervisors participated and completed their ratings (100% response rate). Across the banks, the average response rate was about 74%, with an average of five employees from each unit (ranging from 4 to 7 employees).2 The average age across banks was 36.13 (S.D. = 11.69), 72% were women, and over 90% had completed some college or held a university degree. The average organization tenure was 7.93 years (S.D. = 8.58). 2.2. Measures 2.2.1. Contingent reward transactional leadership The CRT leader behavior was measured by four items from Bass & Avolio's (2000) Multifactor Leadership Questionnaire (MLQ) Form 5x-short (α = .82), but items were changed to refer to the unit supervisor. Employees rated the degree to which they received contingent rewards from their unit supervisor using a 5-point scale ranging from 0 (Not at all) to 4 (Frequently, if not always). Considerable evidence of the validity and reliability of this scale has been compiled (see Judge & Piccolo, 2004). A sample item is: “My unit supervisor makes clear what one can expect to receive when goals are achieved.” 2.2.2. Procedural justice climate To measure procedural justice climate, employees responded to a 7-item measure modified from Colquitt's (2001) individual procedural justice scale. This scale is based on Leventhal (1980) and Thibaut & Walker's (1975) concepts of fairness in social relationships and procedural justice, respectively. Respondents were asked about the procedures used to arrive at rewards in their units—thus the referent point was the unit. A sample item is “Have the procedures been applied consistently in your unit?” Following Colquitt (2001), responses were made on a 5-point scale (1 = to a small extent to 5 = to a large extent). Two items of Colquitt's 7-item scale of procedural justice are from Thibaut & Walker's (1975) process control (e.g., the ability to voice one's views and arguments during a procedure) and decision process (e.g., the ability to influence the actual outcome) and five items are from Levanthal's (1980) procedural justice scale. Because items capturing procedural justice climate came from two different scales, we conducted an exploratory factor analysis on the seven items. Results indicated that a single factor with an eigenvalue greater than 1.00 explained 72 percent of the variance in the items, with an internal consistency (Cronbach α) of the combined scale of .77, suggesting that these items formed a reliable scale. We then aggregated responses to this measure within each unit to generate a unit-level procedural justice climate measure. 2.2.3. Procedural justice climate strength Following Chan's (1998) dispersion composition models, justice climate strength was captured by the variance of the procedural justice climate perceptions of group members (i.e., the within-group dispersion scores). We calculated this variance using the coefficient of variation (Allison, 1978), which corrects for the lack of independence between measures of central tendency and measures of dispersion, for each group. We then created justice climate strength value for each group by computing the standard deviation of procedural justice climate and dividing that value by the group-mean. This value was then standardized and its sign was reversed so that higher values represented higher levels of procedural justice climate strength (see also Colquitt et al., 2002; Roberson, 2006). 2.2.4. Organizational commitment Organizational commitment (α = .92) was measured using 10 items taken from Mowday, Steers, & Porter (1979). This scale assesses the degree to which individuals exert effort as well as their acceptance of an organization's goals and values, and a strong desire to remain an organizational member. This scale of organizational commitment relates highly to Meyer and Allen's concept of affective organizational commitment (Meyer & Allen, 1997). A sample item is “This organization has a great deal of personal meaning for me.” Responses were made on a 5-point scale (1 = Strongly disagree to 5 = Strongly agree). 2.2.5. Satisfaction with the leader We used nine items from Smith, Kendall, & Hulin's (1969) Job Descriptive Index (JDI) to measure satisfaction with the leader (α = .92). Direct reports were asked to respond to a series of statements describing their immediate supervisors. On a 3-point scale, respondents were asked to circle “1” if an item did not describe his/her supervisor, “2” if he/she could not decide whether an item described his/her supervisor, and “3” if an item described his or her supervisor. A sample item is: “My supervisor interferes with my work (reverse coded).” 2 Consistent with Liao and Rupp (2005), we considered employees to be members of work unit when they share a common supervisor and work together in the same unit (i.e., mortgage, teller, purchasing, etc).

F.O. Walumbwa et al. / The Leadership Quarterly 19 (2008) 251–265

257

2.2.6. Organizational citizenship behavior We used 12 items to measure OCB. Due to space limitations and on the recommendation of participating firms, only items measuring helping behavior (also called altruism) and courtesy aspects of OCB were included in the present study. Altruism or helping behavior refers to the “discretionary behaviors that the effect of helping a specific other person with an organizationally relevant task or behaviors.” Courtesy refers to “discretionary behavior on the part of an individual aimed at preventing workrelated problems with others from occurring” (Podsakoff, MacKenzie, Moorman, & Fetter, 1990, p. 115). Leaders provided reports of OCB for each of their direct reports. Helping behavior consisted of seven items from Smith et al. (1983). A sample item is “This employee volunteers to do things not formally required by his or her job. Courtesy was measured using five items developed by Podsakoff et al. (1990). A sample item is “Is mindful of how his or her behavior affects other people's job.” Responses were anchored on a 5-point scale ranging from 1 (Never) to 5 (Always). Results from the exploratory factor analysis indicated that a single factor with an eigenvalue greater than 1.00 explained 74% of the variance in the items, with an internal consistency of .73. The two scales were therefore combined to form a composite OCB index. 2.2.7. Control variable Contingent reward behavior commonly correlates with other subordinate perceptions of leader behavior, such as leader supportiveness and consideration, directiveness, and goal (and separately ‘role’) clarifying behavior (Kim & Yukl, 1995) and transformational leadership (Hofmann & Jones, 2005; Howell & Hall-Merenda, 1999; Judge & Piccolo, 2004; Pillai et al., 1999). In the present study, we used perceived supervisor support as a control variable to rule out possible alternative explanation of our findings. To measure perceived supervisor support, we used Eisenberger, Stinglhamber, Vandenberghe, Sucharski, & Rhoades (2002) 8-item scale. These items were derived from Eisenberger et al.'s (1986) Perceived Organizational Support (POS) scale based on their high factor loadings. For each item, the word organization was replaced with supervisor. Respondents indicated their agreement with each item using a 5-point Likert-type scale (1 = strongly disagree, 5 = strongly agree). 2.3. Level of analysis Kozlowski & Klein (2000) pointed out the importance of specifying the level of analysis at which researchers are assuming variables will operate. In the present study, satisfaction with supervisor, organizational commitment, and rated OCB were examined as individual-level variables. Following Chan's (1998) additive, direct consensus, and dispersion composition models, both justice climate perceptions and justice climate strength were examined as unit-level variables. Although CRT leader behavior can be thought of as a one-on-one aspect of leadership in terms of the formation of individual exchanges, we treated this variable as a group-level variable because we were interested in the behaviors that leaders tend to exhibit across situations and followers— behaviors exhibited to a group as a whole. Secondly, because responses to leadership, evaluations of procedural justice climate perceptions and ratings of work-related attitudes (e.g., satisfaction with supervisor and organizational commitment) came from the same individuals, relationships might be inflated because of common source variance. By treating leadership as a group-level variable, we somehow reduced this risk, because our analyses are based on the relationships between the shared perception among members in a group regarding their leader's behavior and their individual responses to these behaviors. The advantage of this approach is that individual differences in follower reactions to their leader's behavior are treated as error (Bono & Judge, 2003). Others (e.g., Judge et al., 2004) have also argued that because followers are nested within leaders and organizations typically comprised of multiple levels of leaders, a multilevel approach to investigating leadership is certainly a plausible strategy that should be pursued. Thus, consistent with recent theoretical arguments (Yammarino et al., 2005) and several empirical prior studies (Atwater et al., 1997; Yammarino, Spangler, & Dubinsky, 1998), we treated CRT leader behavior as a group-level variable. 2.4. Analysis strategy We utilized hierarchical linear modeling (HLM) to test our hypotheses. We used intercepts-as-outcomes models to test the hypotheses, because such models hypothesize a predictive effect of group-level variables (i.e., CRT leader behavior, procedural justice climate and strength) on individual-level variables (i.e., satisfaction with the leader, organizational commitment, and rated OCB). In all the analyses, we used grand-mean centering for all Level 2 predictors because the approach helps reduce the covariance between intercepts and slopes, thereby reducing potential problems associated with multicollinearity (Hofmann, 1997). According to Kenny, Kashy, & Bolger (1998), to achieve full mediation, four conditions are required. First, the independent variable (i.e., CRT leader behavior) is significantly related to the mediator (i.e., justice climate perceptions and strength). Second, the independent variable is significantly related to dependent variables. Note, however, that Kenny et al. (1998) argued that this relationship could be weak or even non significant in cases where the independent variable is more distal. Third, the mediator must be related to the dependent variables. Fourth, the relationship between the independent and the dependent variables becomes non significant when the mediator variable is introduced into the equation. If the relationship between independent and dependent variables is reduced but remains significant when the mediator is present, there is evidence for partial mediation. Finally, because we collected data from six different banks, we examined whether there were significant differences between the banks in terms of the variables studied using ANOVA tests and found no significant differences. Similarly, age or organizational tenure and gender are commonly controlled in studies involving job attitudes and leadership (Berson, Shamir, Avolio, & Popper, 2001; Riordan, Griffith, & Weatherly, 2003). For example, there is a tendency for older employees to have more favorable attitudes.

258

F.O. Walumbwa et al. / The Leadership Quarterly 19 (2008) 251–265

Zaccaro, Rittman, & Marks (2001) have also suggested that group composition and capabilities may sometimes enhance or severely mitigate leader influence. We also examined the correlations between follower demographics—age, gender, and organizational tenure and group size—and CRT leader behavior, procedural justice climate perceptions, and outcome variables. No significant associations were found. Given these pattern of associations, follower demographics and group size could not serve as a problematic omitted variables in our analyses (James, 1980). 3. Results 3.1. Preliminary analysis To show that CRT leader behavior is distinct from procedural justice climate, we conducted several preliminary analyses at the individual-level of analysis. We first conducted a confirmatory factor analysis (CFA), using AMOS maximum likelihood procedure (Arbuckle & Wothke, 1999) to compare the fit of four different measurement models including CRT leader behavior, procedural justice climate, satisfaction with supervisor, organizational commitment, and rated OCB. For each variable, we randomly averaged items of each measure to form indicators (e.g., 2 for CRT leader behavior, 2 for procedural justice climate, 3 for satisfaction with the leader, 3 for organizational commitment, and 4 for OCB). The CFA results presented in Table 1 show that the 5-factor model fit the data well (χ2 = 112.72; df = 63; CFA = .97; TLI = .96; RMSEA = .06). Against this model, we tested three alternative models: Model 1 was a 4-factor model with CRT leader behavior merged with PJC to form a single factor, satisfaction with the leader, organizational commitment, and OCB; Model 2 was a 3-factor model with satisfaction with the leader, organizational commitment, and OCB merged to form a single factor, CRT leader behavior, and procedural justice climate; Model 3 was a 2-factor model with CRT leader behavior merged with procedural justice climate to form a single factor and satisfaction with the leader, commitment, and OCB merged to form another single factor. As shown in Table 1, the best fitting model was the 5-factor model, providing evidence of the distinctiveness of CRT leader behavior, procedural justice climate, satisfaction with supervisor, organizational commitment, and OCB constructs. In the final test of distinctiveness of CRT leader behavior and procedural justice climate, we compared the correlations of CRT leader behavior and procedural justice climate with rated OCB following the Cohen & Cohen (1983) test of the differences between two Pearson correlations from the same sample. We used rated OCB in this analysis to reduce potential alternative explanations due to common source bias. The t-statistic for the difference between CRT leader behavior-rated OCB and procedural justice climate-rated OCB was 2.00 (df = 209, p b .01). On the basis of these tests, we concluded that CRT leader behavior and procedural justice climate are distinct constructs. 3.2. Aggregation tests Contingent reward leadership, perceived supervisor support (control) and procedural justice climate perceptions are theorized in the present study as unit-level variables. Such conceptual aggregation, however, requires statistical justification (Kozlowski & Klein, 2000). Intraclass correlations, ICC (1) and ICC (2), are statistics commonly used to justify aggregation (Bliese, 2000). The ICC (1) compares the variance between units of analysis (bank branches) to the variance within units of analysis using the individual ratings of each respondent. The ICC (2) assesses the relative status of between and within variability using the average ratings of respondents within each unit. The ICC (1) for CRT leader behavior was .08, .09 for perceived supervisor support, and .12 for procedural justice climate perceptions. The ICC (2) was .68 for CRT leader behavior, .76 for perceived supervisor support, and .83 for procedural justice climate perceptions. Although no strict standards for aggregation using ICC (1) and ICC (2) exist, Bliese (2000) reported values ranging from .05 to .30 for ICC (1) and Glick (1985) recommended an ICC (2) cutoff of .60. The F-tests for the group effect for these variables were significant (F[41, 170] = 1.65, p b .05 for CRT leader behavior, (F[41, 170] = 1.94, p b .01 for perceived supervisor support, and F[41, 170] = 1.81, p b .01 procedural justice climate perceptions). To further justify aggregation, we also calculated rwg (James, Demaree, & Wolf, 1984). The rwg average value was .70 for CRT leader behavior, .74 for perceived supervisor support, and .88 for procedural justice climate perceptions, both meeting the recommended .70 cutoff (James et al., 1984). Taken together, we believed that our aggregation statistics provided sufficient support for aggregation, in addition to theoretical justification and created our group-level variables.

Table 1 Comparison of measurement models χ2

df

CFI

TLI

RMSEA

Five factor model: CRT, PJC, SS, OC, and OCB

112.72

63

–

.97

.96

.06

Four factor model: CRT and PJC combined into one factor, SS, OC, and OCB Three factor model: SS, OC, and OCB combined into one factor, CRT, and PJC Two factor model: CRT and PJC combined into one factor and SS, OC, and OCB combined into one factor

189.35 563.53 742.27

67 70 .72

76.62** 563.53** 629.55**

.92 .62 .58

.90 .50 .47

.09 .20 .21

Models Baseline model Model 1 Model 2 Model 3

Factors

Δχ2

CRT = Contingent reward transactional leader behavior, PJC = procedural justice climate, SS = satisfaction with supervisor, OC = organizational commitment, OCB = organizational citizenship behavior. ** Significant at p b .01 (two-tailed).

F.O. Walumbwa et al. / The Leadership Quarterly 19 (2008) 251–265

259

Table 2 Individual-level means, standard deviations, and inter-correlations for the study variables Variables 1. CRT leader behavior 2. Perceived supervisor support 3. Procedural justice climate 4. Satisfaction with supervisor 5. Organizational commitment 6. Organizational citizenship behavior

Mean

SD

1

2

3

4

5

6

2.63 3.74 4.42 1.90 3.44 3.76

.92 .69 .70 .55 .83 .54

.82 .15** .24** .39** .26** .49**

.72 .24** .20** .40** .49**

.77 .36** .38** .32**

.92 .23** .37**

.92 .47**

.73

Reliabilities reported in bold on diagonal; SD = Standard deviation; CRT = Contingent reward transactional. Procedural justice climate strength only exists as a group-level variable and therefore is not included in this table of individual-level descriptive statistics. n = 212. *p b .05; **p b .01 (two-tailed).

3.3. Hypotheses testing Table 2 presents the individual-level means, standard deviations, and zero-order correlations. Contingent reward leader behavior is positively associated with procedural justice climate (r = .24, p b .01), satisfaction with supervisor (r = .39, p b .01), organizational commitment (r = .26, p b .01), and rated OCB (r = .49, p b .01). Table 2 also shows that procedural justice climate is positively associated with satisfaction with supervisor (r = .36, p b .01), organizational commitment (r = .38, p b .01), and OCB (r = .32, p b .01). A necessary precondition when testing variables at different levels is that there must be a significant within- and betweengroup variance in outcome variables (Hofmann, 1997; Gavin & Hofmann, 2002). Therefore, before testing our hypotheses, we assessed the magnitude of between-group variance in satisfaction with supervisor, organizational commitment, and rated OCB by estimating an HLM model with no Level 1 or Level 2 predictors. This model partitions the variance in the outcome variable into its within-group (i.e., Level 1 residual variance) and between-group (i.e., Level 2 residual variance) components (Raudenbush & Bryk, 2002). Results showed that 45% (χ2(41) = 207.17, p b .001) of the variance in satisfaction with supervisor, 22% (χ2(41) = 77.55, p b .001) of the variance in organizational commitment, and 27% (χ2(41) = 84.34, p b .01) of the variance in OCB reside between groups. The chi-square tests indicated that the between group variances were significant; that is, the intercept terms significantly varied across groups. Table 3 provides a summary of the models and results used to test Hypotheses 1 and 2 following the steps suggested by Kenny et al. (1998) to test mediation relationships. Hypotheses 1 and 2 predicted that CRT leader behavior would be significantly associated with procedural justice climate and justice climate strength, respectively. These hypotheses were tested using Ordinary Least Square (OLS) linear regression because CRT leader behavior, procedural climate justice and strength are at the same grouplevel of analysis. We controlled for perceived supervisor support. This control variable was significant for procedural justice climate (β = .29, p b .01), but not justice climate strength (β = .10, ns). Table 3 also shows that CRT leader behavior was positive and significantly related to procedural justice climate perceptions (β = .56, p b .01) and justice climate strength (β = .59, p b .01), supporting Hypotheses 1 and 2 and Condition 1 for mediation. Table 4 provides a summary of the HLM models and results used to test Hypotheses 3–6, again, following the steps suggested by Kenny et al. (1998). We used grand mean centering for all the Level 2 predictors, because there is some evidence in the literature suggesting that grand mean centering is a better choice of scaling for Level 2 predictors (Hofmann & Gavin, 1998). As shown in Table 4, we controlled for perceived supervisor support in our tests for each hypothesis. This control variable was significant in all the models. Supporting Condition 2 for mediation, HLM results (Table 4, step 2) revealed that CRT leader behavior was positive and significantly related to satisfaction with supervisor (γ02 = .35, p b .01), organizational commitment (γ02 = .27, p b .01), and rated OCB (γ02 = .26, p b .01). Hypotheses 3 and 4 predicted that procedural justice climate and strength would be positively related to satisfaction with supervisor, organizational commitment, and rated OCB. As can be seen in Table 4 (step 3), HLM results revealed that procedural Table 3 Ordinary least squares (OLS) regression results for Hypotheses 1 and 2 Parameter estimates a Step 1 in testing mediation Hypothesis 1 PJCLj = β1 (PSSj) + β2 (CRTj) Hypothesis 2 PJCSj = β1 (PSSj) + β2 (CRTj)

β1

β2

.29**

.56**

.10

.59**

PSS = perceived supervisor support; PJCL = procedural justice climate; PJCL = procedural justice climate strength; CRT = contingent reward transactional leader behavior. **p b .01 (two-tailed). a Results are standardized parameter estimates.

260

F.O. Walumbwa et al. / The Leadership Quarterly 19 (2008) 251–265

Table 4 Hierarchical linear modeling results for Hypotheses 3–6 Model Models and steps in testing mediation a Step 2 L1: SSij = β0j + rij L2: β0j = γ00 + γ01 (PSSj) + γ02 (CRTj) + U0j L1: OCij = β0j + rij L2: β0j = γ00 + γ01 (PSSj) + γ02 (CRTj) + U0j L1: OCBij = β0j + rij L2: β0j = γ00 + γ01 (PSSj) + γ02 (CRTj) + U0j Step 3 Hypotheses 3a and 4a L1: SSij = β0j + rij L2: β0j = γ00 + γ01 (PJCLj) + γ02 (PJCSj) + U0j Hypotheses 3b and 4b L1: OCij = β0j + rij L2: β0j = γ00 + γ01 (PJCLj) + γ02 (PJCSj) + U0j Hypotheses 3c and 4c L1: OCBij = β0j + rij L2: β0j = γ00 + γ01 (PJCLj) + γ02 (PJCSj) + U0j Step 4 Hypotheses 5a and 6a L1: SSij = β0j + rij L2: β0j = γ00 + γ01 (PSSj) + γ02 (CRTj) + γ03 (PJCLj) + γ04 (PJCSj) + U0j Hypotheses 5b and 6b L1: OCij = β0j + rij L2: β0j = γ00 + γ01 (PSSj) + γ02 (CRTj) + γ03 (PJCLj) + γ04 (PJCSj) + U0j Hypotheses 5c and 6c L1: OCBij = β0j + rij L2: β0j = γ00 + γ01 (PSSj) + γ02 (CRTj) + γ03 (PJCLj) + γ04 (PJCSj) + U0j

γ00

γ01

γ02

γ03

γ04

1.90**

.44**

.35**

3.44**

.59**

.27**

3.76**

.53**

.26**

1.89**

.71**

.19**

3.43**

.46**

.27**

3.74**

.29**

.24**

1.89**

.24*

.08

.52**

.11*

3.43**

.46**

.03

.43**

.27**

3.74**

.43**

.19*

.26*

.05

L1 = Level 1; L2 = Level 2; SL = Satisfaction with supervisor; OC = Organizational commitment; OCB = Organizational citizenship behavior; CRT = Contingent reward transactional leader behavior; PSS = Perceived supervisor support; PJCL = Procedural justice climate level; PJCS = Procedural justice climate strength. γ00 = Intercept of Level 2 regression predicting β0j; γ01, γ02, γ03, γ04 = Slope of Level 2 regression predicting β0j. *p b .05; **p b .01 (two-tailed). a Results are unstandardized parameter estimates from the intercepts-as-outcomes models; Level 2 predictors are grand-mean centered.

justice climate (PJCL) and strength (PJCS) are positive and significantly related to satisfaction with supervisor (PJCL: γ02 = .71, p b .01; PJCS: γ02 = .19, p b .01), organizational commitment (PJCL: γ02 = .46, p b .01; PJCS: γ02 = .27, p b .01), and rated OCB (PJCL: γ02 = .29, p b .01; PJCS: γ02 = .24, p b .01). Thus, Hypotheses 3 and 4 are supported. Hypotheses 5 and 6 predicted that justice climate perceptions and strength, respectively, would mediate the relationships between CRT leader behavior and satisfaction with supervisor, organizational commitment, and rated OCB. Given the results for Hypotheses 1–4, the preconditions for mediation were supported (Kenny et al., 1998). To test these hypotheses, we included both procedural justice climate and strength in the same model so that we are not capitalizing on chance by incorporating more tests and whatever shared variation exists are accounted for in the analyses. As shown in Table 4 (step 4), Hypotheses 5 (a, b) and 6 (a, b) were supported. That is, when regressed on satisfaction with supervisor and organizational commitment, both procedural justice climate (satisfaction with supervisor: γ03 = .52, p b .01; organizational commitment: γ03 = .43, p b .01) and strength (satisfaction with supervisor: γ04 = .11, p b .05; organizational commitment: γ04 = .27, p b .01) were significant, whereas CRT leader behavior became non significant (satisfaction with supervisor: γ02 = .08, ns; organizational commitment: γ02 = .03, ns). Results for Hypotheses 5c and 6c are mixed. For example, when regressed onto rated OCB, both CRT leader behavior (γ02 = .17, p b .05) and procedural justice climate (γ03 = .26, p b .01) were significant. However, the magnitude of the coefficient between CRT leader behavior and rated OCB is reduced, suggesting a partial mediation occurred. On the other hand, procedural justice climate strength was not significant in predicting rated OCB (γ04 = .05, ns). To further determine whether significant mediation occurred or not, we conducted a Sobel test (Sobel, 1988) for each model proposed in Hypotheses 5 and 6. Hypotheses 5a, b, 6a and b were supported. Results revealed that procedural justice climate perceptions and strength partially mediated the relationship between CRT leader behavior and rated OCB. Thus, Hypotheses 5c and 6c were partially supported. 3.4. Testing alternative models While the results presented here offer support for the hypothesized relationships, the possibility that a competing model may provide a better description of the data can not be ruled out. For example, whereas work of Schneider and Colleagues (e.g., Schneider et al., 1994, 2003; Schneider & Reichers, 1983) have pointed out the role of leadership and particularly leader reward behavior in creating climate strength, there is a likelihood that climate strength may act as a potential moderator of procedural justice climate effects given that there are differences across units in the extent to which members agree about procedural justice.

F.O. Walumbwa et al. / The Leadership Quarterly 19 (2008) 251–265

261

For example, Colquitt et al. (2002) investigated the interactive effects of procedural justice climate and justice strength in predicting team performance and team absenteeism and found that the effects of climate level were moderated by climate strength, such that the relationships were more beneficial in stronger climates. To provide a comparative support for the hypothesized relationships tested in the present study, we also examined climate strength as a potential moderator of procedural justice climate effects on satisfaction with supervisor, organizational commitment, and rated OCB. Our data did not support moderating effect of procedural justice climate strength in the relationships between procedural justice climate and follower work-related attitudes (e.g., satisfaction with supervisor and organizational commitment), except for OCB, which was marginally significant (p b .10). These results are however consistent with previous work by Lindell & Brandt (2000) who failed to reveal strong moderating effects of climate strength. Secondly, whereas some researchers (e.g., Wayne, Shore, Bommer, & Tetrick, 2002) have argued that justice and contingent reward behavior are antecedents of perceived organizational support and leader–member exchange, others (e.g., Pillai et al., 1999) have also argued that CRT leader behavior is an antecedent of procedural and distributive justice. Therefore, we also examined the effects of changing construct ordering (e.g., CRT leader behavior → procedural justice climate → outcome variables vs. procedural justice climate → CRT leader behavior → outcome variables) to further provide support for our hypothesized relationships. To do this, we conducted two separate hierarchical regression analysis tests and then examined the change in R2 after including procedural justice climate or CRT leader behavior into the equation predicting satisfaction with supervisor, organizational commitment, and rated OCB. We first ran a regression model in which CRT leader behavior was entered into the equation predicting outcome variables in the first step and then entered procedural justice climate in a second step. Results showed that the change in variance explained (ΔR2) when procedural justice climate was included in the equation after CRT leader behavior in predicting satisfaction with supervisor was .04 (p b .05), organizational commitment .03 (p b .05), and rated OCB .07 (p b .01). We then shifted the sequence of the hierarchical regression for outcome variables (e.g., satisfaction with supervisor, organizational commitment, and rated OCB) by entering procedural justice climate in the first step and CRT leader behavior in a second step. Results showed that when CRT leader behavior was included in the regression equation after entering procedural justice climate, CRT leader behavior could not explain any significant incremental variance in outcome variables. Hence, we concluded that CRT leader behavior is best conceptualized as an antecedent to procedural justice climate. 4. Discussion The present study sought to contribute to the literature by testing group-level CRT leader behavior as an antecedent of grouplevel procedural justice climate perceptions and strength, which further influence follower attitudes and behaviors. Our results show that CRT leader behavior contributes to procedural justice climate perceptions and strength. We also found that procedural justice climate perceptions and strength mediate the relationships between CRT leader behavior with satisfaction with supervisor, organizational commitment, and rated OCB. Note, however, that our results did not reveal full mediation for procedural justice climate perceptions and strength in the relationship between CRT leader behavior and rated OCB. We suspect that the OCB measure may have influenced our results. Future research may consider using all OCB measures to replicate our research. Regardless, our results suggest that when unit members collectively feel the procedures used by their unit supervisors are consistent and fair, which could be fostered by leader reward behavior, they are more likely to be satisfied with their supervisor, remain committed to the organization, and display organizational citizenship behaviors. Therefore, unit supervisor's fairness promotion does seem to play a pivotal role in translating the effects of CRT leaders to follower positive attitudes and behaviors. These results are consistent with the argument that the manner in which a leader administers rewards is a critical determinant of the effectiveness of contingent leader behavior (Podsakoff et al., 2006). The present study makes several important contributions to research on emerging area of CRT leader behavior and procedural justice climate. As noted earlier, despite calls for research integrating leadership and organizational justice literatures (Pillai et al., 1999; Scandura, 1999), relatively little theoretical or empirical work has been done to integrate these two important literatures. As an important recent exception, Ehrhart (2004) found that procedural justice climate partially mediated the relationship between servant leadership and unit-level OCB. Thus, both Ehrhart (2004) and our study contribute to the literature by testing group-level CRT leader behavior and servant leadership as antecedents of procedural justice climate, which have rarely been the focus in the leadership literature. Second, our results are consistent with results of Pillai et al. (1999) at the individual-level of analysis. Though they did not test CRT leader behavior as an antecedent of procedural justice, their zero-order correlation indicated that CRT leader behavior was positively related to procedural justice (r = .50). We believe our study extends and contributes above and beyond this study. For example, Pillai et al. (1999) did not test the mediating effects of procedural justice in the relationships between CRT leader behavior and outcome variables; whereas we specifically hypothesized and tested the mediating process of group-level procedural justice climate. This distinction, though slight, is important, and calls for additional research that examines the relationships between justice climate, CRT leader behavior, and work-related attitudes and behaviors at multiple levels of analysis. Third, in light of the few studies that have been published on the antecedents and consequences of procedural justice climate strength, our study adds a valuable contribution to the literature, as it provides empirical evidence on a potential antecedent and consequences of procedural justice climate strength (Colquitt et al., 2002; Cropanzano et al., 2001). Lindell & Brandt (2000) noted, “a number of researchers have recognized the theoretical significance of climate consensus, but little data have been collected on its relationships with either antecedent or outcome variables, let alone whether it mediates the relationship between organizational antecedents and outcomes” (p. 332). Our study joins a growing number of studies by showing that leadership

262

F.O. Walumbwa et al. / The Leadership Quarterly 19 (2008) 251–265

behaviors marked by consistency and contingent reward behaviors are likely to lead to a strong procedural justice climate perceptions and strength. These results also have practical implications for leadership and human resource management, as a better understanding of the relationships between CRT leader behavior, procedural justice climate perceptions and strength, and follower work-related attitudes and behaviors allows the design of more effective training programs. Our results suggest that organizations need to pay more attention to programs and policies that encourage fairness in order to enhance satisfaction with supervisor, commitment with the organization, and employee citizenship behaviors (Costigan, Insinga, Kranas, Kureshov, & Ilter, 2004). For example, one way to foster procedural justice climate is by ensuring that rewards are given out accurately and consistently. Employee surveys or focus groups can provide an accurate picture of whether the processes used to provide reward systems and the way those rewards are viewed by the recipients. In addition, our results emphasize that the supervisor who implements the reward programs and policies must set clear ground rules so that employees are clear on what is expected of them and the reward they will receive for satisfactory effort. Secondly, by showing procedural justice climate perceptions and strength as potential mediators, our results suggest that managers need to consider the mechanism by which CRT leader behavior is related to followers' work attitudes and citizenship behaviors. Specifically, our results suggest that managers need to focus on the fair treatment of a unit as a whole to enhance unit member justice perceptions and strength, which is likely to influence followers' satisfaction with supervisor, levels of organizational commitment, and organizational citizenship behaviors. Interventions designed to clarify the meaning and interpretation of unit treatment may help manage fairness within units (Roberson, 2006). Finally, our results suggest that management selection should not only assess inspiring and charismatic leader behaviors (e.g., transformational leadership behaviors) but also reward leader behaviors such as contingent reward transactional behaviors when selecting individuals for supervisory positions. This may help increase procedural justice climate perceptions and strength, which subsequently influence followers' satisfaction with supervisor, commitment to the organization, and employee citizenship behaviors. 4.1. Limitations and suggestions for future research The present study has some limitations. First, because the study was conducted within a single industry—the banking industry— one could question the generalizability of the results. Future research should extend these findings to non-financial firms to enhance generalizability. Second, because part of our data came from the same individuals and involved perceptual data (i.e., followers provided ratings of their perceptions of supervisors' CRT leader behavior, procedural justice climate, and rated their own workrelated attitudes), there is potential for common source bias, which means we cannot determine causality among the variables, but can only make inferences. We attempted to address this potential problem in several ways. First, by averaging followers' evaluations of their leaders' behavior and justice climate perceptions, and collecting data at different points in time, we somehow reduced the common source bias. Podsakoff, MacKenzie, Lee, & Podsakoff (2003) argued that temporal separation reduces common source variance by allowing previously recalled information to leave short-term memory, in essence diminishing respondent's ability and motivation to use his or her prior responses to answer subsequent questions. Second, we also conducted additional analyses to assess the distinctiveness of CRT leader behavior and procedural justice climate. The first test was a CFA of our measurement model, and the second was the results of the Cohen & Cohen (1983) test of differences between two Pearson correlations (r) from the same sample. Taken together, it seems unlikely that common method bias could account completely for our pattern of results. Regardless, future research should include more behavioral data as well as more objective work performance outcomes obtained from different sources. Researchers may also consider longitudinal experimental designs where both qualitative and quantitative data are collected over repeated observations to provide stronger evidence for causal relationships. Prior studies have controlled for transformational leadership while examining CRT leader behavior (Hofmann & Jones, 2005; Howell & Hall-Merenda, 1999; Judge & Piccolo, 2004; Pillai et al., 1999). We did not control for transformational leadership in the present study, which could be a clear limitation. However, we controlled for the effects of perceived supervisor support (Eisenberger et al., 2002), which is a characteristic of transformational leadership (Bass, 1998). We call for replication of our results in research in which transformational leadership explanations are also controlled. Along the same lines, we encourage future researchers to conduct a more elaborate and integrative studies that include authentic, leader–member exchange, transactional, transformational and servant leadership theories concurrently, and the different aspects of organizational justice, including distributive, informational, and interpersonal justice (Colquitt, 2001; Rupp & Cropanzano, 2002). Such studies are necessary in order to provide a more complete conclusion regarding the relationships among organizational justice, leadership, and workrelated attitudes and behaviors. Despite these limitations, the present study has a number of strengths, including a diverse sample from six different banking organizations, multi-level analyses, data collected at different points in time, and ratings from both supervisors and their direct reports. The multi-level design are particularly important because they increase the confidence in our results by providing greater insight into the relationships between procedural justice climate perceptions and strength, CRT leader behavior, and follower satisfaction with supervisor, organizational commitment, and OCB. Moreover, the findings that CRT leader behavior predicted followers' satisfaction with supervisor, organizational commitment, and supervisor rated OCB, even after controlling for perceived supervisor support effect suggests CRT leader behavior is independently related to followers' work-related attitudes such as satisfaction with supervisor, organizational commitment and rated OCB. Integrating our study in the current literature of organizational justice, we suggest the following future research directions. First, future research should further distinguish the sources of procedural justice climate perceptions and strength and examine

F.O. Walumbwa et al. / The Leadership Quarterly 19 (2008) 251–265

263

their predictability of various outcomes. The multifoci approach in recent organizational justice studies have made a successful attempt to distinguish leader-focused from organization-focused justice perceptions and climate, and concluded that supervisorfocused justice climate better predicted supervisor-focused outcomes, while organization-focused justice climate better predicted organization-focused outcomes (Liao & Rupp, 2005; Rupp & Cropanzano, 2002). Although we paralleled our justice climate measure with organization-focused justice climate, we did not particularly specify the focus. Future research should build on this body of multifoci justice research to further test organization-focused, supervisorfocused, and even coworker-focused justice climate as possible mediating mechanisms between CRT leader behavior and employee outcomes. In addition, because justice climate emerges by team member interactions (Colquitt et al., 2002; Naumann & Bennett, 2000; Roberson & Colquitt, 2005), team characteristics could moderate the CRT leader behavior influence on procedural justice climate perceptions and strength. For example, recent studies have confirmed the main effect of such team characteristics as team size and team collectivism on the level of procedural justice climate and strength (Colquitt et al., 2002). Also, in light of the leadership substitute literature (Howell, Dorfman, & Kerr, 1986), which proposes that team characteristics could serve as enhancers or neutralizers of leadership influence, we suggest that future research integrate leadership influences and team characteristics to test their possible interaction effect on individual work attitudes and work performance mediated by justice climate and justice climate strength. In conclusion, given recent interests in organizational justice, the question of how leaders can directly promote justice climate perceptions and strength in organizations is an important avenue for future research. Our findings provides an important step toward a better understanding of this process by suggesting that CRT leader behavior can be instrumental in creating strong justice climate perceptions, which in turn, promotes employee positive attitudes and behaviors. In doing so, we have also attempted to fill the dearth of research in the CRT leader behavior process and the antecedents and consequences of justice climate perceptions and strength. Acknowledgements We would like thank Leanne Atwater, Timothy Judge, and David Van Fleet for their insights and comments on this article. We would also like to thank Michael Mumford and the three anonymous reviewers whose comments greatly improved the quality of this paper. An earlier version of this manuscript was presented at the 2006 Academy of Management Meeting and appeared in the 2006 Academy of Management Best Paper Proceedings. References Allison, P. D. (1978). Measures of inequality. American Sociological Review, 43, 865−880. Arbuckle, J. L., & Wothke, W. (1999). Amos 4.0 users' guide. Chicago: Smallwaters Corporation. Atwater, L. E., Camobreco, J. F., Dionne, S. D., Avolio, B. J., & Lau, A. N. (1997). Effects of rewards and punishments on leader effectiveness and follower reactions. Leadership Quarterly, 8, 133−152. Avolio, B. J., Bass, B. M., Walumbwa, F. O., & Zhu, W. (2004). Multifactor leadership questionnaire: Manual and sampler test, (3rd ed.) Redwood City, CA: Mind Garden. Bachrach, D. G., & Jex, S. M. (2000). Organizational citizenship behavior and mood: An experimental test of perceived job breadth. Journal of Applied Social Psychology, 30, 641−663. Bass, B. M. (1998). Transformational leadership: Industry, military, and educational impact. Mahwah, NJ: Lawrence Erlbaum. Bass, B. M., & Avolio, B. J. (2000). Multifactor leadership questionnaire: Manual leader form, rater, and scoring key for MLQ (Form 5x-Short). Redwood City, CA: Mind Garden. Bass, B. M., Avolio, B. J., Jung, D. I., & Berson, Y. (2003). Predicting unit performance by assessing transformational and transactional leadership. Journal of Applied Psychology, 88, 207−218. Berson, Y., Shamir, B., Avolio, B. J., & Popper, M. (2001). The relationship between vision, strength, leadership style, and context. Leadership Quarterly, 12, 53−73. Blau, P. M. (1964). Exchange and power in social life. New York: Academic Press. Bliese, P. D. (2000). Within-group agreement, non-independence, and reliability: Implications for data aggregation and analysis. In K. J. Klein, & S. W. J. Kozlowski (Eds.), Multilevel theory, research and methods in organizations: Foundations, extensions, and new directions (pp. 349−381). San Francisco, CA: Jossey-Bass. Bono, J. E., & Judge, T. A. (2003). Self-concordance at work: Toward understanding the motivational effects of transformational leaders. Academy of Management Journal, 46, 554−571. Chan, D. (1998). Functional relations among constructs in the same content domain at different levels of analysis: A typology of compositional models. Journal of Applied Psychology, 83, 234−246. Cohen, S. G., Chang, L., & Ledford, G. E. (1997). A hierarchical construct of self-management leadership and its relationship to quality of work and perceived group effectiveness. Personnel Psychology, 50, 275−308. Cohen, J., & Cohen, P. (1983). Applied multiple regression/correlation analysis for the behavioral sciences, (2nd ed.) Hillsdale, NJ: Erlbaum. Colquitt, J. A. (2001). On the dimensionality of organizational justice: A construct validation of a measure. Journal of Applied Psychology, 86, 386−400. Colquitt, J. A. (2004). Does the justice of the one interact with the justice of the many? Reactions to procedural justice in teams. Journal of Applied Psychology, 89, 633−646. Colquitt, J. A., Noe, R. A., & Jackson, C. L. (2002). Justice in teams: Antecedents and consequences of organizational citizenship justice. Personnel Psychology, 55, 83−109. Colquitt, J. A., Zapata-Phelan, C., & Roberson, Q. M. (2005). In J. J. Martocchio (Ed.), Justice in teams: A literature review and agenda for future researchResearch in personnel and human resource management, Vol. 24. (pp. 53−94) Oxford, UK: Elsevier Science. Costigan, R. D., Insinga, R. C., Kranas, G., Kureshov, V., & Ilter, S. S. (2004). Predictors of employee trust of their CEO: A three-country study. Journal of Managerial Issues, 16, 197−216. Cropanzano, R., Byrne, Z. S., Bobocel, D. R., & Rupp, E. R. (2001). Moral virtues, fairness heuristics, social entities, and other denizens of organizational justice. Journal of Vocational Behavior, 58, 164−209. Dansereau, F., Alutto, J. A., & Yammarino, F. J. (1984). Theory testing in organizational behavior: The variant approach. Englewood Cliffs, NJ: Prentice-Hall. De Cremer, D., van Knippenberg, B., van Knippenberg, D., Mullenders, D., & Stinglhamber, F. (2005). Rewarding leadership and fair procedures as determinants of self-esteem. Journal of Applied Psychology, 90, 3−12. Ehrhart, M. G. (2004). Leadership and procedural justice climate as antecedents of unit-level of organizational citizenship behavior. Personnel Psychology, 57, 61−94. Eisenberger, R., Huntington, R., Hutchison, S., & Sowa, D. (1986). The perceived organizational support. Journal of Applied Psychology, 71, 500−507. Eisenberger, R., Stinglhamber, F., Vandenberghe, C., Sucharski, I. L., & Rhoades, L. (2002). Perceived supervisor support: Contributions to perceived organizational support and employee retention. Journal of Applied Psychology, 87, 565−573.

264

F.O. Walumbwa et al. / The Leadership Quarterly 19 (2008) 251–265

Gavin, M. B., & Hofmann, D. A. (2002). Using hierarchical linear modeling to investigate the moderating influence of leadership climate. The Leadership Quarterly, 13, 15−22. Glick, W. H. (1985). Conceptualizing and measuring organizational and psychological climate: Pitfalls in multilevel research. Academy of Management Review, 10, 601−616. Gouldner, A. W. (1960). The norm of reciprocity: A preliminary statement. American Sociological Review, 25, 161−178. Graen, G. B., & Uhl-Bien, M. (1995). Development of leader–member exchange (LMX) theory of leadership over 25 years: Applying a multi-level multi-domain perspective. The Leadership Quarterly, 6, 219−247. Hofmann, D. (1997). An overview of the logic and rationale of hierarchical linear models. Journal of Management, 6, 723−744. Hofmann, D. A., & Gavin, M. B. (1998). Centering decisions in hierarchical linear Models: Implications for research in organizations. Journal of Management, 24, 623−641. Hofmann, D. A., & Jones, L. M. (2005). Leadership, collective personality, and performance. Journal of Applied Psychology, 90, 509−522. Hogg, M. A. (2001). A social identity theory of leadership. Personality and Social Psychology Review, 5, 184−200. House, R. J., & Aditya, R. N. (1997). The social scientific study of leadership: Quo vadis. Journal of Management, 23, 409−473. Howell, J. M., & Avolio, B. J. (1993). Transformational leadership, transactional leadership, locus of control, and support for innovation: Key predictors of consolidated-business-unit performance. Journal of Applied Psychology, 78, 891−902. Howell, J. P., Dorfman, P. W., & Kerr, S. (1986). Moderator variables in leadership research. Academy of Management Review, 11, 88−102. Howell, J. N., & Hall-Merenda, K. E. (1999). The ties that bind: The impact of leader–member exchange, transformational leadership and transactional leadership, and distance on predicting follower performance. Journal of Applied Psychology, 84, 680−694. James, L. R. (1980). The unmeasured variables problem in path analysis. Journal of Applied Psychology, 65, 415−421. James, L. R., Demaree, R. G., & Wolf, G. (1984). Estimating within-group inter-rater reliability with and without response bias. Journal of Applied Psychology, 69, 85−98. Johnson, R. E., Selenta, C., & Lord, R. (2006). When organizational justice and the self-concept meet: Consequences for the organization and its members. Organizational Behavior and Human Decision Processes, 99, 175−201. Judge, T. A., & Colquitt, J. A. (2004). Organizational justice and stress: The mediating role of work–family conflict. Journal of Applied Psychology, 89, 395−404. Judge, T. A., & Ilies, R. (2004). Is positiveness in organizations always desirable? Academy of Management Executive, 18, 154−155. Judge, T. A., & Piccolo, R. F. (2004). Transformational and transactional leadership: A meta-analytic test of their relative validity. Journal of Applied Psychology, 89, 755−768. Judge, T. A., Piccolo, R. F., & Ilies, R. (2004). The forgotten ones: The validity of consideration and initiating structure in leadership research. Journal of Applied Psychology, 89, 36−51. Katerberg, R., & Hom, P. W. (1981). Effects of within-group and between-group variations in leadership. Journal of Applied Psychology, 66, 218−223. Kenny, D. A., Kashy, D. A., & Bolger, N. (1998). Data analysis in social psychology. In D. T. Gilbert, S. T. Fiske, & G. Lindzey (Eds.), The handbook of social psychology (pp. 233−265)., 4th ed. Boston: McGraw-Hill. Kim, H., & Yukl, G. (1995). Relationships of managerial effectiveness and advancement to self-reported and subordinate-reported leadership behaviors from the multiple-linkage mode. The Leadership Quarterly, 6, 361−371. Konovsky, M. A. (2000). Understanding procedural justice and its impact on business organizations. Journal of Management, 26, 489−511. Korsgaard, M. A., Schweiger, D. M., & Sapienza, H. J. (1995). Building commitment, attachment, and trust in strategic decision-making teams: The role of procedural justice. Academy of Management Journal, 38, 60−84. Kozlowski, S. W. J., & Klein, K. J. (2000). A multilevel approach to theory and research in organizations: Contextual, temporal, and emerging processes. In K. J. Klein, & S. W. J. Kozlowski (Eds.), Multilevel theory, research and methods in organizations: Foundations, extensions, and new directions (pp. 3−90). San Francisco: JosseyBass. Leventhal, G. S. (1980). What should be done with equity theory? New approaches to the study of fairness in social relationship. In K. J. Gergen, M. S. Greenberg, & R. H. Willis (Eds.), Social exchange: Advances in theory and research (pp. 27−55). New York: Plenum. Liao, H., & Rupp, D. E. (2005). The impact of justice orientation on work outcomes: A cross-level multifoci framework. Journal of Applied Psychology, 90, 242−256. Lind, E. A., & Earley, P. C. (1992). Procedural justice and culture. International journal of Psychology, 27, 227−242. Lind, E. A., & Tyler, T. R. (1988). The social psychology of procedural justice. New York: Plenum Press. Lindell, M. K., & Brandt, C. J. (2000). Climate quality and climate consensus as mediators of the relationship between organizational antecedents and outcomes. Journal of Applied Psychology, 85, 331−348. Lowe, K. B., Kroeck, K. G., & Sivasubramaniam, N. (1996). Effectiveness correlates of transformational and transactional leadership: a meta-analytical review of the literature. The Leadership Quarterly, 7, 385−425. Meyer, J. P., & Allen, N. J. (1997). Commitment in the workplace: Theory, research, and application. Thousand Oaks, CA: Sage. Morgeson, F. P., & Hoffman, D. A. (1999). The structure and function of collective constructs: Implications for multilevel research and theory development. Academy of Management Review, 24, 249−265. Mossholder, K. W., Bennett, N., & Martin, C. (1998). A multilevel analysis of procedural justice context. Journal of Organizational Behavior, 19, 131−141. Mowday, R. T., Steers, R. M., & Porter, L. W. (1979). The measurement of organizational commitment. Journal of Vocational Behavior, 14, 224−247. Naumann, S. E., & Bennett, N. (2000). A case for procedural justice climate: Development and test of multilevel model. Academy of Management Journal, 43, 881−889. O'Reilly, C. A., & Chatman, J. A. (1986). Organizational commitment and psychological attachment: The effects of compliance, identification, and internalization in prosocial behavior. Journal of Applied Psychology, 71, 492−499. Organ, D. W. (1988). Organizational citizenship behavior: The good soldier syndrome. Lexington, MA: Lexington Books. Pillai, R., Schriesheim, C. A., & Williams, E. S. (1999). Fairness perceptions and trust as a mediator for transformational and transactional leadership: A two sample study. Journal of Management, 25, 897−933. Podsakoff, P. M., Bommer, W. H., Podsakoff, N. P., & MacKenzie, S. B. (2006). Relationships between leader reward and punishment behavior and subordinate attitudes, perceptions, and behaviors: A meta-analytic review of existing and new research. Organizational Behavior and Human Decision Processes, 99, 113−142. Podsakoff, P. M., MacKenzie, S. C., Lee, J., & Podsakoff, N. P. (2003). Common method biases in behavioral research: A critical review of the literature and recommended remedies. Journal of Applied Psychology, 88, 879−903. Podsakoff, P. M., MacKenzie, S. B., Moorman, R. H., & Fetter, R. (1990). Transformational leader behaviors and their effects on followers' trust in leader, satisfaction, and organizational citizenship behaviors. The Leadership Quarterly, 1, 107−142. Podsakoff, P. M., MacKenzie, S. B., Paine, J. B., & Bachrach, D. G. (2000). Organizational citizenship behaviors: A critical review of the theoretical and empirical literature and suggestions for future research. Journal of Management, 26, 513−563. Raudenbush, S. W., & Bryk, A. S. (2002). Hierarchical linear models: Applications and data analysis methods, 2nd ed. Thousand Oaks, CA: Sage. Riordan, C. M., Griffith, R. W., & Weatherly, E. W. (2003). Age and work-related outcomes: The moderating effects of status characteristics. Journal of Applied Social Psychology, 33, 37−57. Roberson, Q. M. (2006). Justice in teams: The activation and role of sensemaking in the emergence of justice climates. Organizational Behavior and Human Decision Processes, 100, 177−192. Roberson, Q. M., & Colquitt, J. A. (2005). Shared and configural justice: A social network model of justice in teams. Academy of Management Review, 30, 595−607. Rupp, E. R., & Cropanzano, R. (2002). The mediating effects of social exchange relationships in predicting workplace outcomes from multifoci organizational justice. Organizational Behavior and Human Decision Processes, 89, 925−946. Rupp, D. E., Bashshur, M. R., & Liao, H. (2007). Justice climate past, present, and future: Models of structure and emergence. In F. Dansereau, & F. Yammarino (Eds.), Research in Multilevel Issues Vol. 6. (pp. 46−64). Oxford, England: Elsevier. Scandura, T. A. (1999). Rethinking leader-member exchange: An organizational justice perspective. Leadership Quarterly, 10, 25−40.

F.O. Walumbwa et al. / The Leadership Quarterly 19 (2008) 251–265

265

Schneider, B., Gunnarson, A. K., & Niles-Jolly, K. (1994). Creating the climate and culture success. Organizational Dynamics, 23, 17−29. Schneider, B., & Reichers, A. E. (1983). On the etiology of climates. Personnel Psychology, 36, 19−39. Schneider, B., Salvaggio, A. N., & Subirats, M. (2003). Climate strength: A new direction for climate research. Journal of Applied Psychology, 87, 220−229. Schriesheim, C. A., Castro, S. L., Zhou, X., & DeChurch, L. A. (2006). An investigation of path-goal and transformational leadership theory prediction at the individual level of analysis. Leadership Quarterly, 17, 21−38. Simons, T., & Roberson, Q. (2003). Why managers should care about fairness: The effects of aggregate justice perceptions on organizational outcomes. Journal of Applied Psychology, 88, 432−443. Smith, P. C., Kendall, L. M., & Hulin, C. L. (1969). The measurement of satisfaction in work and retirement. Chicago: Rand-McNally. Smith, C. A., Organ, D. W., & Near, J. P. (1983). Organizational citizenship behavior: Its nature and antecedents. Journal of Applied Psychology, 68, 653−663. Sobel, M. E. (1988). Direct and indirect effects in linear structural equation models. In J. S. Long (Ed.), Common problems/proper solutions (pp. 46−64). Beverly Hills, CA: Sage. Thibaut, J., & Walker, L. (1975). Procedural justice: A psychological analysis. Hillsdale, NJ: Erbaum. Tyler, T., Degoey, P., & Smith, H. (1996). Understanding why the justice of group procedures matters: A test of the psychological dynamics of the group-value model. Journal of Personality and Social Psychology, 70, 913−930. van Knippenberg, D., van Knippenberg, B., De Cremer, D., & Hogg, M. A. (2004). Leadership, self, and identity: A review and research agenda. The Leadership Quarterly, 15, 825−856. Vecchio, R. P. (1982). A further test of leadership effects due to between-group variation and within-group variation. Journal of Applied Psychology, 67, 200−208. Vroom, V. H. (1964). Work and motivation. New York: Wiley. Walumbwa, F. O., Wu, C., & Ojode, L. A. (2004). Gender and instructional outcomes: The mediating effects of leadership styles. Journal of Management Development, 23, 124−140. Wayne, S. J., Shore, L. M., Bommer, W. H., & Tetrick, L. E. (2002). The role of fair treatment and rewards in perceptions of organizational support and leader-member exchange. Journal of Applied Psychology, 87, 590−598. Yammarino, F. J., & Bass, B. J. (1990). Long-term forecasting of transformational leadership and its effects among naval officers: Some preliminary findings. In K. E. Clark, & M. B. Clark (Eds.), Measures of leadership (pp. 151−171). West Orange, NJ: Leadership Library of America. Yammarino, F. J., Dionne, S. D., Chun, J. A., & Dansereau, F. (2005). Leadership and levels of analysis: A state-of-the-science review. The Leadership Quarterly, 16, 879−919. Yammarino, F. J., Spangler, W. D., & Dubinsky, A. J. (1998). Transformational leadership and contingent reward leadership: Individual, dyad, and group levels of analysis. The Leadership Quarterly, 9, 27−54. Zaccaro, S. J., Rittman, A. L., & Marks, M. A. (2001). Team leadership. The Leadership Quarterly, 12, 451−483.