Leader Personality Traits and Employee Voice Behavior: Mediating Roles of Ethical Leadership and Work Group Psychological Safety

Journal of Applied Psychology 2009, Vol. 94, No. 5, 1275–1286 © 2009 American Psychological Association 0021-9010/09/$12.00 DOI: 10.1037/a0015848 Le...
Author: Buddy Francis
4 downloads 0 Views 157KB Size
Journal of Applied Psychology 2009, Vol. 94, No. 5, 1275–1286

© 2009 American Psychological Association 0021-9010/09/$12.00 DOI: 10.1037/a0015848

Leader Personality Traits and Employee Voice Behavior: Mediating Roles of Ethical Leadership and Work Group Psychological Safety Fred O. Walumbwa

John Schaubroeck

Arizona State University

Michigan State University

The antecedents and consequences of ethical leadership were examined in a study of 894 employees and their 222 immediate supervisors in a major financial institution in the United States. The leader personality traits of agreeableness and conscientiousness were positively related to direct reports’ ratings of the leader’s ethical leadership, whereas neuroticism was unrelated to these ratings. Ethical leadership influenced followers’ voice behavior as rated by followers’ immediate supervisors. This relationship was partially mediated by followers’ perceptions of psychological safety. Implications for research on ethical leadership and means to enhance ethical behavior among leaders and nonleaders are discussed. Keywords: ethical leadership, leadership, personality, psychological safety, voice

can be predicted using his or her personal characteristics. Identifying trait antecedents will aid in the development of strategies for selecting and developing ethical leaders and determining the best means to reinforce ethical behaviors. Only a few studies have addressed the consequences of ethical leadership behavior (Brown et al., 2005; Detert, Trevin˜o, Burris, & Andiappan, 2007; Mayer et al., 2009). Whereas some reliable evidence from recent studies supports the idea that ethical leadership has a range of favorable outcomes (e.g., Brown et al., 2005; Mayer et al., 2009), Detert et al. (2007) found no relationship between ethical leadership and food shrinkage, an index of counterproductive behavior among restaurant employees. Some psychological mechanisms that may explain the more favorable effects of ethical leadership have been discussed (see Brown & Trevin˜o, 2006), but little empirical attention has been directed toward understanding the psychological processes that may differentiate the behavior of followers of ethical leaders from that of followers of less ethical leaders. A clearer understanding of the mechanisms by which ethical leadership influences outcomes is not only needed for the practical concerns of selecting for, developing, and motivating ethical leadership; such information would also be valuable for determining whether the construct developed by Brown and Trevin˜o and their colleagues contributes something genuinely new to leadership research and practice. With these limitations of the extant literature in mind, we had three aims in the present study. First, we identified individual traits that were expected to influence ethical leadership. We chose to focus on how leader personality relates to follower ratings of the leader’s ethical behavior for two reasons. Brown et al. (2005) and Brown and Trevin˜o (2006) proposed that three personality traits— conscientiousness, agreeableness, and neuroticism—are plausible antecedents of ethical leadership. Personality antecedents may be uniquely suited to predicting ethical leadership, because ethical behavior reflects variation in individuals’ deep-seated values and beliefs; thus, ethical leadership should be a behavioral pattern that is very constant across situations and over time. Second, the present study contributes to the emerging theoretical and empirical research on ethical leadership by examining

In both the mass media and the academic community, there has been a surge in interest in the ethical and unethical behavior of leaders. Although the high-profile corporate scandals in recent years may explain much of the mass media and popular focus (Walumbwa, Avolio, Gardner, Wernsing, & Peterson, 2008), academics’ interest has been piqued by fresh evidence that ethical leadership behavior is associated with both positive and negative organizational processes (e.g., Brown, Trevin˜o, & Harrison, 2005; Mayer, Kuenzi, Greenbaum, Bardes, & Salvador, 2009) and outcomes (De Hoogh & Den Hartog, 2008). In this study, we sought to contribute to this body of knowledge by examining new antecedents and outcomes of ethical leadership. Brown et al. (2005, p. 120) defined ethical leadership as “the demonstration of normatively appropriate conduct through personal actions and interpersonal relationships, and the promotion of such conduct to followers through two-way communication, reinforcement, and decision-making” (for a review, see Brown & Trevin˜o, 2006). Brown et al. argued that ethical leaders not only inform individuals of the benefits of ethical behavior and the cost of inappropriate behavior; such leaders also set clear standards and use rewards and fair and balanced punishment to hold followers accountable for their ethical conduct (see also Trevin˜o, Brown, & Hartman, 2003). Despite the assumed importance and prominence of ethical leadership in organizations, there are still many questions relating to its antecedents and consequences (Brown et al., 2005). For example, researchers know very little about why some leaders engage in the spectrum of ethical leadership behaviors and others do not. One key question is whether the likelihood of an individual being perceived as an ethical leader among subordinates

Fred O. Walumbwa, Department of Management, W. P. Carey School of Business, Arizona State University; John Schaubroeck, The Eli Broad Graduate School of Management and Department of Psychology, Michigan State University. Correspondence concerning this article should be addressed to Fred O. Walumbwa, Department of Management, W. P. Carey School of Business, Arizona State University, Tempe, AZ 85287-4006. E-mail: [email protected] 1275

WALUMBWA AND SCHAUBROECK

1276

individual- and group-level outcomes that have been established to have important implications for work unit functioning: work group psychological safety (Edmondson, 1999) and employee voice behavior (LePine & Van Dyne, 1998). Finally, our study tested whether followers’ perceptions of psychological safety mediated the relationship between ethical leadership and voice behavior. We integrated these various factors in a theoretical model that we tested using data from distinct sources. Figure 1 summarizes the conceptual model examined in the present study.

Hypothesis 1: Ethical leadership is positively related to voice behavior.

Theory and Hypothesis Development Ethical Leadership, Psychological Safety, and Employee Voice Employee voice is defined as “promotive behavior that emphasizes expression of constructive challenge intended to improve rather than merely criticize” (Van Dyne & LePine, 1998, p. 109). It concerns the bottom-up process of rank-and-file employees making innovative suggestions for change and recommending modifications to standard procedures. Voice behavior is an important component of extrarole behavior (i.e., those positive and discretionary behaviors that are not required by the organization but that are necessary to facilitate effective organizational functioning; Organ, Podsakoff, & MacKenzie, 2006). Mayer et al. (2009) found that ethical leadership was a significant predictor of group-level helping behavior, which is a distinct domain of extrarole behavior. Like helping behavior, constructive voice behavior should be valued by leaders because it can reveal problems and solutions to problems as well as point to other ideas that may help work unit functioning (Van Dyne & LePine, 1998). One of the central predictions of ethical leadership theory is that ethical leaders “provide followers with voice” (Brown et al., 2005, p. 120). Ethical leaders speak out publicly against inappropriate organizational actions and behaviors and emphasize doing the right thing. From a social learning perspective (Bandura, 1977), when leaders proactively create a fair work environment, they become a target of emulation (Brown et al., 2005). Because ethical leaders convey high moral standards to employees, they encourage their followers to voice opinions and suggestions, not only about ethical matters but also about other work-related processes and work context. In support of this linkage, Brown et al. found that ethical leadership was significantly related to members’ willing-

Leader Agreeableness

Ethical Leadership

Leader Conscientiousness

Psychological Safety

Voice Behavior

Leader Neuroticism Supervisor self-ratings of personality (Time 1 and group-level)

Direct reports’ ratings of supervisor (Time 1 and grouplevel)

Figure 1.

Direct reports’ psychological safety (Time 2 and group-level weeks)

Hypothesized relationships.

ness to report problems to management. This is only one aspect of Van Dyne and LePine’s (1998) voice behavior construct. In addition, voice behavior includes expressing dissent when employees perceive that certain actions would be inappropriate or unethical, as well as sharing constructive ideas for work unit improvements even when problems have not surfaced. On this basis, we predicted that ethical leadership would promote voice behavior in work units.

Supervisor ratings of voice behavior (Time 2 and individual-level weeks)

Psychological safety refers to shared beliefs among work unit members that it is safe for them to engage in interpersonal risk taking (Edmondson, 1999). According to Edmondson, psychological safety goes beyond perceiving and experiencing high levels of interpersonal trust; it also describes a work climate characterized by mutual respect, one in which people are comfortable expressing their differences. Leaders are pivotal for removing the constraints that often discourage followers from expressing their concerns and other ideas. In environments characterized by high psychological safety, leaders also actively communicate the importance of such behavior and guarantee that it will not have negative repercussions for the individual or the work unit as a whole. In this respect, ethical leadership may be particularly important, and yet this function is underappreciated in the literature. Highly ethical leaders value honest and truthful relationships with their followers (Brown et al., 2005). They are seen to act according to their fundamental values and beliefs, rather than to respond to external pressures or narrow and transitory interests. When leaders interact with followers with openness and truthfulness, interpersonal trust and mutual respect is promoted both between followers and the leader and among the followers themselves. Providing a psychologically secure environment for employees can be a double-edged sword for leaders, however. Whereas favorable solutions may result, leaders often must acknowledge feedback that is not consistent with their plans and wishes. The leaders must allow dissent despite the pressures and complications it entails and resist the temptation to use power and authority when doing so may hinder perceived psychological safety. In addition, ethical leaders openly share information and demonstrate high personal moral standards. When individuals in authority share information and provide a rationale for the benefits of ethical behavior, interpersonal trust among followers is found to increase (Dirks & Ferrin, 2002). This occurs both through role modeling and through reassurance to subordinates that individual rent seeking, social undermining, and other behaviors that reduce trust among employees will not be tolerated. Similarly, when followers trust their leader has sufficient ability, benevolence, and integrity, which are key foundations of employee trust (Mayer, Davis, & Schoorman, 1995), they will be more comfortable about engaging in interpersonal risk taking because they trust that the leader will not unfairly punish them when risk taking leads to an unfavorable outcome (see also Dasborough & Ashkanasy, 2002; Weierter, 1997). This suggests a positive relationship between ethical leadership and psychological safety. We predicted that this linkage with psychological safety would explain the relationship between ethical leadership and voice behavior. According to social exchange theory (Blau, 1964), people develop

LEADER PERSONALITY TRAITS AND EMPLOYEE VOICE BEHAVIOR

trusting or transactional relationships based upon their experiences with others (Cropanzano & Mitchell, 2005). Relationships may be based on economic exchange or social exchange (Bishop, Scott, & Burroughs, 2000). Economic exchange relationships are transactional in nature; they emphasize short-term interactions and quid pro quo exchanges in which obligations are circumscribed to a particular exchange. Social exchanges involve a shared identity, loyalty, and emotional connections. A key feature of social exchange theory is that the quality of the relationship between the parties is the most proximal cause of behavior. In other words, people choose their actions, in large measure, on the basis of the type of attachment they have with the other person. Brown et al. (2005) argued that ethical leadership promotes beneficial employee behavior, such as interpersonal helping. When employees are treated fairly by a leader they trust, they are likely to think about their relationship with the leader in terms of social exchange rather than economic exchange. One way to reciprocate for such treatment is to engage in constructive voice behavior. In addition, it stands to reason that individuals high in psychological safety perceive little risk to their own welfare in engaging in voice behavior. Some evidence for this relationship was provided by Tangirala and Ramanujam (2008). They found that stronger climates of procedural injustice were associated with the “silencing” of dissenting opinions among nurses. Although Tangirala and Ramanujam did not report on a measure of voice, such suppression of voice suggests a very low voice climate. Together, these arguments about the effects of ethical leadership on psychological safety and employee voice and the linkage between psychological safety and employee voice suggest the following hypothesis: Hypothesis 2: The positive relationship between ethical leadership and voice behavior is mediated by employees’ perceptions of psychological safety.

Agreeableness, Conscientiousness, Neuroticism, and Ethical Leadership For decades, researchers have debated whether personality traits are meaningful predictors of leader emergence or effectiveness (Judge, Bono, Ilies, & Gerhardt, 2002; Lord, DeVader, & Alliger, 1986; Zaccaro, 2007). As summarized by Judge et al. (2002, p. 770), “previous research notwithstanding, we have a relatively poor idea of not only which traits are relevant, but why.” Personality concerns aspects of an individual’s thoughts and behavior that are stable over time and relatively consistent across different situations (Roberts & DelVecchio, 2000). Subordinates are less likely to consider a leader ethical if he or she does not behave ethically and promote ethical behavior with a considerable degree of consistency over time and across situations. In other words, ethical leadership is not behavior that a leader takes on to fit a specific situation, as consideration and directiveness have been specified to do in situational theories of leadership (e.g., path– goal theory; House, 1971). Rather, ethical leadership is by definition constant and therefore requires a stable dispositional foundation in the leader. Below we explain how certain distinct personality traits, namely, agreeableness, conscientiousness, and neuroticism, may be expected to reliably predict ethical leader behavior. Agreeableness. Agreeableness refers to a tendency to be accommodative, cooperative, pleasant, trusting, and good-natured

1277

(Costa & McCrae, 1992). We expected that agreeableness would be positively related to ethical leadership for two reasons. First, individuals high on agreeableness are more likely to use constructive tactics to help others (Jensen-Campbell & Graziano, 2001), which is a key component of ethical leadership. Second, as Brown et al. (2005) suggested, individuals high on agreeableness are more concerned about proper and humane treatment of people. They are considerate, helpful, honest, decent, trustworthy, understanding, responsive to the needs and wishes of others, and generally likable (Costa & McCrae, 1992; Tobin, Graziano, Vanman, & Tassinary, 2000). These behaviors are important for being an ethical leader (Brown et al., 2005; Brown & Trevin˜o, 2006), and therefore a trait that predicts stability in such behavior is likely to contribute to perceived ethical leadership. There is also some empirical evidence linking agreeableness to leadership. Judge and Bono (2000) reported that, among various personality predictors in their metaanalysis, agreeableness was most strongly related to the idealized influence dimension of transformational leadership. Brown et al. later found agreeableness to be positively related to ethical leadership. One might alternatively anticipate that agreeableness could in some cases be a hindrance to a leader’s ethical judgment and ability to follow through with ethical intentions. As McClelland and Boyatzis (1982) noted, the need for affiliation can reduce leader effectiveness, because it can potentially undermine objectivity and reduce discretion in situations that require the leader to deviate from what is perceived to be most acceptable to stakeholders when it is necessary to do so. Maintaining an ethical stance often requires a leader to accept being at least briefly unpopular with coworkers and subordinates, and thus it may seem that the ethical orientation of a leader who is high on the agreeableness dimension of personality might be compromised more easily in these situations. However, judgments of ethical leadership are ultimately social evaluations made by a majority of the coworkers who come into contact with the leader (e.g., direct reports). Such majorities, however principled the individuals among them may be, will not always agree that actions that contravene their own interests are honorable based on moral precepts. Therefore, although highly principled ethical decisions represent ethical leadership, evaluations of ethical leadership are more often based on conventional moral thinking that is shared by others. From this perspective, agreeable leaders are more likely to be perceived as engaging in ethical leadership, even though their needs for affiliation may sometimes lead them to make less principled decisions when a more principled decision would go against the grain of popular opinion. Thus, we predicted a positive relationship between agreeableness and ethical leadership on the basis of the behavioral tendencies within the domain of agreeableness that are more consistent with conventionally defined ethical behavior (e.g., interpersonal sensitivity, generosity, trust). Hypothesis 3: Leader agreeableness is positively related to ethical leadership. Conscientiousness. Among other five-factor model personality constructs, conscientiousness has been one of the most commonly studied traits in work psychology (Bono & Judge, 2004). A considerable amount of research indicates that it is one of the best predictors of performance in the workplace (Barrick & Mount,

1278

WALUMBWA AND SCHAUBROECK

1991). Conscientious individuals experience a high degree of moral obligation; they value truth and honesty, are less easily corrupted by others, and maintain a high regard for duties and responsibilities. Individuals high on conscientiousness display traits such as achievement striving, carefulness, dependability, self-discipline, thoroughness, responsibility, deliberation, and persistence (Costa & McCrae, 1992). These tendencies suggest a link between conscientiousness and the behavior pattern that is required to be perceived as an ethical leader. Ethical leaders engage in what Brown et al. (2005, p. 120) termed “normatively appropriate behaviors”; these behaviors include openness and honesty, reliability and truthfulness—the hallmarks of conscientiousness. In addition, because conscientious individuals are goal and detail oriented and typically are well organized, ambitious, and strongwilled (Hogan & Ones, 1997), they may be more likely to communicate clear principles and standards for ethical behavior to their subordinates (Brown & Trevin˜o, 2006). Dollinger and LaMartina (1998) found that conscientiousness was positively associated with higher levels of moral reasoning. Although their finding is indirectly suggestive of a possible link between personality and ethical leadership behavior, we are aware of no previous research that has examined conscientiousness as an antecedent of actual perceived ethical leadership behavior. We therefore advance the following hypothesis: Hypothesis 4: Leader conscientiousness is positively related to ethical leadership. Neuroticism. Neuroticism refers to the tendency to have a negativistic cognitive style and to focus on self-perceptions that are unfavorable. People who score high on neuroticism tend to experience negative emotions, such as anxiety and anger, more often and more intensely. Brown and Trevin˜o (2006) suggested that a more neurotic individual is less likely to be perceived as an ethical leader, because he or she will tend to be “thin-skinned and hostile toward others” (p. 603) and these traits make it difficult for a leader to be an effective role model. Recent research seems to bear out this relationship. Neurotics are more defensive in their social interactions (Barrett & Pietromonaco, 1997) and are prone to interpersonal conflicts (Brissette & Cohen, 2002). Neuroticism converges at very high levels with measures of the broader construct of trait negative affect. Judge et al. (2002) reviewed studies and found that high trait negative affect individuals tend not to emerge as leaders. Moreover, those high trait negative affect individuals who are leaders tend to be rated as less effective than low trait negative affect leaders. Lim and Ployhart (2004) similarly observed that military leaders with high trait negative affect were perceived as uninspiring and unstimulating by their followers. It is therefore plausible that leaders who score higher on neuroticism are likely to be rated lower on ethical leadership, because ethical leadership requires an ability to communicate in a manner that engages followers and influences them to take seriously communications about proper conduct and sanctions for ethical and unethical behavior. Hypothesis 5: Leader neuroticism is negatively related to ethical leadership.

Leader Personality Influences on Employee Voice Our study also examined the indirect effect of leader personality on employees’ psychological safety and voice behavior. Van Dyne and LePine (1998) found that leader agreeableness, which we predicted would be an antecedent of ethical leadership (Hypothesis 3), was positively associated with follower voice behavior. They theorized that this effect could be explained by social exchange motives that are positively linked to ethical leadership. The leader agreeableness– employee voice relationship might be explained more fully by examining the mediating role of ethical leadership and group members’ psychological safety. That is, having a highly agreeable leader seems insufficient for promoting strong norms for voice behavior. The agreeableness trait is associated with leader consideration, but leader consideration alone does not provide the safeguards workers need to experience the psychological safety that is critical for voice behavior. Whereas ethical leaders will by definition be considerate in nature, not all considerate leaders will exhibit the full range of behaviors that meet the criteria of ethical leadership as defined by Brown et al. (2005). Whereas some work groups reporting to considerate, supportive leaders would see the exchange benefits of providing voice, they might be deterred from doing so in some cases because the leader was not manifestly trustworthy or was not able to gratefully accept bad news or suggestions. Therefore, we predicted that whereas leader agreeableness positively influences voice behavior, this effect is indirect and would occur only through the mediating influences of ethical leadership and group members’ psychological safety. Although we are not aware of published research or theory linking leaders’ conscientiousness or neuroticism to psychological safety or to voice behavior, our arguments pertaining to the linkages between these traits and ethical leadership could be applied to link them to these outcomes. For example, the tendency of conscientious individuals to display “normatively appropriate behaviors” (Brown et al., 2005, p. 120) should give employees confidence that their leader will tolerate dissent and allow a fair hearing of diverse observations and viewpoints. Likewise, the tension surrounding neurotic leaders may make their followers hesitant to provide unsolicited input. Thus, to the extent that an individual is perceived to be an ethical leader, such traits should not further enhance (as with leader conscientiousness) or inhibit (as with leader neuroticism) voice behavior. Hypothesis 6: Leader personality (i.e., agreeableness, conscientiousness, neuroticism) is indirectly related to employee voice behavior through the mediating influence of ethical leadership and, in turn, group members’ psychological safety.

Method Participants and Procedures Data for the present study came from 894 employees and their 222 immediate supervisors in a large financial institution, located in the southwestern United States, as part of a leadership development training program. Following George (1990), we considered employees to be members of a work group when they had the same supervisor. Thus, some departments had multiple work groups as defined by supervisor (at least 4 employees per group). The supervisors were, on average, 36 years of age; 54% were

LEADER PERSONALITY TRAITS AND EMPLOYEE VOICE BEHAVIOR

women, 42% were Whites, 38% were Hispanics, and 20% were “other.” The employees were, on average, 27 years of age; 68% were women, 37% were White, 43% were Hispanic, and 20% identified with another ethnic group (e.g., Asian, African American) or did not answer this question. All participants had at least a high school education and had been employed by the bank for at least 6 months. Employees worked in areas that included customer service (e.g., tellers), credit administration, mortgage, retail, personal banking, wealth administration, money markets, operations, new accounts, and commercial loans. Data from both supervisors and direct reports were collected at two points in time separated by approximately 5 weeks. At Time 1, both supervisors and direct reports received a survey packet that contained a cover letter from the researchers requesting their participation. The cover letter explained the purpose of the study, provided assurance of confidentiality, informed participants that they would be receiving another short survey in approximately 5 weeks, and stressed the importance of responding to both surveys. An endorsement from senior management was provided in an accompanying letter. In accordance with the institutional review board’s protocol, we provided written assurances that individual responses would be kept strictly confidential, that only aggregated data would be reported to the bank, and that employees were free to decide not to participate in the study or to terminate their participation at any time without questions or consequences. Employees were provided the option of sending their responses directly to the researchers in a prestamped envelope. Supervisors were given a questionnaire that assessed their personality (e.g., agreeableness, conscientiousness, neuroticism) and were asked about their personal information (e.g., gender). They were asked to indicate their work unit, so that we could match their responses with data to be collected at Time 2. Direct reports were asked to rate their immediate supervisor’s ethical leadership behavior and to provide personal information, including their names, so that we could match data at Time 2. On average, each supervisor was rated by at least 4 direct reports. Both supervisors and their direct reports were given 3 days to complete the survey, which was distributed on-site with the help of the human resources department. The completed surveys were collected by the researchers in sealed envelopes at each bank branch facility. In total, 300 surveys were distributed to supervisors and 1,400 surveys were distributed to direct reports. Surveys were completed by 240 supervisors and 959 direct reports. This translates to response rates of 80% and 69% for the supervisor and direct report sampling frames, respectively. Approximately 5 weeks later, supervisors and direct reports who had completed the Time 1 surveys were given a second survey. Supervisors were given the names of members of their respective work groups and were asked to rate the extent to which each member of the work group demonstrated voice behavior. Similarly, direct reports were asked to complete a measure of psychological safety. Once again, both supervisors and direct reports were asked to provide their names, so we could match data from Time 1. Complete data (Times 1 and 2) were provided by 222 supervisors and 894 direct reports. This represents overall response rates of 74% and 64% for supervisors and subordinates, respectively.

1279

Measures Personality traits of agreeableness, conscientiousness, and neuroticism. We measured agreeableness using a 5-item scale from Goldberg’s (1990) Big Five Inventory (see also Rubin, Munz, & Bommer, 2005). We used 16 items from a brief version of Goldberg’s unipolar Big-Five markers (Saucier, 1994) to measure conscientiousness (8 items) and neuroticism (8 items). The supervisors rated their agreement with the 21 statements (1 ⫽ strongly disagree to 5 ⫽ strongly agree). Sample items include “I have a soft heart,” “I take time out for others,” “I feel others’ emotions” (agreeableness); “I am organized,” “I am efficient,” “I am systematic” (conscientiousness); and “I am moody,” “I am jealous,” “I am envious” (neuroticism). The internal consistency reliabilities (alpha) were .73, .84, and .79 for agreeableness, conscientiousness, and neuroticism, respectively. These reliabilities are similar to those reported in past research (e.g., Liao & Chuang, 2004; Lim & Ployhart, 2004). Ethical leadership. We assessed ethical leadership behavior using the 10-item Ethical Leadership Scale (ELS) developed by Brown et al. (2005). The items (␣ ⫽ .90) are each anchored on a 5-point continuum (1 ⫽ strongly disagree to 5 ⫽ strongly agree). One item (i.e., “conducts his or her personal life in an ethical manner”) was not included in the ELS mean score used for the hypothesis testing because more than a half of subordinates declined to respond to this item. In a space allocated for open-ended responses, some respondents indicated that they perceived this particular item as being inappropriate, difficult to answer, or not in keeping with the corporate culture. Sample ELS items are “Has the best interests of employees in mind,” “Makes fair and balanced decisions,” “Disciplines employees who violate ethical standards,” and “Defines success not just by results but also the way that they are obtained.” Because we seek to explain the work unit interaction behaviors that are strongly influenced by leadership norms and values, we aggregated individual employees’ ratings of ethical leadership to the group level of analysis. Psychological safety. Employees’ perception of psychological safety were measured with seven items (␣ ⫽ .86) taken from Edmondson (1999). Direct reports were instructed to indicate their level of agreement with each of the items on a 5-point scale (1 ⫽ strongly disagree to 5 ⫽ strongly agree). Sample items include “Employees of this work group are able to bring up problems and tough issues” and “Employees of this work group sometimes reject others for being different.” Because psychological safety refers to a shared perception among group members, we aggregated individual employees’ perceptions of safety to the group level to form the measure of psychological safety, as has prior research that has theorized and tested the construct at the work-unit level of analysis (e.g., Edmondson, 1999). Voice behavior. We measured employees’ voice behavior using a six-item scale (␣ ⫽ .81) developed by Van Dyne and LePine (1998). Supervisors provided ratings of each of their respective direct reports’ voice behavior on a 5-point response scale (1 ⫽ strongly disagree to 5 ⫽ strongly agree). Sample items include “[This employee] speaks up and encourages others to get involved in issues that affect this work unit,” “speaks up with ideas for new projects or changes in procedures,” “develops and makes recommendations concerning issues that affect this work group,” and

1280

WALUMBWA AND SCHAUBROECK

“gets involved in issues that affect the quality of work life in this work group.” Control variables. We controlled for leader’s span of control, because larger spans of control can diminish a leader’s ability to influence followers (Rubin et al., 2005). The range of span of control was 4 to 14 direct reports. We also controlled for idealized influence leadership. As noted above, idealized influence is a dimension of transformational leadership behavior that was found to correlate with ethical leadership to a substantial degree (Brown et al., 2005). We measured this construct using four items (␣ ⫽ .80) from the Multifactor Leadership Questionnaire (Bass & Avolio, 2004) anchored on a 5-point scale (0 ⫽ not at all to 4 ⫽ frequently, if not always).

Analytic Strategy and Levels of Analysis Hypotheses 1 and 2 pertained to variables spanning both the individual level (e.g., employee voice behavior) and the group level (e.g., ethical leadership and psychological safety). In addition, each supervisor provided ratings of employees’ voice behavior for an average of 4 employees. Therefore, employees’ reports were nested within supervisory work groups. Because our data came from several groups from different branches of the bank, we performed three-level hierarchical linear modeling (HLM) analyses to decompose the variances of the study variables into withingroup, between-group, and between-branch components to lessen the concern that spurious observed group effects could have emerged due to variance across branches that was not modeled. The results revealed that none of the variables examined (e.g., ethical leadership, psychological safety, voice behavior) had significant between-branch variance. Therefore, we concluded that the unmodeled branch effect did not warrant concern. To account for potential nonindependence effects and crosslevel effects, we used HLM (Raudenbush & Bryk, 2002). Each of the cross-level hypotheses was tested with an intercepts-asoutcomes model. Because we hypothesized effects of group-level variables (e.g., personality traits, leadership, psychological safety) on individual-level outcomes (e.g., voice behavior), we used grand mean centering in all the analyses. This technique helps to reduce the covariance between intercepts and slopes and thereby reduces potential problems associated with multicollinearity (Hofmann & Gavin, 1998). We utilized a structural equation modeling (SEM) approach to test Hypotheses 3–5, because leader personality, ethical leadership, and psychological safety were evaluated at the same level of analysis (e.g., group level). MacKinnon, Lockwood, Hoffman, West, and Sheets (2002) noted that a simultaneous test of the significance of the path from an initial variable to a mediator and the path from the mediator to an outcome (as suggested by SEM) provides, relative to other approaches, the best balance of Type I error rates and statistical power. Hypothesis 6 consisted of variables spanning both the individual level (e.g., voice behavior) and the group level (e.g., personality traits, ethical leadership, psychological safety). To test this hypothesis, we needed to establish the mediating effect of ethical leadership in the relationship between leader personality traits and psychological safety (e.g., leader personality 3 ethical leadership 3 psychological safety). As the latter are all group-level variables, we utilized SEM to test this first part of the hypothesis. The second

part of Hypothesis 6 refers to a cross-level effect, with leader personality traits influencing ethical leadership, which in turn influences group members’ psychology safety and, consequently, supervisory rated employee voice behavior (e.g., leader personality 3 ethical leadership 3 psychological safety 3 voice behavior). We utilized HLM to examine the group-level effects of leader personality, ethical leadership, and psychological safety on individual-level voice behavior.

Results Validity of Measures ELS. Because ethical leadership was measured with 9 items rather than the original 10 items, we conducted analyses to determine whether there might be a systematic bias that led participants not to complete the item in question. First, we created a dummycoded variable representing whether the participant had responded or not responded to this item (0 ⫽ no response, 1 ⫽ responded). Results revealed that this response variable was not related to employees’ perceptions of psychological safety, voice behavior, or the mean of the other 9 items on the ELS. In addition, within the subsample that responded to the “personal life” question, the item–total correlation was roughly equivalent to the other ELS items and the overall coefficient alpha reliability was .82. Next, we conducted a principal-axis factor analysis with varimax rotation. Results showed that a single, one-factor solution in which all the 9 items had high loadings (average loading ⫽ .87) on the single factor. This factor explained 69% of the total variance in the items. Psychological safety. Because psychological safety was measured with self-reports, we assessed the validity of this measure. We examined its dimensionality by conducting a principal-axis factor analysis with varimax rotation. This analysis obtained a one-factor solution in which all of the seven items had high loadings (average loading ⫽ .74) on the single factor. The factor explained 66% of the total variance in the items. We then examined the criterion-related validity of the psychological safety measure by examining its relationship to other measures with which it should be theoretically related. As reported in Table 1, the pattern of correlations was consistent with the nomological network. In particular, at the work-group level, psychological safety was significantly correlated with idealized influence leadership (r ⫽ .44), agreeableness (r ⫽ .41), conscientiousness (r ⫽ .37), ethical leadership (r ⫽ .46), and voice behavior (r ⫽ .44). Similar patterns of correlations were observed at the individual level with idealized influence leadership, ethical leadership, and supervisor-rated voice behavior. These relationships were consistent with theory and evidence concerning psychological safety. Finally, to assess the integrity of the overall measurement model, we conducted a confirmatory factor analysis at the group level that included all the constructs as latent variables and the group-aggregated item means as observed variables. The measurement model demonstrated a good fit to the data, ␹2(1007, N ⫽ 222)⫽ 1,674.47, p ⬍ .01, comparative fit index [CFI] ⫽ .94, root-mean-square residual [RMR] ⫽ .03, root-mean-square error of approximation [RMSEA] ⫽ .05.

Aggregation Issues We then examined the viability of aggregating individual scores of ethical leadership, idealized influence leadership, and psycho-

LEADER PERSONALITY TRAITS AND EMPLOYEE VOICE BEHAVIOR

1281

Table 1 Means, Standard Deviations, and Correlations Variable 1. 2. 3. 4. 5. 6. 7.

Idealized influence leadership Agreeableness Conscientiousness Neuroticism Ethical leadership Psychological safety Voice behavior

M

SD

1

2

3

4

5

6

2.79 3.80 3.87 2.58 3.41 3.57 3.62

0.42 0.43 0.38 0.57 0.48 0.37 0.34

— .28ⴱⴱ .17ⴱ ⫺.18ⴱⴱ .29ⴱⴱ .44ⴱⴱ .41ⴱⴱ

— .28ⴱⴱ ⫺.13ⴱ .43ⴱⴱ .41ⴱⴱ .43ⴱⴱ

— ⫺.07 .39ⴱⴱ .37ⴱⴱ .36ⴱⴱ

— ⫺.12 ⫺.02 ⫺.07

— .46ⴱⴱ .44ⴱⴱ

— .49ⴱⴱ

Note. Correlations are computed at the work unit level (n ⫽ 222). ⴱ p ⬍ .05 (two-tailed test). ⴱⴱ p ⬍ .01 (two-tailed test).

logical safety to the work-group level. This was done by calculating within-group agreement (rwg; James, Demaree, & Wolf, 1984), intraclass correlations (ICC[1]), and the reliability of the means (ICC[2]; Bliese, 2000). For idealized influence leadership, the average rwg was .77 (Mdn ⫽ .83), the ICC(1) was .36, and the ICC(2) was .74. For ethical leadership, the average rwg was .86 (Mdn ⫽ .90), the ICC(1) was .35, and the ICC(2) was .73. For psychological safety, the average rwg was .85 (Mdn ⫽ .88), the ICC(1) was .37, and the ICC(2) was .76. In addition, the analyses of variance on which the ICC(1) values were based all indicated that the group effect was significant ( p ⬍ .01). Taken together, these results indicate that it was statistically appropriate to analyze idealized influence leadership, ethical leadership, and psychological safety at the work-group level.

Hypothesis Tests Hypothesis 1 predicted that ethical leadership would be positively related to employees’ voice behavior. Before testing this hypothesis and other cross-level hypotheses, we examined whether there was significant systematic within- and between-work-group variance in supervisor-rated voice behavior. Results of a null

model, where we partitioned the total variance into within- and between-individuals components, revealed that 28% of variance in supervisor-rated voice behavior was within individuals; betweenindividual differences in average scores were significant and meaningful ( p ⬍ .01). The HLM results from testing Hypothesis 1 are displayed in Table 2 (Model 1). The results reveal that after we controlled for idealized influence leadership and span of control as Level 2 predictors, ethical leadership scores aggregated to the group level significantly predicted followers’ voice behavior. This supports the cross-level main effect of ethical leadership on employee voice behavior. Thus, Hypothesis 1 is supported by our data. Hypothesis 2 predicted that the positive relationship between ethical leadership and supervisor ratings of voice behavior would be mediated by followers’ perceptions of psychological safety. We followed the four-step procedure for testing mediation described by Kenny, Kashy, and Bolger (1998), once again controlling for idealized influence leadership and span of control as Level 2 predictors. These results are shown in Table 2 (Models 2–3). As a first step, ethical leadership should be related to voice behavior. This requirement is supported by results we reported

Table 2 Results of HLM Analyses Testing Cross-Level Effects Variable Intercept Span of control Idealized influence leadership Ethical leadership Intercept Span of control Psychological safety Intercept Span of control Idealized influence leadership Agreeableness Conscientiousness Neuroticism Ethical leadership Psychological safety

Voice behavior (Model 1) 3.62ⴱⴱ 0.01 0.23ⴱⴱ 0.35ⴱⴱ

Voice behavior (Model 2)

3.62ⴱⴱ 0.01 0.50ⴱⴱ

Voice behavior (Model 3)

Voice behavior (Model 4)

3.62ⴱⴱ ⫺0.01 0.17ⴱⴱ

3.63ⴱⴱ 0.00 0.03 0.03 0.06 0.02 0.19ⴱⴱ 0.12ⴱⴱ

0.18ⴱⴱ 0.34ⴱⴱ

Note. These are estimations of Level 2 fixed effects (␥s) with robust standard errors. HLM ⫽ hierarchical linear modeling. ⴱⴱ p ⬍ .01 (two-tailed tests).

WALUMBWA AND SCHAUBROECK

a Agreeableness .37 *

.50**

Conscientiousness

*

.29**

Ethical Leadership

-.07

.54**

Psychological Safety

-.1

1

-.01

Neuroticism

b

Agreeableness .37 *

.51**

Conscientiousness

.28**

-.07 -.05 1

above regarding Hypothesis 1. Step 2 requires that ethical leadership should be related to psychological safety. Because ethical leadership and psychological safety are group-level variables, it was appropriate for us to assess the effect of ethical leadership on psychological safety using an ordinary least squares regression analysis. The results revealed that after we controlled for idealized influence leadership and span of control, ethical leadership predicted psychological safety (␤ ⫽ .39, p ⬍ .01, adjusted R2 ⫽ .22). The third step requires that psychological safety should be related to employees’ voice behavior. As shown in Table 2 (Model 2), after we controlled for span of control (Level 2 predictor), psychological safety (Level 2 predictor) significantly predicted supervisor ratings of employees’ voice behavior, thus meeting the third requirement. In testing the fourth requirement, we included both ethical leadership and psychological safety as Level 2 predictors in the same HLM regression model, once again controlling for Level 2 idealized influence leadership and span of control. The results for this model, shown in Table 2 (Model 3), indicate that psychological safety was significantly related to supervisor ratings of employee voice behavior. Similarly, the effect of ethical leadership on supervisor ratings of employee voice behavior remained significant, but it was reduced in magnitude ( p ⬍ .01) compared with the same effect in Model 1. These results suggest that perceptions of psychological safety partially mediated the relationship between ethical leadership and supervisor ratings of employee voice behavior. A Sobel (1982) test confirmed that the indirect effect was significant (z ⫽ 2.86, p ⬍ .01). Hypotheses 3 and 4 predicted that leader agreeableness and conscientiousness, respectively, would be positively related to ethical leadership, whereas Hypothesis 5 predicted that leader neuroticism would be negatively related to ethical leadership. The SEM results testing these hypotheses are displayed in Figure 2a, ␹2(618, N ⫽ 222) ⫽ 1,047.55, p ⬍ .01, CFI ⫽ .90, RMR ⫽ .03, RMSEA ⫽ .06. As shown in Figure 2a, only agreeableness and conscientiousness were significantly related to ethical leadership, and these effects were in the predicted direction. Thus, Hypotheses 3 and 4 received support from our data, but Hypothesis 5, concerning neuroticism, was not supported. Hypothesis 6 predicted that leader personality (i.e., agreeableness, conscientiousness, and neuroticism) would be indirectly related to employee voice behavior through the mediating influence of ethical leadership and, in turn, group members’ psychological safety. Table 3 presents the total, direct, and indirect effects of leader personality traits on psychological safety. As shown in Table 3, leader agreeableness and conscientiousness were the only personality traits with significant total effects on psychological safety; leader neuroticism did not have a significant total effect on psychological safety. To further establish whether ethical leadership mediates the relationship between leader personality and psychological safety, we compared the fit of the hypothesized model shown in Figure 2a with an alternative model that added direct paths from the personality traits to psychological safety. Our mediation hypothesis would be supported if the fit of our hypothesized model is not improved by the addition of direct paths from the three personality traits to psychological safety. The fit of this alternative model (see Figure 2b) was almost identical to that of Figure 2a, ␹2(615, N ⫽ 222) ⫽ 1,037.90, p ⬍ .01, CFI ⫽ .91, RMR ⫽ .02, RMSEA ⫽ .06,

-.1

1282

.12 *

Ethical Leadership

.36**

Psychological Safety

.27** .05

Neuroticism

Figure 2. a: Structural equation modeling results (Hypotheses 3– 6a). b: Alternative model. Values shown represent ⴱⴱ p ⬍ .01.

and the chi-square difference test was not significant, ␹2(3, N ⫽ 222) ⫽ 9.65, ns. Under rules of model parsimony, we concluded that Figure 2a is a more parsimonious model that achieves the same fit level. Thus, ethical leadership fully mediates the leader personality traits–psychological safety relationships. Furthermore, Sobel (1982) tests revealed that the indirect effects of leader agreeableness and conscientiousness on psychological safety were significant (agreeableness, z ⫽ 3.60, p ⬍ .01; conscientiousness, z ⫽ 3.17, p ⬍ .01). To test the final part of Hypothesis 6 (e.g., leader personality 3 ethical leadership 3 psychological safety 3 voice behavior), we ran an HLM model in which leader personality traits, ethical leadership, psychological safety, and the two controls of idealized influence leadership and span of control, all as Level 2 predictors, all predicted employee voice behavior. These results are reported in Table 2. As shown in Model 4, only ethical leadership and psychological safety significantly predicted employee voice behavior, and the leader personality traits did not predict employee voice behavior. These results suggest that leader personality traits indirectly relate to employee voice behavior through the mediating influence of ethical leadership and, in turn, group members’ psychological safety. Therefore, Hypothesis 6 is supported.

Discussion Antecedents of Ethical Leadership Among the various leadership patterns investigated in the literature, personality may be particularly germane for understanding ethical leadership. Ethically-oriented behavior must be constant

LEADER PERSONALITY TRAITS AND EMPLOYEE VOICE BEHAVIOR

Table 3 Direct, Indirect, and Total Effects of Personality Traits on Ethical Leadership and Psychological Safety

Outcomes of Ethical Leadership

Effect Personality traits Agreeableness Ethical leadership Psychological safety Conscientiousness Ethical leadership Psychological safety Neuroticism Ethical leadership Psychological safety

Total

Direct

.37ⴱⴱ .20ⴱⴱ

.37ⴱⴱ

.29ⴱⴱ .16ⴱ

.29ⴱⴱ

⫺.11 ⫺.06

⫺.11

1283

Indirect

.20ⴱⴱ .16ⴱ ⫺.06

Note. Blank cells indicate data are not applicable. p ⬍ .05. ⴱⴱ p ⬍ .01.



across time and different situations in order for a leader to develop and maintain a reputation for ethical leadership. Agreeableness, conscientiousness, and neuroticism have been proposed as personality antecedents of ethical leader behavior (Brown & Trevin˜o, 2006). Whereas reports of supervisors about their own agreeableness and conscientiousness were positively associated with subordinates’ evaluations of their ethical leadership behavior, we did not find a significant relationship between neuroticism and ethical leadership. But it should be recognized that work context may also influence leaders’ ethical behavior. Brown and Trevin˜o (2006) suggested that ethical leadership becomes more salient to followers in ethical organizational climates, and differences in moral socialization processes within organizations may lead people with similar character traits to differ in ethical leadership later in their careers. It is important for researchers to investigate these possibilities, because the answers will determine whether it is practical for organizations to invest in developing ethical leaders or whether ethical leadership can only be enhanced by more carefully selecting for predictive traits. To fruitfully extend this program of investigation, researchers may examine the influence of distinct ethical climates on the emergence of ethical leaders. We studied a large organization with many fairly homogeneous branches, and we did not find reliable branch differences in means or effect sizes. Thus we suspect that there was little variation in ethical climate that might influence the levels of ethical leadership. Such variations in climate could be particularly relevant to explaining ethical leadership in other studies, however. Schminke, Ambrose, and Neubaum (2005) found that leaders’ moral development, as measured by the Defining Issues Test, was negatively related to instrumental ethical climate and positively related to law and code ethical climate across a range of organizational samples. Investigations of the interactionist underpinnings of ethical leadership, as proposed by Trevin˜o (1986), would be very valuable. As with Schminke et al. (2005), studies that examine interactionist processes could sample multiple organizations that may have distinct ethical climates by using multiple observations over an extended period of time. This approach would enable inferences about the influence of ethical climates on the development of ethical reputations among leaders as well as the impact of ethical leadership on the formation of ethical climates.

Ethical leadership influenced followers’ voice behavior through the mediating influence of followers’ perceptions of psychological safety. Until now, the literature has focused mainly on how ethical leaders reduce employees’ transgressions by role modeling ethical behavior, continually communicating with employees about matters of right and wrong, and utilizing the performance management system to reinforce ethical conduct (Brown & Trevin˜o, 2006). Brown and Trevin˜o (2006, p. 607) noted that “relationships between ethical leaders and their followers are likely to be characterized by social exchange” and that the trust and norms of reciprocity inherent in any social exchange relationship may be expected to promote prosocial behavior. We took this a step further by examining whether ethical leadership liberates employees to engage in more open forms of expression about work processes without fear of censorship or other adverse consequences. A psychologically safe environment is one in which employees feel secure in pointing out problems, new ideas, or suggestions that are intended for the benefit of the work unit as a whole. This is an area of behavior that extends beyond the transactional forms of social exchange characterized by Brown and Trevin˜o. Employees may indeed be more willing to help ethical leaders when they have a positive social exchange relationship with them. However, that type of process is consistent with other, distinctly different approaches to leadership, such as leader–member exchange (Dansereau, Graen, & Haga, 1975). Although there is little that is clearly inspirational or transformational about ethical leadership, our theoretical arguments and findings point to how ethical leadership may unlock creative potential and collective orientation in a manner more conceptually similar to charismatic and transformational leadership than has been considered previously. In addition, our findings show that ethical leadership has substantial effects on psychological safety and voice behavior, even when a separate leadership construct that has been shown to be related to the same outcomes in prior research (idealized influence) is incorporated into the analysis. The indirect influence of psychological safety in the relationship between ethical leadership and employees’ voice behavior only partially explained the total magnitude of this relationship. The direct effect of ethical leadership on voice behavior might be explained by the well-established proclivity of workers to comply with authority figures (Cialdini & Goldstein, 2004). As noted above, an important component of ethical leadership is the solicitation of employees’ observations of wrongdoing and of conditions that may facilitate unethical behavior. Ethical leaders reinforce such behavior through the performance management system. Whereas a climate of psychological safety should make employees more comfortable in expressing a wide range of information about the work unit, subordinates of more ethical leaders may feel more compelled to report transgressions, as they know that failing to do so could have consequences in subsequent personnel actions. This does not point to any missing mediating variable but rather to a proximal effect of ethical leadership on employee voice behavior.

Practical Implications Organizations can change employees’ personalities very little, if at all, and thus our findings concerning ethical leadership anteced-

1284

WALUMBWA AND SCHAUBROECK

ents speak mainly to implications for selection. Personality constraints may limit the extent to which ethical leadership can be developed in organizations through appropriate socialization and training. However, as Brown (2007) observed, the character of leaders is only a partial explanation for the transgressions of some leaders that have been widely reported in recent years. The domain of ethical leadership does not seem to require a highly engaging style as does transformational or charismatic leadership, and thus individuals with different personality types, such as introverts and extroverts, may all have the potential to be quite effective at ethical leadership. This possibility is consistent with the relatively small effects of personality on ethical leadership in this study. A practical implication is that developing climates and systems that are more supportive of ethical behavior will be a far more practical approach than relying on the selection system to identify and attract persons who can be developed into very ethical leaders. Nevertheless, we encourage additional research investigating personality antecedents of ethical leadership. Although we could not identify a rationale for studying the two remaining Big 5 personality traits, Openness and Extraversion, it may be that these or other traits facilitate the ability to convey ethical authority. One may intuit that leader agreeableness is advantageous for ethical leadership because it is associated with benevolence and a desire to be perceived as acting in the best interests of employees. The potential flip side is that the tendency toward a higher need for affiliation among agreeable leaders may make it more difficult for them to make the more principled ethical decisions that can sometimes be unpopular with many employees. Because the majority of persons tend toward a conventional level of moral conduct and reasoning, the followers who rate their leaders on ethical leadership are more likely to agree on behaviors that meet conventional moral criteria. Even among followers who are more prone to principled morality, opinions about behaviors that reflect highly principled leader behavior are likely to vary. Leaders should consider convention when deciding to respond to what they view as very difficult moral dilemmas or situations that engage personal moral principles that may distinguish them from most of their followers. Leaders tend to be biased self-perceivers who are likely to have inflated views of the ethical foundations of their actions (see Brown, 2007). Moreover, if leaders make decisions that are very ethical but that are not perceived as such by followers, the potential benefits of their ethical leadership may be undermined because their ethical reputations may suffer. Future research might profitably examine a more optimistic view of followers. Perhaps followers can “see through” the leader’s being excessively agreeable in morally ambiguous situations and distinguish genuine principled ethical leadership from the leader’s simply following others’ more conventional views, even if these are not highly principled. By encouraging employee voice, ethical leadership may not only develop a more positive and collaborative atmosphere in the work unit, it may also be associated with higher levels of task or process conflict without correspondingly high affective and relationship conflict. It seems likely that a key moderator of the effects of task and process conflict is the effectiveness of the supervisor in managing the conflict, channeling it toward productive outcomes, and minimizing residual relationship conflict. This may be a strength of ethical leaders. An environment in which members feel psychologically safe does not guarantee low interpersonal conflict;

moreover, that task and process conflict can adversely influence work process (e.g., De Dreu & Weingart, 2003). Therefore we encourage researchers to examine the relationship between ethical leadership and the management of conflict within work groups. In conjunction with the theoretical characteristics already associated with ethical leaders (Brown & Trevin˜o, 2006), the present findings should encourage researchers to consider whether the actions and communications of ethical leaders may allow considerable conflict about tasks and processes to develop among members and, also, whether such leaders facilitate swifter resolution of these conflicts and minimize the affectively intense relationship conflicts that often accompany them. Ethical leadership may be part of the solution to the need to balance the conflictual behavior of work groups with their tendencies toward excessive concurrence seeking. Whereas ethical leadership may enhance the conditions that afford high work unit cohesion, the contribution of ethical leadership to voice behavior suggests, at least in theory, that followers reporting to ethical leaders are less likely to experience groupthink. This is because the leader elicits input from employees and discourages self-censorship, confronts petty intellectual tyrannies that promote one-sided thinking (e.g., “mindguards”; Janis, 1972), and challenges assumptions. Much of the leadership literature argues that leaders’ focus should be on cultivating shared values to ensure work unit cohesion and limiting task and relationship conflict. Ethical leadership is more focused on establishing trust among followers, and its presence may enable groups to be more cohesive without the negative by-product of excessive concurrence seeking.

Limitations and Conclusions It is important to consider factors that may distinguish this research setting from others in such a way as to influence the plausible generalizability of the findings. One characteristic of this sample that may meaningfully distinguish it from some others was the fairly close interaction between supervisors and their subordinates. Brown and Trevin˜o (2006) suggested that such close contact with the supervisor may be critical for ethical leadership behavior to become salient to followers, and thus ethical leadership may have a weaker influence on subordinates’ psychological safety and voice behavior when subordinates see little of their supervisors from day to day. Conversely, Brown and Trevin˜o noted that day-to-day contact may make it more difficult for a leader to develop a reputation as an ethical leader. The closer one gets to a leader, the more obvious his or her blemishes may appear, however slight they may be. There are a suite of variables associated with social and physical distance that may influence the salience of ethical leadership (see Hill, Bartol, Tesluk, & Langa, 2009). Such factors warrant investigation as potential moderators. In addition, the personality of the leader may tend to have a greater influence on subordinates’ ratings of him or her as an ethical leader in more collectivistic societal cultures, such as are found in Asia. Branzei, Vertinsky, and Camp (2007) found that in individualistic cultures, people tend to bestow trust—which is an essential aspect of ethical leadership— on the basis of perceived ability and integrity, whereas collectivists trust others when they experience predictable, benevolent patterns of interaction with the other individual. We conceptualized the traits of agreeableness and conscientiousness as influencing ethical leadership primarily on the basis of such predictability and benevolence, but we tested these factors

LEADER PERSONALITY TRAITS AND EMPLOYEE VOICE BEHAVIOR

using a U.S. sample, which is likely to be more individualistic. Future research might compare the power of variables such as leader personality in predicting ethical leadership in individualistic and collectivistic cultural contexts. Like most studies of leadership, ours relied on questionnaire responses, and thus definitive inferences about internal validity are not possible. However, the constructs in the causal chain were separated by time (a 5-week interval) and by source (supervisor vs. direct report). This reduced the likelihood that priming or consistency artifacts inflated the observed relationships. For example, with supervisors reporting about their personality and direct reports reporting on ethical leadership, the effects of personality on ethical leadership were unlikely to be overstated. In addition, ethical leadership and psychological safety represent quite distinct domains that are hard to explain in terms of a response set bias on the part of direct reports. Finally, the effects of these variables on supervisors’ reports of their direct reports’ voice behavior are not plausibly tainted by common method variance. The fairly high level of similarity in the tasks conducted by the various work groups in the sample was advantageous, as it enabled us to hold constant a large number of potential contextual variables that might have influenced voice behavior. Nonetheless, as noted previously, it will be advantageous for researchers to identify and measure relevant contextual influences in more occupationally and organizationally diverse samples. Context-rich studies that examine models of the mutual influences of moral socialization, ethical leadership, and ethical climate over time (i.e., interactionist models such as proposed by Trevin˜o, 1986) will be valuable for understanding how ethical leadership emerges, how it influences behavior in organizations, and how organizations can accelerate this process. Likewise, Judge et al. (2002) suggested that the relationship between leader traits and behavior appears to be context dependent, in that they observed considerable variability in effect sizes across studies in their meta-analysis. Thus neuroticism, which was unrelated to ethical leadership in the current study, might nevertheless be inversely related to ethical leadership in other organizational contexts. In conclusion, the theory and evidence we presented advance the construct validation effort of the construct of ethical leadership that was initiated by Brown et al. (2005) by examining new antecedents (personality variables) and two outcomes (group members’ psychological safety and employee voice behavior). Together, these findings add to the accumulating evidence that ethical leadership has a genuine nomological niche within the leadership literature. The findings suggest that two personality characteristics (i.e., agreeableness and conscientiousness) reported by the leaders themselves independently predict the extent to which their subordinates view them as ethical leaders. Whereas this points to possible constraints on the development of ethical leadership in organizations, the effects are sufficiently small to allow for many other influences, such as socialization processes and organizational climate. Our findings also suggest that ethical leadership should not be viewed simply as a means to reduce misconduct, as it may be a liberating force in work units that promotes psychological safety and, in turn, employee voice behavior. Ethical leadership appears to promote more effective employee behavior because ethical leaders support an environment in which employees feel free to identify problems and opportunities, propose solutions and strategies, and share ideas for improving

1285

work unit or organizational functioning. From this perspective, psychological safety may be part of a broader social exchange process that deserves more attention in theories of leadership.

References Bandura, A. (1977). Social learning theory. Englewood Cliffs, NJ: Prentice-Hall. Barrett, L. F., & Pietromonaco, P. R. (1997). Accuracy of the five-factor model in predicting perceptions of daily social interactions. Personality & Social Psychology Bulletin, 23, 1173–1187. Barrick, M. R., & Mount, M. K. (1991). The Big Five personality dimensions and job performance. A meta-analysis. Personnel Psychology, 44, 1–26. Bass, B. M., & Avolio, B. J. (2004). Multifactor leadership questionnaire: Manual leader form, rater, and scoring key for MLQ (Form 5x–Short). Redwood City, CA: Mind Garden. Bishop, J. W., Scott, K. D., & Burroughs, S. M. (2000). Support, commitment, and employee outcomes in a team environment. Journal of Management, 26, 1113–1132. 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: Jossey-Bass. Bono, J. E., & Judge, T. A. (2004). Personality and transformational and transactional leadership: A meta-analysis. Journal of Applied Psychology, 89, 901–910. Branzei, O., Vertinsky, I., & Camp, R. D., II. (2007). Culture-contingent signs of trust in emergent relationships. Organizational Behavior and Human Decision Processes, 104, 61– 82. Brissette, I., & Cohen, S. (2002). The contribution of individual differences in hostility to the associations between daily interpersonal conflict, affect, and sleep. Personality & Social Psychology Bulletin, 28, 1265– 1274. Brown, M. E. (2007). Misconceptions of ethical leadership: How to avoid pitfalls. Organizational Dynamics, 36, 140 –155. Brown, M. E., & Trevin˜o, L. K. (2006). Ethical leadership: A review and future directions. Leadership Quarterly, 17, 595– 616. Brown, M. E., Trevin˜o, L. K., & Harrison, D. A. (2005). Ethical leadership: A social learning perspective for construct development and testing. Organizational Behavior and Human Decision Processes, 97, 117–134. Cialdini, R. B., & Goldstein, N. J. (2004). Social influence: Compliance and conformity. Annual Review of Psychology, 55, 591– 621. Costa, P. T., & McCrae, R. R. (1992). Revised NEO Personality Inventory professional manual. Odessa, FL: Psychological Assessment Resources. Cropanzano, R., & Mitchell, M. S. (2005). Social exchange theory: An interdisciplinary review. Journal of Management, 31, 874 –900. Dansereau, F., Jr., Graen, G., & 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, 46 –78. Dasborough, M. T., & Ashkanasy, N. M. (2002). Emotion and attribution of intentionality in leader–member relationships. Leadership Quarterly, 13, 615– 634. De Dreu, C. K. W., & Weingart, L. R. (2003). Task versus relationship conflict, team performance, and team member satisfaction: A metaanalysis. Journal of Applied Psychology, 88, 741–749. De Hoogh, A. H. B., & Den Hartog, D. N. (2008). Ethical and despotic leadership, relationships with leader’s social responsibility, top management team effectiveness and subordinates’ optimism: A multi-method study. Leadership Quarterly, 19, 297–311.

1286

WALUMBWA AND SCHAUBROECK

Detert, J., Trevin˜o, L. K., Burris, E., & Andiappan, M. (2007). Managerial models of influence and counterproductivity in organizations: A longitudinal business unit-level investigation. Journal of Applied Psychology, 92, 993–1005. Dirks, K. T., & Ferrin, D. L. (2002). Trust in leadership: Meta-analytic findings and implications for research and practice. Journal of Applied Psychology, 87, 611– 628. Dollinger, S. J., & LaMartina, A. K. (1998). A note on moral reasoning and the five-factor model. Journal of Social Behavior and Personality, 13, 349 –358. Edmondson, A. C. (1999). Psychological safety and learning behavior in work teams. Administrative Science Quarterly, 44, 350 –383. George, J. M. (1990). Personality, affect, and behavior in groups. Journal of Applied Psychology, 75, 107–116. Goldberg, L. R. (1990). An alternative “description of personality”: The Big-Five factor structure. Journal of Personality and Social Psychology, 59, 1216 –1229. Hill, N. S., Bartol, K. M., Tesluk, P. E., & Langa, G. A. (2009). Organizational context and face-to-face interaction: Influences on the development of trust and collaborative behaviors in computer-mediated groups. Organizational Behavior and Human Decision Processes, 108, 187– 201. Hofmann, D. A., & Gavin, M. B. (1998). Centering decisions in hierarchical linear models: Implications for research in organizations. Journal of Management, 24, 623– 641. Hogan, J., & Ones, D. A. (1997). Conscientiousness and integrity at work. In R. Hogan, J. A. Johnson, & S. R. Briggs (Eds.), Handbook of personality psychology (pp. 849 – 870). San Diego, CA: Academic Press. House, R. J. (1971). A path– goal theory of leadership effectiveness. Administrative Science Quarterly, 16, 321–339. James, L. R., Demaree, R. G., & Wolf, G. (1984). Estimating within-group interrater reliability with and without response bias. Journal of Applied Psychology, 69, 85–98. Janis, I. (1972). Victims of groupthink. Boston: Houghton-Mifflin. Jensen-Campbell, L. A., & Graziano, W. G. (2001). Agreeableness as a moderator of interpersonal conflict. Journal of Personality, 69, 323–361. Judge, T. A., & Bono, J. E. (2000). Five-factor model of personality and transformational leadership. Journal of Applied Psychology, 85, 751– 765. Judge, T. A., Bono, J. E., Illies, R., & Gerhardt, M. G. (2002). Personality and leadership: A qualitative and quantitative review. Journal of Applied Psychology, 87, 765–780. Kenny, D. A., Kashy, D. A., & Bolger, N. (1998). Data analysis in social psychology. In D. Gilbert, S. Fiske, & G. Lindzey (Eds.), The handbook of social psychology (4th ed., pp. 233–265). Boston: McGraw-Hill. LePine, J. A., & Van Dyne, L. (1998). Predicting voice in work groups. Journal of Applied Psychology, 83, 853– 868. Liao, H., & Chuang, A. (2004). A multilevel investigation of factors influencing employee service performance and customer outcomes. Academy of Management Journal, 47, 41–58. Lim, B., & Ployhart, R. E. (2004). Transformational leadership: Relations to the five-factor model and team performance in typical and maximum contexts. Journal of Applied Psychology, 89, 610 – 621. Lord, R. G., DeVader, C. L., & Alliger, G. M. (1986). A meta-analysis of the relation between personality traits and leadership perceptions: An application of validity generalization procedures. Journal of Applied Psychology, 71, 402– 410. MacKinnon, D. P., Lockwood, C. M., Hoffman, J. M., West, S. G., &

Sheets, V. (2002). A comparison of methods to test mediation and other intervening variables effects. Psychological Methods, 7, 83–104. Mayer, D. M., Kuenzi, M., Greenbaum, R., Bardes, M., & Salvador, R. (2009). How low does ethical leadership flow? Test of a trickle-down model. Organizational Behavior and Human Decision Processes, 108, 1–13. Mayer, R. C., Davis, J. H., & Schoorman, D. F. (1995). An integrative view of organizational trust. Academy of Management Review, 20, 709 –734. McClelland, D. C., & Boyatzis, R. (1982). The leadership motive pattern and long term success in management. Journal of Applied Psychology, 67, 737–743. Organ, D. W., Podsakoff, P. M., & MacKenzie, S. B. (2006). Organizational citizenship behavior: Its nature, antecedents, and consequences. Thousand Oaks, CA: Sage. Raudenbush, S. W., & Bryk, A. S. (2002). Hierarchical linear models: Applications and data analysis methods (2nd ed.). Thousand Oaks, CA: Sage. Roberts, B. W., & DelVecchio, W. F. (2000). The rank-order consistency of personality from childhood to old age: A quantitative review of longitudinal studies. Psychological Bulletin, 126, 3–25. Rubin, R. S., Munz, D. C., & Bommer, W. H. (2005). Leading from within: The effects of emotion recognition and personality on transformational leadership behavior. Academy of Management Journal, 48, 845– 858. Saucier, G. (1994). Mini markers: A brief version of Goldberg’s unipolar Big-Five markers. Journal of Personality Assessment, 63, 506 –516. Schminke, M., Ambrose, A., & Neubaum, D. O. (2005). The effect of leader moral development on ethical climate and employee attitudes. Organizational Behavior and Human Decision Processes, 97, 135–151. Sobel, M. E. (1982). Asymptotic confidence intervals for indirect effects in structural equation models. In S. Leinhardt (Ed.), Sociological methodology (pp. 290 –312). Washington, DC: American Sociological Association. Tangirala, S., & Ramanujam, R. (2008). Employee silence on critical work issues: The cross level effects of procedural justice climate. Personnel Psychology, 61, 37– 68. Tobin, R. M., Graziano, W. G., Vanman, E., & Tassinary, L. (2000). Personality, emotional experience, and efforts to control emotions. Journal of Personality and Social Psychology, 79, 656 – 669. Trevin˜o, L. K. (1986). Ethical decision making in organizations: A person– situation interactionist model. Academy of Management Review, 11, 601– 617. Trevin˜o, L. K., Brown, M., & Hartman, L. P. (2003). A qualitative investigation of perceived executive ethical leadership: Perceptions from inside and outside the executive suite. Human Relations, 55, 5–37. Van Dyne, L., & LePine, J. A. (1998). Helping and voice extra-role behavior: Evidence of construct and predictive validity. Academy of Management Journal, 41, 108 –119. Walumbwa, F. O., Avolio, B. J., Gardner, W. L., Wernsing, T. S., & Peterson, S. J. (2008). Authentic leadership: Development and validation of a theory-based measure. Journal of Management, 34, 89 –126. Weierter, S. J. M. (1997). Who wants to play “follow the leader?” A theory of charismatic relationships based on routinized charisma and follower characteristics. Leadership Quarterly, 8, 171–193. Zaccaro, S. J. (2007). Trait-based perspectives of leadership. American Psychologist, 62, 6 –16.

Received September 16, 2008 Revision received February 18, 2009 Accepted March 3, 2009 䡲

Suggest Documents