Although there has been a strong awareness of, and efforts to enhance, ethical practices in the global business community over the past decades, periodic corporate scandals (e.g., Wells Fargo’s illegal business practice of opening accounts without customers’ knowledge, Volkswagen’s emissions scandal) continue to remind us of how important it is for organizational decision makers—in other words, leaders—to understand and live up to ethical standards. Unfortunately, it does not appear that the Asia Pacific region is immune to such global concerns on unethical business practices and acts (e.g., Toshiba’s accounting scandal). In fact, according to Transparency International (n.d.), many Asia Pacific countries routinely rank high on the annual Corruption Perceptions Index. This finding underscores the importance of investigating ethical leadership in Asian contexts.

Despite the significance of conducting research in Asian contexts, to date the majority of research on ethical leadership has been conducted in Western contexts, with far fewer studies having focused on Asia (e.g., approximately 30% of the studies included in Ng & Feldman’s, 2015 research). Such lack of evidence from the Asian region is a notable omission because it prevents us not only from understanding the accurate and complete effects of ethical leadership in Asia, but also from untangling the uniqueness (or generalizability) of ethical leadership in this region. For instance, whereas the research on ethical leadership in Western contexts has revealed positive relationships of ethical leadership with followers’ work outcomes, such as task performance and citizenship behavior (e.g., Brown & Treviño, 2006a, b; Brown, Treviño, & Harrison, 2005; Mayer, Aquino, Greenbaum, & Kuenzi, 2012; Ng & Feldman, 2015; Schaubroeck et al., 2012), studies conducted in Asia have often reported a non-significant relationship between ethical leadership and employees’ task performance (e.g., Liu, Kwan, Fu, & Mao, 2013a; Park, Kim, & Song, 2015; Schuh, Zhang, & Tian, 2013; Wu, 2012). These inconsistent findings necessitate replication and consideration of the characteristics of the unique contexts in which employees are embedded. To address this gap in the literature, the current study focuses on the impact of ethical leadership on followers’ work outcomes in a South Korean context.

While maintaining our focus on the effects of ethical leadership in South Korea, we pay special attention to explicating the predictive validity of ethical leadership by teasing apart the effects of similar types of leadership constructs (i.e., the full-range leadership model [FRLM] and authentic leadership). Scholarly efforts to examine the issue of construct validity remain largely unexplored in the field of leadership albeit “Testing the incremental validity of each emerging leadership approach is important to inform the optimal array of leadership forms and evaluate potential construct redundancy” (Hoch, Bommer, Dulebohn, & Wu, 2018: 503). Ng and Feldman (2015) have also noted that the evidence for the incremental validity of ethical leadership is “still far from being strong or definitive” (9). This hinders the quest to expand our knowledge of ethical leadership.

Untangling the incremental validity of a certain leadership style has significant theoretical implications because a great deal of the existing literature posits that leaders demonstrate a wide range of similar leadership behaviors as they seek to influence others (e.g., Bass & Avolio, 1994; Lin, Ma, & Johnson, 2016; Mayer et al., 2012). As such, since different leadership styles tend to result in similar positive or negative work outcomes, identification of incremental validity supports the notion of criterion-related validity, which is critical to understand the uniqueness and unitality of the focal leadership style (Mayer et al., 2012; Ng & Feldman, 2015).

Accordingly, the primary purpose of our study is to investigate the predictive validity of ethical leadership to predict organizational commitment, task performance, and turnover intentions in the presence of the FRLM and authentic leadership. Notably, Ng and Feldman (2015) have demonstrated that ethical leadership, albeit weakly in some cases, significantly predicted task performance even after controlling for the effects of other leadership styles, such as transformational, contingent reward, management by exception. Our study goes beyond Ng and Feldman’s (2015) study by investigating the incremental validity of ethical leadership in the presence of authentic leadership. The prior evidence suggests that ethical leadership has a conceptual redundancy with other leadership constructs and authentic leadership is one such example (Cooper, Scandura, & Schriesheim, 2005; Gardner, Cogliser, Davis, & Dickens, 2011). Thus, examining the incremental validity of ethical leadership while controlling for the effects of authentic leadership will provide additional evidence for the discriminant validity of ethical leadership in predicting follower work outcomes.

In addition, the current study focuses on identifying psychological mediators in ethical leadership process while controlling for the FRLM and authentic leadership. Specifically, drawing on Meyer and Herscovitch’s (2001) model of organizational commitment, we argue that a leader’s ethical behaviors and practices provide a theoretical link to increased affective and normative commitment, which serves as an obliging force that motivates followers to take actions (e.g., meeting work goals, retaining the organizational membership, etc.) that benefit to the specific target that they are committed (i.e., organization). Examining different forms of commitment based on a multi-faceted conceptualization of organizational commitment (e.g., Meyer & Allen, 1991) is important in studies of ethical leadership because it helps us gain more nuanced and sophisticated knowledge about ethical leadership process. Furthermore, in explicating the relationship between ethical leadership and follower work outcomes, the present study uses an objective measure of task performance to address limitations of previous studies that relied on subjective ratings of followers’ performance (e.g., Liu, Loi, & Lam, 2013b; Piccolo, Greenbaum, Den Hartog, & Folger, 2010). Using two separate sets of survey data collected from South Korea, this paper provides valuable insights into the unique influence of ethical leadership on follower work outcomes and the psychological processes behind the ethical leadership-follower outcomes relationships.

Theoretical background

Ethical leadership in Western and non-Western contexts

For more than a decade, in response to the insistent demand for greater awareness of and information on practices related to ethics in the business community, researchers and practitioners have delved into the role and influence of ethical leadership on followers’ work outcomes in organizations (e.g., Brown et al., 2005; Brown & Treviño, 2006a, 2006b; Mayer et al., 2012; Ng & Feldman, 2015; Schaubroeck et al., 2012). Central to the conceptualization of ethical leadership is the leader’s enactment of normative appropriate behaviors with the goal of promoting followers’ ethical conduct: by communicating ethics-related messages; by role-modeling ethical behaviors; by proactively seeking followers’ voices and inputs for ethical concerns; and by incentivizing ethical conducts via rewards and punishments (Brown et al., 2005; Treviño, Brown, & Hartman, 2003).

The cumulative body of research has largely suggested that ethical leadership facilitates positive work outcomes among followers, such as improved job satisfaction, task performance, extra-role performance, voicing behaviors (Brown et al., 2005; Lam, Loi, Chan, & Liu, 2016; Treviño & Brown, 2014; Walumbwa, Mayer, Wang, Wang, Workman, & Christensen, 2011), while deterring negative outcomes, such as unethical conduct and uncivil behaviors (Mayer, Kuenzi, & Greenbaum, 2010; Walsh, Lee, Jensen, McGonagle, & Samnani, 2017). Prior research, often drawing on social learning and social exchange theories, has shown that such positive influence of ethical leadership on followers’ outcomes is due to enhanced task significance, autonomy, psychological empowerment, trust, and self-efficacy (e.g., Ng & Feldman, 2015; Piccolo et al., 2010; Walumbwa et al., 2011; Zhu, May, & Avolio, 2004).

While these positive influences spurred by ethical leadership have been replicated and confirmed in numerous studies in the past decade, these studies have largely focused on the North American context rather than representing the global population (Eisenbeiß & Brodbeck, 2014; Resick, Martin, Keating, Dickson, Kwan, & Peng, 2011). The asymmetric attention to research contexts may be a significant oversight that hinders our understanding of ethical leadership, given that such a leadership is an important practice in both Western and non-Western societies (Resick et al., 2011). Although ethical leadership generally appears to have a positive impact on follower work outcomes (e.g., Bedi, Alpaslan, & Green, 2016; Hoch et al., 2018; Ng & Feldman, 2015), only limited knowledge exists regarding the generalizability of these findings across different parts of the world, due to the lack of research in non-Western contexts (Eisenbeiß & Brodbeck, 2014; Lam, Huang, & Lau, 2012). For instance, Liu et al. (2013a, b), using a sample from a Chinese manufacturing firm, reported that ethical leadership was not significantly related to supervisor ratings of follower task performance. Park et al. (2015), using a sample of nonprofit organizations in South Korea, found that followers’ perceptions of ethical leadership were not associated with the followers’ ratings of their own task performance. Furthermore, a recent study by Wang and colleagues (Wang, Chiang, Chou, & Cheng, 2017) has shown that the Chinese Ethical Leadership Scale, which captures culturally specific behavioral manifestations of ethical leadership in China, explains an additional variance for follower work outcomes (e.g., wrongdoing, task performance) beyond Brown et al.’s (2005) ethical leadership scale.

Identifying the influence of ethical leadership in non-Western contexts would greatly expand our understanding of the generalizable characteristics embedded in this type of leadership. In pursuing the aforementioned research agenda, this study focuses on the incremental validity of ethical leadership on attitudinal and behavioral work outcomes above and beyond the impact of the FRLM and authentic leadership, which were suggested as similar to ethical leadership in prior research. A more detailed discussion of this issue is followed in the next section.

Incremental validity of ethical leadership

Examining incremental validity has meaningful implications in organizational science because demonstrating the unique effects of a certain leadership style above and beyond the effects of other, similar leadership behaviors is directly related to ascertaining construct validity (Hoch et al., 2018). According to Ng and Feldman (2015), such efforts are “theoretically important because it can clarify how far-reaching and how deep the impact of this construct could be” (p. 948). Responding to the growing calls for research in this area, the current study investigates the incremental validity of ethical leadership on followers’ attitudinal and behavioral work outcomes, while controlling for the FRLM and authentic leadership, owing to their behavioral commonalities with ethical leadership.

Ethical leadership and the full-range leadership model

We choose to control for the FRLM because the wide range of leadership behaviors included in the FRLM model (i.e., transformational, transactional contingent reward, management by exception, and laissez-faire leadership) have been found to be significantly associated with ethical leadership (e.g., Bedi et al., 2016; Brown et al., 2005; Hoch et al., 2018; Ng & Feldman, 2015). Specifically, the transformational and contingent reward leadership styles, which are characterized by high levels of consideration, fairness, role-modeling behavior, and clear expectations and rewards (Bass, 1985), share commonalities with the ethical leadership style. In turn, these commonalities might obscure the unique validity of ethical leadership identified in broader-scale research. For instance, a conceptual overlap can be found in the conceptualization of transformational leadership in that this type of leader “can be counted on to do the right thing” and “demonstrates high standards of ethical and moral conduct” (Avolio, 1999: 43). Gini (1997) has argued that ethical leaders set clear standards and dole out rewards and punishments based on the accountability accepted by followers, which is a core tenet of transactional leadership. Such conceptual overlap has also been supported by empirical research. For example, Toor and Ofori (2009), using a sample from Singapore’s construction industry, found that ethical leadership was significantly correlated with transformational leadership (r = .58) and contingent reward leadership (r = .42).

Moreover, we consider negative or ineffective types of leadership behaviors, such as management by exception—active (e.g., closely monitoring followers’ behavior), management by exception—passive (e.g., taking corrective actions), and laissez-faire leadership (e.g., avoiding decisions, not providing behavior-contingent reinforcement), because positive leader behaviors may not explain variability in outcomes beyond negative leader behaviors (Harold & Holtz, 2015). Recent meta-analytic evidence indicates that ethical leadership is significantly correlated with management by exception—passive leadership (ρ = −.63) and laissez-faire leadership (ρ = −.62). This evidence requires further examination to elucidate the incremental predictive power of ethical leadership above and beyond the FRLM.

Ethical leadership and authentic leadership

Increasing scholarly attention has been recently paid to the concept of authentic leadership and its role as “the root of positive forms of leadership” in the field of leadership (e.g., Avolio & Gardner, 2005; Cooper et al., 2005; Gardner et al., 2011; Walumbwa, Avolio, Gardner, Wernsing, & Peterson, 2008). Authentic leaders are defined as “those who are deeply aware of how they think and behave and are perceived by others as being aware of their own and others’ values/moral perspectives, knowledge, and strengths; aware of the context in which they operate and who are confident, hopeful, optimistic, resilient, and of high moral character” (Avolio & Gardner, 2005: 4). As this definition suggests, the conceptualization of authentic leadership is multidimensional, encompassing a leader’s traits, states, behaviors, contexts, and even attributions (Cooper et al., 2005).

Advocates of authentic leadership argue that the ongoing investigation of the failure of ethical practices and the challenges created by a rapidly changing society requires a framework supportive of a fundamentally and profoundly authentic and trustworthy leadership practice (Avolio & Gardner, 2005; Gardner et al., 2011). Despite the initial efforts made by the proponents of this leadership model, several scholars have pointed out the issue of conceptual redundancy between authentic leadership and other similar constructs (Cooper et al., 2005; Gardner et al., 2011). Ethical leadership is one such example, as researchers still need to determine its discriminant validity vis-à-vis authentic leadership.

The possibility of conceptual overlap between authentic leadership and ethical leadership is primarily rooted in the moral component of the conceptualization, which authentic leadership embraces as a core characteristic (Avolio & Gardner, 2005; Ladkin & Taylor, 2010; May, Chan, Hodges, & Avolio, 2003). For example, Avolio and Gardner (2005) contend that authentic leadership is inherently moral, whereas Walumbwa et al. (2008) emphasize taking an internalized moral viewpoint to operationalize authentic leadership. A widely used authentic leadership scale, developed by Walumbwa et al. (2008), also incorporates the internalized moral perspective as a sub-dimension of authentic leadership (e.g., “demonstrates beliefs that are consistent with actions” and “makes decisions based on his/her core beliefs”). Therefore, to capture the true contributions of ethical leadership to followers’ attitudinal and behavioral work outcomes, it would be desirable to probe the incremental validity of ethical leadership by teasing apart the effects of authentic leadership.

Hypotheses development

Ethical leadership and organizational commitment

Although supervision is viewed as one of the critical factors affecting the levels of individuals’ commitment to the organization (Mowday, Steers, & Porter, 1982), we have limited knowledge regardingthe the relationships between ethical leadership and different dimensions of organizational commitment (Zhu et al., 2004). Furthermore, research examining such relationships in a non-western context while controlling for similar leadership styles is much scant. In probing the link between ethical leadership and organizational commitment, we examine both affective and normative commitment because they are theoretically more aligned with the positive influences of ethical leadership process than is continuance commitment. More specifically, affective commitment and normative commitment are established based on a relational motive (e.g., commitment efforts derived from establishing and maintaining the relationship with the current organization), whereas continuance commitment reflects an individual’s self-interest motive. This suggests that continuance commitment (compared to affective commitment and normative commitment) is less likely to be influenced by the significance of relationships individuals have in the organization such as leadership process (Meyer, Stanley, Herscovitch, & Topolnytsky, 2002).

To illuminate the relationships between ethical leadership and affective/normative commitment, it is crucial to understand the theoretical linkage between the behavioral characteristics of ethical leadership and the conceptual nature of organizational commitment. According to Meyer and Herscovitch’s (2001) model of organizational commitment, the essence of commitment is characterized as a stabilizing or obliging force that provides direction to particular actions of employees and this force is directed at specific targets that they are committed to (i.e., organization). In particular, such committed energy based on a relational motive becomes a strong driver for developing the perceptions of what they are obligated to do for the organization (Lam & Liu, 2014). The cognitive element of common mindsets embedded in affective and normative commitment is grounded on value congruence and obligations (Meyer & Allen, 1991) and is further developed to meet goals and maintain their organizational membership (Angle & Perry, 1981; Meyer, Becker, & Van Dick, 2006).

We argue that the behavioral characteristics of ethical leadership play a vital role in promoting the cognitive elements of affective and normative commitment. For example, the strongly weighted morality that leaders emphasize is based on the leaders’ honesty, trustworthiness, caring, and concern for others (Brown & Treviño, 2006a) and such emphasis is explicitly embodied into the practices of normatively appropriate procedures and decision making (Schaubroeck et al., 2012; Treviño, Hartman, & Brown, 2000). We predict that these principled and morally consistent leadership practices that the followers undergo in turn will encourage them to develop and reinforce an emotional bond with the organization because such positive leadership experiences are tied into the work processes in the organization. Thus, experiencing ethical leadership will be positively associated with followers’ affective commitment. In addition to the downward influence that leaders strive for in ethical leadership processes, these leaders inspire the upward influence into organizational procedures in a way that followers’ voices and inputs on ethical concerns are heard and appreciated. We argue that this type of behaviors that ethical leaders display becomes a catalyst for followers to expand emotional attachment to, and involvement in the organization because leaders are viewed as organizational agents who embody the principal focus of what the organization strives to accomplish (Rhoades & Eisenberger, 2002). Indeed, prior research has suggested that when leaders focus on ethical climate, followers tend to be more committed to the organization (Erben & Güneşer, 2008).

Based on this evidence, we predict that the positive relationship between ethical leadership and affective commitment will be supported even in the presence of the leader–member relationships forged by (1) transformational and transactional leaders, whose spectrum of behaviors includes providing clear expectations and rewards contingent on meeting those expectations, being inspiring, considerate, and charismatic, and (2) authentic leadership, whose behaviors mainly focus on authenticity based on its strong emphasis on ethicality in leadership practices, which becomes a primary driving force for affective commitment (Ng & Feldman, 2015).

Additionally, we propose that there will be a positive relationship between ethical leadership and normative commitment. This theoretical link is highly plausible when applying the core behavioral characteristics of ethical leadership (e.g., explicit message to keep ethicality, conducts, feedback, and treatment toward followers) to our research context. The goal of ethical leadership is to promote followers’ ethical awareness and conduct by exposing them to the leader’s enactment of normatively appropriate behaviors. Thus, followers who regularly witness the leader’s moral behaviors and practices will likely be equipped with the desired levels of ethicality (Brown et al., 2005; Brown & Treviño, 2006a, 2006b), which is downward influence from the leader. In particular, normative commitment refers to an individual’s feelings of obligation and responsibility toward the organization created by the internalization of its goals, values, and missions (Allen & Meyer, 1990; Jaros, Jermier, Koehler, & Sincich, 1993). Because this normative aspect of organizational commitment encompasses “the totality of internalized normative pressure to act in a way which meets organizational goals and interests” (Wiener, 1982: 421), normative commitment is further engraved in an individual’s personal beliefs about how others expect him or her to act regarding the right and moral thing to do (Allen & Meyer, 1990; Wiener, 1982). This leads us to argue that despite the conceptual overlap in terms of the moral component that are embraced in ethical leadership and authentic leadership (Avolio & Gardner, 2005), the moral components based on authenticity are not comprehensive enough to capture all of the behavioral characteristics associated with ethical leadership. Accordingly, ethical leadership will be positively associated with normative commitment even after controlling for authentic leadership.

Hypothesis 1a

Ethical leadership is positively related to affective commitment even after controlling for transformational, contingent reward, management-by exception, and laissez-faire leadership behaviors.

Hypothesis 1b

Ethical leadership is positively related to affective commitment even after controlling for authentic leadership behaviors.

Hypothesis 1c

Ethical leadership is positively related to normative commitment even after controlling for authentic leadership behaviors.

The mediating effects of organizational commitment in the relationship between ethical leadership and followers’ work outcomes

The links between organizational commitment and turnover intentions/task performance have been consistently delineated both theoretically and empirically and the core theory is grounded on the conceptualization of commitment regarding a stabilizing force that binds committed individuals to show focal behaviors for remaining with the organization and accomplishing goals (Meyer & Herscovitch, 2001; Mayer & Schoorman, 1998). In line with the theoretical prediction, Stanley, Meyer, Topolnytsky, and Herscovitch (1999) found that both affective commitment and normative commitment were negatively related to turnover intentions and positively related to task performance.

A follower’s feeling of emotional attachment to the organization is derived from the influence processes initiated by ethical leaders. Followers internalize their morality and the ensuing practices that are frequently directed and governed by ethical leaders. We argue that internalization, in turn, contributes to the development of a strong sense of affective commitment and obligation that encourages the individual to execute a series of behavioral practices that reflect “the right thing to do,” as perceived by the ethical leader and the organization. As one way to express their commitment to, and obligations toward these organizational authorities, followers will make substantial efforts to perform their job/tasks based on what is expected or even what goes beyond those expectations. Since affective commitment entails a willingness to exert one’s best efforts on the organization’s behalf (Allen & Meyer, 1990), followers with strong affective commitment tend to generate higher levels of performance (Johnson, Chang, & Yang, 2010). Based on this logic, we predict that affective commitment, derived from ethical leadership, will act as a vehicle for motivating followers to work toward organizational goals and objectives, represented by objective measures of performance.

With regard to the mediating effects of affective commitment on turnover intentions, we argue that committed employees working with ethical leaders are less likely to leave the organization due to their strong emotional and psychological connections with the organization (i.e., affective commitment), with these links being forged via the moral practices prevalent within the organization (Abrams, Ando, & Hinkle, 1998; Mathieu & Zajac, 1990). Recent evidence also suggests that ethical leadership may be an important factor reducing followers’ desire to leave the organization because this type of leadership allows employees’ voices and opinions to be heard (Lam et al., 2016; Resick et al., 2011; Toor & Ofori, 2009). Furthermore, we argue that ethical leadership behaviors, characterized by establishing an explicit two-way communication channel (e.g., clarifying ethical messages, listening feedback related to “the right thing to do”), will be positively associated with a follower’ sense of obligation and responsibility to protect organizational values and goals (i.e., normative commitment). Prior research has suggested that the behavioral implications of organizational commitment derive from the binding of the individual to the organization (Meyer & Herscovitch, 2001). In this case, the behavioral consequences of normative commitment are salient, in that committed followers (working with ethical leaders) are more likely to continue their employment. The lowered turnover intentions among the committed employees arise because those individuals feel obligated to behave in ways that are consistent with organizational shared values and to live up to the moral expectations set by their ethical leader. Based on this evidence and logic, we propose the following hypotheses:

Hypothesis 2a

Affective commitment mediates the relationship between ethical leadership and turnover intentions even after controlling for transformational, contingent reward, management-by exception, and laissez-faire leadership behaviors.

Hypothesis 2b

Affective commitment mediates the relationship between ethical leadership and objective ratings of task performance even after controlling for transformational, contingent reward, management-by exception, and laissez-faire leadership behaviors.

Hypothesis 3a

Affective commitment mediates the relationship between ethical leadership and turnover intentions even after controlling for authentic leadership.

Hypothesis 3b

Normative commitment mediates the relationship between ethical leadership and turnover intentions even after controlling for authentic leadership.

Overview of studies

We examined affective commitment as a mediator in the ethical leadership–outcome relationship while controlling for the FRLM in Study 1 (H1a and H2a/H2b). In Study 2, we replicated the mediating role of affective commitment while controlling for authentic leadership and extended Study 1 by investigating normative commitment as a mediator while controlling for authentic leadership (H1b and H3a/H3b). By employing a two-wave survey design in Study 2, we sought to lessen the concerns about common method variance, which might have existed in the cross-sectional design used in Study 1. Together, the studies examine two types of organizational commitment as intervening mechanisms by which the influence of ethical leadership is transmitted to followers’ outcomes above and beyond the effects of FRLM and authentic leadership.

Study 1

Methods

Participants and procedures

A total of 258 employees working in the headquarters of a company in South Korea were invited to participate in an online survey administered through the company’s intranet; the invitation was sent by the company’s human resources (HR) manager. Participant confidentiality was assured, with the email emphasizing that the survey was being conducted by academic researchers and that the company would not have access to the data collected. Of the 203 employees who were contacted for the survey, 121 employees responded, yielding a response rate of 60%. After deleting responses with missing data, the final sample of 92 was used for analysis. The average age of study participants was 35.76 years (SD = 6.1) and 76.3% of the participants were male. Average supervisory tenure was 2.74 years (SD = 2.79) and average organizational tenure was 6.51 years (SD = 4.80). Approximately 89% of participants had a college degree. The supervisors being referenced in the survey were working at various levels of the organizational hierarchy ranging from frontline leaders to business unit leaders.

Translation of questionnaire items

Since all participants were Korean, the survey questionnaires were provided in Korean. The questionnaire items, which were originally created in English, were translated following the standard and recommended translation and back-translation procedure (Brislin, 1980) to ensure conceptual equivalence and comparability between the English and Korean questionnaire items.

Measures

Employees were asked to assess their supervisor’s ethical leadership as well as their own perceptions of affective commitment and turnover intentions. To minimize the problems associated with common method bias, we obtained objective performance data from the HR department.

Ethical leadership

Ethical leadership was measured using the six items from the Ethical Leadership Scale (ELS) developed by Brown et al. (2005). This scale has been widely used in Asian countries (e.g., Lam et al., 2016; Loi, Lam, Ngo, & Cheong, 2015; Resick et al., 2011). Sample items include “Discusses business ethics or values with employees” and “Sets an example of how to do things the right way in terms of ethics.” A 5-point rating scale (1 = strongly disagree to 5 = strongly agree) was used.

Affective commitment

Affective commitment was assessed using the six items in Meyer, Allen, and Smith’s (1993) measure. Illustrative items are “I really feel as if this organization’s problems are my own” and “This organization has a great deal of personal meaning for me.” A 7-point scale (1 = very strongly disagree to 7 = very strongly agree) was employed.

Turnover intentions

Turnover intentions were measured using two items from the Michigan Organizational Assessment Questionnaire (Cammann, Fichman, Jenkins, & Klesh, 1979). A sample item is “I often think about quitting,” and a 7-point rating scale (1 = strongly disagree to 7 = strongly agree) was used.

Performance ratings

Employees’ objective performance ratings were obtained from the company’s HR department. The performance data were originally recorded with a letter grade (i.e., O = Outstanding; E = Exceeds expectations; M = Meets expectations; P = Partially meets expectations; N = Does not meet expectations) but we converted them to a 5-point rating scale (i.e., 1 = Does not meet expectations to 5 = Outstanding) for further analysis.

Control variables

We controlled for the FRLM (i.e., transformational, transactional contingent reward, management by exception, and laissez-faire leadership) so that we could examine the incremental validity of ethical leadership. Employees were asked to assess their supervisor’s leadership using the Multifactor Leadership Questionnaire (MLQ-5X; Bass & Avolio, 1997), which consists of 36 items (20 items for transformational leadership, 4 items for transactional contingent reward leadership, 4 items for management by exception—active, 4 items for management by exception—passive, and 4 items for laissez-faire leadership) using a 5-point scale (1 = not at all to 5 = frequently or always). Items of the MLQ cannot be reproduced in this paper due to copyright restrictions. We controlled for age, gender, and organizational tenure because these demographics have been reported to be significantly associated with individuals’ affective commitment and turnover intentions (Lee, Carswell, & Allen, 2000; Meyer et al., 2002). Organizational tenure was measured by the number of years the respondents had been employed by the company at the point when they completed the survey.

Data analysis

Data were analyzed using the bias-corrected bootstrap method suggested by MacKinnon, Lockwood, and Williams (2004). This analytical method has been found to yield the most accurate confidence limits and greatest statistical power (MacKinnon et al., 2004; Preacher, Rucker, & Hayes, 2007) and thus, “the bias-corrected bootstrap confidence interval has become the more widely recommended method for inference about the indirect effect in mediation analysis” (Hayes, 2013, p.116). The bias-corrected bootstrap method is one of resampling methods, which adjusts for a possible nonnormal distribution of the mediated effects and corrects for bias in the central tendency of the estimate (Mackinnon et al., 2004). Consistent with the literature (Cheung & Lau, 2008), we specified 5000 bootstrap samples. Statistically significant indirect effects suggest that the relationships between the independent variables and the outcome variables occur through the mediator.

Results and discussion

Table 1 provides the means, standard deviations, Cronbach’s alpha coefficients, and intercorrelations for the study variables. We tested our proposed eight-factor model against five alternative models. A series of confirmatory factor analyses (CFA) were conducted utilizing AMOS version 24. Parcels are preferred when sample sizes are relatively small (Bagozzi & Edwards, 1998; Little, Cunningham, Shahar, & Widaman, 2002). Given the small size of the sample used in Study 1, therefore, the number of observed indicators for each construct was reduced. For a latent construct where construct items were divided into subscales (i.e., transformational leadership), the items of each subscale were averaged to represent an indicator of the latent construct in the CFA (Little et al., 2002). For ethical leadership and affective commitment, using the CFA factor loadings for each construct as a guide, items with a high loading were combined with items with a low loading across parcels until all items were assigned to one of the three indicators for each construct (Landis, Beal, & Tesluk, 2000). For the other FRLM leadership constructs where each construct has four items (i.e., contingent reward, management by exception, and laissez-faire leadership), the number of observed indicators for each construct was reduced to two based on the factor loadings for each construct. For turnover intentions, each of the two items represented a single indicator.

Table 1 Means, standard deviations, coefficient alphas, and intercorrelations for Study 1 (N = 92)
Full size table

The eight-factor model better fit the data (χ2 = 202.5, df = 161, NFI = .88, CFI = .97, RMSEA = .05) than the one-factor model in which all items were loaded onto a single factor (Δχ2 = 592.15, Δdf = 28, p < .001). Further, the eight-factor model better fit the data than the models in which transformational leadership was combined with ethical leadership (Δχ2 = 623.14, Δdf = 8, p < .001); transformational, contingent reward, and ethical leadership were merged and laissez-faire and management by exception were merged (Δχ2 = 477.33, Δdf = 22, p < .001). Therefore, we conducted hypotheses testing using the eight-factor model.

We performed a series of hierarchical regression analyses to test hypotheses; the results are displayed in Table 2. Ethical leadership was positively related to followers’ affective commitment even after controlling for the leadership behaviors included in the FRLM (when controlling for transformational leadership, b = .16; contingent reward leadership, b = .16; management by exception—active, b = .18; management by exception—passive, b = .20; laissez-faire leadership, b = .17). Thus, Hypothesis 1a was supported.

Table 2 Results of mediation analyses (Affective commitment as a mediator) in Study 1
Full size table

Hypotheses 2a and 2b predicted that affective commitment would play a mediating role in the relationship between ethical leadership and followers’ performance and turnover intentions after controlling for the FRLM. The bootstrap analyses demonstrated that the indirect effects of ethical leadership on performance via affective commitment were significant across all the FRLM leadership styles that were used as a control variable (when controlling for transformational leadership, b = .07; contingent reward leadership, b = .07; management by exception—active, b = .07; management by exception—passive, b = .07; laissez-faire leadership, b = .08). Because the confidence interval (CI) did not include zero (95% CI = .01 to .22, depending on the leadership style used as a control variable), the indirect effects were significantly different than zero. Furthermore, the indirect effects of ethical leadership on turnover intentions mediated by affective commitment were significant across all the FRLM leadership styles being controlled for (transformational leadership, b = −.16; contingent reward leadership, b = −.16; management by exception—active, b = −.18; management by exception—passive, b = −.21; laissez-faire leadership, b = −.16). The confidence interval did not include zero (95% CI = −.40 to −.05, depending on the leadership style used as a control variable). Therefore, Hypotheses 2a and 2b were supported. Taken together, these results suggest that affective commitment mediates the relationship between ethical leadership and both followers’ performance and turnover intentions.

Study 2

Study 1 provided evidence supporting our hypotheses, but had the small sample size (N = 92), cross-sectional design, and inability to test alternative mediating mechanisms. Thus, we collected a second data set to replicate some of the findings of Study 1, while addressing these limitations. In Study 2, we also examined normative commitment as a potential mediator in the ethical leadership–turnover intentions relationship.

Methods

Participants and procedures

We invited 195 individuals who had registered for a master of business administration program in a large public university in South Korea to participate in our surveys. These individuals were working in a wide range of professional domains within their organizations (e.g., finance, retail, construction, information technology) at the time when the surveys were conducted. Pencil-and-paper surveys were administered at two time points to create a time lag between survey-taking events to address the reverse-causality issue while testing the posited relationship. The survey waves were spaced one week apart to ensure the independence of the responses.

At Time 1 (T1), 195 participants were asked to assess their supervisor’s ethical leadership, authentic leadership, and demographic information. One hundred eighty-two respondents completed the first survey, yielding a response rate of 93.3%. At Time 2 (T2; one week later after T1), we invited these 182 respondents who completed the T1 survey to assess their affective and normative commitment, as well as their turnover intentions. One week interval has been used in studies on various topics including abusive supervisory behaviors (e.g., Lian, Brown, Ferris, Liang, Keeping, & Morrison, 2014), motivation (e.g., Ferris, Johnson, Rosen, Djurdjevic, Chang, & Tan, 2013), organizational climate (e.g., Luthans, Norman, Avolio, & Avey, 2008), and workplace aggression (Hershcovis, Ogunfowora, Reich, & Christie, 2017). One hundred sixty-one respondents completed both the T1 and T2 surveys, yielding a response rate of 88.5%. Responses that had missing data were removed from the analyses, resulting in a final sample of 142. The average age of respondents was 37.57 years (SD = 8.48) and 71.6% of the respondents were male. Organizational tenure and supervisory tenure were 7.28 years (SD = 6.10) and 3.03 years (SD = 3.62), respectively. All of the respondents had college degrees, and 38 of them (23.6%) had postgraduate (master’s) degrees. The leaders being assessed on their ethical and authentic leadership in the survey were working at various levels in their organization at the time of the surveys.

Measures

As in Study 1, the translation and back-translation procedure was applied to ensure the comparability between the English and Korean items (Brislin, 1980). We also used the identical measures for the variables employed in Study 1 for replication (i.e., ethical leadership, affective commitment, and turnover intentions). Normative commitment was the only new variable added to Study 2.

Normative commitment

Normative commitment was measured using the six items on Meyer et al.’s (1993) organizational commitment scale. Sample items include “I would feel guilty if I left my organization now” and “This organization deserves my loyalty.” A 7-point scale (1 = very strongly disagree to 7 = very strongly agree) was used for this measure.

Control variable

To uncover the incremental validity of ethical leadership on our hypothesized variables, we included authentic leadership as a control variable. Authentic leadership was assessed using Walumbwa et al.’ (2008) eight-item scale, which captures a leader’s self-awareness (e.g., “Seeks feedback to improve interactions with others”), relational transparency (e.g., “Says exactly what he/she means”), internalized moral perspective (e.g., “Demonstrates beliefs that are consistent with actions”), and balanced processing (e.g., “Solicits views that challenge his/her core beliefs”). In addition, age, gender, organizational tenure, and supervisory tenure were included to alleviate the potential confounding effects with the hypothesized variables.

Data analysis

The same statistical analyses conducted in Study 1 were employed to analyze the Study 2 data.

Results

Table 3 provides the means, standard deviations, Cronbach’s alpha coefficients, and intercorrelations for the variables. A series of confirmatory factor analyses (CFA) were conducted utilizing AMOS version 24. Due to the relatively small size of the sample used in Study 2, we used parcels instead of items in CFAs (Bagozzi & Edwards, 1998; Little et al., 2002). The number of observed indicators for each construct was reduced to three by using the method of balancing high- and low-loading items across parcels (Landis et al., 2000). The five-factor model showed a better fit with the data (χ2 = 176.30, df = 94, NFI = .89, CFI = .95, RMSEA = .08) compared with the one-factor model in which all items were loaded onto a single factor (Δχ2 = 534.7, Δdf = 10, p < .001). Further, the five-factor model exhibited a better fit with the data compared with the models in which authentic leadership was combined with ethical leadership (Δχ2 = 53.44, Δdf = 4, p < .001); affective commitment and normative commitment were merged (Δχ2 = 27.93, Δdf = 4, p < .001). Therefore, we conducted hypotheses testing using the five-factor model.

Table 3 Means, standard deviations, coefficient alphas, and intercorrelations for Study 2 (N = 142)
Full size table

Results of the hierarchical regression analyses are displayed in Table 4. Ethical leadership was positively related to followers’ affective commitment after controlling for authentic leadership (b = .36). Ethical leadership was also positively related to followers’ normative commitment after controlling for authentic leadership (b = .41). Thus, Hypotheses 1b and 1c were supported.

Table 4 Results of mediation analyses in Study 2 (Outcome: Turnover intentions)
Full size table

The bootstrap analyses demonstrated that the indirect effects of ethical leadership on turnover intentions via affective commitment were significant (b = −.32, 95% CI = −.63 to −.02), even after controlling for authentic leadership. The indirect effects of ethical leadership on turnover intentions mediated by normative commitment were also significant after controlling for authentic leadership (b = −.22, 95% CI = −.44 to −.07). Therefore, Hypotheses 3a and 3b were supported. Taken together, these results suggest that both affective and normative commitment play a mediating role in the relationship between ethical leadership and followers’ turnover intentions.

Discussion

Ethical leadership has become an increasingly important topic to organizational scholars and practitioners alike, given the revelations about unethical leader behavior that appear in the news on an almost daily basis. Indeed, ethical leadership has assumed a greater significance in the business world since the reports of corporate scandals preceding and following the great recession of 2008–2009, which had global implications and repercussions that echo even today. Scholars, however, have grappled with the challenges of differentiating ethical leadership from similar, though distinct, morally-oriented leadership approaches, such as transformational and authentic leadership. Researchers have argued that ethical leadership is an inherent component of these leadership approaches, raising the prospect of potential construct redundancy (Hoch et al., 2018; Shaffer, DeGeest, & Li, 2016). Moreover, most research on ethical leadership has been conducted in Western contexts, with relatively few studies focusing on how ethical leadership might impact individual and organizational outcomes outside Anglo cultures, such as in Confucian Asian culture. Although mixed findings have been published about the positive association of ethical leadership with followers’ task performance in Asian contexts (e.g., Liu et al., 2013a, b; Park et al., 2015; Schuh et al., 2013), this topic has not been systematically examined.

The purpose of this study was to investigate the unique effects of ethical leadership beyond the FRLM and authentic leadership in South Korea. Drawing on Meyer and Herscovitch’s (2001) model of organizational commitment, we examined the relationships among ethical leadership, organizational commitment, task performance, and turnover intentions. Our findings revealed that both affective commitment and normative commitment had mediating effects on the relationships between ethical leadership and task performance and turnover intentions even in the presence of the FRLM and authentic leadership.

Theoretical implications

Our studies have important theoretical implications for the ethical leadership and organizational commitment literatures. First and foremost, the findings of our research provide important evidence supporting the unique predictive power of ethical leadership on attitudinal and behavioral work outcomes beyond other, similarly related leadership behaviors. Identifying the predictive validity of ethical leadership is critical to lessen the concerns of construct redundancy, and thereby enhancing our accurate understanding of the unique effects of ethical leadership because scholars (e.g., Le, Schmidt, Harter, & Lauver, 2010; Morrow, 1983) have indicated that “the organizational sciences are plagued by construct proliferation and, consequently, construct redundancy” (Hoch et al., 2018: 520). In fact, Hoch et al.’s (2018) recent meta-analytic research showed that the correlation between transformational leadership and ethical leadership was .70. This evidence further demands scholarly efforts to elucidate the empirical distinctiveness among closely related leadership constructs. As such, our work responds to the calls for more research examining the unique effects of ethical leadership (Hoch et al., 2018; Ng & Feldman, 2015) and thus advances our understanding of the predictive validity of ethical leadership in the presence of the FRLM and authentic leadership.

Second, the findings of the studies suggest that the unique effects of ethical leadership on turnover intentions and task performance via followers’ affective/normative commitment, and that this impact is attributable to characteristics of ethical leadership that do not overlap with the FRLM and authentic leadership behaviors. As such, the present studies, using two separate samples from South Korea, confirm the findings of Ng and Feldman (2015) by demonstrating that ethical leadership is significantly related to followers’ affective/normative commitment even after accounting for the effects of transformational, contingent reward, and management-by-exception leadership behaviors. Our research also goes beyond the work of Ng and Feldman (2015) because it accounted for the effects of authentic leadership in examining the incremental validity of ethical leadership. Further, this paper investigated the mediating effects of two types of organizational commitment on objective measures of task performance and turnover intentions. To our best knowledge, our studies are among the first to demonstrate that ethical leadership has unique power to predict two kinds of commitment and the subsequent mediating relationship to performance and turnover intentions beyond the FRLM and authentic leadership in a non-Western context.

Third, our work helps to explain the underlying mechanisms by which ethical leadership enhances followers’ psychological and normative attachment to their organization in a South Korean context. By focusing on the conceptualization of organizational commitment and further expanding the theoretical connections to ethical leadership and the subsequent work outcomes tied into the concept of organizational commitment, this study adds a solid grounding that an ethical leader’s unique behaviors, beyond those behaviors recognized as being inspirational/visionary or authentic, may lead followers to internalize the leader’s value and attached to organization. In particular, the core behavioral characteristics of ethical leadership underline normatively appropriate decision making and procedures that trigger followers’ embrace of accountability based on strong ethical standards. Such an emphasis is aligned with the salient cognitions of obligations and duties toward the collective that were found as prevalent in South Korean (Markus & Kitayama, 1991).

Finally, our studies respond to calls for more research on ethical leadership in non-North American settings (Brown & Mitchell, 2010; Resick et al., 2011). Specifically, Asia presents a multitude of unique contexts that may affect the way in which leadership is manifested (Koo & Park, 2017). Our research, conducted in South Korea, suggests the critical role that affective and normative commitments play in the influence process associated with ethical leadership. The findings of the studies contribute to the literature by offering more nuanced knowledge about the ethical leadership process. As the formal definition of ethical leadership includes “normatively appropriate behavior” (Brown & Treviño, 2006b), it leaves the content of moral norms dependent on the respective cultural context as different cultures have distinct values and mindsets which can shape people’s implicit theories about leadership (Resick et al., 2011). For example, while character and accountability (e.g., conducts his/her life in an ethical manner, disciplines employees who violate ethical standards) are central to employee perceptions of ethical leadership in the U.S., leaders who demonstrate respect for others and look out for employees’ interest tend to be perceived as an ethical leader in the Confucian Asian cultures (Resick et al., 2011). These aspects of ethical leadership (consideration and respect for others) might have a less impact on task performance of followers compared to accountability which “reflects the U.S. endorsement of performance orientation as a cultural value” (Resick et al., 2011: 447). More importantly, we found significant indirect effects of ethical leadership on task performance. This result suggests that ethical leaders may exert a positive influence on followers indirectly through elevating their emotional and obligatory commitment to the organization. Therefore, focusing solely on the direct impact of ethical leadership on follower performance may impede researchers from unveiling a more complete picture of the ethical leadership process. This might be particularly true for studies conducted in a non-West context because the mediation mechanisms illuminate the issue of why ethical leadership may operate differently in a certain culture. Thus, the current studies advance our knowledge of ethical leadership in Asia by shedding new light on leadership mechanism at work. In doing so, the paper responds to calls for more empirical research on leadership in Asia to better understand contextual differences which may affect the way in which leadership is manifested (Koo & Park, 2017; Lam et al., 2012; Loi et al., 2012). Although the existing Western ethical leadership measure (Brown et al., 2005) has been found to be applicable across cultures (Resick et al., 2011) and thus our studies used this measure to capture a universal factor of ethical leadership, emerging research findings also suggest that leaders in a different culture manifest distinct sets of culturally specific ethical leadership behaviors. For example, a culturally specific ethical leadership measure in China has been developed and used to predict Chinese follower work outcomes beyond the Western ethical leadership measure (Wang et al., 2017). We call for continued research on context-specific research on ethical leadership to fully understand universal as well as unique aspects of ethical leadership process in different national and cultural contexts.

Practical implications

Knowledge of the incremental predictive power of ethical leadership in a non-Western setting has several important implications for ethical management across different cultures and in terms of the practical value of ethical leadership to organizations in such a region. First, our research suggests that ethical leaders foster affective and normative commitment among their followers, which in turn leads to better performance and lower intentions to quit. Enhanced employee performance and lowered turnover intentions resulting from ethical leadership could bring a number of benefits to the organization, including lower recruiting costs and increased productivity. Recent evidence also suggests that ethical leadership has far-reaching effects. Zhu, Sun, and Leung (2014) found that ethical leadership reinforces the relationship between corporate social responsibility and firm performance in Chinese firms. As one way of improving employee retention rates and enhancing employees’ performance, organizations in the Asian region might focus their efforts and resources on promoting followers’ loyalty and psychological attachment to the organization by having their managers display ethical leadership.

Second, as more organizations begin to operate globally, providing ethical leadership across different cultures may prove challenging, especially given that ethics can be relative across cultures (i.e., practices such as hiring one’s family members may be perceived as ethical in some cultures and as nepotism in others). A more nuanced understanding of followers and contexts might help organizations design and implement more effective human resources practices related to expatriate selection and leadership training.

Third, our research indicates that, compared with transformational, contingent reward, and authentic leaderships, which represent positive forms of leadership, ethical leadership can benefit followers and organizations. Accordingly, while organizations continue to encourage these positive leadership behaviors, they might be well advised to pay particular attention to cultivating ethical leadership within their own ranks. As leadership often trickles down to lower levels of the organizational hierarchy, senior managers’ ethical leadership can serve as a vehicle for creating an ethical climate within the overall organization. This can be achieved by selecting managers who demonstrate integrity and providing training programs focusing on ethical leadership.

Limitations and suggestions for future research

Like all studies, our research is not without limitations. First, the self-reported data collected in our studies might have inflated correlations among variables, resulting in potential common method bias. To mitigate this problem, we employed objective performance data provided by the company (while responses on leaderships, commitment, and turnover intentions were assessed by followers) in Study 1; thus, this study is one of the few to use an objective performance measures to study ethical leadership. Second, although turnover intentions have been used as an effective predictor of turnover in prior research (Hom, Mitchell, Lee, & Griffeth, 2012), it would be helpful for future researchers to use objective turnover data. Third, in Study 2, we were unable to control for the FRLM leadership behaviors as well as alternative mediating mechanisms (e.g., trust, psychological empowerment) suggested in the prior literature. However, by accounting for the effects of the FRML and authentic leadership in two data sets, which were individually gathered through various methodological techniques (i.e., online versus paper-and-pencil surveys, cross-sectional versus two-wave design, single organization versus multiple organizations), our two studies provide a rigorous testing of the unique validity of ethical leadership in a non-Western culture. Future research should extend the scope of investigation to not only other nations with Confucian cultures, but also other cultural clusters, to see if the findings hold up. Fourth, we did not directly integrate cultural values into the hypotheses although the samples used in the study were obtained from South Korea. Prior research suggests that employees with different levels of power-distance orientation may have varying views on supervisors’ leadership behaviors and, therefore, react to those behaviors differently (Kirkman, Chen, Farh, Chen, & Lowe, 2009; Lian, Ferris, & Brown, 2012). In a high power distance culture, because the entitlement to power and privilege is not derived mainly from supportive behaviors, but rather from the norm of reverence and obedience toward higher-status superiors (Bochner & Hesketh, 1994), leadership behaviors and initiatives might have a weaker impact on follower behavior than in a low power distance culture. As such, because Korean culture belongs to Confucian Asian cluster, which is characterized as high power distance, collectivism, and traditionality (Gupta & Hanges, 2004), examining the moderating roles of culture-related variables may further illuminate the boundary conditions under which ethical leadership operates. This topic was outside the scope of the current study, as our goal was to examine the unique validity of ethical leadership in predicting followers’ organizational commitment and work outcomes. We believe that it continues to be an important topic for researchers to probe leadership issues in a cross-cultural setting. Finally, it would be interesting to study different types of outcomes affected by ethical leadership. For example, employees in a collectivist and high-power-distance culture might be more likely to engage in unethical, pro-organizational behaviors and less likely to be whistleblowers owing to their strong ties to their organization. This is a rich area for future research.

Conclusion

Ethical leadership is an important practice for contemporary organizations not only to sustain organizational success, but also to take social responsibility for their workforces. The current research, using two separate samples from South Korea, reveals the evidence supporting the cross-cultural generalizability of the unique power of ethical leadership to predict followers’ performance and turnover intentions by disentangling the effects of the closely related leadership behaviors. In addition, the findings of our research suggest that affective and normative commitments serve as a psychological mechanism that connects ethical leadership with follower work outcomes. We hope that our work encourages future researchers to replicate and expand this line of research by investigating the unique validity of ethical leadership as a predictor of various types of individual and organizational outcomes and its intervening mechanisms underlying the relationships.