Treviño and Nelson (2004) highlighted that ethical phenomena have existed ever since the dawn of humanity. This is because the human being is, by nature, a moral being, with the capacity to distinguish good from evil, regardless of their birthplace, the historical moment in which he/she lives (21st century) or the religion he/she professes. However, in organizations, having a role model that shows the appropriate path forward is critical in fostering workplace ethical behavior. Even though individuals, as human beings, may question themselves as to “what should they do,” the guidance and direction of leaders play an important role in directing their decisions and actions toward an ethical direction (Trevino et al., 2000). Thus, the challenge of any person in the organization who occupies a relevant management position is to transmit its ethical essence to others in order to build a reputation for ethical leadership. Being seen as an ethical leader implies that others think of you as a person who has ethical qualities, who participates in ethical actions, and makes decisions based on ethical principles. The true ethical leader “walks the talk” and, by doing so, impacts the ethical lives and behaviors of others in the organization (Trevino et al., 2000; Brown and Treviño, 2006).

For the purpose of maintaining effective work situations and ethical standards in organizations, Brown et al. (2005) proposed the concept of “ethical leadership behavior.” This form of leadership recognizes the importance of virtuousness in action, including truthfulness and honesty, and can be defined as “the demonstration of normatively conducted behavior through personal actions and interpersonal relations” (Brown et al., 2005, p. 120). In practice, this means demonstrating aspects such as honesty, integrity, and principled decision-making (Brown et al., 2005).

Brown et al. (2005) highlighted that managers serve as representatives of organizations, meaning they are an important influence for employees in terms of behavior, for social learning (Bandura, 1977) and social exchange (Blau, 1964) motives. According to social learning theory (Bandura, 1977), employees learn appropriate behaviors by paying attention to and emulating the behaviors of attractive, credible role models such as managers. As such, if ethical leadership is practiced by managers, they are likely to influence employee ethical behavior positively (Ruiz et al., 2011; Ruiz-Palomino and Martínez-Cañas, 2011; Ruiz-Palomino and Linuesa-Langreo, 2018; Al Halbusi et al., 2020a). In line with social exchange theory (Blau, 1964), the fair and honest behavior demonstrated by these leaders toward their employees could engender in employees a willingness to trust their leaders (Mayer et al., 2009) and an obligation to reciprocate with positive, ethical behavior (cf., Gouldner, 1960). Thus,

H1. Ethical leadership of (senior) managers positively relates to employee ethical behavior.

Managers are conferred with the legal power to manage their employees and with taking charge of organizational resources; they are looked upon as carrying out a principal function in their organizations (Loi et al., 2009) and as being in an exclusive position to administer justice (Brown et al., 2005; Ruiz-Palomino et al., 2012). Thus, given that managers are representatives of the organizations (Hou et al., 2018), insofar as employees perceive their managers as ethical, justice, which is implicit in the concept of ethicality (cf., Colquitt, 2001), will be likely perceived as something that is present in their organizations as well. This reasoning become firmer in the case of managers who practice ethical leadership in senior management levels, as these leaders are more likely to be perceived as representative of the organization. In particular, inasmuch as senior managers are perceived as ethical, employees will see that justice is present in the (a) outcomes achieved (distributive justice), (b) procedures realized (procedural justice), (c) relationships established (interpersonal justice), and (d) information received (informational justice) in their organizations (Colquitt, 2001). This is likely to have positive effects on their ethical behavior (Ruiz-Palomino et al., 2012), by activating social exchange processes (El Akremi et al., 2010), as described below.

Rooted in the equity theory developed by Adams (1963) and associated with the equal distribution of outcomes based on the performance of each employee (Burney et al., 2009), distributive justice and its perception in the organization should lead employees to more ethical behaviors. When distributive justice is perceived, the conditions at work should be perceived as more positive (Oshio and Kobayashi, 2009) and reciprocity and social exchange processes should be developed (Gouldner, 1960; Blau, 1964), thus leading employees to respond with positive behaviors toward the leader and organization, such as ethical behavior. In a similar vein, procedural justice, which emphasizes the perceived fairness of the processes followed to make decisions (i.e., Greenberg, 2001; procedures and policies used to determine outcomes or resource distributions, Colquitt, 2001), should lead to more ethical behavior among employees. In effect, a higher level of perceived procedural justice will be accompanied by employees perceiving that they have some voice over the outcome (Lind and Tyler, 1988). According to social exchange processes (Gouldner, 1960; Blau, 1964), this is likely to lead employees to engage in positive behaviors toward the leader and the organization, such as ethical behavior (McCain et al., 2010). As in the procedural and distributive cases, interpersonal justice, defined as the extent to which one perceives that they are being treated with dignity, respect, and fairness (Colquitt, 2001), is also likely to lead to ethical behavior. If employees are well-treated, the norm of reciprocity (Blau, 1964) suggests that these employees will respond to the organization with positive behaviors. Finally, informational justice, which refers to the extent to which communication is developed honestly, fairly (Bies, 2001, and clearly (Colquitt, 2001), is also likely to lead to higher levels of ethical behavior. Employees will see themselves as valued and appreciated and so are likely to respond with positive behavior, in line with reciprocity processes (Gouldner, 1960). For example, in downsizing contexts, informational justice perceptions have been shown to make employees develop trusting attitudes (Brockner et al., 1995), which is one important driver of social exchange processes (Colquitt et al., 2007).

Overall, the extent to which each of these justice dimensions (distributive, procedural, interpersonal, informational) can be shaped by managers' ethical leadership, organizational justice, and each of its different dimensions, is likely to mediate the managerial ethical leadership to employee ethical behavior relationship. Accordingly, we predict the following,

H2. Perceived organizational justice and its dimensions of distributive (a), procedural (b), interpersonal (c), and informational (d) justice each mediate the positive relationship between the ethical leadership of (senior) managers and employee ethical behavior.

Figure 1 shows all the hypotheses proposed in the form of a research model. It includes two main hypotheses (H1, H2) and four sub-hypotheses (H2a, H2b, H2c, H2d).

After using Brislin's (1980) back-translation procedure, the survey questions were pretested using cognitive interviews with 18 employees from the manufacturing industry. They suggested slight semantic adjustments and confirmed the questionnaire's clarity, readability, comprehension, and suitability. To obtain reliable responses, the questionnaires were distributed to employees with an organizational tenure of at least 6 months. The questionnaires were thus distributed to 350 employees who reported directly to upper/middle managers in 12 manufacturing firms located in Selangor (Malaysia), worked full-time, and had frequent contact with their immediate managers. These questionnaires were distributed in two waves to reduce the occurrence of common method variance (CMV), as recommended (Podsakoff et al., 2012). In the first wave, respondents provided demographic information, together with perceptions of their managers' ethical leadership and organizational justice. In the second wave, after 3 weeks had passed, these respondents rated their own level of ethical behavior. A cover letter was also distributed to respondents, assuring them of total confidentiality and informing them about how important their participation was in this study, which could help reduce social desirability bias (SDB) and CMV (Podsakoff et al., 2003). Both sets of surveys were coded to confirm that the responses could be matched, resulting in 270 valid responses and a 77% response rate. 74.4% of the respondents were men (25.6% women), and the largest group (41.9%) fell in the range between 31 and 40 years. In terms of educational level and job experience, most participants had a bachelor's degree (55.9%), and had been working for the same company for 6–10 years (75%).

The nature of the variables for this study enabled us to differentiate first- (i.e., ethical leadership of senior managers, employee ethical behavior, social desirability bias) and second-order constructs (i.e., organizational justice), captured in Mode A (reflective) constructs, as recommended (Hair et al., 2017). All measures relied on five-point Likert response formats, and apart from organizational justice perceptions (1= “to a small extent” to 5= “to a large extent”), all were based on the respondents' level of agreement on each item surveyed (1 = “strongly disagree”; 5 = “strongly agree”). Table 1 shows the items of these variables.

In addition to the above-described ex-ante procedural remedies, one post-hoc test was conducted to evaluate whether CMV could have biased our findings. It revealed no problems of this type. We ran in AMOS v.24 the common factor tests to compare the difference in standardized weights with and without a common latent factor (CLF); if this difference is >0.2, CMV could be a problem in the data (Podsakoff et al., 2003, 2012). In our study, all the differences were lower than 0.2. Furthermore, the CLF model did not yield a significantly better fit (χ2/df =1.514; CFI = 0.967; RMSEA = 0.0413) compared with the full measurement model (χ2/df = 1.511; CFI = 0.964; RMSEA = 0.0416). This confirms that CMV is not likely to be a serious concern in this study (Podsakoff et al., 2003, 2012).

To assess the measurement model, we examined individual item reliability, internal consistency, convergent validity, and discriminant validity (Tables 1, 2). In terms of item reliability, most items exceeded the recommended 0.707 level; otherwise, items generally were above the minimum 0.5 threshold (Hair et al., 2017; see Table 1). To evaluate the internal consistency of the measurement scale, Cronbach's Alpha (CA) and Composite Reliability (CR) were utilized in this study. The ranging of both indices was far above the acceptable level of 0.707, so internal consistency reliability can be supported (Hair et al., 2017). Average Variance Extracted (AVE) was conducted to measure the convergent validity of the study constructs. The evidence of convergent validity was confirmed because the AVE for constructs exceeded the threshold of 0.50, as recommended (Hair et al., 2017). Discriminant validity was also confirmed by HTMT inference and Fornell-Larcker criterions; HTMT values were significantly different from 1, and the square roots of AVE for each variable were greater than the correlation of each variable with the others (Hair et al., 2017; see Table 2).

None of the demographic variables (i.e., gender, age, education, and job experience) showed a significant impact on employee ethical behavior (Figure 2). The influence of SDB was not significant either (β = 0.04 ns, Figure 2), thus showing no influence of this bias in this study. Regarding hypothesis testing, the ethical leadership of (senior) managers was observed to influence employee ethical behavior positively (β = 0.239, p < 0.001; Table 3, Figure 2), in support of H1. Our findings also reveal support for our prediction in H2 and its sub-hypotheses. First, a significant indirect effect of the ethical leadership of (senior) managers on employee ethical behavior through organizational justice exists (indirect effect= 0.170, p < 0.001; Table 3, Figure 2), in support of H2; in terms of effect size (f²), this effect is large (f² = 0.550; Table 3, Figure 2). Second, the results also give wide support to the sub-hypotheses derived from H2. As Table 3 shows, distributive justice mediates between managerial ethical leadership and employee ethical behavior (indirect effect = 0.066, p < 0.01), in support of H2a. Procedural justice mediates in this relationship as the indirect effect is positive and significant (indirect effect = 0.049, p < 0.01). Table 3 also supports mediation for interpersonal justice, by revealing a significant, positive indirect effect of ethical leadership on employee ethical behavior (indirect effect = 0.324, p < 0.001). Finally, informational justice was also a significant mediator in the relationship (indirect effect = 0.242, p < 0.001, Table 3). Thus, each of the dimensions of organizational justice (distributive, procedural, interpersonal, and informational) are significant mediators, and H2a to H2d can be supported (Table 3, Figure 2). Importantly, however, some justice dimensions were more important mediators than others. Whereas, interpersonal and informational justice showed the strongest size effects (f² = 0.116 for interpersonal justice; f² = 0.063 for informational justice), procedural and distributive justice showed the weakest effect sizes (small to moderate for procedural justice, f² = 0.045, and small for distributive justice, f² = 0.017).

In terms of explanatory power, the model explains moderate to substantial variance (R² = 0.587) of employee ethical behavior (Hair et al., 2017). The Stone-Geisser blindfolding sample reuse technique revealed Q²-values larger than zero, thus indicating good predictive power for overall organizational justice (Q² = 0.161) and each of its dimensions (Q²_distributive = 0.037; Q²_procedural = 0.129, Q²_{interpersonal} = 0.134, Q²_{informational} = 0.126), and employee ethical behavior (Q² = 0.222) (Hair et al., 2017). Finally, in terms of overall goodness-of-fit (GoF), the SRMR index (standardized root means square residual) offered a value of 0.041, which is far below the 0.08 cut-off (Henseler, 2017); the SRMR's 95% bootstrap quantile was 0.054 and therefore higher than the SRMR value, indicating that the model has a good fit (Hair et al., 2017). The discrepancy indices dULS (unweighted least squares discrepancy) and dG (geodesic discrepancy) were also below the bootstrap-based 95% percentile (dULS = 1.421 < HI 95 of dULS = 2.521; dG = 0.562 < HI 95 of dG = 0.989), thus confirming good model fit and indicating that the model tested in this study is likely to be valid (Henseler, 2017).

This study makes an important contribution to the ethical leadership and ethical behavior literature. Prior studies have found ethical leadership to have a positive effect on employees' ethical behavior (e.g., Mayer et al., 2009; Ruiz-Palomino and Linuesa-Langreo, 2018), but the mechanisms connecting the ethical leadership of senior managers to employees' ethical behavior are still not well-established (Treviño et al., 2014). This study fills this gap to some extent and puts forward organizational justice perceptions as a possible mechanism between ethical leadership of senior managers and employee ethical behavior. In particular, an interesting advancement of the literature is showing the strong connection between the ethical leadership phenomenon and the development of justice perceptions regarding the procedures (i.e., procedural justice), the outcomes received (distributive justice), the relationships that are held with the manager (interpersonal justice), and the information received within the organization (informational justice), as well as the differentiated mediation role of each of these justice dimensions in the ethical leadership–employee ethical behavior relationship. Thus, for example, the interpersonal and informational dimensions of organizational justice have the strongest individual mediating roles in this relationship. This is in line with previous research showing that, compared to procedural and distributive justice, interpersonal and informational justice tend to have stronger effects on employee attitudes and behaviors in general (Williams et al., 2002; Ambrose and Schminke, 2003). This is probably because procedural and distributive justice, which are about the fairness of decision-making processes and its outcomes, are linked to “resource exchange” contexts, whereas interpersonal and informational justice involve “encounters” (Bies, 2005). The former are more infrequent than the latter, as these occur between managers and employees, suggesting that interpersonal and informational justice have a stronger “day-to-day” significance than procedural and distributive justice. They could therefore exert a stronger influence on employees' workplace behavior, including ethical behavior.

Another important implication is that the current investigation brings ethical leadership, organizational justice—including all its dimensions—and employee ethical behavior together in a non-Western context (Malaysia). Existing ethical leadership research has typically over-emphasized Western countries (Resick et al., 2011), but new investigations focusing on other cultural contexts are necessary to generalize the robustness of the theory underlying the ethical leadership phenomenon and its positive outcomes in the workplace. This study is set in a multi-racial country (Malaya, Chinese, Indian; Weintraub, 2011), and reveals the influence of ethical leadership on employee ethical behavior via the way it shapes organizational justice. This represents a positive development regarding the generalization of the direct and mediated relationship between ethical leadership and employee ethical behavior across distinct ethnicities. Malaysia is known for its extremely “high power” distance compared to the US and other Western cultural contexts (Hofstede Center, 1967–2010). Malaysia is also a country where Islam is predominant (Weintraub, 2011), a religion on which ethical leadership research is still not abundant (Metwally et al., 2019). Thus, by demonstrating the positive effect of ethical leadership on employee ethical behavior via organizational justice, and the stronger mediation size effect of interpersonal and informational justice dimensions in such a relationship, this study helps give robustness to the theory of ethical leadership and its outcomes across national cultural geographies.

From a managerial point of view, this study has a number of important implications. First, in identifying that ethical leadership of senior managers is important in fostering ethical behavior in the workplace, we make an important contribution to managerial practice. All efforts to understand the mechanisms that help such behavior to emerge are interesting from a managerial perspective due to the important benefits generally associated with having ethical employees. Ethical employees are more willing to put in extra effort to finish their work on time and follow the ethical standards indicated by their employer; they try to be a representative of the company in terms of public relations even when they are not at work, and make efforts to promote the organization's development (Treviño et al., 2014). Therefore, given that the practice of ethical leadership at all managerial levels can serve to spread ethical behavior within the organization, we believe that Human Resource (HR) managers should devote time and attention to hiring and training candidates for managerial positions. In other words, considering the benefits of ethical leadership in increasing employee ethical behavior, HR managers need to try their best to recruit, select, and train suitable persons who are willing to practice and develop an ethical leadership approach. Organizations should use a variety of assessment tools such as focus groups, in-basket exercises, structured interviews, and business games that focus on or involve relevant ethical issues to evaluate whether candidates have core traits of ethical leaders (e.g., honesty, fairness, caring, Brown et al., 2005). Organizations can also design training programs oriented toward developing managers' ethical leadership behaviors in their day-to-day activities. These could include strategies to support trainees (specifically, current or future managers) on how to make ethical decisions or act ethically, communicate about ethics and values, and/or set up ethical examples for others (e.g., subordinates or team members) in their respective organizations.

Second, although the practice of ethical leadership by managers impacts the ethical behavior of employees directly, there is also an indirect influence through enhancing employees' organizational justice perceptions. Thus, organizational justice becomes an important mechanism through which ethical managers prompt ethical behavior among their employees. Managers should thus become aware that all efforts directed to make employees perceive that justice is present in the outcomes achieved, the procedures realized, the supervisor–employee relationships that are established, and the information that is received can have a significant role in fostering ethical behavior in the workplace. Of special importance is that managers can be trained in providing their followers with proper rationales of their decisions and on treating them with high levels of dignity and respect (interpersonal justice). Also, managers could increase their followers' employee ethical behavior by ensuring that smooth top–down communication channels become established, which can help employees to obtain first-hand information about what is of relevance to their interests (informational justice).

The first limitation of the current study concerns our cross-sectional data design, which makes it difficult to provide definitive conclusions about causality. We recommend that future research uses experimental or longitudinal designs that will help reinforce the causality findings of the current study. In addition, our conclusions are limited by the cultural context of the study (Malaysia), so we recommend that future studies replicate our study in other cultures to improve the study's external validity and generalizability.

The second limitation relates to the data we used. In this study, our data came from a single source. Although we conducted two waves of surveys (Podsakoff et al., 2012), we cannot rule out CMV completely, and the rigor of our empirical findings could have been negatively affected, albeit minimally. However, the post-hoc test conducted (i.e., common latent factor, Podsakoff et al., 2003) revealed no serious concerns regarding CMB. In addition, because data was not collected from multiple sources, we could not rule out social desirability response bias completely. However, we included in our analysis a variable that tried to measure the extent to which respondents were more prone to answer the questionnaire in a socially responsible manner, so our findings were controlled by this potential bias, as recommended (Randall and Fernandes, 2012). Future studies, however, could collect data from multiple sources. Although ethical behavior is not easy to observe (Randall and Fernandes, 2012), future researchers could obtain more reliable data by having different informants (Treviño et al., 2006): close workmates rather than supervisors, as the former will likely offer more reliable information. Also, using laboratory research designs such as in-basket exercises and computer simulations could help to obtain highly reliable information (Treviño et al., 2006); these techniques allow respondents to be presented with multiple scenarios, where only some have ethical implications, so respondents are less likely to note that their ethical decision-making or behavior is being measured.

Third, ethical leadership and organizational justice perceptions appear to be related to employee ethical behavior, this association is likely to be dependent on certain boundaries. The workplace environment has an enormous influence on ethical decision-making processes; however, individual differences also play an important role (O'Fallon and Butterfield, 2005; Craft, 2013). In this regard, recent research exploring cognitive processing of moral cues suggests that employees differ in the extent to which they pay attention to moral issues and therefore differ in the moral attentiveness they show in their day-to-day activities (Reynolds, 2008). Thus, the moral cues offered by either ethical leadership or the fair outcomes, processes, interactions, and communication perceived (i.e., organizational justice) may be interpreted or captured in a different way by individuals who have higher rather than lower moral attentiveness. With greater moral attentiveness, the moral clues perceived thanks to the presence of ethical leaders and organizational justice may be more salient. This offers an interesting future line of research to advance on how ethical leadership can become more effective at work.

Fourth, there are major variations in Malaysia in terms of the beliefs, spirituality and religion of each ethnic group (Weintraub, 2011). Future researchers could thus implement studies that could evaluate if these differences in beliefs, religiosity, and spirituality may affect our findings (Whetten, 2009; Ribberink et al., 2018). Importantly, since this study was conducted in the context of an “Islamic” perspective, variables such as “Islamic work ethic” and “shariah compliance” may constitute critical conditional variables to be explored.

Finally, in this study, we did not examine other leadership styles (e.g., transformational leadership, transactional leadership) and did not devote efforts to assessing their influence on employee ethical behavior. However, some partial overlap can exist between ethical leadership and other leadership styles (Brown et al., 2005; Brown and Treviño, 2006). For instance, Brown et al. (2005) suggest that ethical leadership could include a series of transformational behaviors, such as engaging in ethical decision-making and showing genuine concern for subordinates' well-being. Managers may also display transactional behaviors to promote employee ethical behavior, and these behaviors may include aspects such as communicating ethical standards or (punishing) rewarding (un)ethical behavior, which is very characteristic of ethical leaders (Brown et al., 2005). Therefore, future inquiries could explore differences vs. similarities of the ethical leadership approach compared to other leadership approaches in explaining employee ethical behavior via organizational justice.