Ethical Leadership Supports Voice Behavior: Evidence from Vietnamese Service Firms

The research of ethical leadership has emerged as an important topic in relation to understanding the effects of leadership within an organization. Besides, in the last decades, there are numerous studies have been done that investigate some extra role behaviors, their antecedents, and consquences. Though voice behavior is not required in one’s job, it can contribute directly to the organizational contextual performance and indirectly to the job performance of individual employee as well as the performance of the whole organization. In this study, we propose that leader-member exchange (LMX) as a mechanism reflecting how ethical leadership affects the voice behavior of employees. We develop a mediation model of the psychological processes linking perceptions of ethical leadership and employee voice. This study examined the relationship between ethical leadership, leader-member exchange and the two components of voice behavior, defined as promotive voice and prohibitive voice, using a sample of 1238 supervisor-subordinate dyads from Vietnamese service firms. Results showed that ethical leadership related positively to promotive voice and prohibitive voice. In addition, leader-member exchange mediated the relationship between ethical leadership and promotive voice as well as prohibitive voice. Limitations of the study, directions for future research, and implications of the findings are discussed.


Introduction
Voice behavior has been the subject of research for the past decades. Employee Voice behavior was defined as speaking up with suggestions as well as a concern [1]. Voice is the expression of constructive opinions, concerns, or ideas about work-related issues [1]. Voice is essential for organizations to enhance their current status and to achieve sustainable development [2]. We follow voice as behavior that is not in an employee's job description and may or may not is rewarded by organizations. Though voice behavior is not required in one's job, it can contribute directly to the organizational contextual performance and indirectly to the job performance of individual employee as well as the performance of the whole organization [3].
Voice behavior can be classified into two categories: promotive voice and prohibitive voice. Promotive voice refers to employees' expression of new ideas or suggestions for improving the overall functioning of their work unit or organization [4]. Promotive voice is challenging because it proposes ways of changing the status quo. In contrast, prohibitive voice refers to employees' expressions of concern about work practices, incidents, or employee behavior that are harmful to their organization [4]. Prohibitive voice plays an important function for organizational health because such alarming messages place previously undetected problems on the collective agenda to be resolved or prevent problematic initiatives from taking place [4]. Yuan et al., (2018) pointed out that although the prior study had focused on the individual-level antecedents of voice behavior, organizational learning, social exchange structures, and values as an organizational contextual factor explained more variance in employee voice promotive and prohibitive voice. They concluded that by paying more attention to organizational context, researchers can improve their predictions on employee voice [5].
In the literature, various types of leadership, as organizational contextual factors, have been related to employees' attitudes and behaviors. For example, authentic leadership was found related to workplace deviance behavior [6]. Transformational leadership was found related to employee well-being and performance [7]. Researchers have specifically concerned themselves about the effects of ethical leadership on employee voice behavior. Ethical leadership is different from other types of leadership in its emphases on moral management, an explicit attempt of the leaders to affect followers' ethical conduct by setting ethical standards and using reward and punishment to hold the followers accountable for the standards [8]. Through transformational, spiritual, or authentic leaderships share the same concern for others and ethicality in decision-making, they nonetheless typify their unique characteristics. Transformational leadership stresses the leader's role in providing vision, values, and intellectual stimulation to employees [8]. Spiritual leadership emphasizes the leader's role in the provision of vision, hope, and faith to employees [8]. Authentic leadership refers to the authenticity of leader and the self-awareness of employees [8]. Different from ethical leadership, these leaderships are less involved with the transaction process of ethical management [8]. The focus of ethical leadership on the practical management of subordinates' ethical conduct makes it more relevant to employee promotive voice and prohibitive voice than other leadership styles because it can enhance caring of follower about their coworkers and organizations.
To understand the positive influence of ethical leadership on employee voice behavior, researchers have relied on the explanations of social learning theory and social exchange theory [8,9]. According to the social learning theory, ethical leaders serve as role models for the well-being of others, and their employees will emulate the leaders' exemplary behaviors and become prosocial toward their coworker and organizational [10,11]. On the other hand, the social exchange theory emphasizes that as ethical leaders care for the benefit of their employees and organizations, their followers are likely to perform well and feel more comfortable speaking up [12]. According to several previous studies, there is a positive relationship between ethical leadership and employee voice [13,14]. However, there remains a lack of research regarding the mechanisms by which ethical leadership influences employee voice behavior. This study aims to address this research gap by exploring the mechanisms that link ethical leadership to employee voice behavior.
Recently, some scholars used a different perspective-the social comparison theory-to explain the psychological processes that generate employee voice behavior. Because employees experience a favorable comparison against their coworkers on job achievement, the treatment received from their leaders [15], the exchange relationships with their leaders [16], they will engage more in speaking up. Vidyarthi et al. (2010) found that the positive effect as a consequence of favorable social comparisons also had a positive effect on employee work behaviors. A central premise of leadermember exchange theory is that differentiated social exchange relationships within a work unit act as the motive behind employees' reciprocal behaviors. Because employees often compare with each other the treatment that they receive from their leaders [17], the leadership of their leaders may affect their voice behavior through the process of social comparison. The perspective of social comparison may provide an explanation for why ethical leadership can affect employee voice. In the literature, there is a lack of research conducted to validate this explanation. The purpose of the present study is to test the influence of ethical leadership on employee voice behavior based on the social comparison perspective. We will used leader-member exchange as a mechanism to explain the influence of ethical leadership on employee voice when verifying the social comparison explaination. By anchoring on leader-member exchange we can validate more precisely the explanatory power of the social comparison theory on the effect of ethical leadership on employee voice.

Perceived Ethical Leadership and Employee Voice Behavior
Recently, ethical behavior of leaders has received more attention in both the mass media and the business literature. The concept of ethical leadership which was advanced by [8] demonstrates normatively appropriate conduct through personal actions and interpersonal relationships. Ethical leadership promotes followers through two-way communication, reinforcement, and decision-making. In proposing the theory of ethical leadership [18] suggested that ethical leadership behavior plays an important role in promoting enhanced employee attitudes and behaviors. The effect of ethical leadership has been a frequent subject of research in recent years [10,19,20]. Researcher have found that ethical leadership can not only increase task performance of followers [21][22][23] but also enhance prosocial behavior beyond the followers' immediate role, such as employee voice [12,24]. First, ethical leaders are interested in followers' thoughts and feelings and are concerned about followers' opinions. This attitude encourages followers to speak up suggestions relating not only to work-related issues but also to ethical matters [13,18]. Second, ethical leaders attempt to set up appropriate follower conduct and therefore stress the importance of ethical behavior. This issue is a good example for employees and encourages them to give voice. Besides, ethical leaders who encourage and recompense employees who give voice reinforce this learning process [13,14]. Third, ethical leaders interact truthfully and openly with employees. The result will enhance interpersonal trust between leaders and employees [18]. When employees trust that they are treated fairly and respectfully by their leaders, they are likely to reciprocate through constructive voice behavior [25]. Finally, ethical leaders improve other aspects of the work environment by broadcasting the importance of voice, issuing incentives for employee voice, and increase voice legitimacy, among other practices. These practices bring about a supportive environment in which employee feel that stating their opinions is both safe and meaningful [13,26]. Several empirical research have found that ethical leadership is significantly and positively associated to employee voice [27,28], and employees' willingness to report problems to management [18].
Based on these arguments, we proposed the following hypothesis: Hypothesis 1. Perceived ethical leadership has a positive effect on employee promotive voice and prohibitive voice.

Perceived Ethical Leadership and Leader-Member Exchange
Leader-member exchange theory has received extensive attention in organizational sciences [29,30]. The foundation of leader-member exchange is the degree of emotional support and exchange of valued resources [31] between an employee and his or her immediate supervisor. Hence, leader-member exchange is a social exchange relationship between a supervisor and his or her direct report [32]. According to social exchange theory, employees tend to develop high-quality relationships based upon whom they interact with, how they interact with them, and their experiences with them. This makes leadership an important currency in social exchanges [33]. Thus, we argue that leader-member exchange leaderships are developed through a series of interactions or exchanges between leaders and employees. Immediate supervisors are critical in increasing the leader-member exchange relationship because of their proximity to employees.
There are many ways that ethical leaders can increase highquality leader-member exchange with their employees. First, ethical leaders are seen as moral persons who are trustworthy and honest; They are also viewed as principled decision makers who care more about the stronger good of followers, the organization, and society [8,18]. When employees perceive that leaders act in their best interests and are caring, employees infer that leaders are committed to them. The results are increased high-quality leader-member exchange because ethical leadership makes to enhance high levels of loyalty, emotional connections, and mutual support [34]. Also,  argued that "Because ethical leaders are caring…relationships with ethical leaders are built upon social exchange and norms of reciprocity" [35] (p. 967). Ethical leaders inform their employees of the benefits of ethical behavior and the cost od inappropriate behavior and then use balanced punishment to hold employees accountable [18]. As a result, ethical leaders can develop meaningful interpersonal relationships that go beyond specified economic exchange agreements by encouraging employees' opinions [8], therefore facilitating high-quality LMX.
Based on these arguments, we proposed the following hypothesis: Hypothesis 2. Perceived ethical leadership has a positive effect on leader-member exchange.

Leader-Member Exchange and Employee Voice Behavior
The notion of voice stems from the idea that dissatisfied employees may desire to actively express their ideas and advocate changes to improve their situation or to improve organizational well-being [36]. Although voice research mainly considers verbal suggestions and concerns, voice can include nonverbal endeavors [37]. Voice behavior, being discretionary, and not listed informal job duties, is not a job requirement [4].
Voice can be risky because it intends to change policies or procedures [38]. Employees often calculate costs and benefits before voicing concerns [39]. Besides, to such cognitive calculations, affective forces, such as feelings of obligation, may encourage them to voice as well [40]. The desire to correct organizational problems prompts employees to consider potential impact, willingly assume risks, voice their concerns, and act to improve the status quo.
Leader-member exchange may facilitate employee voice. In high-quality leader-member exchange relationships, both parties have a fondness, loyalty, and respect for each other, encouraging mutual obligation and reciprocity [41]. Employees, likely to be broadly and deeply invested in the relationship, work beyond their formal job descriptions in return, for example, by choosing to help others [42] or initiating changes [43]. Moreover, high-quality relationships may reduce perceptions about potential risks in expressing voice. Even if their leaders do not support their ideas, employees are assured that the leaders will avoid criticizing them for expressing their concerns and will adopt suggestions that seem beneficial. Under such relationship contexts, employees willingly express their ideas. Thus we propose: Hypothesis 3. Leader-member exchange has a positive effect on employee promotive voice and prohibitive voice.

Mediation Effect of Leader-Member Exchange on the Perception of Ethical Leadership -Employee Voice Behavior Link
Voice behaviors that challenge the status quo in organizations often encompass personal costs and risks. The goals of voice behaviors are usually direct or indirect managers. Therefore, the relationship between an employee and his/her leader will affect his/her evaluation of the costs and benefits of voice behavior. The higher quality leadermember exchange relationship employees have, the higher trust in and the more understanding of their supervisors they have. As a result, they can express their ideas and positions more openly and freely, and they do not need worry about being misunderstood. Moreover, high-quality leader-member exchange expresses that the leader-member relationship progresses to a partnership level [44]. Followers apprehend their leaders' position and perspective; therefore, they can combine their leaders' concerns and try to achieve the collective good [45]. In this type of relationship, followers may not consider their personal costs and risks as a primary concern, and thus have higher intrinsic motivation and obligations to make constructive suggestions and improve organizational practices [46]. Leadership behaviors may shape the dyadic relationship between leaders and members and further affect members' behavior. Given that ethical leadership helps develop high-quality leader-member exchange, and high-quality leader-member exchange encourages employee voice behavior. Thus, leader-member exchange quality should play a mediating role in the relationship between ethical leadership and voice behaviors.
Based on these arguments, we proposed the following hypothesis: Hypothesis 4. Leader-member exchange mediates the positive influence of perceived ethical leadership on employee promotive voice and prohibitive voice; that is, an increase in perceived ethical leadership increases leadermember exchange, and the increase in leader-member exchange can then result in an increase in promotive voice and prohibitive voice.

Sample and Procedures
Participants in this study were employees of service firms in Vietnam with about 3000 employees of 172 service firms. Separate questionnaires were developed for supervisors and subordinates. The supervisor questionnaires were distributed to 545 supervisors, and the subordinate questionnaires were distributed to 1635 immediate subordinates of these supervisors. In other words, three immediate subordinates of each supervisor received the subordinate questionnaires. Those who completed the supervisory questionnaire were asked not to fill in the subordinate questionnaire. The number of questionnaires returned was 457 questionnaires and 1318 subordinate questionnaires, representing response rates of 83.9% and 80.6%, respectively. After deleting records with unmatched supervisor-subordinate pairs, a total of 1238 supervisor-subordinate dyads (1238 subordinates and 432 supervisors) remained and constituted the sample for this study.

Measures
The two questionnaires contained the measures used in the present study. First, the supervisory questionnaires contained the voice behavior measures, in which supervisors were asked to evaluate the promotive voice and prohibitive voice behavior of their immediate subordinates. Second, the subordinate questionnaire contained measures of ethical leadership, leader-member exchange, and demographic variables. All items used in the present research were in Vietnamese service firms. To assure equivalence of the measures in the Vietnamese and the English versions, a standard translation and back-translation procedure was performed. All items were measured on a 5-point Likert-Scale (1="Strongly Disagree; 5="Strongly Agree").
Ethical leadership. Ethical leadership was measured by a 10 item scale taken from [18]. These items were also adopted in the studies of [5]. A sample item is "My supervisor makes fair and balanced decisions." The Cronbach alpha for this scale was 0.92.
Leader-Member Exchange. We assessed leader-member exchange with a scale originally developed by [47]. These items were also adopted in the studies of Yuan et al. (2018) [5]. A sample item is "My supervisor understands my job problems and needs." The Cronbach alpha for this scale was 0.91.
Promotive voice and Prohibitive voice. These scales were measured by a 10-item scale taken from Liang et al. (2012) [48], with five items to measure each construct. Sample items are "This subordinate raises suggestions to improve the unit's working procedure" (promotive voice) and "This subordinate dares to point out problems when they appear in the unit, even if that would hamper relationships with other colleagues" (prohibitive voice). The Cronbach alpha was 0.85 for promotive voice and 0.87 for prohibitive voice.
Subordinate demographics. Four demographic attributes were measured and used as control variables. Age, education, tenure, and gender. Age, gender, and tenure were measured by the number of years. Gender was coded with 0 designating man and 1 designating women. These variables were included because demographics may influence the voice behavior of employees.

Construct Validity of the Measurement
We performed confirmatory factor analyses (CFA) on the four variables in our study (i.e., perceived ethical leadership, leader-member exchange, promotive voice, and prohibitive voice) to establish their discriminant and convergent validities. Model fit was evaluated using the following indices: the comparative fit index (CFI), Tucker-Lewis index (TLI), standardized root mean square residual (SRMR), and root mean square error of approximation (RMSEA), CFI and TLI values greater than 0.95 indicate a good fit [49]. SRMR values less than 0.08 indicate a good fit [49]. RMSEA values of 0.06 or less indicate relatively good fit [49]. The results in Table 1 showed that the four-factor model demonstrated good fit to data (χ 2 [371]=510.94; p=0.000; CFI=0.97; NFI=0.96; TLI=0.97; SRMR=0.03; RMSEA=0.04). The Chi-square difference tests also showed that this four factor model was a better fit compared to other plausible alternative models (See models 2-6, Table 1). The results suggested that satisfactory discriminant validity was attained on the measurement of the four variables. Moreover, the factor loadings of the items in each of the four variables were all greater than 0.50 and significant at the statistical level of 0.01, a result indicating the attainment of satisfactory convergent validity on the measurement of the variables [50].

Hypothesis Testing
Hypothesis 1 proposed that perceived ethical leadership has a positive effect on employee promotive voice and prohibitive voice. According to the results from Model 2 and Model 4 of Table 3, perceived ethical leadership was positively related to employee promotive voice (β=0.07, p<0.01), and employee prohibitive voice (β=0.29, p<0.001), thereby supporting Hypothesis 1. Hypothesis 2 predicted that perceived ethical leadership has a positive effect on leadermember exchange. The results from Model 1 of Table 3 showed that perceived ethical leadership was positively related to leader-member exchange (β=0.29, p<0.001), thus supporting Hypothesis 2. Hypothesis 3 stated that leader-member exchange has a positive effect on employee promotive voice and prohibitive voice. The results form Model 3 and Model 5 of Table 3 showed that leader-member exchange was positively related to employee promotive voice (β=0.12, p<0.001), and employee prohibitive voice (β=0.18, p<0.001). Hence, Hypothesis 3 was supported.
To examine the mediation effect of leader-member exchange on the relationship between perceived ethical leadership and employee voice behavior suggested in Hypothesis 4, the first step requires a significant relationship between perceived ethical leadership and employee promotive voice and prohibitive voice, which we have found. Following the procedures of [51], we entered both perceived ethical leadership and leader-member exchange simultaneously into the regression equation in Model 3 and Model 5 of Table 3. We found that the regression coefficient (β) of perceived ethical leadership on employee promotive voice and prohibitive voice dropped from 0.07 to 0.03 and from 0.29 to 0.24, respectively, suggesting that there could be a mediation effect of leader-member exchange. To examine further whether the mediation was significant, we used the bootstrapping procedures of [52] to calculate the mediation effect and used a 95% confidence interval (CI) to examine its significance. The bootstrapping tests confirmed that the mediation effects of perceived ethical leadership on employee promotive voice and prohibitive voice through leader-member exchange were significant. Specifically, for employee promotive voice, the 955 CI of the mediation effect was (0.010, 0.054), not containing zero; for employee prohibitive voice, the 95% CI of the mediation effect was (0.012, 0.053), which also contained no zero. Together, these above results supported hypothesis 4.

Conclusions Discussion
Most researchers in the literature have relied on the arguments of social learning and social exchange theories to explain the effects of ethical leadership on employee voice. Though the social comparison in leader-member exchange was found to be related to employee voice [53], few researchers have applied this perspective to explain the influence of ethical leadership on employee voice. The present study showed that leader-member exchange, a consequence of social exchange, can mediate the effects of ethical leadership on employee voice-a finding that fills the gap of the literature concerning the scarcity of research using social learning and social exchange perspective to study the ethical leadership-employee voice behavior relationship.
Ethical leadership and leader-member exchange were found as psychological mechanisms to stimulate employee voice [5,54], but no research has examined how ethical leadership and to affect as a consequence of social exchange affects two dimensions of employee voice behavior (promotive voice and prohibitive voice). The finding of the positive relationship between leader-member exchange and employee voice also enriches the literature by showing that leader-member exchange mediates the relationship between perception of ethical leadership and two dimensions of employee voice behavior.
At the level of interpersonal interaction, employees' attitudes can be shaped by the leadership style of superiors, and these attitudes can further affect their behavior. Simultaneously, at the level of organizational context, subordinates' attitudes that related to their behavior can be determined by leadership. For example, the good leadermember exchange under the consideration-oriented leadership can create in subordinates the intent to speak up to their leaders. At the same time, because the caring atmosphere nurtured by the consideration-oriented leadership highlights the importance of interpersonal mutual concern in organizations, the leadership can strengthen the influence of the subordinates' reciprocation intention on their voice behavior. In other words, leadership can work as an antecedent of employee attitudes and behaviors. In using the social exchange theory to test the influence of ethical leadership on two dimensions of employee voice behavior, the mediation of leader-member exchange on the link between ethical leadership and two dimensions of employee voice behavior has been neglected by researchers. By showing that perceived ethical leadership can attenuate the positive effects of leader-member exchange on both employee promotive voice and prohibitive voice, the present study contributes to the literature by filling the above gap concerning the lack of research on the mediation of leadermember exchange.

Implications for Management
Our findings provide important managerial implications for Vietnamese organizations. Companies should be aware that both social exchange relationships and social learning are important if Vietnamese employees are to be encouraged to voice. Although Vietnamese employees accept unequal power distribution structures, they may less willingly accept their lower hierarchical positions. While many firms have adopted measures to encourage followers express opinions (e.g., am open door policy and anonymous voice mailbox), these measures usually do not have satisfactory effects. By examining the process from ethical leadership to two dimensions of voice behavior, this study reveals a way to increase employees' opinion expression. Voice behavior can be seen as a manifestation of ethicality. To initiate employees' sincerity and honesty, organizational leaders should first display their ethicality and act as role models. Besides, firms should be recognized that ethical leadership is an essential component of a sustainable organization. Given the positive influence of ethical leadership on employee voice, organizations are encouraged to select or and train ethical leaders. For example, organizations should select or promote supervisors based on their moral standards, integrity, and concern for others. In addition, organizations should emphasize the ethicality of managers' and supervisors' behavior and establish procedures to reward ethical behavior and to punish misconduct. On the other hand, by showing that leader-member exchange is proximal antecedents to employee voice, the results indicate that supervisors who seek to enhance employee voice should develop processes to increase leader-subordinate relationship quality.

Limitations and Directions for Future Research
We chose employees in service firms as the subjects of our study because leader-member exchange, voice behavior, and ethical leadership were highly relevant to their work situations. Our findings may be applicable to other organizations whose employees are working in situations similar to those of employees in service firms, for example, manufacturing firms. However, the generalization of the findings to other organizations which have different working environment needs to be done cautiously.
The results of our study can not rule out the possibility of the reversed causality of variables in the hypotheses. For example, after employees speak up, their colleagues are likely to respond with similar behaviors. Because the beneficiaries of this reciprocity, the employees may have more level of leader-member exchange. When they feel that having more level of leader-member exchange, they may attribute the felling to the fairness atmosphere nurtured by their managers and perceive the managers as ethical leaders. Besides, when employees perform less speaking up, they may become less level of leader-member exchange, as a consequence, may attribute their jealousy to the unfairness of their leaders, a perception of unethical news in their leaders. An enhanced ethical leadership may first work to increase leader-member exchange and result in more voice behavior of employees. In the next cycle of causality, this increase in voice behavior may work to increase leader-member exchange, and enhance their perceived ethical leadership. Longitudinal research is needed to test whether the cyclic causality is plausible.
As transformational, spiritual, and authentic leaderships share the same characteristics with ethical leadership on leader's concern for others, role modeling, or ethical decision-making [8], they may also work to affect employee voice behavior through the social comparison processes. These leaderships can be used as control variables in future studies while trying to manifest more clearly the influence of ethical leadership on employee voice behavior through the leader-member exchange.

Conclusions
This research tested the impact of ethical leadership on two dimensions of the voice behavior of service firms` employees in Vietnam, and has made an important contribution by bringing out the relationship that exists between ethical leadership, LMX, and voice behavior of employees. The findings of this research show that issues relating to leadership and job design should be studied from an interactional perspective to promote employee voice behavior. Thus, this research tried to draw an overall picture of how ethical leadership style can promote employee voice behavior through building social exchange between leader and employees in service firms. From our results, we expect to inspire more research to find out the effect of ethical leadership on other individual-referenced outcomes, as well as to explore the possible mechanisms underlying the relationship.