Linking authentic leadership,moral voice and silence—A serial mediation model comprising follower constructive cognition and moral efficacy

Authentic leadership and its moral core

The concept of authentic leadership was introduced by Luthans and Avolio (2003) and can be traced back to theoretical works on authenticity (Gardner et al., 2011). On one hand authenticity comprises a deeply-rooted reflection and understanding of one’s self and on the other hand behaviors that correspond to this conception of oneself (Gardner et al., 2011). This conceptualization of authenticity already points to self-awareness and self-concordance as central and unique aspects of authentic leadership (Lemoine et al., 2019). Going beyond the scope of mere authenticity, authentic leadership also encompasses the way of how a deep understanding of one’s self develops. For example, it can be strengthened through a preferably non-biased processing of self-relevant information from different sources such as self-observation or feedback from others (Gardner et al., 2005; Kernis, 2003). Developing openness and trust in relationships is crucial (Kernis, 2003), as close others provide valuable feedback in order to stimulate self-reflection and develop self-awareness. The self-reflective nature of authentic leadership therefore also means that leaders put an emphasis on participation of followers by seeking feedback, accepting inputs and discussing decisions (Avolio et al., 2009; Gardner et al., 2011; Gill et al., 2018). It is part of their social role as a leader. The way authentic leaders construct and interpret their leadership role also provides the rationale for the moral component of authentic leadership (May et al., 2003). Whether or not a social role, as a subset of a person’s moral identity, is activated in a morally relevant situation can be decisive for moral action to occur (Hannah et al., 2011b). May et al. (2003) point out that authentic leaders are highly aware that their social role as a leader is associated with a high level of responsibility to act morally. Their moral behavior sends a strong message and gives orientation to followers on how to act in certain situations. Moreover, Lemoine et al. (2019) state that the moral core of authentic leadership is rooted in virtue ethics emphasizing that authentic leaders’ morality does not depend on external expectations but is strongly based on their own moral compass. Authentic leaders value their own beliefs and convictions when addressing moral issues (Guignon, 2004) and engage in self-concordant behaviors that match these beliefs and convictions (May et al., 2003). It is theorized that authentic leaders also strive to build self-awareness and self-concordance in their followers in order to develop their autonomy to make own moral judgments (Lemoine et al., 2019). Following May et al. (2003), the moral component of authentic leadership also comprises the leader’s ability for perspective-taking in dealing with moral dilemmas. By acknowledging the diverse interests present in a given situation, including their own, they are able to better understand the complexity and moral implications of situations, which is necessary to find solutions that are oriented towards high moral standards and eventually to take action (Hannah et al., 2011b; May et al., 2003). Therefore, in the context of leadership, authenticity is not limited to the leader as an individual in the sense of self-reflection. It also emphasizes the importance of social interactions, the social role as a leader and moral role modeling. These aspects find their expression in the presumably most widely accepted definition of authentic leadership by Walumbwa et al. (2008: 94), who describe it as

a pattern of leader behavior that draws upon and promotes both positive psychological capacities and a positive ethical climate, to foster greater self-awareness, an internalized moral perspective, balanced processing of information, and relational transparency on the part of leaders working with followers, fostering positive self-development.

Walumbwa et al. (2008) propose four behavioral components of authentic leadership (cf. Frömmer et al., 2014): Self-awareness refers to self-reflection, the deepened understanding of one’s own strengths, weaknesses, and effect on others. In essence, relational transparency means to show one’s true self to others without concealing or pretending thoughts or feelings. Consequently, this promotes trust in relations. Balanced processing means the objective analysis of relevant information before making a decision. Authentic leaders solicit opinions from employees and put their own (also moral) basic beliefs and convictions up for discussion. Finally, the component internalized moral perspective describes a form of self-regulation that is guided by internal moral norms and values that leads to decisions and behaviors that comply with these norms and values.

In accordance with Walumbwa et al. (2008), in this study we conceptualize authentic leadership as a higher order construct overarching the four behavioral components. Although, empirical studies on authentic leadership measures do not provide clear evidence for this assumption, the use of an overall score reflecting authentic leadership as a higher order construct still seems warranted (Franke-Bartholdt et al., 2018).

Moral voice and silence as moral outcomes

In her review Morrison (2014: 375) conceptualizes voice as the “. . .discretionary communication of ideas, suggestions, concerns, or opinions about work-related issues with the intent to improve organizational or unit functioning.” This also shows that the possible content of voice is widespread. Thus, when embedding voice, as well as silence, within the moral domain, the content that is or is not conveyed by employees needs to be specified (cf. Knoll et al., 2016; Morrison, 2014). In this study we focus on voice and silence on moral issues. Following the definition of Frömmer et al. (2012), we understand moral issues as incidents, practices, or behaviors within an organization which violate principles of fairness, justice, and equal treatment and negatively affect the well-being of persons, for example, ostracizing team members, disrespectful treatment in the workplace, favoritism, data manipulation, and violation of standards.

Voice behavior challenges the status quo or even the behavior of important stakeholders and is therefore seen as potentially risky (Morrison, 2014). It may go along with formal sanctions, a negative public image, damage of relationships or even suffering retaliation (Miceli et al., 2008; Milliken et al., 2003; Morrison, 2014). Therefore, it is assumed that employees weigh up the costs against the benefits when deciding to engage in voice behavior or not (Detert and Burris, 2007).

Employee silence is generally defined “. . . as purposeful withholding of ideas, questions, concerns, information, or opinions by employees about issues relating to their jobs and the organizations in which they work” (Brinsfield et al., 2009: 20). Along the lines of moral voice, in this study we investigate silence on moral issues, which is deemed a frequent phenomenon in organizations (Kish-Gephart et al., 2009; Milliken et al., 2003; Morrison and Milliken, 2000). Employee silence is driven by different motives (Brinsfield et al., 2009; Knoll and Van Dick, 2013a). We take quiescent (driven by fear) and acquiescent silence (driven by resignation) into account, which are the theoretically most substantiated types of silence (Brinsfield, 2013; Knoll et al., 2016; Pinder and Harlos, 2001). According to Pinder and Harlos (2001), acquiescence silence describes a submissive acceptance of organizational circumstances, which goes hand in hand with a reduced awareness of alternatives to silence. Employees feel that their opinion is not valued and conformity is promoted while dissent is suppressed (Morrison and Milliken, 2000). This condition is marked by resignation and hopelessness and represents disengaged employee behavior (Knoll and Van Dick, 2013a). Quiescence silence describes an aversive, conscious state in which employees deliberately keep silent, even if they see alternative ways of how the situation could be changed for the better (Pinder and Harlos, 2001). This is triggered by an individual’s fear that the consequences of speaking up can be personally uncomfortable (Pinder and Harlos, 2001). This condition can be described as “suffering in silence (Knoll and Van Dick, 2013a: 351)” as employees are still interested in their work.

Knoll et al. (2016) argue that voice and silence should not be understood in a bipolar way. The relationship can better be pictured as a continuum where motives for both, voice and silence, are active but in certain situations sometimes motives for voice and sometimes motives for silence are dominating (Brinsfield, 2013; Knoll et al., 2016). Van Dyne et al. (2003) state that it comes with more difficulties and ambiguity for observers to interpret the behavioral cues of silence compared with voice. Furthermore, Knoll et al. (2016) point out that organizational tools for overcoming silence may differ from tools that facilitate voice. Thus, in line with recent contributions on the topic (Brinsfield, 2013; Brinsfield et al., 2009; Kish-Gephart et al., 2009; Knoll et al., 2016), we believe that it is valuable to treat moral silence and moral voice as distinct constructs in order to gain a more differentiated picture of the silence-voice-threshold.

Authentic leadership and its influence on moral voice and silence

We posit that authentic leadership fosters moral voice and mitigates moral silence of employees by means of role-modeling and processes of social learning (Bandura, 1977, 1991). Authentic leaders are described as persons of high moral character exhibiting honesty, integrity, openness, and a desire to do the right thing (Gardner et al., 2011; Walumbwa et al., 2008), making them attractive role models for followers in the context of social learning. Furthermore, they are expected to behave in a way that is perceived as showing consistency between their internal core moral values and beliefs and their actual actions (Avolio et al., 2004). They “practice what they preach”, thus they stay true to their moral convictions, build up self-confidence to do the right thing and signal their follower to behave in the same way (Hannah et al., 2011a). The continuous role-modeling of the leader also increases the likelihood that moral behavior will be emulated and implemented when confronted with moral issues (Bandura, 1977). Followers learn to align their actions according to their internal moral standards and stand up for their values (cf. Hannah et al., 2011a), which should become manifest through increased moral voice behavior of followers. Empirical support for this assumption can be inferred from a study by Liu et al. (2015) that revealed a positive effect of perceived authentic leadership on supervisor-rated whistleblowing intentions of employees. Furthermore, the moral perspective of authentic leadership entails a strong orientation towards dialogue with followers concerning one’s own thoughts and feelings. It is postulated that authentic leaders encourage their employees to reduce discrepancies between their values and actions through transparent discussions concerning moral issues, their own values and line of thinking (Avolio and Gardner, 2005). Through balanced processing an authentic leader solicits feedback from employees and invites them to challenge his or her opinions, beliefs, and convictions (cf. Walumbwa et al., 2008). That way authentic leaders are able to effectively involve employees in processes of moral decision-making and can also clarify their own (even abstract) values and principles (cf. Hannah et al., 2011a). Under these circumstances it seems plausible that employees make the experience that their moral views mean something, are valued and may lead to subsequent changes. Therefore, motives for acquiescent moral silence should be reduced. Moreover, we argue that withholding one’s own opinions and thoughts out of fear of negative personal consequences is less likely when employees are led by authentic leaders. It is theorized that authentic leaders deal with their own strengths and weaknesses in a confident and transparent manner, have no trouble admitting own mistakes and are open for feedback (Cropanzano and Walumbwa, 2010; Walumbwa et al., 2008). Following the principle of self awareness, critical remarks and moral concerns of employees are not labeled as “being offensive”, but rather as benevolent and useful. Openness in communication is therefore encouraged and not punished. We argue that a norm of openness and discourse is established through the exemplary behavior of authentic leaders in terms of constructively dealing with errors, problems, weaknesses, values, and moral concepts. Thus, quiescent moral silence should decrease. In conclusion we propose:

Authentic leadership and moral behavior of employees: Mediating roles of follower constructive cognition and moral efficacy

An important goal of our research is to shed light on mediating mechanisms. Gardner et al. (2011) call to devote more attention to authentic followership. They argue that the mutual influence of leaders and followers may be an important driver for the development of authentic relationships and authentic leadership itself. We view employee moral voice as an important indicator of authentic followership, because it presumably reflects congruence between internal moral values of employees and their subsequent actions. Regarding potential mediators, we further focus on the role of followers. As mentioned above, modeling self-regulatory behaviors is argued to be a core mechanism through which authentic leadership in turn fosters self-regulation within followers (Lemoine et al., 2019). Hannah et al.’s (2011b) capacity approach on moral action provides an excellent theoretical framework to infer key mechanisms of follower self-regulation specific to the moral domain. It distinguishes between moral maturation capacities and moral conation capacities. The former basically comprise cognitive mechanisms that deal with sensitivity, awareness, and processing pertaining to moral cues. The latter comprise mechanisms dealing with ownership and motivational resources pertaining to moral issues and actions. Based on this differentiation, we suggest a serial multiple mediator model using follower constructive cognition, representing a moral maturation capacity, and moral efficacy, representing a moral conation capacity, as mediators (see Figure 1).

OPEN IN VIEWER

Figure 1.

Graphical depiction of the hypothesized serial mediation model.

In order to gain a more differentiated picture on the linkage between authentic leadership and follower moral behavior, our model not only considers the sequential effect of follower constructive cognition and moral efficacy, but also their unique indirect influence as single mediators. Firstly, we focus on the mediating role of follower constructive cognition. Constructive cognition is embedded in the broader concept of self-leadership, which entails behavioral and cognitive strategies related to the theoretical framework of self-regulation (Neck and Houghton, 2006). According to Neck and Houghton (2006: 271), “Self-leadership is a self-influence process through which people achieve the self-direction and self-motivation necessary to perform.” Houghton et al. (2012) identified three dimensions that “encapsulate the heart of the classic self-leadership strategy dimensions” (Houghton et al., 2012: 224): Behavior awareness and volition involves, for example, self-goal setting and self-observation. Task motivation reflects feelings of competence, visualizing successful performance, and self-reward. Constructive cognition represents the notion of positive thinking by reshaping key mental processes.

Behavior awareness and volition and task motivation emphasize aspects of motivation and performance (e.g. goal-setting, task competence, self-reward, and visualizing success) that go beyond the scope of moral maturation and the moral domain in general. Therefore, based on the conceptualization of moral maturation, as a driver of moral cognition rather than motivation, we focus exclusively on constructive cognition as a proxy for moral maturation. Constructive cognition represents “thought self-leadership” (Neck and Houghton, 2006: 273) and clearly overlaps with what Hannah et al. (2011b) describe as metacognitive ability, a moral maturation capacity that “is composed of monitoring and regulating cognitive processes” (Hannah et al., 2011b: 669). Constructive cognition can be considered a highly self-reflective thought process that encompasses to reflect on one’s own thought patterns, identifying dysfunctional beliefs and to challenge and evaluate one’s own beliefs and assumptions when encountering difficult situations or issues (Houghton et al., 2012). It is therefore plausible that constructive cognition supports individuals to build more complex representations of a given situation, issue or knowledge domain, which should also be helpful when dealing with moral issues or dilemmas. Moral issues and dilemmas are rather complex and cognitively demanding (Hannah et al., 2011b; Strobel et al., 2017) and more cognitively complex individuals should be better in resolving such issues (Hannah et al., 2011b). Empirical research also underlines the significance of thought processes in fostering moral behavior and decision-making. For example, Strobel et al. (2017) argue that thought-oriented individuals tend to have a deeper information processing and engage in more elaborated decision-making. They revealed in their study that an individual’s need for cognition (Cacioppo et al., 1996) positively predicted moral behavior. Furthermore, Steinbauer et al. (2014) revealed that employee self-leadership was related to moral judgment when they used active reasoning as opposed to mere intuition.

Therefore, we argue that the thought process of constructive cognition is especially helpful for gaining a more comprehensive overview, better orientation, and personal judgment when dealing with moral issues. This should support one’s desire and willingness to “do the right thing” and thus, enhance voice behavior. Furthermore, we argue that an elaborated way of dealing with moral issues entails a more detailed understanding of possible ways to resolve a moral issue and increases one’s ability to infer a more realistic preview of consequences that may result when engaging in voice behavior. Consequently, follower constructive cognition should reduce acquiescent and quiescent silence as well.

It is likely that authentic leaders trigger constructive cognition regarding themselves by displaying balanced processing (e.g. solicit feedback, invite others to challenge one’s own beliefs and convictions). Therefore, authentic leaders function as salient role models and should enhance follower constructive cognition through processes of social learning (cf. Hannah et al., 2011a, 2011b). In addition, follower constructive cognition, understood as a moral maturation capacity, can also be strengthened by exposing followers to moral conflicts and different moral perspectives in order to challenge their thought patterns (cf. Hannah et al., 2011b). Authentic leaders are theorized to do that by involving followers in moral decision-making, discussions on moral issues, dilemmas, values, and beliefs (Hannah et al., 2011a; Walumbwa et al., 2008). That way, cognitive representations of moral issues and moral complexity may be fostered. In sum, this leads to the following hypotheses:

Secondly, we provide the rationale for the mediating role of moral efficacy. Hannah et al. (2011b: 675) define moral efficacy as “. . .an individual’s belief in his or her capabilities to organize and mobilize the motivation, cognitive resources, means, and courses of action needed to attain moral performance, within a given moral domain, while persisting in the face of moral adversity.” Hence, moral efficacy is conceptualized as a moral conation capacity reflecting processes of moral motivation (Hannah et al., 2011b). To build up this capacity in followers social learning processes and role-modeling of leaders are again being considered as crucial (Hannah et al., 2011b; Lee et al., 2017). Hannah et al. (2011a: 562) state that the internalized moral perspective of authentic leaders encompasses that their behaviors “. . .are guided by internal moral standards and values as opposed to being driven by external pressure from peers, higher level leaders, or other social forces”. Thus, authentic leaders engage in exemplary behavior by acting in line with their internal value structure (cf. Hannah et al., 2005). By observing such behaviors, followers can learn about strategies their leaders use, adapt them for themselves and therefore widen their potential and latitude when dealing with moral issues (Bandura, 1997). Furthermore, we argue that authentic leaders bolster the confidence of followers through means of participation by actively asking for and valuing their viewpoints, involving them in moral decision-making and discussions (cf. Hannah et al., 2011a; Walumbwa et al., 2008). Given the above arguments, authentic leadership should increase the moral efficacy of employees. Moral efficacy, in turn, should influence moral behaviors. Meta-analytical research has convincingly shown the crucial role of self-efficacy as a driver of intention and behavior (Judge and Bono, 2001; Stajkovic and Luthans, 1998). As self-efficacy is considered to be domain-specific (Bandura, 1997), moral efficacy relates to the domain of moral action (Hannah et al., 2011b; Mitchell and Palmer, 2010). Moral efficacy suggests that individuals possess the confidence and belief that they are capable to address a moral issue and furthermore have the necessary means at hand to act successfully (Hannah et al., 2011b). Mitchell and Palmer (2010) contend that efficacy beliefs help individuals to control their behaviors to meet personal moral standards. Thus, efficacy beliefs are important for converting an individual’s moral intentions into subsequent moral behavior (May et al., 2003; Mitchell and Palmer, 2010). They instill the needed confidence to strive for what one thinks is right and morally appropriate (Mitchell and Palmer, 2010). We conclude that moral efficacy enhances moral voice behavior. Empirical support for our assumption can be found in a recent study by Lee et al. (2017) that showed the mediating influence of moral efficacy for the relationship between ethical leadership, another type of moral leadership, and employee moral voice. Moral efficacy comprises a higher level of trust in one’s abilities to overcome hinderances, obstacles, and pressures (Hannah et al., 2011b; May et al., 2014). Therefore, efficacy beliefs are considered crucial in coping with fear-laden situations and overcoming the tendency toward silence (Kish-Gephart et al., 2009). Thus, we argue that it furthermore reduces fear-driven motives for silence. Self-confidence in the ability to successfully perform in a given moral situation constitutes a core aspect of moral efficacy (Bandura, 1997; Hannah et al., 2011b). From our point of view this entails the belief and conviction that one’s contributions and opinions are of relevance and may enable change. Motives resulting from resignation should therefore be mitigated through higher levels of moral efficacy. In conclusion, we propose that moral efficacy is a valuable resource in overcoming the threshold between silence and voice and that it is promoted by authentic leadership.

After looking at each mediator separately, we propose a third pathway where authentic leaders influence follower moral behavior via follower constructive cognition and moral efficacy as sequentially connected mediators. We provide a brief step-by-step rationale for the serial interaction by looking at each single path. The first path pertains to the positive connection of authentic leadership and follower constructive cognition. As outlined above (see Hypothesis 2), authentic leadership should enhance follower constructive cognition through social learning and by involving followers into moral discussions and decision-making.

The second path concerns the assumption that follower constructive cognition is positively connected to moral efficacy. The moral capacity approach of Hannah et al. (2011b) incorporates Rest’s well-established model of moral behavior (Rest, 1986) that comprises four sequenced components: (1) moral sensitivity, which relates to interpreting a moral situation and being aware that there is a moral issue, (2) moral judgment, which entails ways to determine the proper moral actions, (3) moral motivation, which means to commit oneself to a course of action and create intentions, and (4) moral action, which means to take appropriate action and to show persistence when facing challenges. Hannah et al. (2011b) group moral sensitivity and judgment primarily into the maturation capacities and moral motivation and action primarily into the conation capacities. Follower constructive cognition, understood as a moral maturation capacity and moral efficacy, understood as a moral conation capacity, basically build a logical bridge between moral cognition and moral motivation. According to Hannah et al. (2011b) moral efficacy, i. a., refers to the belief of being capable to organize and mobilize cognitive resources and motivation for moral action. As mentioned above, constructive cognition, as a thought process, should foster active reasoning and a more elaborated representation of a given moral issue (cf. Steinbauer et al., 2014; Strobel et al., 2017). In turn, this may help to get a better overview on one’s latitude in order to strengthen one’s self-belief. Furthermore, Neck and Houghton (2006) argue that constructive thought strategies have the potential to improve self-regulatory effectiveness. It seems plausible, that constructive cognition leads to identifying and eliminating dysfunctional or misleading assumptions that otherwise would mitigate one’s self-belief to be capable to successfully resolve a difficult moral issue.

Thus, constructive cognition should strengthen moral efficacy, because it enables employees to cognitively structure and organize moral contents, potentially eliminates hindering assumptions and therefore fosters their self-belief in own capabilities.

The third path connects moral efficacy with moral behavior. Given the above arguments (see Hypothesis 3), we argue that moral efficacy enhances moral behavior by instilling the confidence necessary to engage in moral action.

Combining the rationales for each path we finally propose a serial mediation:

Sample and procedure

In order to qualify for participation, employees had to have a direct supervisor and find themselves in a full-time position. They were recruited trough both professional and social networks and by marketing the study’s website. Our study also partnered with regional institutions (printing company, health care institute), which received result reports (descriptive data) in return for their participation. Study participants filled out a web-based survey (www.unipark.com), which took about 15 minutes to complete. All participants were automatically instructed about their participation being voluntary and received background knowledge on the research setting. A total number of 866 people arrived on the landing page of the survey. Of these, 34.1% completed the survey. Thus, the final sample involved 295 participants. Of these, 173 (58.7%) were woman and 121 (41.1%) were men (n_missings = 0.3%). The average age was 39.3 years (SD = 10.6), with 32.2% of them having less than 5 years of experience in their current field of work, 47.5% between 5 and 20 years and 20.3% more than 20 years. Respondents represented organizations of different sizes with 15.6% of them working in small organizations (up to 50 employees), 34.2% working in mid-sized organizations (between 51 and 500 employees), and 50.2% being employed in large scale institutions (more than 500 employees). Study participants were from different industries (50 from healthcare, 47 from IT, 41 from public sector, 25 from human resources, 21 from marketing, 19 from sales, 18 from accounting, and 74 from diverse others). All variables were assessed using the same method. In order to reduce potential common method bias, we assured anonymity of participants, used means of methodological separation (e.g. separation of predictor and criterion measures on the questionnaire, use of different scale properties) and integrated proper survey instructions to avoid feelings of evaluation apprehension and enhance motivation of participants to provide optimal response (Podsakoff et al., 2003, 2012).

Measures ¹

Data preparation and analytical procedures

We performed Little’s (1988) missing completely at random test (MCAR) for all items, which revealed that data were missing randomly, χ² (964) = 1030.90, p = 0.066. Out of all scale-based item values 0.43% were missing. The highest rate for missing values was 2.7% referring to an item belonging to the DIAF. Therefore, we applied the expectation-maximization (EM) algorithm (Little and Rubin, 1989) to the data set to impute missing values. Hayes and Rockwood (2017: 45) outline that in serial mediation “. . .mediators are linked together in a causal chain, with one mediator allowed to influence one or more mediators causally downstream.” To test all direct, indirect and total effects of our serial mediation model, we applied structural equation modeling (SEM) based on latent factors using AMOS 25.0. SEM allows to test mediational effects regarding multiple outcomes simultaneously. We used 95% bias-corrected confidence intervals (CI) based on 10,000 bootstrap samples to test for statistical significance. If the reported CI of an effect does not contain zero, it is considered statistically significant. Generally, a serial mediation model with two mediators comprises three different indirect effects: the mediating influence of each single mediator and their joined influence as sequentially connected mediators. Therefore, the model allows a comparison of all indirect effects (cf. Hayes and Rockwood, 2017). To compare the size of indirect effects we used the integrated AMOS-procedure of user-defined estimands, which allows to calculate differences of indirect effects by defining difference functions. It also provides 95% bias-corrected CIs based on bootstrapping. A difference in size is considered significant, if the reported CI does not contain zero.

Descriptive statistics and preliminary analysis

We conducted confirmatory factor analyses (CFAs) with AMOS 25.0 to examine the distinctiveness of the measures we used in our study. Authentic leadership was integrated as a higher order construct with the first order factors representing the four components of authentic leadership. No parcels were used. Firstly, we tested our hypothesized 6-factor-model, where all items loaded on their intended factor, each representing a construct of our hypothesized model (χ² = 977.46, p < 0.01; df = 508; CFI = 0.91; RMSEA = 0.056, 90% CI [0.05, 0.06]; SRMR = 0.06). As alternative models we tested (a) the baseline model where all items loaded on a single factor (χ² = 2,841.13, p < 0.01; df = 527; CFI = 0.87; RMSEA = 0.122, 90% CI [0.12, 0.13]; SRMR = 0.13) and (b) a 5-factor-model with constructive cognition and moral efficacy representing one factor (χ² = 1,048.73, p < 0.01; df = 514; CFI = 0.90; RMSEA = 0.059, 90% CI [0.05, 0.07]; SRMR = 0.07) and (c) a 5-factor-model with moral efficacy and moral voice representing one factor (χ² = 1,144.00, p < 0.01; df = 513; CFI = 0.88; RMSEA = 0.065, 90% CI [0.06, 0.07]; SRMR = 0.07) and (d) a 7-factor-model with moral voice split into two factors representing moral courage and moral ownership (χ² = 966.65, p < 0.01; df = 502; CFI = 0.91; RMSEA = 0.056, 90% CI [0.05, 0.06]; SRMR = 0.06). Referring to the cut-off values of Hu and Bentler (1999), our measurement model exhibited an acceptable fit and superior fit compared to models (a) to (c) (Δ χ² vs model a: Δ χ² = 1863.67, p < .01; vs model b: Δ χ² = 71.27, p < 0.01; vs model c: Δ χ² = 166.54, p < 0.01). Model (d) exhibited a very similar fit with no significant difference (Δ χ² = 10.82, p > 0.05). Furthermore, according to Chen (2007), comparison of CFI showed no relevant change in model fit (Δ CFI < 0.01). Following the rule of parsimony, we preferred to use the 6-factor-model instead of the 7-factor alternative.

Furthermore, we checked our measurement model for potential common method variance (Podsakoff et al., 2003). Harman’s single-factor test (Podsakoff et al., 2003) showed that an assumed general factor did not account for the majority of covariance (30.3%). Table 1 presents the means, standard deviations and intercorrelations for all study variables. Correlations between the construct variables of the hypothesized model were low to large in size. One exception is the non-significant correlation between follower constructive cognition and quiescent moral silence.

OPEN IN VIEWER

Table 1.

Means, standard deviations, and correlations among research variables.

		M	SD	1	2	3	4	5	6	7	8	9	10
1	Age	39.34	10.60	–
2	Gender a	1.58	0.50	−0.09	–
3	Professional tenure b	1.88	0.72	0.63	−0.08	–
4	Size of organization c	2.35	0.74	0.11	0.06	0.13*	–
5	Authentic leadership	3.39	0.83	0.00	−0.11*	−0.12*	0.02	0.94
6	Follower constructive cognition	3.71	0.80	0.19	0.04	0.05	0.01	0.18	0.66
7	Moral efficacy	3.93	0.64	0.11	−0.04	0.06	0.09	0.36	0.31	0.81
8	Moral voice	3.63	0.68	0.11	0.00	0.09	0.05	0.29	0.34	0.54	0.84
9	Acquiescent moral silence	3.40	1.70	0.07	0.10	0.12*	0.03	−0.57	−.14*	−0.27	−0.38	0.88
10	Quiescent moral silence	3.35	1.50	−0.10	0.09	−0.07	0.08	−0.21	−.10	−0.27	−0.38	0.53	0.84

N = 295. Bold-faced correlations are significant at p < 0.01 (two-tailed); correlations marked with an asterisk are significant at p < 0.05. Reliabilities (Cronbach’s alpha) are listed on the diagonal.

a

coded: 1 = male, 2 = female.

b

coded: 1 = < 5 years, 2 = 5–20 years, 3 ⩾ 20 years.

c

coded: 1 = small (n < 51), 2 = medium (n = 51–500), 3 = large (n ⩾ 500).

Hypotheses test

We tested all hypotheses using a structural latent factor model accounting for all indirect effects and for correlations between the three outcomes. The model showed the exact same fit as the measurement model (χ² = 977.46, p < 0.01; df = 508; CFI = 0.91; RMSEA = 0.056, 90% CI [0.05, 0.06]; SRMR = 0.06) and revealed a positive correlation between quiescence and acquiescence silence (r = 0.56, p < 0.05) and negative correlations between the silence outcomes and moral voice (each r = −0.33, p < 0.05). All path coefficients are shown in Figure 2. All indirect effects and parameters can be found in Table 2. Hypothesis 1 predicted that authentic leadership is positively related to employee moral voice (a) and negatively to acquiescent (b) and quiescent (c) moral silence of employees. Correlation coefficients and total effects of our model show a positive connection between authentic leadership and employee moral voice (r = 0.29, p < 0.01; β = 0.30, p < 0.001) and negative relationships between authentic leadership and acquiescent moral silence (r = −0.57, p < 0.01; β = −.57, p < 0.001), as well as quiescent moral silence (r = −0.21, p < 0.01; β = −0.27, p < 0.001) of employees. According to Bosco et al. (2015) the results indicate a medium to large effect regarding moral voice, a large effect regarding acquiescent moral silence and a medium effect regarding quiescent moral silence. Hypotheses 1a through 1c could be confirmed. Next, we tested serial mediation for each outcome.

OPEN IN VIEWER

Figure 2.

Depiction of mediation results based on the structural SEM model.

OPEN IN VIEWER

Table 2.

Overview of mediation effects and parameters (10,000 bootstrap samples).

DV	M1	M2	Indirect effect b (β)	SE	p	LL BootCI	UL Boot CI
Moral voice	FCC	–	0.03 (0.04)	0.02	<0.05	0.00	0.07
	–	ME	0.11 (0.17)	0.03	<0.05	0.05	0.17
	FCC	ME	0.03 (0.04)	0.01	<0.05	0.01	0.06
Aquiescent moral silence	FCC	–	0.00 (0.00)	0.02	n.s.	−0.06	0.04
	–	ME	−0.04 (–0.03)	0.04	n.s.	−0.12	0.02
	FCC	ME	−0.01 (–0.01)	0.01	n.s.	−0.04	0.00
Quiescent moral silence	FCC	–	0.00 (0.00)	0.02	n.s.	−0.05	0.05
	–	ME	−.11 (–.08)	.05	<0.05	−.22	−.04
	FCC	ME	−.03 (–.02)	.02	<0.05	−.08	−.01

N = 295. b: unstandardized effect; β: standardized effect; DV: dependent variable; M1: mediator 1; M2: mediator 2; FCC: Follower constructive cognition; ME: Moral efficacy; SE: standard error; LL Boot CI: Lower limit bootstrapped 95% unstandardized confidence interval; UL Boot CI: Upper limit bootstrapped 95% unstandardized confidence interval.

Firstly, we looked at the relationship between authentic leadership and employee moral voice. Indirect effects reflecting the unique influence of each mediator were found to be significant for follower constructive cognition (β = 0.04, p < 0.05) and moral efficacy (β = 0.17, p < 0.05). Therefore, hypotheses 2a and 3a could be confirmed. Results also showed a significant indirect effect of serial mediation for the relationship between authentic leadership and employee moral voice (β = 0.04, p < 0.05) with all path coefficients reflecting the intended directions of influence (see Figure 2). Therefore, hypothesis 4a was supported. Comparing the size of all three indirect effects, we found that the unique indirect effect of moral efficacy was significantly stronger compared to the other two (Δ b (β) = 0.08 (0.13), 95% CI [0.02, 0.15] compared to serial mediation effect; Δ b (β) = 0.08 (0.13), 95% CI [0.01, 0.15] compared to unique effect of follower constructive cognition). The direct effect of authentic leadership remained significant.

Second, we looked at the relationship between authentic leadership and acquiescent moral silence of employees. Results showed no significant indirect effects. Hypotheses 2b, 3b, and 4b were not supported. The direct effect of authentic leadership remained significant.

Finally, we looked at the relationship between authentic leadership and quiescent moral silence of employees. An indirect effect reflecting the unique influence of moral efficacy was found to be significant (β = −0.08, p < 0.05). Hypothesis 3c could be confirmed. The unique indirect effect of follower constructive cognition was not significant. Therefore, hypothesis 2c was not supported. Results also showed a significant indirect effect of serial mediation (β = −0.02, p < 0.05) with all path coefficients reflecting the intended directions of influence (see Figure 2). Hypothesis 4c was supported. Comparing the size of the two significant indirect effects, we found that the unique indirect effect of moral efficacy was significantly stronger compared to the serial mediation effect (Δ b (β) = 0.08 (0.06), 95% CI [0.02, 0.18]). The direct effect of authentic leadership remained significant.

Theoretical implications

Although authentic leadership is theorized as being a form of moral leadership (e.g. Gardner et al., 2005; May et al., 2003, Walumbwa et al., 2008), fundamental assumptions specifically pertaining to the moral aspect of authentic leadership have received insufficient attention in empirical research. Thus, this study mainly aimed to connect theory with empirical evidence in order to advance our knowledge regarding the effectiveness and mechanisms of authentic leadership in influencing employee moral behavior.

Firstly, there is the core assumption that authentic leadership is positively connected to moral employee behavior (Gardner et al., 2011; May et al., 2003; Walumbwa et al., 2008), which has rarely been tested so far. Hannah et al. (2011a) point out that a main reason for this is that the use of more direct behavioral moral outcomes is highly missing in leadership research. We addressed this issue. Our study results revealed a positive relationship between authentic leadership and self-reported moral behavior. As moral behavioral outcomes we focused on self-reported employee voice and (quiescent and acquiescent) silence on moral issues. Moral issues were specified using a definition by Frömmer et al. (2012) that draws on moral principles such as fairness, justice, and well-being of persons. Results support the notion that perceived authentic leadership positively influences employees to speak up on moral issues in their workplace and potentially mitigates fear-based and resignation-based motives to intentionally remain silent. Although not hypothesized, our results indicate that the relationship of authentic leadership with the three moral outcomes differs in strength. Small to medium effects were found for the relationships of authentic leadership with moral voice and quiescent moral silence while a large effect was found regarding acquiescent moral silence. This shows that authentic leadership seems to be especially effective in helping employees to overcome feelings of resignation and disengagement. In other words, authentic leadership has the potential to “reactivate” employee participation in creating moral working climates. In sum, this evidence lends support for the important theoretical assumption that authentic leadership positively influences employee moral behavior. In addition, study results enhance our understanding of the differentiated effects of authentic leadership on moral voice and silence.

Secondly, there is the proposition that the moral approach of authentic leadership is specifically based on self-awareness and self-concordance (Lemoine et al., 2019). This goes along with the assumption that modeling of self-regulatory behaviors to followers, which in turn should enhance self-regulation of followers themselves, is the core mechanism of influence of authentic leaders with respect to follower moral behavior. Focusing on the abovementioned proposition, this study extends existing research on moral authentic leadership by adding empirical insights into different self-regulatory mechanisms that elucidate how authentic leadership may foster employees’ moral voice and mitigate employees’ moral silence. Sticking consistently to the moral domain, we drew on the moral capacity approach by Hannah et al. (2011b) to derive relevant mechanisms of follower self-regulation. Based on the framework of moral maturation and moral conation (Hannah et al., 2011b), we developed a more complex model of serial mediation focusing on the mediating roles of follower constructive cognition and followers’ moral efficacy. Our analyses revealed differing results concerning the three moral outcomes. With regards to employee moral voice our findings indicate that follower constructive cognition and moral efficacy serve as mediators through which authentic leadership may motivate employees to engage in moral voice. As expected, we found a serial mediation involving both follower constructive cognition and moral efficacy, and unique effects of each mediator. This corroborates our theoretical assumptions. Thus, follower constructive cognition may function as a moral maturation capacity (Hannah et al., 2011b) that is fostered by authentic leadership through social learning and enables employees to figure out moral issues in a differentiated way, leading to more elaborated judgments and willingness to speak up. Moral efficacy may function as a moral conation capacity (Hannah et al., 2011b) that enables employees to follow through with moral intentions and mobilize the necessary motivation to eventually act. Additionally, both mediators had an effect when interconnected in a sequence, with follower constructive cognition positively influencing employees’ moral efficacy. This supports the assumption that follower constructive cognition potentially bolsters moral efficacy by promoting self-belief and by helping employees to gain a more elaborated understanding of moral issues. Comparisons further revealed that the unique effect of moral efficacy explained significantly more variance than the other two indirect effects. This provides empirical support for the importance of efficacy beliefs within the moral domain (Hannah et al., 2011b; Mitchell and Palmer, 2010).

Concerning quiescent moral silence, the strongest mediating effect was solely based on moral efficacy. We also found a serial mediation effect based on both mediators. Our results are in line with recent research that conceptualizes moral efficacy as an important resource to overcome the tendency toward silence in situations that are experienced as threatening (Kish-Gephart et al., 2009). Moreover, our analyses again revealed a serial mediation via follower constructive cognition and moral efficacy by which perceived authentic leadership may attenuate the tendency to resort to silence in the presence of moral issues. This underlines the relevance of follower constructive cognition in enhancing efficacy beliefs when dealing with moral issues. In contrast to the results regarding moral voice, no unique mediating effect of follower constructive cognition was found for the relationship between authentic leadership and quiescent moral silence. Correlational results also showed no significant influence of follower constructive cognition in predicting quiescent moral silence, whereas it is important to note that, following a modern view on mediation, such a connection is not necessary to infer a mediation effect (Hayes and Rockwood, 2017). We argued that follower constructive cognition leads to a more comprehensive understanding of moral issues and realistic preview of consequences that possibly go along with speaking up. Our results indicate that this may not be sufficient to influence moral silence when the motive of fear is involved. Therefore, authentic leadership potentially reduces quiescent moral silence through follower self-leadership only when constructive cognition is paired up with an individual’s moral efficacy beliefs. That reaffirms the essential role of moral efficacy. Surprisingly, with regards to acquiescent moral silence we found no mediating effects. We briefly outline some ideas to explain these unexpected results.

Acquiescence silence is rooted in feelings of resignation and hopelessness (Knoll and Van Dick, 2013a). Employees are not willing to get involved (Pinder and Harlos, 2001) and resort to silence to detach themselves (Kish-Gephart et al., 2009). Under such circumstances, employees are perhaps less willing to engage in thinking thoroughly about moral issues or to mobilize self-resources for action as there seems to be no benefit. Thus, constructive cognition and moral efficacy may become less effective. Moreover, when comparing the two types of silence, quiescent silence is rather self-focused as it emphasizes perceived personal risks and challenges (Kish-Gephart et al., 2009). By contrast, acquiescent silence is rather based on relational aspects such as repeated experiences of rejection and disinterest by others (Kish-Gephart et al., 2009; Knoll and Van Dick, 2013b; Pinder and Harlos, 2001). Thus, creating relationships and working climates where employees feel appreciated and attached may be much more relevant to buffer acquiescent silence. As authentic leaders are theorized to value relationships with followers that are based on trust and openness (Gardner et al., 2005, 2011), it is plausible that creating such relationships functions as a mediating mechanism. Useful concepts in this context may be leader member exchange (Graen and Scandura, 1987) or team psychological safety (Edmondson, 1999).

Altogether, our results lend support for the proposition that self-regulatory processes play a crucial role regarding the effectiveness of authentic leadership within the moral domain. Furthermore, results indicate the differentiating role of self-regulation of followers with regard to moral voice and silence.

Limitations and directions for future research

There are several limitations regarding our research design.

First, we used cross-sectional data, which prohibits to draw conclusions about causality. Future studies should pursue longitudinal approaches to address this limitation. We shortly address the issue of possible reverse causality: Reviews on voice, silence, and moral leadership (Knoll et al., 2016; Lemoine et al., 2019; Morrison, 2014) reveal that employee voice and silence are usually conceptualized as outcomes of leadership as leaders are considered as salient and influential role models. Furthermore, results of longitudinal and two-wave studies (Bormann and Rowold, 2016; Detert and Burris, 2007; Lee et al., 2017) provide evidence that leadership functions as a causal predictor of employee voice and silence. Therefore, we argue that our approach to view voice and silence as outcomes of authentic leadership is in accordance with current findings. With regard to our serial mediation model, it can be argued that Hannah et al.’s (2011b) moral capacity approach postulates a step-wise sequence of moral action that describes cognitive processes as necessary antecedents to enable motivational processes for moral action. This theoretical notion is matched by our hypothesized sequence. However, our results should be interpreted with caution as without longitudinal proof the sequence can only be postulated causally.

Second, assessment of variables was based solely on employee ratings raising concerns about common method variance. However, using CFA we provided evidence for the distinctiveness of our measures and realized a-priori steps (e.g. assuring anonymity) to minimize common method variance. It is further important to note that according to Siemsen et al. (2010) method bias may affect the size of bivariate relationships, but it cannot inflate interaction effects, meaning it cannot be used to account for statistically significant interaction effects. In addition, there are also reasonable arguments that the use of self-reports is appropriate. Regarding silence, previous research points out that silence behaviors can be quite ambiguous to observers (Van Dyne et al., 2003) and therefore results based on observer-ratings can lead to misinterpretations (Knoll et al. (2016). Pertaining to employee voice, for example, supervisor-ratings can only be based on situations where voice behavior was actually observed by the supervisor. This may exclude multiple of relevant voice situations, which could lead to distorted ratings. Furthermore, regarding morality Lemoine et al. rightly state that “exactly what is moral is often subjective” (Lemoine et al., 2019: 175). Thus, within the moral domain, self-reports seem to be even necessary to cover this subjectivity. Notwithstanding, self-reports bear risks such as effects of social desirability (Podsakoff et al., 2003) and therefore future studies might include self-ratings and ratings by others. Third, our study was restricted to the individual level of analysis. As leaders nowadays usually operate in a team context, it seems necessary to investigate authentic leadership as a group level phenomenon. For example, Hsiung (2012) showed cross-level effects of authentic leadership on employee voice regarding continuous improvement processes and organizational performance. Therefore, multilevel-designs to replicate and extend such findings within the moral domain appear to be promising.

Fourth, our study did not control for competing styles of moral leadership though there is an ongoing debate on construct redundancy (Banks et al., 2018; Hoch et al., 2018). Authentic, ethical and servant leadership can be considered the most prominent concepts with an explicit focus on morality (Lemoine et al., 2019). Thus, we want to share a few arguments concerning the potential incremental validity of authentic leadership in comparison with servant and ethical leadership when investigating effects on moral voice, silence and mechanisms of self-regulation.

With regard to promoting employee voice, it is plausible that servant and authentic leadership may both be effective as they similarly aim to stimulate followers’ personal growth by means of empowerment and involvement (Lemoine et al., 2019). However, in view of current conceptualizations of servant leadership (e.g. Van Dierendonck and Nuijten, 2011), empowerment mainly targets personnel development and improving performance by enhancing self-organized working. That way, employee voice will be rather limited to topics such as continuous improvement and performance. Operationalizations of authentic leadership (Neider and Schriesheim, 2011; Walumbwa et al., 2008) go beyond that scope by practicing balanced processing, where employees are encouraged to challenge core beliefs of leaders which enables open discussions about (moral) principles. Furthermore, we argue that authentic leadership clearly entails a more initiating role of the leader when trying to trigger critical feedback and attenuate silence. This is reflected by behavioral anchors such as encouraging others to voice opposing views or soliciting personal feedback. Most importantly, authentic leadership differs from servant leadership in that it emphasizes moral consistency (Lemoine et al., 2019). Leaders signal their followers to act according to the principle “practice what you preach” and openly stand up for their moral convictions. To sum up, we believe that the wider scope of empowerment, the stronger feedback-initiating role and the focus on morally consistent behavior help authentic leaders to be more effective than servant leaders in promoting moral voice and mitigating silence.

When looking at authentic and ethical leadership, there certainly exists overlap, for example, concerning moral consistency (Lemoine et al., 2019), but there are clear differences which may lower the effectiveness of ethical leaders to build a culture of open discourse. Following ethical leadership theory (Brown et al., 2005; Brown and Mitchell, 2010), its core and unique moral focus is the emphasis on compliance with norms, standards, and rules. Furthermore, it is based on a transactional approach which entails to discipline employees who violate ethical standards (Brown et al., 2005). Therefore, unlike authentic leadership, ethical leadership may strengthen a follower’s sensitivity for authority which makes differences in hierarchy and power more salient. As a consequence, overcoming the tendency to be silent due to fear-based motives is inhibited (Kish-Gephart et al., 2009). This seems in line with a recent study from Bormann and Rowold (2016) which revealed that ethical leadership was only related to acquiescent silence and not to quiescent silence. In contrast, our study exhibited a significant influence of authentic leadership on acquiescent and quiescent silence. Moreover, “the ethical leader is less open to criticism of company standards and policies” (Lemoine et al., 2019: 165) and strongly fosters duty orientation within followers (Hannah et al., 2014). Thus, the pronounced importance of norm compliance may bear the risk that ethical leaders impede moral voice when rules and norms themselves are put into question (cf. Lemoine et al., 2019). As moral dilemmas are rather complex (Hannah et al., 2011b; Strobel et al., 2017), it seems conceivable that a wider perspective, including the open discourse about rules, may be needed to resolve an issue. Authentic leaders should enhance a wider perspective based on balanced processing as they, for example, engage in differentiated discussions on the pros and cons of company policies (Gill et al., 2018). Another key difference pertains to the aspect that authentic leaders place emphasis on enhancing followers’ growth to develop their ability, confidence and autonomy to make own moral judgments (Lemoine et al., 2019). This, in turn, should foster a more genuine and balanced discourse on moral issues between leaders and followers. On the contrary, we argue that ethical leadership depicts a more leader-centered approach focusing more on conformity rather than empowerment. For example, Lee et al. (2017) found that ethical leadership fosters employee moral voice (mediated by moral efficacy) only in case of high leader-follower value congruence. This may also trigger resignation-based silence motives as voice regarding discrepancies in values may be perceived as being futile. “Value incongruence thus may lead to ethical leadership being considered as simply dogmatic or impractical” (Lee et al., 2017: 51). This is in line with a study by Stouten et al. (2013), who found a curvilinear effect that followers can feel judged and morally reproached by high-level ethical leaders, revealing a large leader-follower distance, which, in turn, undermines extra-role behaviors. Consequently, Stouten et al. (2013) recommend to allow more employee voice on ethical standards and to show respect and recognition of employees’ perspectives and values. We argue that these behaviors correspond more to authentic leadership behaviors of empowerment and enhancing follower growth. To sum up, we propose that authentic leaders (compared to ethical leaders) can be more effective in dealing with voice and silence based on empowerment, enabling discourse on rules and norms and putting less emphasis on authority, hierarchy and conformity.

Furthermore, Lemoine et al. (2019: 156) point out that, compared to servant and ethical leadership, authentic leaders “uniquely demonstrate self-awareness and actively seek feedback for personal growth”. This pertains to root aspects of authentic leadership comprising self-reflection, self-understanding and seeking self-relevant information (Gardner et al., 2005, 2011). Moreover, authentic leaders are postulated to be transparent regarding their own strengths, weaknesses and mistakes (Walumbwa et al., 2008). In contrast, ethical leadership follows a rather idealized conceptualization of the leader as mistakes and weaknesses of the leader or the need of personal feedback are not addressed and are also missing in its operationalization (e.g. the scale by Brown et al., 2005). Self-regulation seems to be no central aspect of ethical leadership. Regarding servant leadership, Van Dierendonck and Nuijten (2011) included humility and authenticity as dimensions which encompass important aspects of self-regulation, but do not comprise the initiating role of the leader to actively seek personal feedback. In conclusion, authentic leadership places much more emphasis on leader behaviors demonstrating self-reflection, self-regulation and openness for feedback. Therefore, concerning these aspects authentic leaders should be more efficient role-models.

Regarding moral leadership approaches in general, Lemoine et al. (2019: 175) state that current concepts “are quite normative and morally absolutist in that they suggest one optimal moral behavioral pattern”. As ethical, servant and authentic leadership can be distinguished in that they each are based on different philosophical roots of normative morality, there may be a need for more acceptance that the three styles can exist in parallel, but may differ in their effectiveness within specific moral contexts (Lemoine et al., 2019). Therefore, future research on authentic leadership within the moral domain should control for other forms of positive leadership to gain insights on incremental validity and should stick to the specifics of authentic leadership. It is noteworthy that in their meta-analytic review Banks et al. (2018) point out that authentic leadership appears to distinguish itself from the other constructs (i.e. transformational, ethical, and servant leadership) especially when the outcome is organizational citizenship behavior. This lends support to the idea to specifically focus on followership, positive employee behaviors and interactions when investigating authentic leadership.

In addition, we argue that research on authentic leadership and moral behaviors such as voice and silence would benefit from including the “dark side” of leadership. Detert et al. (2007) for example showed that counterproductive behavior was only predicted by abusive supervision (Tepper, 2000) while simultaneously examining ethical leadership (Brown and Mitchell, 2010). Studies that directly contrast moral and immoral leadership are needed to gain a more comprehensive picture (Schyns and Schilling, 2013).

Last but not least, our results revealed no mediation effects regarding acquiescent silence. Consequently, we recommend to further explore possible mediators and moderators. As has been noted above, we believe that leader member exchange (Graen and Scandura, 1987) or team psychological safety (Edmondson, 1999) are mechanisms that channel the influence of authentic leaders. Regarding potential moderators, for example, Knoll and Van Dick (2013b) showed a moderating influence of organizational identification on the relationship between authenticity and acquiescent silence. Moreover, we think that it is a valuable approach to use existing theory on moral behavior, such as the moral capacity approach by Hannah et al. (2011b), to infer further potential mediating mechanisms specific to the moral domain.

Implications for practitioners

Open communication about moral issues and concerns is crucial in order to develop a positive and moral work climate in organizations. This study showed that authentic leadership is positively related to employee moral voice and may reduce employee silence on moral issues that is based on fear or resignation. Therefore, organizations should benefit from strengthening authentic leadership within their management culture. Gill et al. (2018) make the case that authentic leadership behaviors especially help to close the gap between intended HR practices and their actual implementation and the way they are eventually perceived by employees. This is because, authentic leaders assumingly do not blindly accept or reject HR policies, but seek to engage in dialogue about potential benefits, difficulties or inconsistencies (Gill et al., 2018). Authentic leadership in organizations can be enhanced by means of leader selection, training and development, and promotion. For example, in job interviews HR practitioners may ask candidates to reflect on moral challenges and dilemmas in vignette scenarios in order to gain insights on a candidate’s moral perspective (Pohling and Strobel, 2017). Gill et al. (2018) also suggest, to supplement selection processes using questionnaire measuring on authentic leadership. Regarding training and development of current supervisors and managers, authentic leader behaviors could be fostered by developing a leader’s self-awareness and self-reflective skills using coaching and 360° feedback. Higher levels of self-awareness, also emphasizing the social role as a leader, should enable leaders to better align their actions according to their own values and convictions.

Furthermore, our study revealed follower constructive cognition and moral efficacy as potential self-regulatory mechanisms to overcome silence and make employee voice more likely. Involving employees in moral discourse on difficult decisions or using case-analysis and role-taking in a training setting can improve cognitive reflection and moral reasoning of employees (cf. Hannah et al., 2011b). Moral efficacy may be facilitated by improving communication and feedback skills. When employees practice how to appropriately frame issues and give feedback in a constructive way, this should enhance their confidence in the ability to successfully cope with a difficult voice situation (Kish-Gephart et al., 2009). Furthermore, authentic leaders influence their employees’ moral efficacy by means of role-modeling and social learning. More generally speaking, we want to point out that instruments of organizational participation (e.g. employee opinion surveys, annual appraisal interviews, employee suggestion systems) are helpful for motivating employee voice behavior by providing voice opportunities and allowing employees to get involved (Wegge et al., 2010).

In conclusion, we believe that the development of a moral work climate and open feedback culture in organizations needs a multifaceted approach—fostering authentic leadership should be a promising starting point.