In modern organisations, many employees make the decision to remain silent about potentially important issues that arise (Tangirala & Ramanujam, Reference Tangirala and Ramanujam2008). Despite the fact that employees may have disagreements about work-related issues, such as the others’ rudeness or underlying perils within their organisation, they may not express their opinions and disagreements to their supervisors (Morrison, Reference Morrison2011). A survey by Milliken, Morrison, and Hewlin (Reference Milliken, Morrison and Hewlin2003) found that 85% of employees had deliberately withheld important work information, and 50% admitted that talking about organisational problems would make them uncomfortable. In some ways, the silence can be beneficial if it decreases excess information, reduces interpersonal conflicts, and increases information privacy among co-workers (Morrison & Milliken, Reference Morrison and Milliken2000). However, in most cases, the employee silence not only hinders information feedback and decision effectiveness, but it also reduces employee satisfaction and job involvement (Dyne, Ang, & Botero, Reference Dyne, Ang and Botero2003; Milliken et al., Reference Milliken, Morrison and Hewlin2003).
Expressing or supressing work information is a decision that every employee has to face every day, and a growing interest has been observed in the reasons for the decisions (Brinsfield, Reference Brinsfield2013; Xu, Loi, & Lam, Reference Xu, Loi and Lam2015). Previous studies have found that the behaviours or characteristics of leaders are important situational factors that determine employee silence (Edmondson, Reference Edmondson2003; Detert & Burris, Reference Detert and Burris2007). Milliken et al. (Reference Milliken, Morrison and Hewlin2003) argued that some interactions with leaders (such as unsupported leadership, lack of intimacy) led to negative psychological expectations and then affected an employee's silence behaviour. Therefore, in order to explore the impact of destructive leadership on employee silence, the current study aimed to understand the underlying mechanism behind employee silence more thoroughly and to provide guidance for improved management practices.
Hobfoll's (Reference Hobfoll1989) conservation of resources (COR) theory showed subordinates caught in a stress situation may engage in regulative tactics (i.e., attempt to maintain relationships by avoiding contact) and feedback avoidance behaviours (i.e., attempt to intentionally evade feedback from the supervisor) to select or create a comfortable environment to minimise the loss of their resources (Tepper, Moss, Lockhart, & Carr, Reference Tepper, Moss, Lockhart and Carr2007; Whitman, Halbesleben, & Holmes, Reference Whitman, Halbesleben and Holmes2014). In an employee's daily routine, apathy, ridicule and abuse from a supervisor will significantly reduce the subordinate's perceptions of their leader's consideration and participation, and subsequently increase their role stress (Teas, Reference Teas1983; Dale & Fox, Reference Dale and Fox2008). At this point, they need to use significant emotional resources to cope with their supervisors and deal with the stress, and keeping silence is a major way to preserve their emotional resources and achieve effective survival (Xu et al., Reference Xu, Loi and Lam2015). In addition, although high job complexity means that subordinates need to obtain ample work information (Hunter, Schmidt, & Judiesch, Reference Hunter, Schmidt and Judiesch1990), they are likely to suffer from a lack of information when they are faced with destructive leadership (Schyns & Schilling, Reference Schyns and Schilling2013). Meanwhile, conflicting information or not knowing what's expected of them may result in subordinates experiencing unpredictability or possible role stress (Teas, Reference Teas1983; Dale & Fox, Reference Dale and Fox2008). Therefore, we postulate that destructive leadership in high job complexity may be more imperilling to subordinates’ resources, which culminates in heightened role stress and silence behaviour. Figure 1 presents our conceptual model.
Figure 1 Conceptual model.
Our research makes the following contributions. First, our study attempted to establish the relationship between destructive leadership and employee silence by examining why and in which contexts such leadership behaviour affects employees’ silence behaviour in the workplace (Park, Carter, DeFrank, & Deng, Reference Park, Carter, DeFrank and Deng2016). We answered the call of a previous study for more research on subordinates’ passive responses by conducting an experimental study that examined the influence process of destructive leadership on employee silence via role stress (Tepper et al., Reference Tepper, Moss, Lockhart and Carr2007; Restubog, Scott, & Zagenczyk, Reference Restubog, Scott and Zagenczyk2011). It is one of the few studies that has examined the three dimensions of role stress as mediators in the effect of such a negative leadership behaviour on employee silence (Mackey, Frieder, Brees, & Martinko, Reference Mackey, Frieder, Brees and Martinko2015). Second, our research makes important contributions to the extant silence literature by investigating the antecedents and contextual factors of employee silence and identifying the potential moderating effects of job complexity. Most studies suggest that job complexity is a positive factor for employee behaviour (Jung, Nam, Lee, & Kim, Reference Jung, Nam, Lee and Kim2016; Mohan & Mulla, Reference Mohan and Mulla2013; Tierney & Farmer, Reference Tierney and Farmer2002), but we found a different result when simultaneously taking destructive leadership and job complexity into account to examine their interactive impact on employee behaviour. Third, our study broadens the current literature on the deleterious impacts of destructive leadership. Although destructive leadership is common and highly dysfunctional in the workplace, the research on the consequences of destructive leadership needs to be further explored (Einarsen, Aasland, & Skogstad, Reference Einarsen, Aasland and Skogstad2007). We responded to the call to pay attention to the crippling impact of abusive supervision and provide some suggestions for organisations to help prevent silence behaviour (Xu et al., Reference Xu, Loi and Lam2015).
Theoretical Background and Hypotheses
Destructive Leadership and Employee Silence
Scholars have been discussing the negative characteristics of personal charisma for a long time, and more researchers have been paying attention to various forms of destructive leadership and their negative effects (Einarsen et al., Reference Einarsen, Aasland and Skogstad2007). When employees experience destructive leadership, they have to spend more time and energy to struggle, survive, and preserve their limited resources, and thus they remain silent (Park et al., Reference Park, Carter, DeFrank and Deng2016). Because a leader is the recipient and disposer of the employees’ opinions and voice within organisations, various motivations may cause individuals to rarely report or express their concerns and opinions on job-related issues to their supervisors (Dedahanov et al., Reference Dedahanov, Dedahanov, Lee, Lee, Rhee and Rhee2016). From the perspective of COR theory (Hobfoll Reference Hobfoll1989), when subordinates are under a destructive leadership, they need to consume time and energy to cope with the stressful workplace event, instead of focusing on finding work problems and reporting them. Meanwhile, when destructive leadership depletes the employees’ resources, they will be more inclined to stay silent to protect their own resources (Halbesleben, Neveu, Paustian-Underdahl, & Westman, Reference Halbesleben, Neveu, Paustian-Underdahl and Westman2014; Park et al., Reference Park, Carter, DeFrank and Deng2016). Furthermore, employees tend to remain silent if the leaders are not interested in or hold a negative attitude to the advice and comments the employees offer (Vakola & Bouradas, Reference Vakola and Bouradas2005). When employees are aware that their voice may breed some risks or have an impact on their interpersonal relationships, they may become reluctant to speak up (Fuller, Marler, & Hester, Reference Fuller, Marler and Hester2006). Therefore, we believe that destructive leadership contributes to individuals’ decisions to share their concerns or keep silent. Thus, we offer the following hypothesis:
Hypothesis 1: Destructive leadership is positively related to employee silence.
The Mediating Effect of Role Stress
Role theory can interpret employees’ mentality, behaviour, and transition through their social role, and can reveal the individuals’ different roles in the organisational context (Tubre & Collins, Reference Tubre and Collins2000; Walker, Churchil, & Ford, Reference Walker, Churchill and Ford1975). When the perceptions of individual roles are uncontrolled, employees’ roles may become ineffective and create role stressors (Walker et al., Reference Walker, Churchill and Ford1975). Role stressors are conceptualised as role conflict, role ambiguity, and role overload using three constructs (Belasco, Reference Belasco1966; Behrman & Perreault, Reference Behrman and Perreault1984). Role conflict occurs when an employee perceives that the ‘real role’ and ‘expected role’ are incompatible; role ambiguity refers to the confusion of role expectations because of absent information or uncertainty about different role expectations; role overload is related to employees’ cumulative role tasks that exceed his or her abilities and it means they may be unable to finish the tasks in time (Conley & You, Reference Conley and You2014; Singh, Reference Singh1998).
The role episode model of Kahn, Wolfe, Quinn, Snoek, and Rosenthal (Reference Kahn, Wolfe, Quinn, Snoek and Rosenthal1964) suggests that employees who are unable to clearly distinguish the appropriate expectations and standards that are held by others or caused from their own behaviour may develop role stress. Employees’ perceptions of stress may be influenced by their cognitive ability, mental ability, social factors, and personal characteristics; while the instructions and requirements of destructive leadership can cause role conflict, role ambiguity, and role overload for employees (Singh, Reference Singh1998). For instance, when the expectations of destructive leadership are inconsistent with organisational goals, employees’ role conflicts will be increased (Teas, Reference Teas1983; Tubre & Collins, Reference Tubre and Collins2000). Subordinates who are ridiculed and belittled by their supervisors may not have enough information available to them. They may not clearly understand the expectations of their leader, which can result in a lack of expectations for their own behaviours and an increase in role ambiguity (Singh, Reference Singh1993; Tubre & Collins, Reference Tubre and Collins2000). Employees who are believed to be incompetent or stupid by leaders are often unable to complete difficult tasks by a set time or in specific conditions, and if they are given more role tasks than other employees, they may feel role overload (Peterson et al., Reference Peterson, Smith, Akande, Ayestaran, Bochner, Callan and Hofmann1995). Thus, we consider that destructive leadership is positively related to role stress of employees.
Moreover, COR theory provides an explanation for the mediating mechanism of role stress on destructive leadership and employee silence (Ng & Feldman, Reference Ng and Feldman2012; Xu et al., Reference Xu, Loi and Lam2015). Past consensual evidence has demonstrated that role stressors have a negative effect on employees’ attitudes and behaviours (Brown & Peterson Reference Brown and Peterson1993). Stressed individuals may be motivated to preserve their limited resources by avoiding actions (i.e., silence behaviour) to prevent their remaining resources being depleted (Hobfoll, Reference Hobfoll1989, Reference Hobfoll2001). Employees in role conflict may worry about being marked as complainers or troublemakers and thus lose desirable personal resources or career opportunities and, as a result, they do not speak up (Detert & Treviño, Reference Detert and Treviño2010). The employees in role ambiguity due to being abused by destructive leadership may remain silent because their voice will take more of their available resources and they will face more risk of future resource depletion (Ng & Feldman, Reference Ng and Feldman2012). Further, employees who experience role overload will try to retain their remaining resources when they are exhausted, and silence is a natural and safe way to protect themselves (Morrison, Reference Morrison2011). Accordingly, employees with role stress are motivated to minimise the adverse effects of stressors (Xu et al., Reference Xu, Loi and Lam2015). They will not consume their remaining resources to change the status quo, even though it may improve their work circumstances (Ng & Feldman, Reference Ng and Feldman2012). Moreover, even when they are conscious of the fact that silence is detrimental to the development of organisations, they still adopt silence to isolate themselves from the work to protect their limited valuable resources and prevent further resource loss (Bolton, Harvey, Grawitch, & Barber, Reference Bolton, Harvey, Grawitch and Barber2012). Thus, we suggest the following hypotheses:
Hypothesis 2a: Role conflict mediates the positive relationship between destructive leadership and employee silence.
Hypothesis 2b: Role ambiguity mediates the positive relationship between destructive leadership and employee silence.
Hypothesis 2c: Role overload mediates the positive relationship between destructive leadership and employee silence.
The Moderating Effect of Job Complexity
Job complexity is characterised by high autonomy, identity, feedback, skill variety and significance, which constitute the job characteristics model (Hackman & Oldham, Reference Hackman and Oldham1975, Reference Hackman and Oldham1980). Complex tasks make employees tend to express greater intrinsic motivation to develop work devotion and feel that their job is meaningful and important (Wang, Tsai, & Tsai, Reference Wang, Tsai and Tsai2014). Meanwhile, destructive leadership denotes specific supervisor behaviour that is perceived as hostile and/or obstructive by the employee (Schyns & Schilling, Reference Schyns and Schilling2013). Thus, job complexity and destructive leadership can coexist as different perspectives in the perceptions and actions of the employee (Coelho, Augusto, & Lages Reference Coelho, Augusto and Lages2011). Furthermore, although employees with different levels of job complexity may be affected by destructive leadership, those with more complex tasks may be more likely to perceive and remember destructive leadership (Wang et al., Reference Wang, Tsai and Tsai2014). Therefore, our study supported destructive leadership and job complexity as two independent variables, and job complexity was a moderating variable.
According to the results of Tepper (Reference Tepper2000) and Tepper et al. (Reference Tepper, Moss, Lockhart and Carr2007), the influence of negative leadership on subordinate behaviour and attitudes may be affected by job characteristics. From the perspective of COR theory, employees in complex jobs have more resources and higher levels of intrinsic motivations, and they are more likely to speak up than keep silent under the same conditions (Bakker & Demerouti, Reference Bakker and Demerouti2007). However, if they speak up and simultaneously encounter destructive leadership, they may lose those resources or be transferred to a less complex job, and they will therefore tend to keep silent in this situation (Hakanen, Bakker, & Demerouti, Reference Hakanen, Bakker and Demerouti2005). On the contrary, although those who work in less complex jobs possess relatively fewer resources and weak internal motivation, they tend to keep a lower frequency of silence as a way to get additional resources when they face destructive leadership (Einarsen et al., Reference Einarsen, Aasland and Skogstad2007; May, Wesche, Heinitz, & Kerschreiter, Reference May, Wesche, Heinitz and Kerschreiter2015). As a result, employees who are engaged in a high level of job complexity are more likely to remain silent than employees who are engaged in less complex jobs when they experience destructive leadership. Thus, we put forward the following hypothesis:
Hypothesis 3: Job complexity moderates the positive relationship between destructive leadership and employee silence, and this relationship is stronger when job complexity is high rather than low.
Incompatibile information sources may form an uncontrollable and unpredictable work context (Greenbaum, Mawritz, & Piccolo, Reference Greenbaum, Mawritz and Piccolo2015), where employees’ role stress will more likely occur (Tubre & Collins; Reference Tubre and Collins2000). In an organisation, job complexity can motivate employee performance (Wang et al., Reference Wang, Tsai and Tsai2014; Nurmi & Hinds, Reference Nurmi and Hinds2016), but it can also create hostility or obstructive behaviour (May et al., Reference May, Wesche, Heinitz and Kerschreiter2015). When these two situations occur simultaneously, employees may experience the psychological uneasiness of cognitive dissonance and feel greater role conflict (Coelho et al., Reference Coelho, Augusto and Lages2011). Similarly, with destructive leadership, employees with complex jobs may be more likely to experience role ambiguity than those with less complex jobs (Jackson & Schuler, Reference Jackson and Schuler1985) because job complexity entails more autonomy, and if there is a lack of focus and work formalisation, it may lead to increased levels of role ambiguity (Organ & Greene, Reference Organ and Greene1981; Rousseau, Reference Rousseau1978; Sorenson & Sorenson, Reference Sorensen and Sorensen1974). In addition, the motivational force of the job complexity may cause an overstimulation effect under destructive leadership (Singh, Reference Singh1998). Too much complexity in jobs can be perceived as information overload and make employees feel role overload (Singh, Reference Singh1998). Consequently, a high level of job complexity with destructive leadership may result in more role stress than those in less complex jobs. Accordingly, we propose the following hypotheses:
Hypothesis 4a: Job complexity moderates the positive relationship between destructive leadership and role conflict, and such a relationship is stronger when job complexity is high rather than low.
Hypothesis 4b: Job complexity moderates the positive relationship between destructive leadership and role ambiguity, and such a relationship is stronger when job complexity is high rather than low.
Hypothesis 4c: Job complexity moderates the positive relationship between destructive leadership and role overload, and such a relationship is stronger when job complexity is high rather than low.
From these assumptions, we further put forward the moderated mediation model analysis of job complexity; specifically, that the three dimensions of role stress (i.e., role conflict, role ambiguity, and role overload) mediate the positive effect of destructive leadership on employee silence. Job complexity not only moderates the relationship between destructive leadership and the three dimensions of role stress, but also moderates the relationship between destructive leadership and employee silence. Accordingly, we propose a moderated mediation model using the method of Preacher, Rucker, and Hayes (Reference Preacher, Rucker and Hayes2007). We expect that destructive leadership's indirect effect on employee silence through the three dimensions of role stress also depends on the level of job complexity. Thus, we offer the following hypotheses:
Hypothesis 5a: Job complexity moderates the strength of the indirect effect of destructive leadership on employee silence via role conflict, such that the mediated relationship is stronger when job complexity is high rather than low.
Hypothesis 5b: Job complexity moderates the strength of the indirect effect of destructive leadership on employee silence via role ambiguity, such that the mediated relationship is stronger when job complexity is high rather than low.
Hypothesis 5c: Job complexity moderates the strength of the indirect effect of destructive leadership on employee silence via role overload, such that the mediated relationship is stronger when job complexity is high rather than low.
Method
Sample and Procedure
The data were collected in Shanghai Municipality and Hangzhou City in China from October to December 2016. With the assistance of the management committees of two industries, we randomly selected 12 manufacturing firms to conduct a questionnaire survey. Our team contacted the human resources management department of each company, and received their support. With the help of the HR managers, we selected permanent staff under a direct leader as the research target. In order to avoid common method variance, we paired the sample by randomly selecting 3–6 leaders with 3–8 of the direct subordinates in each firm. The leader filled out the questionnaire first and marked the employees’ number, and then our research members found the corresponding employees to complete their questionnaires. Meanwhile, in order to reduce social desirability bias, we recorded the purposes and details of the survey and the voluntary nature of participation in the cover page of the questionnaire. We also assured anonymity and confidentiality to the participants and provided the contact information of the first author. Respondents were required to complete the questionnaires alone during working hours and return them in sealed envelopes. In this survey, all questionnaires of the model variables were filled out by subordinates. The leader only needed to complete basic information to serve as a reference for the subordinates’ evaluation. We distributed questionnaires to 96 supervisors and 426 subordinates, and of these, 82 and 352 respectively were returned. After excluding invalid and obvious abnormal questionnaires, there were 318 completed and valid samples from the 352 returned, representing an overall response rate of 74.65%. In this sample, 58.18% were male, 68.55% were aged from 26 to 40, the mean working time was 5.49 years, and 44.97% had received a university degree or higher.
Measures
Destructive leadership
We used a five-item scale developed and validated by Mitchell and Ambrose (Reference Mitchell and Ambrose2007) to assess destructive leadership. Sample items are ‘My supervisor ridicules me’ and ‘My supervisor tells me my thoughts or feelings are stupid’. Response options range from 1 = never, to 5 = always. Cronbach's alpha for this scale is 0.84.
Role stress
We referred to the role stress measurement index system designed by Peterson et al. (Reference Peterson, Smith, Akande, Ayestaran, Bochner, Callan and Hofmann1995) and measured role conflict with three items, role ambiguity with five items, and role overload with five items. A sample item of role conflict is ‘I often have to face some conflicting situations’. A sample item of role ambiguity is ‘I know what my job is’. A sample item of role overload is ‘I feel a heavy burden on my work’. Response options range from 1 = never, to 5 = always. It is worth mentioning that the scale of the role ambiguity are reverse items, and we carried out the reverse score processing in order to ensure the internal consistency of the role stress scale. The scale's reliability of role conflict, role ambiguity, and role overload are 0.82, 0.90, and 0.75 respectively.
Job complexity
We measured job complexity using a four-item scale adapted from a well-validated German scale (Zapf, Reference Zapf1993) that is widely used in Western countries. Sample items include: ‘Do you receive tasks that are extraordinary and particularly difficult?’ and ‘Can you use all your knowledge and skills in your work?’. Response options range from 1 = very little, to 5 = very much. The scale's reliability is 0.72.
Employee silence
We used a five-item scale designed by Dyne (2003) to assess employee silence. Sample items include: ‘Although I have ideas or suggestions to improve my work, I don't say it’ and ‘The problems are caused by the communication between leaders and employees, and I will not mention’. Response options range from 1 = strongly disagree to 5 = strongly agree. Cronbach's alpha for this scale is 0.86.
Control variables
Because all the respondents in our research were frontline employees, we controlled participants’ gender, age, education, and working time to be consistent with previous research (Hofmann, Schmeichel, & Baddeley, Reference Hofmann, Schmeichel and Baddeley2012; Tangirala & Ramanujam, Reference Tangirala and Ramanujam2008). Gender was a dummy variable, and males were 0 and females were 1. Age was coded as 1 = 25 or below, 2 = 26–35, 3 = 36–45, 4 = 46 or above. Working time was coded as 1 = 1 year or below, 2 = 2–5 years, 3 = 6–9 years, 4 = 9 years or above. Education was coded as 1 = high school or under, 2 = vocational school, 3 = university, 4 = graduate school.
Results
Confirmatory Factor Analysis (CFA)
We applied AMOS 21.0 to conduct confirmatory factor analysis of all variables for ensuring the rationality of the data structure. The discriminant validity of each scale was tested by comparing χ²/df and the comparative fit index (CFI), Tucker Lewis index (TLI), incremental fit index (IFI), goodness-of-fit index GFI, and root mean square error of approximation (RMSEA). As shown in Table 1, compared with other models, the six-factor model with 27 items fit our data better than the alternative models, which yielded a poor fit to the data. The data for χ² of 788.59 (df = 297, p < .01), CFI of 0.92, TLI of 0.91, IFI of 0.92, and RMSEA of 0.07 with the inspection of factor loadings and factor covariance suggest that all factor loadings were significant, demonstrating convergent validity. In addition, the baseline six-factor model has a significant improvement in χ² (model 1, Δχ²(3) = 253.03, p < .01; model 2, Δχ²(5) = 456.89, p < .01; model 3, Δχ²(6) = 458.41, p < .01; model 4, Δχ²(7) = 780.44, p < .01), which were a better fit to our data than the other alternative measurement models. Thus, the discriminant validity of the constructs was clear and suitable for the correlation analysis.
Table 1 Results of Confirmatory Factor Analysis of the Measurement Models
Note: DL = destructive leadership; RC = role conflict; RA = role ambiguity; RO = role overload; JC = job complexity; ES = employee silence; “+” means two factors are combined into one factor. **p < .01; *p < .05.
Descriptive Statistics
Table 2 reports means, standard deviations, and correlations of the variables. As predicted above, destructive leadership was positively related to employee silence (r = .20, p < .01), and significantly affects role conflict (r = .24, p < .01), role ambiguity (r = .15, p < .01), and role overload (r = .26, p < .01). Hypothesis 1 is supported initially and provides a basis for further analysis. Otherwise, consistent with our expectations, job complexity was not related to destructive leadership (r = .04, p > 0.05) but positively related to role conflict r = .34, p < .01), role ambiguity (r = .21, p < .01), and role overload (r = .33, p < .01).
Table 2 Means, Standard Deviation, and Correlations of Variables
Note: M = mean, SD = standard deviation. **p < .01; *p < .05. Cronbach's α reliability coefficients appear on the diagonal.
Hypotheses Tests
Table 3 shows the results of the mediating effect and the moderating effect generated by the PROCESS macro suggested by Hayes (Reference Hayes2013). We used model 4 in a PROCESS macro to test the mediation effect. According to Baron and Kenny's (Reference Baron and Kenny1986) suggestions, as can be seen from the test of mediation effect, destructive leadership is positively related to employee silence (β = 0.22, p < .01), which further supports hypothesis 1. When we include the mediator in the model, we can see that role conflict is still positively associated with employee silence (β = 0.50, p < .01), whereas the relationship between destructive leadership and employee silence is no longer significant (β = 0.06, p > .05); role ambiguity is still positively associated with employee silence (β = 0.45, p < .01), whereas the relationship between destructive leadership and employee silence significantly declines (β = 0.16, p < .05); and role overload is still positively associated with employee silence (β = 0.69, p < .01), whereas the relationship between destructive leadership and employee silence is no longer significant (β = 0.06, p < .05). Also, the Sobel test proved that the indirect effect (ab) is statistically significant (role conflict: effect size = 0.17, SE = 0.04, z = 4.21, 95% CI [0.09, 0.25], p < .01; role ambiguity: effect size = 0.07, SE = 0.03, z = 2.32, 95% CI [0.02, 0.14], p < .05; role overload: Effect size = 0.17, SE = 0.04, z = 4.22, 95% CI [0.09, 0.26], p < .01). The results with a bootstrapped 99% confidence interval around the indirect effect not containing zero (role conflict: 0.07 ~ 0.28; role ambiguity: 0.01 ~ 0.17; role overload:0.07 ~ 0.29) confirmed the Sobel test (MacKinnon, Lockwood, & Williams, Reference MacKinnon, Lockwood and Williams2004). Therefore, role conflict and role overload fully mediated the positive relationship between destructive leadership and employee silence, and role ambiguity partly mediated the positive relationship between the two. Thus, hypotheses 2a, 2b, and 2c are supported.
Table 3 Conditional process analysis
Note: **p < .01; *p < .05. Unstandardised regression coefficients were reported. Bootstrap sample size = 5,000. LL = lower limit, UL = upper limit, CI = confidence interval. Bias-corrected CI is reported. Low = 1 SD below the mean; High = 1 SD above the mean.
We used model 1 in a PROCESS macro to test moderation effect. From the test of moderation effect shown in Table 2, the interaction of destructive leadership with job complexity had a significant effect on employee silence (β = 0.31, p < .01), role conflict (β = 0.23, p < .01), role ambiguity (β = 0.23, p < .01), and role overload (β = 0.17, p < .05). This illustrates that hypotheses 3, 4a, 4b, 4c are supported. Furthermore, in order to obtain a more intuitive response, the interaction effects of job complexity at different levels (i.e., one standard deviation above or below the mean level), we plotted the moderating effect figure following the recommendation of Aiken and West (Reference Aiken and West1991). Figures 2, 3, 4, and 5 show that the positive relationship between destructive leadership and employee silence, and the positive relationship between destructive leadership and the three dimensions of role stress are relatively stronger in high job complexity rather than low.
Figure 2 Job complexity as a moderator of destructive leadership and employee silence.
Figure 3 Job complexity as a moderator of destructive leadership and role conflict.
Figure 4 Job complexity as a moderator of destructive leadership and role ambiguity.
Figure 5 Job complexity as a moderator of destructive leadership and role overload.
The above analysis has laid the foundation for the development of a moderated mediation model. In the last part of Table 3, we used model 8 in a PROCESS macro to test the moderated mediation effect. The conditional indirect effect of destructive leadership on employee silence through role conflict, role ambiguity, and role overload are stronger and significant for high job complexity (effect size = 0.21, 0.13, 0.21, 95% bias-corrected confidence interval from 0.11 to 0.32, 0.05 to 0.23, 0.11 to 0.34, without zero) but weaker and not significant for low job complexity (effect size = 0.07, −0.02, 0.06, 95% bias-corrected confidence interval from −0.02 to 0.16, −0.08 to 0.05, −0.01 to 0.16, containing zero). Consequently, our theoretical model satisfies conditions of the moderated mediation effect, and hypotheses 5a, 5b, 5c are supported and validated.
Discussion
Theoretical Implications
The theoretical contributions in our study include the following two aspects. First, our study attempted to extend the silence literature by shedding light on the mediating mechanism of role stress on the relationship between destructive leadership and employee silence. Although previous studies have reviewed the negative outcomes and effects of destructive leadership (e.g., Schyns & Schilling, Reference Schyns and Schilling2013; Thoroughgood, Padilla, Hunter, & Tate, Reference Thoroughgood, Padilla, Hunter and Tate2012), the literature about the antecedents and underlying mechanism of silence triggered by destructive leadership is still limited (Martinko, Harvey, Brees, & Mackey, Reference Martinko, Harvey, Brees and Mackey2013; Mackey et al., Reference Mackey, Frieder, Brees and Martinko2015). Some studies have examined the mediating mechanism of destructive leadership on employee silence and its detrimental psychological consequences (e.g., psychological distress, fear, emotional exhaustion; Kiewitz, Restubog, Shoss, Garcia, & Tang, Reference Kiewitz, Restubog, Shoss, Garcia and Tang2016; Xu et al., Reference Xu, Loi and Lam2015), but the linkage of the ‘destructive leadership role’ and ‘stress-employee silence’ has not been established, and this study is the first to investigate the mediating mechanism of role conflict, role ambiguity, and role overload between destructive leadership and employee silence. Accordingly, our findings indicate that, drawing on COR theory and role theory, destructive leadership depletes employees’ resources by shaping a high level of role conflict, role ambiguity and role overload, which can result in employee silence. The research provides a strong basis to understand more deeply the motivating mechanism behind employee silence and expands the path of the existing research on the relationship between destructive leadership and employee silence.
Second, another contribution of this work is that the examination of job complexity as a new explanatory variable enriches the extant silence literature and provides a different perspective to understand the moderating mechanism on the relationship between destructive leadership and role stress (Farh & Chen, Reference Farh and Chen2014). We found that complex jobs do not motivate employee behaviours when they experience destructive leadership. That is to say, destructive leadership for high-level job complexity is more beneficial to role stress than for less complex jobs. This conclusion is not consistent with the argument that job complexity has a positive influence on employee behaviour (Mohan & Mulla, Reference Mohan and Mulla2013; Jung et al., Reference Jung, Nam, Lee and Kim2016). We developed previous research by combining COR theory and the job characteristics model to link destructive leadership concurrently with job complexity to other organisational consequences (i.e., role stress, employee silence). Accordingly, we further inferred that job complexity may not necessarily promote positive outcomes. Therefore, in order to obtain comprehensive research of how a leader exerts their influence on employee behaviour, more research on the interactive role of leadership behaviours and context boundary conditions should be undertaken (Restubog et al., Reference Restubog, Scott and Zagenczyk2011; Xu et al. Reference Xu, Loi and Lam2015). In addition, our studies enrich the theoretical interpretation of the leader influence on employees’ choice to speak up or remain silent, and provides a new perspective for future research to uncover the underlying mechanism.
Practical Implications
In recent years, with the more intense competition of internal and external environments faced by organisations, and as organisational tasks have become more complicated, the frequency of destructive leadership is also increasing. Employee silence is an important tactic in workplace survival under destructive leadership, and it brings serious threats for organisations’ development, especially for important issues that need to be resolved in a timely manner, as well as for key information and creativity. Employees still remain silent when they are aware of the pressure brought by the destructive leadership, which greatly affects the operation and development of the organisation, and it can have wide-ranging destructive effects on all organisational levels. Therefore, the practical implications of our research for management is to have reasonable intervention strategies and to effectively prevent and inhibit the harm and consequences of destructive leadership in the organisation.
Specifically, first, the organisation needs to construct a control system and punishment mechanism for destructive leadership. With regard to the situation when employees facing destructive leadership tend to keep silent and not report the problem, the organisation should establish a safety complaint channel to avoid employees suffering retaliation and encourage their timely feedback for problems in the workplace. Second, the organisation should provide adequate conditions and resources protection for employees, alleviate role stress brought by destructive leadership, and reduce the negative emotion and cognitive embodiment from role conflict, role ambiguity, and role overload. Meanwhile, the organisation should also explore how it can achieve the best combination of work requirements and resources so that employees can work better and maintain a good conditions so as to reduce the perception of role stress. Finally, because the influence of job complexity on destructive leadership and employee silence behaviour is consistent, managers need to fully consider the complexity of the work content and reasonably assign different tasks for different employees. With the help of job design and organisational management measures to standardise the work requirements, employees can formulate flexible work plans and have a voice, no matter what situations they are facing.
Limitations and Future Research
Although this study achieved some results, there were certain limitations that need further exploration. On the one hand, our study focused on the influence of destructive leadership on employee behaviour in the workplace, but it did not explore destructive leadership when it has an impact on the negative behaviour of employees outside the workplace. Recently, some experts in the field of organisational behaviour have pointed out that the encounter of employees in the workplace may lead to deviant behaviours outside the workplace, which can have an impact on their work-family relationship and work-life balance. Conversely, the encounter of employees in their social life may cause negative behaviours in the workplace and may affect their relationships with colleagues and organisations. Therefore, future research should focus on expanding the boundary of destructive leadership and analysing the role mechanism of destructive leadership on employees’ behaviours in different contexts (such as internally and externally to the organisation). In particular, we advise researchers to consider the impact of destructive leadership on employees’ behaviour outside the workplace, using various mediating and moderating factors.
On the other hand, future research could also carry out a comparative study of cross-cultural backgrounds; for instance, by comparing the data in a Chinese cultural background with the data from developed countries in Europe and America, and analysing differences in the data collected from various Eastern cultures in mainland China and other Eastern countries or areas (e.g., Hongkong, Macao, Taiwan, Singapore, South Korea, Japan) that are deeply affected by the Chinese cultural circle.
Conclusion
As a hot issue in the field of negative organisational behaviour, destructive leadership has been given much attention by many scholars and entrepreneurs. The article was inspired by the present literature and the circumstances of China. Combining the COR theory, role theory and job characteristics model, our research adopted both qualitative and quantitative methods to discuss the influence results and role paths of destructive leadership in the Chinese cultural background and constructed a theoretical model of the impact of destructive leadership on employee silence. After collecting and analysing the data to validate the model, we obtained the following conclusions. (1) Disruptive leadership has a significant positive impact on employee silence; the more destructive the leadership, the more the employee silence in the organisation. (2) The three dimensions of role stress had a mediating effect on destructive leadership and employee silence, and destructive leadership usually motivates employees to produce a situation of role conflict, role ambiguity and role overload, and then form employee silence. (3) Job complexity moderates the positive relationship between destructive leadership and employee silence, as well as the positive relationship between destructive leadership and the three dimensions of role stress. The role conflict, role ambiguity, and role overload from destructive leadership are more likely to cause employee silence when the job complexity is high. (4) Job complexity moderates the strength of the indirect effect of destructive leadership on employee silence via role conflict, role ambiguity and role overload, such that the mediated relationship is stronger when job complexity is high rather than low.
Open access
