Authoritarian leaders control every organizational decision, demand absolute compliance, and threaten sanctions for disobedience. Authoritarian leaders may evoke negative reactions in employees by depriving them of their autonomy; generating ambiguity; inducing fear, perceptions of threat, and stress; neglecting their social-emotional needs; or disrupting dyadic relationships (Chan et al., 2013; Chen et al., 2014; Farh & Cheng, 2000; Guo et al., 2018; Schaubroeck et al., 2017; Zhang & Xie, 2017; Zheng et al., 2021). Anecdotal evidence, however, shows that authoritarian leadership continues to be widely used by managers worldwide (Harms et al., 2018; Huang et al., 2015). Farh and Cheng (2000), in their seminal work, suggested that authoritarian leadership may be conducive to organizational effectiveness, which may help explain its ubiquity (see also Chen et al., 2014; Harms et al., 2018; Huang et al., 2015). Authoritarian leadership may be used to press employees to meet high task demands, such as harsh deadlines and heavy workloads. Previous studies, however, have reported a weak or mixed relationship between authoritarian leadership and work performance (Chen et al., 2014; Chou et al., 2015; Gu et al., 2020).

Scholars underline that it is essential to understand how and when authoritarian leadership enhances or undermines work performance (Chen et al., 2014; Huang et al., 2015; Schaubroeck et al., 2017). To this end, we draw from social information processing theory (Griffin, 1983; Salancik & Pfeffer, 1978) and work stressors literature (Cavanaugh et al., 2000; Podsakoff et al., 2007) to propose that authoritarian leadership may simultaneously generate challenge stressors, which are stressful demands that must be met in order to achieve performance goals, and hindrance stressors, which are stressful demands that thwart attainment of performance goals attainment. Challenge and hindrance stressors may yield counterbalancing pathways underlying the relationship between authoritarian leadership and work performance, causing, at least in part, the mixed results observed in the literature.

Social information processing theory suggests that employees learn from social cues and make sense of the work environment via social interactions in the workplace. Leaders are “meaning makers” (Ashford et al., 2009) and construct work experiences for employees by setting deadlines, assignments, and responsibilities and by signaling their implicit expectations to employees (Griffin, 1983; Salancik & Pfeffer, 1978; Thomas & Griffin, 1983; Zheng et al., 2022). Authoritarian leaders expect absolute compliance with their instructions and signal possible punishments for disobedience (Cheng et al., 2014; Farh & Cheng, 2000), thereby generating work stressors, defined by Parasuraman and Alutto (1981) as “job demands, constraints, and job-related events or situations that may affect an individual’s role fulfillment” (p. 51). More importantly, leaders may direct their authoritarian behaviors at different work domains and therefore produce contrasting types of work stressors.

On the one hand, authoritarian leaders may demand compliance in meeting deadlines, fulfilling duties, and taking responsibility for tasks, which are the key features of challenge stressors (LePine et al., 2005; Zhang et al., 2014). Challenge stressors may signal a positive relationship between effort expended on these demands and the likelihood of meeting them; thus, as authoritarian leaders convey their expectations firmly, they drive employees to achieve high work performance (LePine et al., 2005; Wallace et al., 2009). On the other hand, authoritarian leaders may highlight their authority by demanding compliance in fulfilling their personal preferences (Farh & Cheng, 2000; Lau et al., 2007; Shen et al., 2019). They may demand that employees strictly follow their improvised instructions, driven by self-interests, and evaluate employees based on the leader’s personal preferences, which may produce ambiguity, job insecurity, politics, red tape, and career bottlenecks for their employees. These hindrance stressors may signal that no reasonable level of effort will be adequate to address the leader’s demands (LePine et al., 2005). As a result, authoritarian leaders may hinder employees from exerting efforts to perform. Therefore, we propose that authoritarian leadership has both positive and negative impacts on work performance by producing both challenge and hindrance stressors.

Further, we attempt to address why some authoritarian leaders generate more challenge stressors and fewer hindrance stressors, leading to better work performance. Because personal value may influence how individuals choose and apply from the available modes, styles, and means of behaviors (Fu et al., 2010; Kluckhohn, 1951), leader’s personal value may influence the way leaders use authoritarian-style of behaviors. So, we depart from previous research that focuses on how employees’ personal values, such as employees’ power distance orientation, determine employees’ reactions toward authoritarian leadership (Li & Sun, 2015; Schaubroeck et al., 2017; Shen et al., 2019) and propose that leaders’ power distance orientation may dictate which domain leaders may direct their authoritarian behaviors at and corresponding types of stressors they may produce to affect performance.

Leaders’ power distance orientation refers to the extent to which leaders endorse an unequal distribution of power between themselves and their employees (Clugston et al., 2000; Hofstede, 1980, 2001; Kirkman et al., 2009). High power distance leaders may rationalize the unequal power distribution by seeing leaders as superior and employees as inferior and incapable (Dorfman & Howell, 1988; Farh et al., 2007), and thus may use authoritarian behaviors to satisfy their own preferences and demonstrate superiority (Shen et al., 2019) rather than using employees’ potential to address challenging demands. So, authoritarian leaders high on power distance orientation may generate more hindrance stressors than challenge stressors, which in turn harm performance. In contrast, low power distance leaders may not view employees as inferior but believe that employees have potential to perform (Farh et al., 2007; Hu et al., 2018; Kirkman et al., 2009), and thus may use authoritarianism to press employees with challenges (Chen et al., 2014) rather than imposing unmanageable obstacles. So, authoritarian leaders low on power distance orientation may generate more challenge stressors than hindrance stressors, which in turn promote performance.

Our research makes three contributions to the literature. First, we contribute to authoritarian leadership literature by showing that authoritarian leadership is not entirely negative and that its positive effects may be offset by its concomitant negative effects. We therefore reconcile the previous inconsistent findings of the effects of authoritarian leadership by disentangling both positive and negative effects of it through differential pathways (Chen et al., 2014; Pellegrini & Scandura, 2008; Schaubroeck et al., 2017). More importantly, we extend the current understanding by theorizing and testing the positive pathway of authoritarian leadership, which supports previous theoretical prediction that it drives employee performance by demanding compliance (Chen et al., 2014; Farh & Cheng, 2000). Second, we build on social information processing theory (Salancik & Pfeffer, 1978) to suggest that authoritarian leadership may shape work experiences and generate work stressors for employees. We advance two new explanatory mechanisms—challenge stressors and hindrance stressors—underlying the effects of authoritarian leadership on employees’ work performance. Finally, we contribute to the literature by theorizing that leaders’ power distance orientation influences how leaders exercise their authoritarian behaviors. We offer new insights into the role of leaders’ personal values in determining how they direct their authoritarian leadership behaviors to different domain of work, generating either beneficial or detrimental effects on performance (Fig. 1).

Fig. 1
figure 1

Theoretical model

Full size image

Theory and hypotheses

Authoritarian leadership

Authoritarian leaders engage in behaviors that impose substantial control over employees, including exercising strict discipline, compelling employees to obey all instructions, and using the threat of punishment to push employees to deliver their best performance (Chen et al., 2014; Farh & Cheng, 2000; Schaubroeck et al., 2017). Compared to leaders who engage in other types of controlling leadership, such as directive leadership and initiating structures (Judge et al., 2004), authoritarian leaders exert stronger authority and absolute control over their followers and do not permit challenges to their authority. Both directive leadership and initiating structures provide employees with specific guidance and mechanisms for achieving desired organizational goals (Judge et al., 2004). In contrast, authoritarian leaders focus on regulating work behaviors, and they achieve organizational goals through punishment and forced compliance rather than by offering clear guidance (Farh & Cheng, 2000). Authoritarian leadership is distinct from abusive supervision (Li et al., 2021; Thau et al., 2009). Authoritarian leadership is characterized by the exercise of punishment and control conditional on employees’ compliance, whereas abusive supervision is characterized by a sustained display of hostility and degradation of employees, sometimes irrespective of their compliance (Mackey et al., 2017; Tepper, 2000, 2007).

Research on the effects of authoritarian leadership on work performance has produced inconsistent results. Researchers suggest that authoritarian leadership has negative impacts: it deprives employees of autonomy and self-determination (Den Hartog and De Hoogh, 2009), shows a lack of respect for employees’ input (Chen et al., 2014), undermines employees’ self-worth (Chan et al., 2013), generates uncertainty about how to interact with the leader (Zheng et al., 2021), and induces feelings of powerlessness (Li et al., 2021), fear, threat, and stress (Gu et al., 2020; Guo et al., 2018), and therefore demotivates employees to achieve better performance (Chan et al., 2013; De Cremer, 2006). From a social exchange perspective, authoritarian leaders who threaten employees with punishment do not “offer the socio-emotional benefits needed to initiate reciprocal interrelations” (Chen et al., 2014, p. 6). Thus, authoritarian leadership undermines work performance because it fails to induce employee trust in leaders, which is critical for a high-quality social exchange relationship (Chen et al., 2014; Shen et al., 2019; Wu et al., 2012). Also, past research has not shown consistent empirical evidence regarding the effects of authoritarian leadership on in-role performance, that is, the completion of required tasks and work roles, or extra-role performance, that is, performing expanded work responsibilities (Williams & Anderson, 1991). Some studies have found that authoritarian leadership is negatively associated with in-role performance (Chan et al., 2013; Schaubroeck et al., 2017) and extra-role behaviors (Chan et al., 2013; Chen et al., 2014; Gu et al., 2020; Zheng et al., 2021). Chen and colleagues (2014) found that authoritarian leadership had no significant effect on in-role performance and was detrimental to extra-role performance.

Other studies suggest a potential positive effect of authoritarian leadership but have provided limited theoretical reasoning or empirical evidence for this observation. Authoritarian leaders signal that employees’ disobedience could lead to sanctions. Therefore, employees tend to abide by work instructions, rules, and demands of authoritarian leaders and put forth effort to avoid punishment (Cheng et al., 2014; Farh & Cheng, 2000; Zheng et al., 2020). Following this logic, Zheng et al. (2020) found that authoritarian leaders deter employee deviance because of the threat of punishment. Although several researchers argue that an authoritarian leadership style may improve employee performance by securing employees’ compliance (Chen et al., 2014; Farh & Cheng, 2000; Huang et al., 2015), no empirical evidence provides direct support for this assertion. Instead, there is evidence that under certain circumstances, authoritarian leadership can contribute to performance at the team (De Hoogh et al., 2015), firm (Huang et al., 2015), and individual levels (Chou et al., 2010; Wang & Guan, 2018).

Building on their findings of the positive effects of authoritarian leadership on performance at the firm level, Huang et al. (2015) called for further theoretical development on the functionality of authoritarian leadership at both the firm and individual levels. Chen et al. (2014) and Takeuchi et al. (2020) also pointed out the need to develop alternative frameworks for a deeper understanding of the impact of authoritarian leadership on work performance. Therefore, we propose and examine a dual-path model of authoritarian leadership and provide evidence of its functionality by drawing on social information processing theory and a challenge/hindrance stressor framework.

The roles of dual stressors from the perspective of social information processing

Social information processing theory suggests that employees’ work experiences are socially constructed. Salancik and Pfeffer (1978) noted that individuals make judgments and meaning of their work environment based on processing social information in the workplace. Workplace social information indicates how employees should act at work by shaping their perceptions and sense-making of work environments, from which they may learn to adapt their attitudes and behaviors. Social interactions in the workplace shape employees’ work experiences because in these interactions employees receive social cues about what is expected, what should be done, what is salient, and what is appropriate. Employees may “test and confirm their perception of reality” by comparing and aligning their perceptions, attitudes, and actions with these social cues (Brass & Burkhardt, 1993, p. 447). Therefore, social information controls the process of constructing the characteristics of work demands, tasks, and stressors (Chen et al., 2013; Lin et al., 2020; Piccolo & Colquitt, 2006; Salancik & Pfeffer, 1978; Shamir et al., 1993).

Leaders, for example, as the center of attention in a work context, play a critical role in sending social information cues to help employees make sense of their work environments (Griffin, 1983; Piccolo & Colquitt, 2006; Thomas & Griffin, 1983). Leaders may shape employees’ sense-making by using their authority to construct observable and explicit work conditions, such as setting rules and procedures, stipulating job requirements, and formalizing behavioral expectations; they may also signal and impose unobservable and implicit expectations and norms through their actions, decision-making patterns, or interactions with employees (Griffin, 1983; Lu et al., 2019; Smircich and Morgan, 1982; Zhang & Xie, 2017). Scholars have demonstrated that transformational leadership, such as intellectual stimulation and visionary motivation, tends to provide powerful informational cues that prompt employees to view their jobs as having high levels of autonomy, importance, meaning, and challenge (Fernet & Austin, 2014; Fernet et al., 2015; Gillet & Vandenberghe, 2014; Piccolo & Colquitt, 2006; Purvanova et al., 2006).

Taking a social information processing perspective (Griffin, 1983; Thomas & Griffin, 1983), we contend that as authoritarian leaders demand absolute compliance, they are likely to shape employees’ sense-making by imposing different types of work stressors, which have implications for work performance. In a controlling fashion, authoritarian leaders may exacerbate the demands associated with employees’ work. If such leaders assign excessively tough goals and deadlines, scrutinize employees’ slightest actions, and discipline employees who exhibit minor disobedience, employees may perceive their work demands as overwhelming and stressful (Che et al., 2017; Fernet & Austin, 2014). These stressful demands are work stressors, which may be of different types and have correspondingly different effects on work performance (Cavanaugh et al., 2000; LePine et al., 2005).

In the challenge/hindrance stressor framework, challenge stressors signal that there is an association between effort expended on these demands and the likelihood of accomplishing them, thereby driving employees toward addressing work demands and achieve better performance. In contrast, hindrance stressors, such as role ambiguity and job insecurity, are unmanageable and hold little promise to address them; thus they are detrimental to work performance (Cavanaugh et al., 2000; Crawford et al., 2010; LePine et al., 2005; Wallace et al., 2009). Empirical studies (LePine et al., 2016; Rodell & Judge, 2009; Wallace et al., 2009; Webster et al., 2011) and meta-analyses (Crawford et al., 2010; LePine et al., 2005; Podsakoff et al., 2007) have generally shown that although both types of stressors generate strain, challenge stressors are associated with positive work performance and hindrance stressors are associated with negative work performance.

We propose that authoritarian leaders may simultaneously generate both types of stressors for employees, which may have differential effects on work performance. Authoritarian leadership behaviors, such as making all decisions, demanding compliance, and disciplining disobedience, may be directed at different work domains, produce different sense-making for employees, and therefore generate challenge and hindrance stressors that have contrasting effects on work performance. On the one hand, authoritarian leaders may promote work performance by generating challenge stressors. The strong message conveyed by authoritarian leaders to meet the performance expectations may manifest in various ways, such as assigning large projects to employees and setting high performance standards, goals, and deadlines that are difficult to meet. Also, authoritarian leaders may make employees’ responsibilities more explicit and salient by imposing punishment if employees fail to fulfil their duties (Chen et al., 2014; Farh & Cheng, 2000). Hence, employees who work under authoritarian leaders may be forced to take on more responsibilities and meet high performance standards. These stressful demands constitute challenge stressors at work. Challenge stressors represent an implicit “deal” between authoritarian leaders and employees, signaling that by meeting the demands, they could avoid punishment. Employees may also experience a sense of accomplishment for achieving performance goals and meeting leaders’ expectations (LePine et al., 2005; Wallace et al., 2009; Zhang et al., 2014). Furthermore, because the implicit deal entailed in challenge stressors may convey to employees that their organization treats them equitably, they may devote themselves to achieving organizational goals in return, which may result in positive work performance (McNeely & Meglino, 1994; Organ & Konovsky, 1989; Zhang et al., 2014).

On the other hand, authoritarian leaders may undermine work performance by generating hindrance stressors. They may highlight their own authority by demanding compliance in fulfilling personal preferences and self-interests (Dorfman et al., 1997; Farh & Cheng, 2000; Lau et al., 2007; Schuh et al., 2013; Shen et al., 2019). They may demand that employees strictly follow their improvised instructions without providing sufficient explanations, causing confusion about their exact expectations and standards and how to accomplish tasks properly and thereby generating ambiguities for employees (Stellmacher & Petzel, 2005; Thau et al., 2009; Zhang & Xie, 2017). Also, authoritarian leaders may signal that their evaluations of employees are based on their personal preferences rather than objective indicators and that meeting their personal agenda, preferences, and expectations is the key to survive in the workplace, driving employees to engage in political behaviors to retain their job or secure advancement (Ferris et al., 1996). Authoritarian leaders can hinder employees’ career progress because they focus more on their own preferences (Shen et al., 2019) than on employees’ growth needs. Further, employees may experience a lack of control and security when they feel that their authoritarian leader’s behaviors and instructions strengthen the leader’s own superiority and are driven by the leader’s quest for personal benefits (Shen et al., 2019). Experiences of ambiguities, politics, career bottlenecks, and insecurity constitute hindrance stressors, which may distract employees from their work, drain their personal resources, and render them unable to maintain high in-role performance (Webster et al., 2011). In addition, hindrance stressors communicate reduced promise of future gains; signal unfair, devalued, and disrespectful organizational practices, and, in turn, hinder them from engaging in extra-role performance (Crawford et al., 2010; Eatough et al., 2011; Rodell & Judge, 2009; Zhang et al., 2014).

We use in-role performance and extra-role performance as two important indicators of work performance (Rotundo & Sackett, 2002; Wallace et al., 2009; Zhang et al., 2014). A central emphasis of leadership research is to examine the influence of leadership on employee performance, both, in-role and extra-role (Chan & Mak, 2012; Chan et al., 2013; Xu et al., 2012). Most studies examining the relationship between challenge and hindrance stressors and work performance have considered both in-role and extra-role performance and supported the notion that these two stressors have contrasting effects on work performance (LePine et al., 2005; Ohly & Fritz, 2010; Rodell & Judge, 2009; Wallace et al., 2009; Webster et al., 2010; Zhang et al., 2014). We propose that authoritarian leadership may compel employees to improve in-role and extra-role performance by generating challenge stressors and may hinder employees from performing well on these two performance indicators by shaping hindrance stressors.

Hypothesis 1

Authoritarian leadership has a positive indirect effect on in-role performance (1a) and extra-role performance (1b) via challenge stressors.

Hypothesis 2

Authoritarian leadership has a negative indirect effect on in-role performance (2a) and extra-role performance (2b) via hindrance stressors.

The roles of leader power distance orientation

Personal values influence “the selection from available modes, means, and ends of actions” (Kluckhohn, 1951; 395) and thus influence the way leaders exercise their leadership behaviors. As a type of personal value (Farh et al., 2007; Hofstede, 1980; Kirkman et al., 2009), power distance orientation may dictate the domains leaders exercise their authoritarian behaviors (Brockner et al., 2001; Fu et al., 2010), thereby affecting the types of stressors they produce (Salancik & Pfeffer, 1978). High power distance leaders may underscore the unequal power distribution between leaders and employees, highlight their own superiority, and underestimate employees’ capacities (Dorfman & Howell, 1988; Farh et al., 2007). They may use authoritarian leadership behaviors to satisfy their own preferences and show their superiority rather than using employees’ potential at work (Farh et al., 2007; Kluckhohn, 1951). So, these leaders may not direct authoritarian behaviors at demanding that employees handle heavier workloads, meet tighter deadlines and higher performance standards, and take on more responsibilities; their employees may experience fewer challenge stressors (LePine et al., 2005, 2016). Instead, they are more likely to give improvised instructions, neglect employees’ growth needs, and make evaluations based on their own preferences (Chan et al., 2013; Shen et al., 2019); their employees may experience more hindrance stressors (LePine et al., 2005, 2016). Thus, authoritarian leaders high on power distance orientation may generate fewer challenge stressors and more hindrance stressors, which ultimately lead to lower in-role and extra-role performance (LePine et al., 2005, 2016).

By contrast, low power distance leaders are less likely to highlight their own superiority and underestimate employees’ capacities (Dorfman & Howell, 1988; Farh et al., 2007). Therefore, they are less likely to erect unmanageable obstacles for employees, such as giving improvised instructions that cause employees unable to clearly understand what is expected at work. These leaders are more likely to seek employees’ input and contributions by imposing more challenge stressors (Chen et al., 2014), such as giving heavier workloads, tighter deadlines, and a broader scope of responsibility (LePine et al., 2005). Thus, authoritarian leaders low on power distance orientation may generate more challenge stressors and fewer hindrance stressors, which ultimately lead to higher in-role and extra-role performance (LePine et al., 2005, 2016).

Hypothesis 3

Leaders’ power distance orientation moderates the indirect relationships between authoritarian leadership and in-role performance (3a) and extra-role performance (3b) via challenge stressors, such that the indirect relationships are more positive when leaders’ power distance orientation is low rather than high.

Hypothesis 4

Leaders’ power distance orientation moderates the indirect relationships between authoritarian leadership and in-role performance (4a) and extra-role performance (4b) via hindrance stressors, such that the indirect relationships are more negative when leaders’ power distance orientation is high rather than low.

Overview of our research

We conducted two studies to test our hypotheses. In Study 1, using samples from China and the United Kingdom, we examined the positive indirect effects of authoritarian leadership on objective performance and extra-role performance as rated by peers via challenge stressors (Hypotheses 1a and 1b) and the adverse indirect effects of authoritarian leadership via hindrance stressors (Hypotheses 2a and 2b). In Study 2, using a different Chinese sample, we verified the mediating effects of the two stressors and further investigated which type of authoritarian leader is more likely to push employees toward better performance by generating challenge stressors rather than hindrance stressors. Specifically, we examined the moderating role of leader power distance orientation in the indirect effects of authoritarian leadership on supervisor-rated in-role performance and extra-role performance via shaping dual stressors (Hypotheses 3a, 3b, 4a, and 4b).

Study 1 Methods

Participants and procedure

We tested our mediation hypotheses in Study 1. To enhance cultural generalizability, we included samples from different cultures, namely China (Sample 1: catering chain) and the United Kingdom (Sample 2: police force). In Sample 1, we measured employee in-role performance using objective indicators and tested Hypotheses 1a and 2a. We were unable to collect in-role performance data for Sample 2 because the human resources department in the targeted force felt that the items were not applicable in the policing context. Thus, for Sample 2, we only tested the hypotheses related to extra-role performance (Hypotheses 1b and 2b).

Sample 1

We surveyed four province-level subsidiaries of a large chain catering company located in Guangdong, Guangxi, Fujian, and Zhejiang provinces in China. At the time of our survey, each of the four subsidiaries had about 500 restaurants. Each restaurant employed approximately 70 employees, including one restaurant manager, two or three deputy managers, and about 60 subordinates. We treated restaurant managers and deputy managers as “employees” and regional managers (supervisors of restaurant managers), as “leaders.” The restaurant managers were responsible for recruitment, overseeing restaurant operations, service quality, and setting and meeting sales goals. Each regional manager managed five to six restaurants and the performance of the restaurant managers. Regional managers routinely inspected the restaurants, evaluated the performance of restaurant managers, and bridged communication between restaurant managers and headquarters. Regional managers visited the restaurants under their supervision at least twice a month.

With the assistance of senior human resources managers at the head office, we randomly invited 480 restaurant managers and 115 supervising regional managers to participate in our study. We collected the data at three time points at two-month intervals. At Time 1, the restaurant managers were asked to report the frequency of authoritarian leadership behaviors of their direct supervisors (i.e., regional managers) and their demographic information. Two months later, at Time 2, we asked the restaurant managers to report their ratings of challenge and hindrance stressors and asked the regional managers to provide demographic information. At Time 3, another two months later, we collected information on the objective performance (net profits) of each restaurant. We visited the participants in person to brief them about the study’s purposes and to explain the survey procedures. All participants received a cover letter explaining the study, a questionnaire, and a return envelope. To ensure confidentiality, the participants sealed their completed questionnaires in envelopes and returned them directly to us on site.

The final sample consisted of 402 restaurant managers and 115 regional managers. The effective response rates were approximately 83.75% for restaurant managers and 100% for regional managers. Among restaurant managers, 62% were women and the mean age and tenure were 28.91 years and 16.31 months, respectively. Most of the restaurant managers had a junior college degree or above. Among regional managers, 57% were women. The mean age was 34.29 years. The majority had a junior college degree or above. In average, restaurant managers worked under their regional managers for 11.23 months.

Sample 2

Sample 2 included 369 police officers and staff of a police force in the United Kingdom. Their primary job responsibilities included maintaining public order, responding to emergencies, and making arrests. The research team packaged the paper-and-pencil surveys with pre-prepared, self-addressed, sealable envelopes and sent them to the human resources department, which sent out the surveys through the internal mail system.

We sent out 1,000 questionnaires and obtained a response rate of 36%. We asked the participants to rate authoritarian leadership of their superiors and challenge/hindrance stressors they have experienced in the workplace and to provide demographic information. We also asked each participant to randomly select a coworker and provide the coworker’s contact information. Two weeks later, the coworker was asked to assess the original participant’s extra-role performance (i.e., citizenship behavior toward the organization). To match the responses from the two sources, we assigned each questionnaire an identification number. Of the final matched sample of 360, 61.2% were men, 66% were police officers, and 34% were police staff. The average age was 42.6 years.

Measures

The survey for Sample 1 was in Mandarin Chinese. We used Brislin’s (1980) translation–back–translation procedure to translate English scales into Mandarin Chinese. For Sample 2, we used the original English scales.

Authoritarian Leadership

For both samples, we used the 9-item scale developed by Cheng et al. (2004) to measure supervisors’ authoritarian leadership behaviors. Example items are “My immediate supervisor asks me to obey his/her instructions completely” and “We have to follow his/her rules to get things done. If not, he/she punishes us severely.” The employees were asked to indicate the frequency with which their supervisors exhibited the behavioral style described on a six-point Likert scale, ranging from 1 (never) to 6 (very frequently). Cronbach’s alpha was 0.89 for Sample 1 and 0.86 for Sample 2.

Challenge Stressors and Hindrance Stressors

For both samples, we measured challenge and hindrance stressors using Cavanaugh et al. (2000). The employees were asked to indicate the extent to which they had experienced stress from nine work stressors on a 10-point Likert scale, ranging from 1 (not stressful at all) to 10 (extremely stressful). The challenge stressors scale included six items that captured workload, work time, time pressure, task complexity, and responsibility. Example items include “The volume of work that must be accomplished in the allotted time” and “Time pressures I experience.” Cronbach’s alpha was 0.80 for Sample 1 and 0.90 for Sample 2. The hindrance stressors scale included five items related to role ambiguity, red tape, job insecurity, organizational politics, and career bottlenecks. Example items are “The inability to clearly understand what is expected of me on the job” and “The lack of job security I have.” Cronbach’s alpha was 0.86 for Sample 1 and 0.78 for Sample 2.

In-role Performance

For Sample 1, in-role performance was measured using the restaurants’ objective performance indicator––net profits––one month after the administration of the questionnaires. Achieving more net profits is one of the primary tasks or in-role goals of restaurant managers (Koene et al., 2002; Peterson et al., 2012). Net profits are calculated by subtracting total costs (including production, distribution, labor, and operational costs) from total sales. The net profits of each store were indicated as a percentage of total sales. The net profits of the restaurants ranged from -76% to 41%. We standardized the net profits.

Extra-role Performance

For Sample 2, we measured extra-role performance using eight items developed by Lee and Allen (2002). We focused on how authoritarian leadership is associated with citizenship behavior towards organization because this extra-role performance is likely to be influenced by the work environment, work conditions, and leadership behavior (Chan & Mak, 2012; Chan et al., 2013; Huang et al., 2010; Skarlicki & Latham, 1996, 1997; Williams and Anderson, 1991). An example item is “[This employee] attends functions that are not required but help the organizational image.” The supervisors were asked to rate their employees’ extra-role performance on a five-point Likert scale, ranging from 1 (totally disagree) to 5 (totally agree). Cronbach’s alpha was 0.92 for Sample 2.

Control Variables

Following previous leadership research (Chen et al., 2014; Schaubroeck et al., 2017), we controlled for managers’ age, gender, education, and work tenure (in months) and supervisors’ gender, age, and education in Sample 1, and employees’ age, gender, and job role (0 = police officer; 1 = police staff) in Sample 2.

Study 1 Results

Tables 1 and 2 present the means, standard deviations, and correlations for all variables in both samples. Before testing the hypotheses, we conducted confirmatory factor analyses (CFA) to assess the fit of our data to a measurement model. As we used leader-member exchange (LMX; Graen & Scandura, 1987) in supplementary analyses, we included it in the CFA. We used parcels to maintain a better sample size to parameter ratio and decrease the likelihood of identification problems in the CFA (Bagozzi & Edwards, 1998; Williams & O’Boyle, 2008). Challenge stressors, hindrance stressors, authoritarian leadership, LMX, and extra-role performance were modeled with three parcels. We assigned items to the parcels randomly.

Table 1 Study 1: Descriptives and Correlations of Variables for Sample 1
Full size table
Table 2 Study 1: Descriptives and correlations of variables for Sample 2
Full size table

In Sample 1, Model 1 (the hypothesized model) fit the data well (χ2[60] = 112.09, comparative fit index [CFI] = 0.98, Tucker–Lewis index [TLI] = 0.97, root mean square error of approximation [RMSEA] = 0.05) and it was also superior to two alternative models, including a four-factor model in which two stressors were set to load on a single factor (∆χ2[3] = 311.20, p < 0.01, CFI = 0.82, TLI = 0.78, RMSEA = 0.12) and a three-factor model in which two stressors and LMX were set to load on one factor (∆χ2[5] = 827.05, p < 0.01, CFI = 0.57, TLI = 0.49, RMSEA = 0.18). The results for Sample 2 also showed that the proposed model fit the data well (χ2[67] = 125.12, CFI = 0.98, TLI = 0.97, RMSEA = 0.04) and that it was also superior to three alternative models, including a four-factor model in which two stressors were set to load on a single factor (∆χ2[4] = 204.74, p < 0.01, CFI = 0.89, TLI = 0.86, RMSEA = 0.08), a three-factor model in which two stressors and LMX were set to load on a single factor (∆χ2[7] = 501.09, p < 0.01, CFI = 0.77, TLI = 0.71, RMSEA = 0.12), and a two-factor model in which authoritarian leadership and extra-role performance were set to load on one factor and other variables were set to load on one factor, respectively (∆χ2[9] = 1,078.08, p < 0.01, CFI = 0.53, TLI = 0.43, RMSEA = 0.17). In summary, the CFA results supported the adequacy of the measures used to test the hypotheses.

Study 1 aimed to examine the mediation model of Hypotheses 1 and 2. We tested the hypotheses using Mplus 7.4 (Muthén & Muthén, 2015). All of the parameters were estimated simultaneously for the two outcome variables. In Sample 1, multiple restaurant managers were nested in each regional manager. We therefore used structural equation modeling (SEM) and the “Cluster” and “Type = Complex” Mplus syntax to account for non-independence. With this approach, the standard errors are adjusted using a sandwich estimator to account for non-independence due to employees’ clustering within supervisory groups (Muthén & Muthén, 2015). In Sample 2, we used SEM at the individual level. Because the survey was not completed by the supervisors and most of the participants’ positions were mobile, there was no nested structure, and we did not control for nesting effects.

We assumed that data were missing completely at random, used pairwise deletion (Muthén & Muthén, 2015), and computed standard errors for parameter estimates based on the observed information matrix (Greco & Kraimer, 2020; Kenward & Molenberghs, 1998). Significance tests of indirect effects were conducted using a bootstrap approach, which produced 1,000 bootstrap samples and allowed us to construct bias-corrected confidence intervals for each significance test (Edwards & Lambert, 2007; Mooney et al., 1993). The bootstrap approach is more advantageous than the more commonly used Sobel test as it overcomes the high Type I error rate due to the violation of normal distribution assumptions (Shrout & Bolger, 2002).

Indirect effects via dual stressors

We first tested the hypothesized paths of the indirect effects and assessed how well the data fit the hypothesized model. As illustrated in Figs. 2a and 2b, authoritarian leadership was positively related to both challenge stressors (Sample 1: B = 0.29, SE = 0.11, p = 0.01; Sample 2: B = 0.13, SE = 0.06, p < 0.05) and hindrance stressors (Sample 1: B = 0.29, SE = 0.13, p < 0.05; Sample 2: B = 0.25, SE = 0.06, p < 0.001). Challenge stressors were positively related to objective performance (Sample 1: B = 0.09, SE = 0.04, p < 0.05) and extra-role performance (Sample 2: B = 0.28, SE = 0.10, p < 0.01), whereas hindrance stressors were negatively related to objective performance (B = -0.10, SE = 0.04, p < 0.05) and extra-role performance (Sample 2: B = -0.39, SE = 0.11, p < 0.001). The model fit indices suggest that the mediation models fit the data well for both samples (Sample 1: χ2[79] = 117.49, CFI = 0.96, TLI = 0.94, RMSEA = 0.04; Sample 2: χ2[75] = 109.09, CFI = 0.98, TLI = 0.98, RMSEA = 0.04). Overall, in Sample 1, our mediation model explained 5% of the total variance in objective performance (net profits). In Sample 2, our mediation model explained 12% of the total variance in extra-role performance.

Fig. 2
figure 2

A Mediation model for Sample 1 in Study 1. B Mediation model for Sample 2 in Study 1. Note: Unstandardized coefficient; standard errors appear in brackets; dashed lines represent p > 0.05; solid lines represent p < 0.05

Full size image

We applied the Monte Carlo resampling approach (Lorinkova et al., 2013; Selig & Preacher, 2008) to calculate the indirect effects of authoritarian leadership on performance. We used 20,000 resamples for each confidence interval. The Monte Carlo approach improves power problems introduced by non-normal sampling distributions (MacKinnon et al., 2004; Preacher et al., 2010).

The indirect effects of authoritarian leadership via challenge stressors were statistically significant for both types of performance outcomes, as indicated by 95% confidence intervals (CIs) that did not include zero (Sample 1: objective performance, indirect effect = 0.03, SE = 0.02, 95% CI = [0.001, 0.06]. Sample 2: extra-role performance, indirect effect = 0.04, SE = 0.02, 95% CI = [0.003, 0.08]). Furthermore, in both samples, the indirect effects of authoritarian leadership via hindrance stressors were statistically significant for both performance indicators, as indicated by 95% CIs that did not include zero (Sample 1: objective performance, indirect effect = -0.03, SE = 0.02, 95% CI = [-0.06, -0.001]. Sample 2: extra-role performance, indirect effect = -0.10, SE = 0.03, 95% CI = [-0.19, -0.04]). As such, Hypotheses 1a, 1b, 2a, and 2b were supported in Study 1.

We also compared the indirect effects via challenge stressors with the indirect effects via hindrance stressors for both samples. The results showed that in the Chinese sample, the two paths canceled each other out (B = -0.003, SE = 0.02, n.s.), whereas in the U.K. sample, the path via hindrance stressors (the negative path) was stronger than the path via challenge stressors (the positive path; B = -0.06, SE = 0.03, p < 0.05).

Supplementary analysis: Controlling for the social exchange mechanism

Most studies have attempted to explain the negative effects of authoritarian leadership using a social exchange perspective. They have demonstrated that authoritarian leaders tend to ignore employees’ needs and to undermine the exchange relationship and trust between employees and supervisors, leading to poor employee performance (Chen et al., 2014; Pellegrini & Scandura, 2008; Wu et al., 2012). Thus, it is crucial to examine whether our proposed model can explain the relationship between authoritarian leadership and work performance above and beyond the predominant social exchange mechanism. We therefore performed a set of supplementary analyses for both samples. We tested our mediating hypotheses after controlling for the mediating effect of LMX, a key construct referring to the quality of social exchange between leaders and employees (Graen & Scandura, 1987; Graen & Uhl-Bien, 1995). We used seven items from Scandura and Schrieshiem’s (1994) LMX-7 scale for Sample 1 (collected at Time 2). For Sample 2, we selected two items from the scale (collected at Time 1): “My working relationship with my supervisor is effective” and “My supervisor has enough confidence in me that he/she would defend and justify my decisions if I am not present to do so.”

We integrated the three mechanisms in our model and tested the indirect effects of authoritarian leadership on in-role performance (Sample 1) and extra-role performance (Sample 2) simultaneously through challenge stressors, hindrance stressors, and LMX. The SEM results showed that in both samples, authoritarian leadership was positively related to both challenge stressors (Sample 1: B = 0.29, SE = 0.11, p < 0.01; Sample 2: B = 0.13, SE = 0.06, p < 0.05) and hindrance stressors (Sample 1: B = 0.29, SE = 0.13, p < 0.05; Sample 2: B = 0.25, SE = 0.06, p < 0.001) and was negatively related to LMX (Sample 1: B = -0.27, SE = 0.05, p < 0.01; Sample 2: B = -0.39, SE = 0.07, p < 0.001). Challenge stressors were positively related to objective performance (Sample 1: B = 0.09, SE = 0.04, p < 0.05) and extra-role performance (Sample 2: B = 0.28, SE = 0.10, p < 0.01); hindrance stressors were negatively related to objective performance (Sample 1: B = -0.09, SE = 0.04, p < 0.05) and extra-role performance (Sample 2: B = -0.38, SE = 0.11, p < 0.001); and LMX was not significantly related to objective performance (Sample 1: B = 0.08, SE = 0.10, n.s.) or extra-role performance (Sample 2: B = 0.10, SE = 0.07, n.s.). Overall, in Sample 1, our mediation model explained 5% of the total variance in objective performance. In Sample 2, our mediation model explained 13% of the total variance in extra-role performance.

The indirect effects of authoritarian leadership attributable to challenge stressors were significant (Sample 1, objective performance: indirect effect = 0.03, SE = 0.02, 95% CI = [0.002, 0.06]; Sample 2, extra-role performance: indirect effect = 0.04, SE = 0.02, 95% CI = [0.006, 0.09]), and the indirect effects attributable to hindrance stressors were also significant (Sample 1, objective performance: indirect effect = -0.03, SE = 0.01, 95% CI = [-0.07, -0.002]; Sample 2, extra-role performance: indirect effect = -0.10, SE = 0.03, 95% CI = [-0.17, -0.04]). Thus, Hypotheses 1a, 1b, 2a, and 2b were supported after controlling for LMX as a mediator.

Study 1 Discussion

The results of Study 1 indicated that authoritarian leadership was positively related to in-role performance (Sample 1) and extra-role performance (Sample 2) via challenge stressors. In contrast, authoritarian leadership was negatively related to in-role performance (Sample 1) and extra-role performance (Sample 2) via hindrance stressors. These findings support our argument that as a controlling type of leadership, authoritarian leadership can act as a source of stressors and can induce high performance by generating challenge stressors.

The fact that our results were cross-validated across two samples increases cross-cultural generalizability. The consistent findings of the time-lagged design also bolster our confidence in our interpretation. The results of the supplementary analyses offer additional support for the hypothesis that stressor mechanisms can explain the relationship between authoritarian leadership and performance beyond the exchange mechanism, which has been a major explanatory mechanism used in the literature on authoritarian leadership (e.g., Chen et al., 2014; Wu et al., 2012). However, in Study 1, we did not test the complete model as we did not examine the moderating role of leaders’ power distance orientation. We therefore conducted Study 2 using an independent sample to test the whole model.

Study 2 Methods

Sample and procedure

We collected data from three organizations in the electric power industry located in Beijing, Fujian, and Sichuan provinces, China. The reason why we included these three organizations rather than others was convenience. The study participants were frontline leaders in the organizations and their corresponding employees. The frontline leaders participated in a series of five-day leadership training programs. We told the frontline leaders that we were conducting a scientific study about leadership and that all the information they provided would only be used in our research. We obtained permission from the instructors of the training programs to contact these frontline leaders during the training, sent the participants links to the surveys via WeChat, and asked the frontline leaders to provide contact information for their employees; we then invited one or two of each leader’s employees to participate in the surveys. The participants who completed the survey were rewarded with a small amount of money (Approximately RMB 5 yuan). We also offered to provide feedback to the participants about the study results. We instructed the frontline leaders to send the links to their employees first and to complete the supervisor survey after the employees had completed their surveys. All English scales were translated into Mandarin Chinese following Brislin’s (1980) procedure. We provided a cover letter to explain the purpose of the surveys and to assure the participants that their answers would not be seen by anyone in their company and would be sent directly to the researchers.

The supervisor survey collected data on employees’ in-role and extra-role performance, leader power distance orientation, and supervisors’ demographic information. The employee survey collected data on authoritarian leadership, challenge/hindrance stressors, employee power distance orientation, and employees’ demographic information. After excluding incomplete dyads, the final data set comprised 195 employees supervised by 123 supervisors, giving an effective response rate of 90.28%. The employees were predominantly men (69.3%), their average age was 33.25 years (SD = 8.00), and 93.8% had a junior college degree or above. The supervisors’ average age was 39.51 years (SD = 6.44), 76.9% were men, and 96.4% had a junior college degree or above.

Measures

We assessed power distance orientation using Dorfman and Howell’s (1988) 6-item Power Distance Orientation Scale (see also Farh et al., 2007). Both employee and supervisor participants indicated the extent to which they agreed with such statements as “It is frequently necessary for a manager to use authority and power when dealing with subordinates” and “Employees should not disagree with management decisions” on a five-point Likert scale, ranging from 1 (strongly disagree) to 5 (strongly agree). Cronbach’s alpha was 0.87 for leader power distance orientation and 0.91 for employee power distance orientation. We measured in-role performance using a four-item scale developed by Chen et al. (2002). Example items include “[This employee] always completes job assignments on time” and “The performance of [this employee] always meets the expectations of the supervisor.” The supervisors rated the in-role performance of their employees on a five-point Likert scale, ranging from 1 (totally disagree) to 5 (totally agree). Cronbach’s alpha was 0.86. Authoritarian leadership (Cronbach’s alpha = 0.92), challenge stressors (Cronbach’s alpha = 0.95), hindrance stressors (Cronbach’s alpha = 0.86), and extra-role performance (Cronbach’s alpha = 0.86) were assessed using the same instruments as those in Study 1. As for Sample 1 in Study 1, we controlled for employees’ age, gender, education, and tenure (in years), as well as leaders’ age, gender, and education.

Study 2 Results

Table 3 presents the means, standard deviations, and correlations for all variables. In CFA, we used the item parceling method (Bagozzi & Edwards, 1998). All of the variables were modeled with three parcels. We assigned items to the parcels randomly. The proposed model fit the data well (χ2[168] = 274.54, CFI = 0.96, TLI = 0.95, RMSEA = 0.06) and was also superior to five alternative models, including a six-factor model in which in-role and extra-role performance were set to load on a single factor (∆χ2[6] = 13.77, p < 0.05, CFI = 0.96, TLI = 0.95, RMSEA = 0.06), another six-factor model in which two stressors were set to load on a single factor (∆χ2[6] = 194.55, p < 0.01, CFI = 0.89, TLI = 0.87, RMSEA = 0.09), a five-factor model in which authoritarian leadership and two stressors were set to load on a single factor (∆χ2[11] = 313.98, p < 0.01, CFI = 0.85, TLI = 0.82, RMSEA = 0.11), a four-factor model in which authoritarian leadership, leader power distance orientation, and two stressors were set to load on a single factor (∆χ2[15] = 587.79, p < 0.01, CFI = 0.75, TLI = 0.71, RMSEA = 0.14), and a three-factor model in which authoritarian leadership, leader power distance orientation, employee power distance orientation, and two stressors were set to load on a single factor (∆χ2[18] = 845.73, p < 0.01, CFI = 0.65, TLI = 0.61, RMSEA = 0.16).

Table 3 Study 2: Descriptives and correlations of variables
Full size table

In Study 2, employees were nested within supervisors. We used the “Cluster” and “Type = Complex” Mplus syntax to account for non-independence. We performed SEM (Muthén & Muthén, 2015) at the individual level and controlled for the effects of age, gender, and education on all endogenous variables (Edwards & Lambert, 2007). We applied the Monte Carlo resampling approach (Lorinkova et al., 2013; Selig & Preacher, 2008) to calculate the indirect effects of authoritarian leadership.

Indirect effects via dual stressors

First, we tested the mediation model. As shown in Fig. 3a, authoritarian leadership was positively related to both challenge stressors (B = 0.83, SE = 0.25, p < 0.01) and hindrance stressors (B = 0.96, SE = 0.22, p < 0.01). Challenge stressors were positively related to both in-role (B = 0.10, SE = 0.04, p < 0.05) and extra-role performance (B = 0.08, SE = 0.04, p < 0.05). Hindrance stressors were negatively related to both in-role performance (B = -0.09, SE = 0.04, p < 0.05) and extra-role performance (B = -0.07, SE = 0.03; p < 0.05). The mediation model fit the data well (χ2[160] = 232.01, CFI = 0.96, TLI = 0.95, RMSEA = 0.04). Overall, our mediation model explained 13% and 10% of the total variance in in-role performance and extra-role performance, respectively.

Fig. 3
figure 3

A Results of mediation model in Study 2. B Results of moderated-mediation model in Study 2. Note: Unstandardized coefficient; standard errors appear in brackets; dashed lines represent insignificant relationships; * p < 0.05, **p < 0.01

Full size image

The indirect effects of authoritarian leadership via challenge stressors were significant for both in-role performance (indirect effect = 0.09, SE = 0.05, 95% CI = [0.01, 0.19]) and extra-role performance (indirect effect = 0.06, SE = 0.03, 95% CI = [0.02, 0.12]). Furthermore, the indirect effects via hindrance stressors were significant for both in-role performance (indirect effect = -0.08, SE = 0.04, 95% CI = [-0.17, -0.01]) and extra-role performance (indirect effect = -0.07, SE = 0.03, 95% CI = [-0.13, -0.01]). Thus, Hypotheses 1a, 1b, 2a, and 2b were supported. In addition, these two paths could cancel each other out (in-role performance: B = 0.004, SE = 0.03, n.s.; extra-role performance: B = -0.002, SE = 0.03, n.s.), indicating that neither challenge stressors nor hindrance stressors were a more dominant path. Further, the direct effects of authoritarian leadership on both in-role performance (B = 0.004, SE = 0.03, n.s.) and extra-role performance (B = -0.03, SE = 0.05, n.s.) were not significant, and the total effects (i.e., indirect effects + direct effect) of authoritarian leadership on both in-role performance (B = 0.004, SE = 0.06, n.s.) and extra-role performance were also not significant (B = -0.002, SE = 0.03, n.s.).

Conditional indirect relationships

Next, we tested the first-stage moderated mediation (Edwards & Lambert, 2007) with latent moderated structural equations, as described by Cheung and Lau (2017), to test Hypotheses 3 and 4 (Bamberger & Belogolovsky, 2017; Wayne et al., 2017). We evaluated the model in Fig. 3b with one latent interaction between authoritarian leadership and leaders’ power distance orientation, as well as the estimations of the two paths from the latent interactions to challenge stressors and hindrance stressors, respectively. The results showed that the interaction term of authoritarian leadership and leader power distance orientation was negatively but not significantly related to challenge stressors (B = -0.04, SE = 0.24, n.s.). Thus, Hypotheses 3a and 3b were not supported. In contrast, the interaction term was positively related to hindrance stressors (B = 0.68, SE = 0.26, p < 0.01). Overall, our moderated-mediation model explained 32% and 24% of the total variance in in-role and extra-role performance, respectively.

We then examined the conditional indirect relationships between authoritarian leadership and in-role and extra-role performance through hindrance stressors at high and low values of leader power distance orientation (Cheung & Lau, 2017), using the Monte Carlo resampling method (Bauer et al., 2006). When leader power distance orientation was high, the indirect relationships via hindrance stressors were negatively significant (in-role performance, indirect effect = -0.20, SE = 0.08, 95% CI = [-0.37, -0.07]; extra-role performance, indirect effect = -0.15, SE = 0.06, 95% CI = [-0.28, -0.04]). When leader power distance orientation was low, the indirect relationships via hindrance stressors were not significant (in-role performance, indirect effect = -0.04, SE = 0.05, 95% CI = [-0.15, 0.05]; extra-role performance, indirect effect = -0.01, SE = 0.04, 95% CI = [-0.11, 0.04]; see Figs. 4a and 4b). Thus, Hypotheses 4a and 4b were supported.

Fig. 4
figure 4

A Indirect effect of authoritarian leadership on in-role performance via hindrance stressors at different level of leader power distance orientation (-1 SD to + 1 SD) in Study 2. B Indirect effect of authoritarian leadership on extra-role performance via hindrance stressors at different level of leader power distance orientation (-1 SD to + 1 SD) in Study 2

Full size image

Supplementary analysis: Controlling for the moderating effect of employee power distance orientation

Most studies have treated employee power distance orientation as the boundary of the impacts of authoritarian leadership on employees (e.g., Li & Sun, 2015). Scholars have posited that because employees with a high rather than low power distance orientation are more likely to comply with their leaders’ orders, they are more accepting of their leaders’ demanding leadership style and therefore authoritarian leadership is less likely to harm their work performance (e.g., Gu et al., 2018; Schaubroeck et al., 2017; Shen et al., 2019). To demonstrate the unique variance attributable to the moderating effect of leader power distance orientation, it is crucial to control for the moderating effect of employee power distance orientation. Therefore, in a supplementary analysis, we tested our hypothesized conditional indirect relationships via hindrance stressors after controlling for employee power distance orientation and its interaction with authoritarian leadership.

The results showed that the interaction term of authoritarian leadership × leader power distance orientation was positively related to hindrance stressors (B = 0.72, SE = 0.32, p < 0.05), whereas the interaction term of authoritarian leadership × employee power distance orientation was negative but not significant (B = -0.38, SE = 0.24, n.s.; see Fig. 5a). Overall, the moderated-mediation model explained 42% and 34% of the total variance in in-role performance and extra-role performance, respectively. When leader power distance orientation was high, the indirect effects of authoritarian leadership via hindrance stressors were negatively significant (in-role performance, indirect effect = -0.15, SE = 0.07, 95% CI = [-0.32, -0.03]; extra-role performance, indirect effect = -0.12, SE = 0.06, 95% CI = [-0.25, -0.03]). When leader power distance orientation was low, the indirect effects of authoritarian leadership via hindrance stressors were not significant (in-role performance, indirect effect = -0.01, SE = 0.04, 95% CI = [-0.10, 0.08]; extra-role performance, indirect effect = -0.01, SE = 0.03, 95% CI = [-0.08, 0.06]; see Figs. 5b and 5c). Thus, Hypotheses 4a and 4b were supported.Footnote 1Footnote 2Footnote 3

Fig. 5
figure 5figure 5

A Results of moderated-mediation model in supplementary analysis in Study 2. Note: Unstandardized coefficient; standard errors appear in brackets; dashed lines represent insignificant relationships; * p < 0.05, **p < 0.01. B Indirect effect of authoritarian leadership on in-role performance via hindrance stressors at different level of leader power distance orientation (-1 SD to + 1 SD) in supplementary analysis in Study 2. C Indirect effect of authoritarian leadership on extra-role performance via hindrance stressors at different level of leader power distance orientation (-1 SD to + 1 SD) in supplementary analysis in Study 2

Full size image

General discussion

Across two studies, we found consistent support for our hypothesis that authoritarian leadership has both positive and negative effects on employee performance. Authoritarian leadership is positively related to in-role and extra-role performance via challenge stressors and negatively related to these two performance criteria via hindrance stressors. Furthermore, the mediated effects are moderated by leaders’ power distance orientation. Specifically, authoritarian leaders high on power distance orientation are more likely to harm employees’ performance via hindrance stressors. Below, we discuss the theoretical and practical contributions and suggest future research directions.

Theoretical contributions

Our research makes several important contributions to authoritarian leadership research and the literature on leadership in general. First, we extend the current understanding of authoritarian leadership by revealing that it can have both positive and negative influences on employee performance. Our approach of dual effects of authoritarian leadership helps reconcile the previous inconsistencies in understanding its effects. Most research theorizes authoritarian leadership is detrimental to performance but does not provide consistent evidence (Bass and Bass, 2009; Chan et al., 2013; Harms et al., 2018; Schaubroeck et al., 2017; Zhang & Xie, 2017; Chen et al., 2014; Cheng et al., 2004; Gu et al., 2020). One important reason to explain these inconsistent findings may be authoritarian leadership is not entirely negative and that its positive effects may be offset by its concomitant negative effects. We echo most previous research to find the negative effects of authoritarian leadership on employees (Chan et al., 2013; Schaubroeck et al., 2017; Wu et al., 2012). More importantly, we extend previous literature by building a second pathway that transmits its positive effects on employee performance. We empirically confirm both positive and negative effects of authoritarian leadership on performance through different pathways, which help reconcile the previous inconsistent findings. Further, supporting previously unconfirmed predictions that authoritarian leadership may promote performance by driving employees to comply with leaders’ performance instructions (Chen et al., 2014; Farh & Cheng, 2000; Huang et al., 2015), we deepen the theoretical basis of authoritarian leadership’ positive mechanism that it stresses accomplishing heavy workloads, meeting tight deadlines, and a large scope of responsibilities, thereby promoting employee performance.

Second, drawing from social information processing theory and stressor literature (LePine et al., 2005, 2016; Salancik & Pfeffer, 1978; Shamir et al., 1993), we extend the literature on the mechanisms underlying the impacts of authoritarian leadership on employee performance. One dominant underlying mechanism is the social exchange lens; the literature emphasizes its negative effects on work performance. It has been posited that authoritarian leadership stresses employees’ compliance without considering their socio-emotional needs and is detrimental to dyadic relationship quality, resulting in lower employee performance. (Chen et al., 2014; Shen et al., 2019; Wu et al., 2012). Using culturally diverse samples, we showed that the challenge stressors and hindrance stressors could act as important mediating mechanisms to explain the influences of authoritarian leadership on employee performance after controlling for the social exchange mechanism. We also show that the mediating effects of dual types of stressors are only present in authoritarian leadership and not in other leadership styles such as transformational and benevolent leadership. Hence, we contribute to the literature by introducing two new mechanisms—challenge stressors and hindrance stressors—underlying the effects of authoritarian leadership on employee performance.

Third, we contribute to authoritarian leadership literature by showing how leaders’ espoused values, in particular, power distance orientation, affect how they exercise authoritarianism. Specifically, authoritarian leadership was not related to in-role or extra-role performance via hindrance stressors when leader power distance orientation was low. Research has treated employee personal values, especially employee power distance orientation, authoritarianism, and traditionality (Cheng et al., 2004; Li & Sun, 2015; Schaubroeck et al., 2017; Shen et al., 2019), as the boundaries that influence the effects of authoritarian leadership on employees. Yet little is known about how leaders’ personal values may dictate the work domain that authoritarian behaviors are directed at and thus impact the effects of authoritarian leadership on work performance. Our research focuses on leader power distance orientation and offers insights into when the negative effects of authoritarian leadership on employee performance via hindrance stressors can be mitigated.

Notably, we found a small effect size of the correlation between authoritarian leadership and leader power distance orientation (r = 0.29, p < 0.01; see Table 3; Cohen, 1988), which suggests that although authoritarian leadership is related to a leader power distance orientation, authoritarian leaders may vary in their power distance orientation. Further, Study 2 provided evidence that leader power distance orientation may serve as a critical boundary condition regulating the indirect effects of authoritarian leadership via two types of stressors after controlling for the moderating role of employees’ power distance orientation. Also, we replaced leader power distance orientation with employee power distance orientation as the moderator in our model. The results showed that the moderating effects of employee power distance orientation are not significant on either the relationship between authoritarian leadership and challenge stressors or the relationship between authoritarian leadership and hindrance stressors. Therefore, these supplementary analyses provided more evidence that leader power distance orientation could be a better moderator than employee power distance orientation in the relationships between authoritarian leadership and challenge and hindrance stressors.

More generally, we highlight that leaders may send cues and messages that shape employees’ perceptions of dual stressors. Our discussion is theoretically grounded in the notion that social information transmitted by leaders has a strong influence on employees’ interpretations of their work experiences (Goffman, 1974; Piccolo & Colquitt, 2006; Schutz, 1967; Smircich & Morgan, 1982). It is well known, for example, that leaders influence how followers interpret job characteristics (Piccolo & Colquitt, 2006; Piccolo et al., 2010). Unfortunately, scholars have yet to demonstrate the role that leadership plays in influencing other types of work experiences and conditions perceived by employees, and how these perceptions influence work performance. To this end, we draw from social information processing theory (Goffman, 1974) and the literature on stressors (LePine et al., 2005) to show how one particular style of leadership, authoritarian leadership, as a source of social information, can shape the employee experience of work stressors. By doing this, we offer new insights that, beyond job characteristics, employees’ other important work experiences, especially work stressors, may also be generated by leaders and transmit leadership effects on employee performance.

Strengths

Our research has two major strengths. First, we replicated our findings using multiple samples from different cultures––China and the United Kingdom. Across three samples in two countries, we consistently found that authoritarian leadership can shape both challenge stressors and hindrance stressors, which have contrasting effects on work performance. Research on authoritarian leadership has predominantly used Chinese samples (Cheng et al., 2004; Li & Sun, 2015; Schaubroeck et al., 2017; Shen et al., 2019). The replication of our results across three samples from different cultures increases our confidence in both the robustness of our theory and the generalizability of our findings. We also compared the indirect effects via challenge stressors with the indirect effects via hindrance stressors for all three samples from the two cultures. In both Chinese samples, the indirect effects via challenge stressors (the positive pathway) and the indirect effects via hindrance stressors (the negative pathway) cancelled each other out. By contrast, in the U.K. sample, the negative pathway via hindrance stressors was stronger than the positive pathway via challenge stressors. These results suggest that although authoritarian leadership could yield positive effects on work performance via inducing challenge stressors across cultures, it may be more accepted in Chinese culture than in Western cultures because Chinese culture is characterized by high power distance and collectivism (Hofstede, 1980, 1991), which is consistent with the previous literature on authoritarian leadership (Farh & Cheng, 2000; Li & Sun, 2015).

Second, our data still supported our hypotheses after controlling for the alternative mechanisms reported in previous research on authoritarian leadership. Our stressor mechanisms explaining the effects of authoritarian leadership on work performance still held after controlling for the social exchange mechanism in Study 1. In Study 2, leaders’ power distance orientation moderated the effects of authoritarian leadership on performance after controlling for employees’ power distance orientation. In supplementary analyses, our mediating mechanisms still held after controlling for other styles of leadership.

Limitations and future research

Despite its strengths, this study has some limitations. First, across all three samples, both authoritarian leadership and stressors were self-reported. We chose self-ratings because this method is aligned with our theory. We applied social information theory suggesting that authoritarian leadership impacts employees’ sense-making of work experiences (i.e., challenge and hindrance stressors). Employees themselves are the targets of authoritarian leadership behaviors, and how they make sense of work experiences and demands are private constructs that others, such as leaders and coworkers, may not rate accurately (Conway & Lance, 2010; Fox et al., 2007; Kossek et al., 2012; Leslie et al., 2012). Therefore, consistent with prior empirical research that also examined the impacts of leadership on employees’ work experiences (Fernet et al., 2015; Piccolo & Colquitt, 2006; Piccolo et al., 2010), we used self-ratings of leadership and stressors. However, applying self-reports may imply that the relationship between authoritarian leadership and work stressors contains common method variance (CMV; Podsakoff et al., 2003). We took some precautions during data collection to decrease CMV, such as promising participants anonymity to reduce the influence of a wish for social desirability, using a temporally lagged design between authoritarian leadership and work stressors, and employing multiple samples (Conway & Lance, 2010).

To detect the CMV’s influence, we followed the procedure of Harman’s single factor specified by Podsakoff et al (2003), which uses exploratory factor analysis where all items of authoritarian leadership and work stressors are loaded onto a single factor to see whether most of the variance can be accounted for by one general factor. The results show that the single factor in all our samples explains less than 50% variance (Study 1a: 22.44%; Study 1b: 21.06%; Study 2: 42.33%), suggesting that CMV has a limited influence. To further boost our confidence, we used unmeasured latent method factor (Podsakoff et al, 2003) and conducted CFAs to compare the model fit of the three-factor model (authoritarian leadership, challenge and hindrance stressors) with an alternative model including an additional latent factor to account for self-report method with all the items measured by employees as its indicators. We set all of factor loadings of common method factor to equality for achieving convergent solution (Podsakoff, et al., 2003; Porter et al., 2016). The results show that the common method factor did not improve the model fit significantly (Study 1a: ∆χ2[1] = 0.00, n.s.; Study 1b: ∆χ2[2] = 2.18, n.s.; Study 2: ∆χ2[1] = 0.00, n.s.), again suggesting that the effect of CMV is limited. Nevertheless, we cannot totally exclude this risk. We thus suggest that future research apply more complex designs, such as longitudinal design or experimental design that manipulates authoritarian leadership and work stressors to minimize this concern.

Second, the correlational design precludes causal conclusions and leaves open the possibility of alternative explanations, such as reverse causality. However, our predictions are still acceptable for the following reasons. First, they are based on a strong theoretical foundation and our theoretical model cannot be refuted only because of the possible existence of alternative explanations. Second, Study 1 used a time-lagged design, and the findings of mediating effects across the three samples from two studies give us confidence in our results. Nevertheless, to draw causal conclusions and rule out feedback loops, experimental or longitudinal data are needed to address this concern.

Third, we did not find a moderating effect of leader power distance orientation on the relationships between authoritarian leadership and challenge stressors. Our participants were only frontline leaders in the organizations in the same industry who participated in a series of 5-day leadership training programs and their corresponding subordinates. In this context, it may be less likely to find a moderating effect because there may be sample-based range restriction effects regarding the leader power distance orientation because organizations may select trainees who hold similar individual orientations to attend this leadership training program. We highly recommend replicating our models with other samples with a broader range of industries and power distance orientation, which would help address this potential concern and strengthen the generalization of our findings.

While authoritarian leaders low on power distance orientation tend to show more respect to their subordinates, they may not necessarily press them to meet high task-related demands. We suggest a more proximal construct, conscientiousness, which is more likely to moderate the effects of authoritarian leadership on challenge stressors. Conscientiousness refers to the extent to which individuals are dutiful, persevering, and disciplined and tend to fulfill performance requirements as their top priority (Barrick & Mount, 1991; Barrick et al., 1993; Costa et al., 1991; Gellatly, 1996). Leaders high on conscientiousness may use authoritarian behaviors to persuade employees to fulfill their duties and achieve high performance. Therefore, they may direct authoritarian behaviors at setting high expectations of performance standards, tight time pressures, and heavy workloads and generate more challenge stressors. By contrast, as authoritarian leaders low on conscientiousness do not regard the fulfillment of performance requirements as their top priority, they are less likely to produce challenge stressors to secure high performance.

Further, we have theorized that authoritarian leaders with a low power distance orientation may generate fewer hindrance stressors and more challenge stressors because they are less likely to emphasize their own authority and superiority or to show no interest in followers’ needs and growth potential. Our results show that authoritarian leadership with a low power distance orientation generates fewer hindrance stressors. Future research could develop intervention studies that decrease the level of authoritarian leaders’ power distance.

Contributions to managerial practices

We suggest the specific ways through which managers may effectively use authoritarian behaviors to achieve high performance. Managers should direct authoritarian behaviors to generating work challenges for employees to accomplish, such as setting high performance standards for employees to achieve, enforcing tight deadlines, and assigning demanding workloads. Authoritarian leadership is likely to fail to drive employees to contribute to the organization, when managers use their authority to highlight their own superiority and to fulfill their self-interest, which results in generating obstacles for employees’ work and career development. These insights can be incorporated in leadership training programs to help managers understand the “two faces” of authoritarianism and develop action plans to make use of this controlling style of leadership to enhance employee performance.

We also show that authoritarian leaders low on power distance orientation generate fewer hindrance stressors. Some research has suggested that power distance orientation could be malleable to some extent (e.g., McGrath et al., 1992). Organizations can offer leadership training programs that help managers develop employee-oriented mindsets and interpersonal skills that reduce managers’ social and psychological distance (Lacerenza et al., 2017; Martin et al., 2013).

Conclusion

We found that authoritarian leadership generated both challenge and hindrance stressors, which were respectively beneficial and detrimental for performance, and that authoritarian leaders low on power distance orientation did not negatively affect performance via hindrance stressors. We have taken an initial step toward drawing an integrated picture of the effects of authoritarian leadership and encourage future research to extend understanding of both the positive and negative influences of authoritarian leadership.