Supervisors use various leadership styles to exercise their formal authority and position power (Hoel & Salin, Reference Hoel, Salin, Einarsen, Hoel, Zapf and Cooper2003). Authoritarian leadership, as one such style, allows leaders to make unilateral decisions (Uhl-Bien & Maslyn, Reference Uhl-Bien and Maslyn2005) and prevail over subordinates (Pelligrini, Scandura, & Jayaraman, Reference Pelligrini, Scandura and Jayaraman2010). In response, subordinates will follow directions (Cheng, Chou, Wu, Huang, & Farh, Reference Cheng, Chou, Wu, Huang and Farh2004) and adhere to job requirements or demands, but generally ignore discretionary extra-role behaviors (Bergeron, Reference Bergeron2007). Thus some researchers believe that authoritarian leadership reduces employee organizational citizenship behavior (e.g., Chan, Huang, Snape, & Lam, Reference Chan, Huang, Snape and Lam2013; Chen, Eberly, Chiang, Farh, & Cheng, Reference Chen, Eberly, Chiang, Farh and Cheng2014). Nevertheless, others suggest that authoritarian leadership has no direct effect (Aryee, Chen, Sun, & Debrah, Reference Aryee, Chen, Sun and Debrah2007; Wu, Huang, Li, & Liu, Reference Wu, Huang, Li and Liu2012). Consequently, we ask, whether and how do authoritarian behaviors affect employee OCB?
Scholars have used social exchange and justice perspectives to explain how authoritarian leadership affects subordinate OCB. They have focused on the quality of leader – member social interactions, such as the quality of leader – member exchange (Liang, Ling, & Hsieh, Reference Liang, Ling and Hsieh2007), trust in leaders (Chen et al., Reference Chen, Eberly, Chiang, Farh and Cheng2014; Wu et al., Reference Wu, Huang, Li and Liu2012), and interactional justice (Wu et al., Reference Wu, Huang, Li and Liu2012), failed to consider the hierarchical supervisor – subordinate structure that defines work roles for both parties regarding obligations, responsibilities, rights, performance expectations, and behavior patterns (Biddle, Reference Biddle1979). Assuming superior roles in the structure, supervisors may strongly shape subordinates’ role perceptions, responsibilities, and work behaviors (Peterson et al., Reference Peterson, Smith, Akande, Ayestaran, Bochner, Callan, Cho, Jesuino, D'Amorim, Francois, Hofmann, Koopman, Leung, Lim, Mortazavi, Munene, Radford, Ropo, Savage, Setiadi, Sinha, Sorenson and Viedge1995).
Based on fundamental role structures in organizational hierarchies, we use role theory to explain how authoritarian behavior affects subordinate OCB. The theory posits that role structures generate role perceptions (Peterson et al., Reference Peterson, Smith, Akande, Ayestaran, Bochner, Callan, Cho, Jesuino, D'Amorim, Francois, Hofmann, Koopman, Leung, Lim, Mortazavi, Munene, Radford, Ropo, Savage, Setiadi, Sinha, Sorenson and Viedge1995). Supervisors, as role senders in supervisor – subordinate structures, define roles and create role expectations. Subordinates, as role receivers, are expected to follow the expectations. Furthermore, supervisor – subordinate structure allows supervisors to exercise power in an authoritarian manner. Subordinates may view authoritarian supervision as legitimate in-role supervisory behavior to frame or define organizational and job realities (Karambayya, Brett, & Lytle, Reference Karambayya, Brett and Lytle1992; Smircich & Morgan, Reference Smircich and Morgan1982). Through work interactions, authoritarian supervisors are likely to shape subordinate perceptions regarding job realities, role expectations, and relevant behaviors. We suggest that authoritarian supervisors may affect subordinates’ extra-role behaviors depending on how subordinates perceive their work roles: whether they perceive roles conflicting, ambiguous, or overloading (Katz & Kahn, Reference Katz and Kahn1978; Rizzo, House, & Lirtzman, Reference Rizzo, House and Lirtzman1970).
Authoritarian leadership depends on general organizational understandings regarding the nature of power, respect for authority, and behavioral expectations (Pye & Pye, Reference Pye and Pye2009), and role perceptions are rooted in organizational role structures (Peterson et al., Reference Peterson, Smith, Akande, Ayestaran, Bochner, Callan, Cho, Jesuino, D'Amorim, Francois, Hofmann, Koopman, Leung, Lim, Mortazavi, Munene, Radford, Ropo, Savage, Setiadi, Sinha, Sorenson and Viedge1995). Consequently, we believe that authoritarian leadership universally affects subordinate role perceptions and extra-role behavior in organizations across cultures. However, cultural viewpoints may differ regarding power distance and may determine whether authoritarian behavior is encouraged or inhibited (Robert, Probst, Martocchio, Drasgow, & Lawler, Reference Robert, Probst, Martocchio, Drasgow and Lawler2000). Chinese culture strongly emphasizes power and hierarchy (Hofstede, Reference Hofstede1980), so both Chinese supervisors and subordinates are more likely to legitimize authoritarian leadership (Saufi, Wafa, & Hamzah, Reference Saufi, Wafa and Hamzah2002). Consequently, China is a suitable research context for studying perceptions of roles and behaviors in relation to authoritarian leadership.
Our primary objective is to assess whether authoritarian leadership impacts subordinates’ role perceptions and subsequent extra-role behaviors in China. We hope to show that leaders are role-makers affecting subordinates’ role perceptions that mediates whether authoritarian leadership indirectly inhibits pro-organizational behaviors.
We contribute to the literature in three important aspects. First, role perspectives offer new insights into how leadership behaviors affect subordinate work behaviors. Previous leadership studies focused on how leaders influence subordinate self-esteem (Chan et al., Reference Chan, Huang, Snape and Lam2013), self-efficacy (Gong, Huang, Farh, Reference Gong, Huang and Farh2009), job evaluations, including core job characteristics (Piccolo & Colquitt, Reference Piccolo and Colquitt2006), or attitudes towards supervisors such as trust (Chen et al., Reference Chen, Eberly, Chiang, Farh and Cheng2014), largely ignored how leaders influence subordinates’ role perceptions. Consequently, we use a role perspective to explain how leadership affects subordinates.
Second, we integrate leadership and work role theories, and include leadership behaviors as determining work role perceptions. The literature has mainly considered role conflict, ambiguity, and overload as role stressors, and has investigated consequences such as emotional exhaustion, organizational commitment, job satisfaction, and job performance (see meta-analysis by Örtqvist & Wincent, Reference Örtqvist and Wincent2006), and lack of investigaing the determinants. Although scholars theorize that employees perceive their roles according to work events and supervisors’ expectations in the role structure (Peterson et al., Reference Peterson, Smith, Akande, Ayestaran, Bochner, Callan, Cho, Jesuino, D'Amorim, Francois, Hofmann, Koopman, Leung, Lim, Mortazavi, Munene, Radford, Ropo, Savage, Setiadi, Sinha, Sorenson and Viedge1995), we propose that supervisory behaviors deliver social cues that may also affect their perceptions.
Moreover, rather than focusing on effective leadership behaviors, such as transformational leadership, we focus on traditional authoritarian leadership to explain whether and how such ‘in-role’ behavior may decrease extra-role behavior. If the relationship can be negative, the implications will be profound for managerial practice.
THEORETICAL BACKGROUND AND HYPOTHESES
Authoritarian Leadership and Organizational Citizenship Behavior
Employee behaviors that contribute to ‘the maintenance and enhancement of the social and psychological contexts that support task performance’ are called organizational citizenship behaviors (OCBs) (Organ, Reference Organ1997: 91). OCBs exceed job descriptions (Tepper & Taylor, Reference Tepper and Taylor2003) and are less likely to be rewarded or punished (Organ, Podsakoff, & MacKenzie, Reference Organ, Podsakoff and MacKenzie2006), although they benefit employees (Allen & Rush, Reference Allen and Rush1998), work groups (Podsakoff, Ahearne, & MacKenzie, Reference Podsakoff, Ahearne and MacKenzie1997), and organizations (Bolino, Turnley, & Bloodgood, Reference Bolino, Turnley and Bloodgood2002). Supervisors largely affect subordinate discretionary behavior (Podsakoff, MacKenzie, Bachrach, Reference Podsakoff, MacKenzie and Bachrach2000). For example, transformational leadership (Podsakoff, MacKenzie, Moorman, & Fetter, Reference Podsakoff, MacKenzie, Moorman and Fetter1990) and supervisory monitoring (Niehoff & Moorman, Reference Niehoff and Moorman1993) are known to enhance OCB.
In contrast, authoritarian leadership emphasizes ‘absolute authority and control over subordinates and demands unquestionable obedience from subordinates’ (Cheng et al., Reference Cheng, Chou, Wu, Huang and Farh2004: 91). Authoritarian leaders strictly control the hierarchical order requiring subordinates to be submissive, dependent, and obedient (Pelligrini, Scandura, & Jayaraman, Reference Pelligrini, Scandura and Jayaraman2010). They enforce rules, determine rewards and punishments (Aryee et al. Reference Aryee, Chen, Sun and Debrah2007), and stress ‘personal dominance’ (Tsui, Wang, Xin, Zhang, & Fu, Reference Tsui, Wang, Xin, Zhang and Fu2004). Authoritarian leaders also make all important decisions (Tsui et al., Reference Tsui, Wang, Xin, Zhang and Fu2004; Uhl-Bien & Maslyn, Reference Uhl-Bien and Maslyn2005): assign tasks, give orders (Bond & Hwang, Reference Bond, Hwang and Bond1993), and set high requirements (Cheng et al., Reference Cheng, Chou, Wu, Huang and Farh2004; Wang, Chiang, Tsai, Lin, & Cheng, Reference Wang, Chiang, Tsai, Lin and Cheng2013). Under hierarchical structures, subordinates perceive that they must obey the legitimate dominance inherent in authoritarian leadership (Aryee et al., Reference Aryee, Chen, Sun and Debrah2007), or else they must be punished. Consequently, they feel uneasy, oppressed, and often erupt in negative subordinate – supervisor social exchanges (Wu et al., Reference Wu, Huang, Li and Liu2012). Lacking socio-emotional benefits, they confine their behaviors to explicit in-role requirements to be ‘good’ workers but are unmotivated to work beyond their duties (Chen et al., Reference Chen, Eberly, Chiang, Farh and Cheng2014).
Literature has revealed that authoritarian leadership negatively affects subordinates. Early social psychology studies showed that authoritarian leadership increases spontaneous aggression and hostile behaviors (Lewin, Lippitt, & White, Reference Lewin, Lippitt and White1939). Recent management studies suggested it reduces subordinate job satisfaction (Smither, Reference Smither1993), organization-based self-efficacy (Chan et al., Reference Chan, Huang, Snape and Lam2013), trust in leaders (Chen et al., Reference Chen, Eberly, Chiang, Farh and Cheng2014), interactional justice (Wu et al., Reference Wu, Huang, Li and Liu2012), voice (Chan, Reference Chan2013; Li & Sun, Reference Li and Sun2015), task performance, and conscientious behavior (Wang et al., Reference Wang, Chiang, Tsai, Lin and Cheng2013). Later contingency theories stated that specific contextual factors such as uncertainties (Rast III, Hogg, & Giessner, Reference Rast III, Hogg and Giessner2013) may increase authoritarian leadership effectiveness (e.g., Vroom & Yetton, Reference Vroom and Yetton1973; Yukl, Reference Yukl, Bryman, Collinson, Grint, Jackson and Uhl-Bien2011), dependence and compliance (Chou, Sibley, Liu, Lin, & Cheng, Reference Chou, Sibley, Liu, Lin and Cheng2015), and productivity (Smither, Reference Smither1993). Although authoritarian leadership may change employees’ self-evaluations, justice judgment, and attitudes towards the leader and the job, we know little about the effects on their role perceptions.
Role Theory and Role Perceptions
Roles specify patterns of behaviors in various social contexts. In organizations, employees undertake work roles and use assigned obligations, rights, and performance expectations to guide their behavior and expectations regarding others’ behaviors (Biddle, Reference Biddle1986). Role theory, concerning the study of the patterns of role behaviors (Biddle, Reference Biddle1979), is highly useful for studying individuals and the collective within a single conceptual framework and can thus help us better understand how organizational factors affect employees’ job roles and behaviors.
As role theory suggests, job descriptions formally define work roles, but formal organizational policies and procedures, individual backgrounds and experiences, and treatment by coworkers and supervisors affect role perceptions (Peterson et al., Reference Peterson, Smith, Akande, Ayestaran, Bochner, Callan, Cho, Jesuino, D'Amorim, Francois, Hofmann, Koopman, Leung, Lim, Mortazavi, Munene, Radford, Ropo, Savage, Setiadi, Sinha, Sorenson and Viedge1995). Supervisors have legitimate power to redefine jobs, revise job requirements, change job characteristics, and provide job-related social cues when they evaluate work progress and output (Griffin, Reference Griffin1983). Therefore, immediate supervisors are role-makers and role-senders, shaping subordinate role perceptions (Peterson et al., Reference Peterson, Smith, Akande, Ayestaran, Bochner, Callan, Cho, Jesuino, D'Amorim, Francois, Hofmann, Koopman, Leung, Lim, Mortazavi, Munene, Radford, Ropo, Savage, Setiadi, Sinha, Sorenson and Viedge1995).
Role Perceptions as Mediating Mechanisms
Supervisors send various work expectations and social cues to subordinates (Dale & Fox, Reference Dale and Fox2008). As role theory states, when expectations are unclear, conflicted, or overwhelming, employees then perceive their work roles to be ambiguous, conflicting, or overloaded (Katz & Kahn, Reference Katz and Kahn1978; Rizzo et al., Reference Rizzo, House and Lirtzman1970). The negative perceptions then affect their work attitudes and behaviors (Jex, Reference Jex1998; Thompson & Werner, Reference Thompson and Werner1997) such as job satisfaction (Johnston, Parasuraman, & Futrell, Reference Johnston, Parasuraman and Futrell1989), emotional exhaustion (Örtqvist & Wincent, Reference Örtqvist and Wincent2006), and job performance (Tubre & Collins, Reference Tubre and Collins2000). We expect that authoritarian leadership will cause employees to perceive conflict, ambiguity, and overload in their roles, with consequent negative effects on employee OCB.
Role Conflict as a Mediating Mechanism
Role conflict is ‘incompatibility between the expectations of parties or between aspects of a single role’ (Peterson et al., Reference Peterson, Smith, Akande, Ayestaran, Bochner, Callan, Cho, Jesuino, D'Amorim, Francois, Hofmann, Koopman, Leung, Lim, Mortazavi, Munene, Radford, Ropo, Savage, Setiadi, Sinha, Sorenson and Viedge1995: 430). Employees experience role conflict when they undertake several roles requiring different behaviors, when they are confronted with incompatible expectations from multiple persons or standards, or when they perceive that role expectations and requirements conflict with their internal standards, values, time, resources, or capabilities (Katz & Kahn, Reference Katz and Kahn1978; Rizzo et al., Reference Rizzo, House and Lirtzman1970).
Authoritarian supervisors may provoke acute perceptions of role conflict by making unilateral decisions and centralizing authority on themselves (Tsui, Wang, Xin, Zhang, & Fu, Reference Tsui, Wang, Xin, Zhang and Fu2004). To make decisions independently and highlight their authority, they control resources and rules (Chan, Reference Chan2013) while tightly withholding or even concealing key information (Cheng & Wang, Reference Cheng and Wang2015; Farh & Cheng, Reference Farh, Cheng, Li, Tsui and Weldon2000; Wang et al., Reference Wang, Chiang, Tsai, Lin and Cheng2013). Simultaneously, they insist on adherence to high work standards (Wang et al., Reference Wang, Chiang, Tsai, Lin and Cheng2013). Subordinates are then conflicted: they lack work-relevant information and resource support but are expected to meet high expectations.
Second, authoritarian leaders tend to make decisions unilaterally, maintaining supervisor/subordinate distance (Aryee et al., Reference Aryee, Chen, Sun and Debrah2007) and striving to enhance their personal status (Schuh, Zhang, & Tian, Reference Schuh, Zhang and Tian2013). They believe that they know more than other employees (Ertureten, Cemalcilar, & Aycan, Reference Ertureten, Cemalcilar and Aycan2013), so they can ignore, disrespect, devalue, and discount subordinates’ suggestions and contributions (Aryee et al., Reference Aryee, Chen, Sun and Debrah2007). Their controlling behaviors further signal strong disregard for subordinates’ interests, perspectives and abilities (De Cremer, Reference De Cremer2006; Foels, Driskell, Mullen, & Salas, Reference Foels, Driskell, Mullen and Salas2000), which then impairs their sense of competence and self-evaluation (Chan et al., Reference Chan, Huang, Snape and Lam2013). Employees feel devalued and discouraged, while pressured to meet high work standards. The incompatible expectations produce role conflicts. We propose:
Hypothesis 1a: Supervisors’ authoritarian behaviors will positively affect subordinates’ role conflict.
Under high role conflict, subordinates may spend much work time struggling to balance incompatible expectations. The pressure generates psychological strain such as increased depression and frustration (Beehr, Jex, Stacy, & Murray, Reference Beehr, Jex, Stacy and Murray2000; Jex & Beehr, Reference Jex, Beehr, Ferris and Rowland1991). Employees may feel exhausted and lack cognitive and emotional resources (Lankau, Carlson, & Nielson, Reference Lankau, Carlson and Nielson2006) to perform discretionary activities. Instead, they must focus on mandatory tasks (Thompson & Werner, Reference Thompson and Werner1997), directly aligning their behavior with the evaluation standards or explicit requirements (Bergeron, Reference Bergeron2007; Eatough, Chang, Miloslavic, & Johnson, Reference Eatough, Chang, Miloslavic and Johnson2011). In summary, subordinates under authoritarian supervisors will suffer role conflict and will be less likely to perform OCB:
Hypothesis 1b: Subordinates’ role conflict will mediate the relationship between supervisors’ authoritarian behaviors and subordinates’ OCB. Authoritarian behaviors will positively affect role conflict; role conflict will further negatively affect OCB.
Role Ambiguity as a Mediating Mechanism
Role ambiguity is ‘uncertainty about what actions to take to fulfill a role’ (Peterson et al., Reference Peterson, Smith, Akande, Ayestaran, Bochner, Callan, Cho, Jesuino, D'Amorim, Francois, Hofmann, Koopman, Leung, Lim, Mortazavi, Munene, Radford, Ropo, Savage, Setiadi, Sinha, Sorenson and Viedge1995: 430). When role senders fail to define or communicate role expectations clearly, employees have unclear expectations and consequently perceive role ambiguity (Kahn, Wolfe, Quinn, & Snoek, Reference Kahn, Wolfe, Quinn and Snoek1964; Katz & Kahn, Reference Katz and Kahn1978; Lyons, Reference Lyons1971).
Authoritarian supervisors may increase subordinates’ perceptions of role ambiguity. They withhold important information and resources (Chan, Reference Chan2013; Wang et al., Reference Wang, Chiang, Tsai, Lin and Cheng2013) for control, subordinates feel they cannot get work support. Simultaneously supervisors set high standards and deliver error-free expectations. The hierarchical distance prevents subordinates from explicitly dissenting or daring to seek help (Chan et al., Reference Chan, Huang, Snape and Lam2013; Farh & Cheng, Reference Farh, Cheng, Li, Tsui and Weldon2000; Flood et al., Reference Flood, Hannan, Smith, Turner, West and Dawson2000; Pellegrini & Scandura, Reference Pelligrini and Scandura2008). They are unclear about how to reach high standards.
Moreover, authoritarian leadership allows little room for subordinates to perform their assigned roles independently and autonomously. With independence and autonomy, can subordinates ignore obscure or vague messages, modify or even redefine their roles according to their own expectations (Jackson & Schuler, Reference Jackson and Schuler1985; Shenkar & Zeira, Reference Shenkar and Zeira1992). Conversely, authoritarian leaders demand unquestioned obedience, compliance, and dependence (Chan, Reference Chan2013), so employees must accept and process ambiguous information:
Hypothesis 2a: Supervisors’ authoritarian behaviors will positively affect subordinates’ role ambiguity.
Subordinates who are uncertain about job expectations are uncertain about the extent of their own authority or the standards that will be used to judge them. They experience anxiety, depression, and futility that lower their self-esteem (Jackson & Schuler, Reference Jackson and Schuler1985; Van Sell, Brief, & Schuler, Reference Van Sell, Brief and Schuler1981), so they avoid job involvement. Instead, they rely on passive trial-and-error approaches for meeting supervisor expectations (Rizzo et al., Reference Rizzo, House and Lirtzman1970), perhaps without knowing that they are working counterproductively from the organizational perspective (Van Sell et al., Reference Van Sell, Brief and Schuler1981). Role ambiguity is known to reduce voluntary efforts to do quality work (Beehr, Walsh, & Tabler, Reference Beehr, Walsh and Taber1976, Van Sell et al., Reference Van Sell, Brief and Schuler1981). Hence, we expect authoritarian supervisors may cause more role ambiguity and further reduce OCB:
Hypothesis 2b: Subordinates’ role ambiguity will mediate the relationship between supervisors’ authoritarian behaviors and subordinates’ OCB. Authoritarian behaviors will positively affect role ambiguity, and role ambiguity will further negatively affect OCB.
Role Overload as a Mediating Mechanism
Role overload describes ‘lack of the personal resources needed to fulfill commitments, obligations, or requirements’ (Peterson et al.: 430). It occurs when employees perceive that they have too many responsibilities compared with their time, resources, abilities, and other constraints; when they are overwhelmed by role demands exceeding the available resources; or when they are confronted by multiple role senders with various expectations (Rizzo et al., Reference Rizzo, House and Lirtzman1970; Singh, Reference Singh2000).
Authoritarian supervisors may increase role-overload perceptions by centralizing work decisions and requiring high performance without excuses, exceptions, or subordinate participation in decision-making (Tsui et al., Reference Tsui, Wang, Xin, Zhang and Fu2004). Lacking adequate information exchange, subordinates cannot tell supervisors about practical problems they face in completing work assignments. Supervisory control of information and resources further exacerbates their practical difficulties. Even worse are the high performance standards (Cheng et al., Reference Cheng, Chou, Wu, Huang and Farh2004) without bilateral communications that would allow supervisors to know whether subordinates have the ability, time, and resources to meet expectations. Moreover, authoritarian supervisors transmit salient information cues that employees lack job autonomy (Chan et al., Reference Chan, Huang, Snape and Lam2013) and must obey without question. Employees are then less likely to seek help from supervisors and other outside resources. Consequently, they feel overloaded:
Hypothesis 3a: Supervisors’ authoritarian behaviors will positively affect subordinates’ role overload.
From a resource allocation perspective, overloaded and overwhelmed employees (Bolino & Turnley, Reference Bolino and Turnley2005) must spend significant time and resources to alleviate overload perceptions and assure in role performance. Consequently, role overload may reduce their sense of achievement and confidence (LePine, Podsakoff, & LePine, Reference LePine, Podsakoff and LePine2005). Subordinates are less satisfied with their jobs (Eatough et al., Reference Eatough, Chang, Miloslavic and Johnson2011), and demotivated to perform discretionary extra-role efforts (Bergeron, Reference Bergeron2007). Thus:
Hypothesis 3b: Subordinates’ role overload will mediate the relationship between supervisors’ authoritarian behaviors and subordinates’ OCB. Authoritarian behaviors will positively affect role overload, and role overload will further negatively affect OCB.
Sample and Procedure
We collected leader – subordinate paired data from seven technical support and service private companies in North China, whose businesses include e-business, Internet technology, telecommunications, wind power products, and insurance. All the companies have formal hierarchical organizational structures. Participants included all supervisors and their subordinates in the same teams in a large company unit/branch. The teams engage in technical work or consumer service. Contact persons in each firm provided information about participating employees and their immediate supervisors, which allowed us to develop codes to match them. The contact persons administered paper surveys onsite. Respondents returned completed surveys in sealed, preaddressed envelopes to ensure confidentiality.
We surveyed participants twice because temporal separation between an independent variable and mediation variables allows previously recalled information to leave short-term memory (Podsakoff, Mackenzie, Lee, & Podsakoff, Reference Podsakoff, MacKenzie, Lee and Podsakoff2003). At time 1, subordinates rated supervisors’ authoritarian leadership and their demographic information. After three weeks, a time lag that is long enough to eliminate short-term memory influence but will not mask pre-existing relationships (Gong et al., Reference Gong, Huang and Farh2009; Podsakoff et al., Reference Podsakoff, MacKenzie, Lee and Podsakoff2003), they reported their role perceptions and control variables (leader – member exchange and procedural justice perceptions), and their supervisors assessed their OCB. We distributed 786 employee questionnaires; 731 were returned at time one; 722 were returned at Time 2; 705 returned both questionnaires (a response rate of 89.7). We excluded those who missed first- or second-time data and excluded invalid questionnaires such as those missing a majority of blanks, giving all items the same scores, or showing zigzag patterns, for a final sample of 613 subordinates. Each supervisor had an average of 6.19 subordinates (SD = 2.59).
In the subordinate sample, 59.6 percent were men; 87.3 percent had college or above education. The mean age was 27.6 (SD = 6.60) and mean organizational tenure was 2.2 (SD = 2.08) years.
We used a five-point Likert-type scale for the questionnaires: 1 = strongly disagree and 5 = strongly agree. All relevant variables used measures that had good reliabilities in previous literature. For the English-version measures, we generated Chinese versions following the translation-back-translation procedure (Brislin, Reference Brislin, Triandis and Lambert1980).
Following Zhang, Tsui, and Wang (Reference Zhang, Tsui and Wang2011), we used five items to measure supervisors’ authoritarian behavioral tendency to make directive decisions and maintain hierarchical order: ‘My supervisor has personal control of most matters’, ‘My supervisor centralizes decision on him/herself’, ‘My supervisor makes unilateral decisions and takes individual actions’, ‘My supervisor always behaves commandingly in front of employees’, ‘We must follow his/her rules to get things done. If not, he/she punishes us severely’. Cronbach's alpha was 0.92.
We adopted role perception measurements from Peterson et al.’s (Reference Peterson, Smith, Akande, Ayestaran, Bochner, Callan, Cho, Jesuino, D'Amorim, Francois, Hofmann, Koopman, Leung, Lim, Mortazavi, Munene, Radford, Ropo, Savage, Setiadi, Sinha, Sorenson and Viedge1995) scale based on role conflict, role ambiguity (Rizzo et al., Reference Rizzo, House and Lirtzman1970), and role overload scales (Pareek, Reference Pareek1976). For role conflict we used three items ‘I often get involved in situations in which there are conflicting requirements’, ‘I receive incompatible requests from two or more people’, ‘I have to do things that should be done differently under different conditions’. For role ambiguity we used five reverse-coded items: ‘I (do not) have clear planned goals and objectives for my job’, ‘I (do not) have clear planned goals and objectives for my job’, ‘I (do not) know exactly what is expected of me’, ‘I (do not) know what my responsibilities are’, ‘I (do not) feel certain about how much responsibility I have’, ‘My responsibilities are (not) clearly defined’. For role overload we used five items: ‘There is a need to reduce some parts of my role’, ‘I feel overburdened in my role’, ‘I have been given too much responsibility’, ‘My work load is too heavy’, ‘The amount of work I have to do interfere with the quality I want to maintain’. Cronbach's alpha coefficients were 0.92, 0.89, and 0.87, respectively.
We adopted a 15-item scale from Song, Tsui, and Law (Reference Song, Tsui and Law2009). For example, ‘He/she is willing to help new employees adapt to the environment though it's not part of his/her job’ and ‘He/She is willing to help his/her colleagues with their work problems’. Cronbach's alpha coefficient was 0.91.
We controlled for leader – member exchange and procedural justice perceptions because they reflect two alternative mechanisms for explaining how authoritarian leadership affects subordinate OCB (Liang et al., Reference Liang, Ling and Hsieh2007; Wu et al., Reference Wu, Huang, Li and Liu2012). We adopted a seven-item scale (Scandura & Graen, Reference Scandura and Graen1984) to measure leader – member exchange. For example, ‘I feel that my immediate supervisor recognizes my potential’ and ‘I feel that my immediate supervisor understands my problems and needs at work’. Cronbach's alpha coefficient was 0.92. We adapted a four-item scale from Moorman (Reference Moorman1991) to measure procedural justice: ‘My supervisor makes job decisions in an unbiased manner’, ‘My supervisor provides opportunities to appeal or challenge job decisions’, ‘My supervisor explains why decisions are made’, ‘My supervisor has developed procedures designed to generate standards so that decisions could be made with consistency’. Cronbach's alpha coefficient was 0.92. We used a five-point Likert scale from 1 = strongly disagree to 1 = strongly agree for the two measures. We also controlled for age, gender, and job types because they affect work perceptions and behaviors (e.g., Chen et al., Reference Chen, Eberly, Chiang, Farh and Cheng2014; Wu et al., Reference Wu, Huang, Li and Liu2012). Age was measured in years. Gender was a categorical variable with 1 for women and 0 for men. Job type was a categorical variable with 1 for technical work and 0 for consumer service work. Seven of the firms differ in their business and organizational characteristics, but their teams focus mainly on one job type. To avoid multicollinearity, we controlled only for job type.
We used Mplus 7.31 to examine the hypothesized model. To account for the non-independence in OCB ratings, we used MSEM because it accounts for hierarchical data and avoids inaccurate standard errors and biased statistical conclusions because of non-independence (Bliese, Reference Bliese, Klein and Kozlowski2000). We treated our dyadic-level research constructs at the individual level and controlled team identity at the team level.
Before we conducted confirmatory factor analysis and tested the structural models, we averaged items into OCB dimensions and treated the dimension scores as separate indicators of OCB. In addition, we randomly assigned the items into three parcels for each unidimensional latent variable with more than three items, namely authoritarian leadership, role overload, and role ambiguity, because parcels have psychometric merits relative to items and are particularly effective for unidimensional constructs (Little, Cunningham, Shahar, & Widaman, Reference Little, Cunningham, Shahar and Widaman2002) and such approach can maximize the degree of freedom and avoid model identification problems (Hau & Marsh, Reference Hau and Marsh2004). We then used item scores as indicators of role conflict because it has only three items. Thereafter we did confirmative factor analysis to test the construct validity, ran the hypothesized structural model and alternative models, and compared model fit indices to evaluate model fit (Bollen, Reference Bollen1989; Kelloway, Reference Kelloway1998). Next, we reported the indirect effects for mediating hypotheses, which, in Mplus, is estimated using maximum-likelihood and is based on counterfactuals (causal inference). Potential outcomes using expectations were previously developed (c.f., Muthén, Reference Muthén2011; Pearl, Reference Pearl2001; Robins, Reference Robins, Green, Hjort and Richardson2003; Robins & Greenland, Reference Robins and Greenland1992).
Table 1 summarizes means, standard deviations, correlations, and reliability coefficients (Cronbach's alpha) of all the variables. Authoritarian behavior was negatively associated with OCB (r = −0.15, p < 0.01) and positively associated with role conflict (r = 0.35, p < 0.01), role ambiguity (r = 0.16, p < 0.01), and role overload (r = 0.39, p < 0.01). Role conflict, role ambiguity, and role overload were each significantly correlated with OCB (r = −0.24, p < 0.01; r = −0.09, p < 0.05; r = −0.22, p < 0.01). Those correlations preliminarily aligned with the hypotheses.
Note: *p < .05, **p < .01; numbers in the bracket are Cronbach alpha reliability; OCB = organizational citizenship behavior.
We used multilevel confirmatory factor analysis (MCFA) to test the construct distinctiveness between authoritarian leadership, role conflict, role ambiguity, role overload, and OCB (Table 2). The hypothesized five-factor model had a good fit with all the fit indices at the acceptable level (χ 2 = 627.24, df = 185, χ 2/df = 3.39, CFI = 0.94, TLI = 0.93, RMSEA = 0.06). To check the discriminant validity of the hypothesized five-factor measurement model, we compared it with four alternative models. In the four-factor model, role conflict and role overload were merged into one factor (Δχ 2 (4) = 804.22, p < 0.01). In the three-factor model, three role perceptions were merged into one factor (Δχ 2 (7) = 2840.90, p < 0.01). In the two-factor model, authoritarian behavior was combined with the three role stresses (Δχ 2 (9) = 4757.84, p < 0.01). In the one-factor model, all five variables were combined as one factor (Δχ 2 (10) = 4969.00, p < 0.01). Based on those results, we concluded that the five-factor model showed better fit with data.
Note: ***p < .001. AB = authoritarian behavior; RO = role overload; RA = role ambiguity; RC = role conflict; OCB = organizational citizenship behavior.
We ran a full mediation model: supervisors’ authoritarian behavior affected subordinate OCB fully through three role perceptions. It had good fit indices: χ 2 = 1546.36, df = 501, χ 2/df = 3.09, CFI = 0.91, TLI = 0.90, RMSEA = 0.06, SRMR = 0.08. We also ran a nested partial mediation model, adding a direct path from authoritarian behavior to subordinate OCB. The fit indices were χ 2 = 1546.23, df = 500, χ 2/df =, CFI = 0.91, TLI = 0.90, RMSEA = 0.06, SRMR = 0.08. The partial mediation model was not significantly better than the full mediation model (Δχ 2 (1) = 0.13, p < 0.05), so we used the full mediation model to test the hypotheses following the parsimony principle (see Figure 1).
In the full mediation model, the standardized estimates of the paths from authoritarian behavior to role conflict, to role ambiguity, and to role overload were 0.42 (p < 0.001), 0.09 (p < 0.05) and 0.34 (p < 0.001), supporting H1a, H2a, and H3a respectively. The path estimates from role conflict, role ambiguity, and role overload to OCB were −0.16 (p < 0.05), −0.07 (n.s.) and −0.15 (p < 0.01), demonstrating that both role conflict and role overload were negatively related with OCB, and role ambiguity was insignificant to OCB. Thus Hypothesis 2b was rejected.
To test the mediation hypotheses (H1b and H3b), we examined the indirect effects. Particularly, the indirect effect mediated by role conflict from authoritarian behavior to OCB was −0.067, p < 0.05 supporting H1b. The indirect effect mediated by role overload was −0.050, p < 0.05, supporting H3b.
In this study, we propose that authoritarian supervisors cause employees to perceive their job roles as conflicted, ambiguous, and overloaded, with consequent negative effects on their OCB. We find that role conflict and overload perceptions, but not role ambiguity, are significant mediators. We conclude that role perceptions greatly explain the influences of authoritarian behavior on subordinates’ OCB in Chinese organizations.
Unexpectedly, perceptions of role ambiguity did not affect employee OCB in SEM results. Nevertheless, the two were negatively associated in the correlation analysis (r = −0.09, p < 0.05). The structural model simultaneously considers the effects of the three role perceptions on the dependent variable. Compared with role conflict and role overload, role ambiguity has a weaker association with OCB and a moderately high correlation with role conflict (r = 0.49, p < 0.001). Thus role conflict effects may partial out role ambiguity effects on OCB.
Our role perspective contributes to leadership and OCB literature by exploring whether role conflict, overload, and ambiguity perceptions mediate authoritarian leadership effects on extra-role behaviors. First, leadership scholars have predominantly focused on characteristics or behaviors of effective/successful leaders (Hunter, Bedell-Avers, & Mumford, Reference Hunter, Bedell-Avers and Mumford2007), but have largely ignored how authoritarian leadership may negatively affect subordinates, and have rarely considered the authoritarian style rooted in supervisory roles. Making unilateral decisions and maintaining hierarchical distance even become regular supervisory role activities, but simultaneously produce subtle social influences on subordinates. Our research deepens our understanding of how authoritarian leadership affects subordinates.
We confine our investigation to authoritarian supervision to show that authoritarian supervision indirectly harms OCB through complex role-making processes. The role perspective will help scholars better understand how leadership shapes role perceptions and resulting extra-role behaviors. Leaders who set high requirements and act as superiors in their leadership roles ensure that subordinates see themselves as playing lower roles; they are forced to focus too much on their in-role responsibilities and ignore extra-role work (Bergeron, Reference Bergeron2007).
Third, we provide insights into work role literature by proposing that leadership helps to determine role perceptions. The role-stress literature has mainly considered role conflict, ambiguity, and overload as exogenous forces, and rarely considered their predictors. However, work events, and senders’ expectations do not cause negative role perceptions in isolation (Peterson et al., Reference Peterson, Smith, Akande, Ayestaran, Bochner, Callan, Cho, Jesuino, D'Amorim, Francois, Hofmann, Koopman, Leung, Lim, Mortazavi, Munene, Radford, Ropo, Savage, Setiadi, Sinha, Sorenson and Viedge1995). Supervisory behavior also sends powerful social informational cues. By linking the leadership and role stress literature, we show a new way to understand how organizational leaders can interpersonally amend role perceptions.
This research provides practical implications for organizational management. Unilateral decision-making and hierarchical distance conveys authoritarian style leadership which then negatively affects employees’ role perceptions. To counterbalance such effects, leaders might consult with subordinates before making decisions, obtain their input, and explain reasons for higher-level decisions. Firms may also want to design and implement leadership programs that train supervisors in leadership skills beyond authoritarianism. To increase role clarity and decrease role conflict and overload, firms should comprehensively analyze and clearly define jobs so that supervisors will have less negative social influence on subordinate work role perceptions. If subordinates are clear about in-role requirements, they may be more likely to perform extra-role behaviors.
Limitations and Future Research Directions
This study has several limitations that provide suggestions for future research directions. We conducted our study in China and should be cautious about generalizability. Although authoritarian leadership is a form of paternalistic leadership, the style is universal (Pelligrini & Scandura, Reference Pelligrini and Scandura2008). Because of the Western cultural emphasis on equality, authoritarian leadership may be less popular in Western countries. Nevertheless, Western supervisors must make final decisions and maintain certain position status. Consequently, authoritarianism might have more salient effects on role perceptions in the West because employees are sensitive to unequal treatments. Future research can compare whether the current research model is more salient in Western contexts.
Second, focusing only on authoritarian leadership may narrow our lens to a pure negative aspect. Indeed, other leader behaviors approach power differently, and have different effects on role perceptions. For example, transactional supervisors ‘clarify the role and task requirements for followers’ (Bass & Avolio, Reference Bass, Avolio, Woodman and Pasmore1990: 236). By increasing job clarity, they reduce role conflict (MacKenzie, Podsakoff, & Rich, Reference Mackenzie, Podsakoff and Rich2001). Another supervisory behavior – initiating structure – also reduces role stress and ambiguity (Dale & Fox, Reference Dale and Fox2008). Future research should further demonstrate that role perceptions are important consequences of leadership behaviors. Moreover, the research might investigate conditions that will allow authoritarian leaders to cause less role conflict, overload, and ambiguity. For example, benevolent leadership may buffer the negative effects on role perceptions. When subordinates are highly traditional, older, or have more work experience, they may be more likely to accept authoritarian behavior and less likely to have negative role perceptions.
Third, we used OCB to represent extra-role behavior. We did not consider alternative proactive, voice, and creative behavior because we do not expect authoritarian leadership to similarly influence those behaviors. For example, high work expectations but low work resources and support may force subordinates to creatively accomplish their tasks. Future research could investigate whether incompatible situations may trigger creativity.
Fourth, we proposed cognitive overload, information processing, and emotional exhaustion in explaining how role perceptions affect OCB, but did not test those mechanisms. Future research could empirically investigate whether they mediate between role perceptions and OCBs.
Fifth, methodologically, we failed to separately collect role perceptions and OCB variables although they come from different sources. Ideally, time lags should occur between measures of supervisory behavior, role perceptions, and OCB. Nevertheless, the findings align with our logic so our results may still be valid. For example, we assume that increasing role overload reduces OCB, and empirical results show a negative relationship, but OCB has empirically been shown to increase role overload (Bolino & Turnley, Reference Bolino and Turnley2005). Additionally, our questionnaire surveys may have caused mono-method bias and may fail to show causal relationships.
Supervisory behaviors greatly determine subordinate performance. Companies in China's rapidly changing and highly competitive environment are especially pressed to discover the most effective management styles. Our study contributes to the leadership literature by identifying how China's typically authoritarian supervisors affect subordinates’ role perceptions and their consequent OCB.