The Influence of Authoritarian-Benevolent Leadership on Subordinates’ Work Engagement: A Social Information Processing Perspective

Introduction

COVID-19 has proved to be a game of changers in an era of new perils and new hardships, in which organizations call for ambidextrous leadership to cope with the contradictions and tensions in the dynamic and complex external and internal environment.¹ Ambidextrous leadership, defined as the ability of flexibly switching between different but complementary leadership behaviors,² is currently raising scholars’ interest.³ In the post-COVID-19 era, managers still inevitably need to balance various mutually exclusive needs,⁴ such as expectation of high performance and consideration, and strict obedience requirement and flexible management.⁵ Scholars suggest that effective leaders need to follow both/and thinking, adopt an ambidextrous leadership approach, and integrate multiple conflicts according to the specific situation.⁶ As a kind of traditional management strategy, the carrot-and-stick leading thought requires leaders to show both favor and threat, which means to conduct authoritative leadership or benevolent leadership behavior according to the need, so as to create a loose and tight management climate.⁷ Authoritarian leadership refers to leaders strictly supervising and controlling subordinates, and benevolent leadership refers to leaders showing concern and satisfying personalized needs of subordinates.⁸ Authoritarian-benevolent leadership is in turn defined as leaders taking use of these two complementary leadership behaviors across different conditions.⁹ Authoritarian-benevolent leadership has been proved to be an effective ambidextrous leadership style which positively affects subordinates’ psychological status and behaviors, such as promoting employees’ paradox mindset,¹⁰ creativity¹¹ and work performance.⁹ Although a growing attention has been paid on authoritarian-benevolent leadership, several important gaps still exist in understanding its impacts on subordinates’ work outcomes.

Firstly, despite increasing research on authoritarian-benevolent leadership, scholars have neglected its dark side. Requiring leaders to skillfully switch between two types of contradictory leadership behaviors according to situations may do harm to subordinates’ psychological well-being and work outcomes, such as reducing subordinates’ subjective well-being¹² and increasing employees’ unethical pro-organizational behaviors.¹³ These findings challenge the basic assumption that ambidextrous leadership always brings positive effects, and call for research on its potential negative impacts.^9,14 Work engagement as an important variable in organizational research arouses employees’ full potential that increases organizational performance and gives an organization competitive advantage.¹⁵ Leadership is a fundamental factor that determines subordinates’ work engagement, since it is closely associated with employees’ psychological safety that is related to work engagement.¹⁶ However, the inconsistency of leadership behaviors easily causes tensions and distress¹⁷ that may threaten subordinates’ work engagement. At present, the potential negative relationship between authoritarian-benevolent leadership and work engagement is not yet clear, and empirical research is necessary to explore its influencing mechanism. Therefore, to bridge this gap, the current study seeks to uncover the underlying mechanism between authoritarian-benevolent leadership and work engagement.

Secondly, different from single leadership styles, authoritarian-benevolent leadership shows its uniqueness in switching leadership behaviors in line with conditions.² However, few studies have explored in which way such switch improves or hinders employee outcomes. Social information processing theory¹⁸ as an appropriate theoretical framework was introduced to clarify the mechanism of authoritarian-benevolent leadership on work engagement by figuring out a mediating variable of leader-member exchange ambivalence. According to social information processing theory, individuals adjust their attitudes, behaviors and beliefs according to the social environment they are in, their past and present experiences and their states.¹⁸ Leaders’ behaviors influence employees’ cognitive functions and sensemaking processes, since an employee’s mental state is directly linked to the information conveyed to him/her.¹⁹ The inconsistent information delivered by authoritarian-benevolent leadership easily brings contradictory psychological experiences to subordinates and causes negative effects.²⁰ Authoritarian leadership tends to lower subordinates’ evaluations of their status,²¹ while benevolent leadership behaviors tend to strengthen subordinates’ evaluations of leader-member exchange quality and their status.²² The ambivalent evaluation orientation of leader–follower relationship is defined as leader-member exchange ambivalence (LMX ambivalence), which refers to the subjective experience of individuals perceiving both positive and negative judgments towards their leader–follower relationship.²³ Subordinates need relationship information about the interpersonal interaction qualities from their superiors to ensure their position and status in the leader–follower relationship, while the inconsistency in leadership behaviors or decision-makings may reduce subordinates’ self-evaluation and increase subordinates’ uncertainty about their interpersonal relationships with their leaders.²⁴ Troubled subordinates may find it hard to get motivated from the insider identity and devote themselves to work.²⁵ In short, authoritarian-benevolent leaders tend to produce subordinates’ LMX ambivalence, thus reducing their work engagement. However, few studies have employed a comprehensive theoretical framework to examine the above mediating mechanism. Hence, drawing on social information processing theory, it is necessary to investigate the mediating role of LMX ambivalence on the relationship between authoritarian-benevolent leadership and work engagement.

Thirdly, previous studies have rarely examined the moderating effect of subordinates’ trait mindfulness as a potential boundary condition of authoritarian-benevolent leadership. According to social information processing theory, individual resources (individual abilities and traits) can influence individuals’ information judgment and processing.¹⁸ This study assumes that one way to interpret how information is processed and whether it works depends on trait mindfulness. Trait mindfulness is defined as the ability to concentrate on and perceive what is happening in the present moment.²⁶ It enables employees to focus more on the task at hand.²⁶ Despite the external interference, individuals with high trait mindfulness pay more attention to the meaning of the tasks themselves, and engage in promoting the quality of their own experience.²⁷ Therefore, it may restrain the decline of work engagement brought by the ambivalent experience. Hence, this study proposes to examine one of the boundary conditions of authoritarian-benevolent leadership on subordinates’ work engagement through LMX ambivalence by testing the moderating role of subordinates’ trait mindfulness. In sum, the current study proposed the following research questions:

By addressing these research questions, this study contributes to the research of authoritarian-benevolent leadership in three aspects. First, building on the social information processing theory, this research provides a more comprehensive understanding of the mechanism underlying the effect of authoritarian-benevolent leadership on subordinates’ work engagement, which answers the recent calls to focus on the negative outcomes of ambidextrous leadership.^9,14 Second, this study believes that the difficulty in processing complex social information can result in subordinates’ ambivalent experience as LMX ambivalence, and this view provides a powerful mediating explanation for the impact of authoritarian-benevolent leadership on subordinates’ work engagement, and enriches the antecedent studies of work engagement. Meanwhile, it fills the gap of research on the antecedents of LMX ambivalence, and expands its outcomes in the work settings.²⁸ Third, by exploring the moderating effects of trait mindfulness, this study not only broadens the boundary conditions of the relationship between authoritarian-benevolent leadership and subordinates’ work engagement but also provides implications for practitioners to mitigate the negative side of authoritarian-benevolent leadership.

Literature Review and Hypotheses

Authoritarian-Benevolent Leadership and Leader-Member Exchange Ambivalence

Authoritarian-benevolent leaders refer to the ones who can continuously and flexibly adjust benevolent leadership and authoritarian leadership according to the situational requirements.¹² Such leaders tend to emphasize both the rules and leader–member relationships. When a leader intends to exert power, his/her authoritative behaviors can be divided into two dimensions: personnel orientation and task orientation. From the personnel orientation perspective, leaders advocate personal authority, belittle subordinates’ contributions, and make subordinates obey and fear of leaders; from the task orientation perspective, leaders strictly monitor the work tasks and work procedures of subordinates, and require subordinates to firmly obey the organizational norms.²⁹ Authoritarian leadership behaviors of exploiting and controlling subordinates, and focusing on the task and self-interests often lead to subordinates’ unspoken pressure, insecurity and burnout,^30,31 which in turn lead to counterproductive work behaviors.³² However, authoritarian-benevolent leaders emphasize acting according to circumstances. When a leader intends to show benevolence, he/she will show concern and care to his/her subordinates and satisfy their emotional and relational needs,³³ which helps erase their psychological distress,³⁴ and is beneficial for promoting employees’ loyalty.³⁵ The two contradictory leadership behaviors deliver different signals, which increases the difficulty for subordinates in processing the vague information and figuring out leaders’ true intentions.^20,24

Social information processing theory argues that individuals, as self-adaptive organisms, adjust their attitudes, behaviors and perceptions according to the social environment they are in, their past and present experiences and their states.¹⁸ In the workplace, employees receive social information mainly through interactions and communications with their supervisors or co-workers.³⁶ Leaders are important sources of social information for individuals because they are in a higher organizational position and have more direct interactions with subordinates.^37,38 By processing cues from leaders, subordinates can understand what is expected, punished, and rewarded.³⁹ In the process of individuals interacting and communicating with leaders in the workplace, leaders will constantly provide environmental information, which leads individuals to adjust their attitudes and behaviors along with the information received.³⁶ The social information from leaders engenders a socially supportive climate for employees, which may positively or negatively affect employee emotional outcomes.⁴⁰

Since the social information conveyed by an authoritarian-benevolent leader is always complex, vague and difficult to predict, it is difficult for subordinates to use their experience to guess the leader’s intention and possible leadership behaviors,⁴¹ and is even more difficult to judge their own relationship qualities with the leader.¹³ Leader-member exchange (LMX) is a kind of exchange behavior in nature, in which one party transmits an exchange signal through attitude and behavior, and the other receives the signal and gives feedback, and when both expectations are met, the degree of LMX would increase.⁴² However, when a subordinate recalls and compares his/her experience and predicts that a leader will show him/her concern and compassion in a certain situation, but the superior acts in an authoritative way, the subordinate may break his/her insider identity perception and suffer from a strong psychological gap.⁴³ This complex information with conflicts and tensions can trigger emotional tension, hesitation, and ambivalence for subordinates,⁴⁴ leading to LMX ambivalence.⁴⁵ Subordinates with the ambivalent experience about their superior–subordinate relationships have both positive and negative views of LMX, which means that they sometimes feel that they are insiders of the superior, but sometimes feel that the superior does not value their LMX.²³ Based on the above analyses, this study believes that authoritarian-benevolent leadership makes it difficult for subordinates to judge the leader’s attitude and LMX, and finally leads to LMX ambivalence. In conclusion, this study proposes the following hypothesis:

Hypothesis 1: Authoritarian-benevolent leadership is positively associated with subordinates’ LMX ambivalence.

Authoritarian-Benevolent Leadership, Leader-Member Exchange Ambivalence and Work Engagement

Social information processing theory posits that before enacting behaviors, individuals engage in several mental steps of social information processing, including encoding social cues, interpreting cues to recognize others’ intentions or conduct self-evaluations, clarifying goals, accessing and deciding responses, and these steps finally lead to behavioral outcomes.⁴⁶ On this basis, this study proposes that authoritarian-benevolent leadership may not only trigger subordinates’ LMX ambivalence, but also further damage subordinates’ work engagement through the LMX ambivalence. Work engagement refers to the state in which individuals are physically, cognitively, and emotionally engaged in their work roles, experiencing work meaning, psychological security, and fulfillment, and is characterized by vigor (energy and resilience), dedication (significance and enthusiasm), and absorption (concentration).⁴⁷ Employees’ work engagement is mainly influenced by work resources (such as perceived organizational support and performance feedback) and personal resources (such as individuals’ abilities to positively self-regulate and control the environment).⁴⁸ In organizations, leaders exert important influences on subordinates, since leaders control the access of the subordinates’ work resources, and leaders’ own leadership styles and the LMX also belong to work resources that directly affect subordinates.¹⁵ However, it has been shown that highly interdependent interpersonal relationships are often prone to cause ambivalence, leading to disordered or negative outcomes.⁴⁹ Such ambivalence is hard to avoid because the relationship between superior and subordinate is inherently contradictory, since leaders and subordinates are often trapped in a cycle of seeking interpersonal intimacy while maintaining a power distance⁵⁰ that is more remarkable under authoritarian-benevolent leadership. The relationship contradiction is easy to lead to an individual’s conflicting thought and loathing mood, thus reduces his/her work vigor and hinders him/her from concentrating on work tasks, because individuals with high-level ambivalent experiences are more prone to be influenced by negative information.⁵¹ Empirical studies have shown that LMX ambivalence has negative impacts on individual work outcomes, such as negative impacts on work performance²³ and proactive behavior.⁵² The social information processing theory suggests that individuals’ processing of complex and ambiguous information will affect their attitudes and behaviors.¹⁸ For subordinates who experience LMX ambivalence, the sometimes close and sometimes aloof sense of distance will affect subordinates’ judgment of their status and identity, doubt the degree to which they are valued by their superiors, and worry about their future career prospects.⁵³ These consequences reduce subordinates’ psychological security and make them difficult to devote themselves cognitively and emotionally into work, thus reduce work engagement.⁴³

Based on the above analyses, this study believes that leadership can increase the experience of subordinates’ LMX ambivalence, and the LMX ambivalence can further decrease work engagement. In other words, LMX ambivalence, as the channel connecting authoritarian-benevolent leadership and subordinates’ work engagement, plays a mediating role. Specifically, authoritarian-benevolent leadership makes subordinates unable to accurately confirm their insider identity, and difficult to predict the intentions and ideas of superiors, resulting in the LMX ambivalence. This kind of ambivalence makes subordinates feel less accessible to work resources, resulting in massive pressure and worries for the future work,⁵³ thus damaging subordinates’ work engagement. Therefore, this study proposes the following hypothesis:

Hypothesis 2: LMX ambivalence mediates the relationship between authoritarian-benevolent leadership and subordinates’ work engagement.

The Moderating Role of Trait Mindfulness

The concept of mindfulness originated in Buddhism and has been widely used in psychology, social science and other fields in recent years.⁵⁴ At present, mindfulness is generally defined as an individual’s awareness (the process of concentrating on a particular event or experience) and perception (the perception of the on-going experience) of the present experience, and broadly speaking, mindfulness refers to the ability to concentrate on and perceive what is happening in the present moment.²⁶ Mindfulness can refer not only to an individual’s consciousness and state, but also to an individual trait.⁵⁵ Studies of mindfulness tend to focus on its effects on individuals’ work outcomes, health conditions, and mental well-being,⁵⁶ such as increasing employee work engagement and slowing down emotional exhaustion.⁵⁷ According to the social information processing theory, individual characteristics and external environment jointly affect individuals’ processing and judging information, and their behaviors are the comprehensive result of the interactions.¹⁸ Individual traits are stable individual differences that affect people’s interpretations and reactions to their job experiences.⁵⁸ Individuals with certain personality traits may be easier to deal with the conflicting behaviors and attitudes of leaders.⁵⁹ As an important individual resource,⁶⁰ mindfulness can help individuals focus on analyzing and solving current problems without being distracted by external objects,⁶¹ thus resisting the negative effects brought about by the LMX ambivalence. Based on the above analyses, this study focuses on the moderating role of individual trait mindfulness in the relationship between authoritarian-benevolent leadership and work engagement.

When subordinates are trapped in the contradictory and repeated consideration on the quality of LMX, the sense of ambivalence and anxiety will make it difficult for individuals to be motivated by their insider identity and hinders their commitment to work.⁶² However, faced with work tasks, individuals with higher levels of trait mindfulness pay more attention to the meaning of the tasks themselves, and try to enhance the quality of their own experience without external interference.²⁷ Although ambivalence may be at risk of impairing subordinate work engagement, the trait mindfulness can help individuals focus on the on-going experience, help them quickly make decisions, and regulate emotions, so as to respond appropriately in different social contexts.⁶³ Subordinates with higher levels of trait mindfulness are more likely to regulate their emotions, reduce divergent thinking, and avoid bringing their emotions to work,⁶⁴ thereby enjoying the joy and challenges of the work itself, and take an active part in the work.⁵⁵ To sum up, when subordinates have LMX ambivalence, the ones with high trait mindfulness tend to focus more on the sense of meaning given by the work itself and pay more attention to experiencing the current mood, and weaken the impact of the LMX ambivalence on work engagement. Thus, this study proposes the following hypothesis:

Hypothesis 3: Subordinates’ trait mindfulness moderates the relationship between LMX ambivalence and work engagement, such that the relationship is stronger when subordinates’ trait mindfulness is low than high.

The Moderated Mediating Effect

As previously discussed, social information processing theory indicates that after receiving social information, individuals’ processing and interpretation of information will affect their choices of attitudes and behaviors, in which individual characteristics work.¹⁸ Trait mindfulness enables individuals to better process information and behaves in a more rational way.⁶⁵ Following this viewpoint, this study proposes a moderated mediation model. Specifically, subordinates’ trait mindfulness moderates the indirect effect of authoritarian-benevolent leadership on subordinates’ work engagement through LMX ambivalence. Authoritarian-benevolent leaders change their leadership behavior flexibly according to the requirement of the context, which leads to the subordinates’ ambiguous and inconsistent ambivalent experiences about the attitudes and intentions of leaders.²⁰ Subordinates with high trait mindfulness can flexibly process social information,⁶⁶ concentrate more on their tasks and are easier to relieve workplace stress.⁶⁷ In this case, even if suffering from the downsides of authoritarian-benevolent leadership, subordinates can still easily maintain concentration, enthusiasm and dedication to work, and circumvent the negative impact of situational factors.⁶⁶ In contrast, individuals with lower trait mindfulness tend to be stressed and depressed and find it difficult to engage in work in a highly dynamic and stressful organizational environment.⁵⁶ In conclusion, this study proposes the following hypothesis:

Hypothesis 4: Subordinates’ trait mindfulness moderates the indirect effect of LMX ambivalence on the relationship between authoritarian-benevolent leadership and work engagement, such that the indirect effect is stronger when subordinates’ trait mindfulness is low than high.

Based on the above analyses, the theoretical model of this study is shown in Figure 1.

Figure 1 Research Framework.

Methods

Samples and Data Collection

This study used a questionnaire method and collected employee data from three enterprises located in Beijing, Dalian and Changchun in IT, civil aviation, and machinery manufacturing industries in China. To reduce the impact of common method bias, researchers did a time-lagged survey in two waves. At first, the researchers contacted the heads of human resource management departments of the three companies to clarify the matters needing attention when filling out questionnaires and informed them that the data collection process would strictly follow the principles of confidentiality, anonymity, and voluntariness, ensuring that the data collected would be used only for academic research purpose. Before collecting data, researchers numbered the participants, and paper questionnaires were sent to 397 employees at T1 time point. At this time point, participants were asked to rate their immediate superiors’ authoritarian-benevolent leadership, as well as employees’ own trait mindfulness. After eliminating the wrong, missed, and random questionnaires, 355 valid questionnaires remained, and the response rate was 89.42%. One month later, at the T2 time point, participants who had completed the previous round of survey were invited to rate their LMX ambivalence, work engagement, and to fill in some demographic information. A total of 309 effective questionnaires were obtained in this round, and the response rate was 87.04%. In the final sample, 58.9% of the participants were men (182 participants) and 41.1% of the participants were women (127 participants). The average age of the sample was 31.52 years old, and the average tenure was 6.82 years.

Research Tools

In this study, the variables were measured by well-tested scales with good reliability and validity. To ensure the semantic accuracy of the scales, the study followed the standard translation and back-translation procedure.⁶⁸ The variables were measured using 5-point Likert scale, with “1” (1 = strongly disagree) to “5” (5 = strongly agree).

Authoritarian-Benevolent Leadership

Authoritarian leadership was measured by an 18-item scale of Authoritarian Leadership Scale,²⁹ which included two dimensions of “juan-chiuan leadership” (8 items) and “shang-yan leadership” (10 items). Sample items include “His/her ideal subordinate must obey his/her words”, “He/she will supervise my job schedule and ask me to spare no pains”, and the Cronbach’s α value of the scale was 0.897. Benevolent leadership was measured by an 11-item subscale of Paternalistic Leadership Scales.³³ The scale was divided into two dimensions, namely “individual care” (6 items) and “consideration and tolerance” (5 items). Sample items include “My supervisor meets my needs according to my personal requests”, “My supervisor tries to understand what the cause is when I don’t perform well”, and the Cronbach’s α was 0.945. Referring to the previous research, the authoritarian-benevolent leadership was measured by multiplying the mean values of authoritarian leadership with benevolent leadership.^13,69

Leader-Member Exchange Ambivalence

This research used the 7-item LMX Ambivalence Scale²³ to measure the experience of subordinates’ LMX ambivalence. Sample items were “I have conflicting thoughts: sometimes I think that my working relationship with my manager is very good, while at other times I don’t”, “I have conflicting thoughts: sometimes I think my manager understands my problems and needs, while at other times I don’t.” The Cronbach’s α of the scale was 0.806.

Work Engagement

The 9-item version of the Utrecht Work Engagement Scale (UWES) was used to measure subordinates’ work engagement,⁴⁷ which includes three dimensions: vigor (3 items), dedication (3 items) and absorption (3 items). Sample items were “At my work, I feel bursting with energy”, “My job inspires me.” The Cronbach’s α of the scale was 0.796.

Trait Mindfulness

This study used the 15-item Mindful Attention Awareness Scale (MAAS) to measure subordinates’ trait mindfulness.²⁶ Sample items were “I found it difficult to stay focused on what’s happening in the present”, “I could be experiencing some emotion and not be conscious of it until sometime later.” The Cronbach’s α of the scale was 0.906.

Control Variables

Previous research has shown that demographic variables including gender, age, education and tenure could influence the model,^13,70 because these are important confounders that affect employees’ work emotions and behaviors.⁷¹ This study controlled for these variables.

Statistical Analysis

In this study, R 4.2.1, SPSS 26.0 and SPSS PROCESS macro were used for data processing. R 4.2.1 was used for common method bias test and confirmatory factor analysis. SPSS 26.0 and SPSS PROCESS macro were used for descriptive statistical analysis, correlation analysis, reliability test and hypothesis tests. To test Hypotheses 1 and 3, a hierarchical regression analysis was performed in use of SPSS 26.0. And to test Hypotheses 2 and 4, Model 4 and Model 14 of the SPSS PROCESS macro were used to examine the mediating and moderated mediating effects.⁷²

Results

Confirmatory Factor Analysis

Through the software of R 4.2.1, the confirmatory factor analysis (CFA) was conducted to test the distinctiveness of the variables: benevolent leadership, authoritarian leadership, LMX ambivalence, work engagement, and trait mindfulness. As shown in Table 1, the five-factor model exhibited adequate fit to the data, χ² = 568.114, df = 367, χ²/df = 1.548, comparative fit index (CFI) = 0.945, Tucker-Lewis coefficient (TLI) = 0.939, incremental fit index (IFI) = 0.945, root mean square error of approximation (RMSEA) = 0.042, standardized root-mean-square residual (SRMR) = 0.049. Four alternative models were examined, and the hypothesized five-factor model was significantly better than the other models, indicating that there was good discriminant validity between variables.

Table 1 Confirmatory Factor Analysis Results

Common Method Bias Test

Although this study used a multi-source, time-lagged approach in data collection, variables were self-reported, which was still possible to produce common method bias. Harman’s single-factor test was used to test the common method bias.⁷³ The results showed that the first factor only explained 19.751% of the total variance, which was less than the standard value of 50%, indicating that the homologous deviation of the data collected in this paper was controlled within a reasonable range. In addition, this study further tested the common method bias by controlling for an unmeasured latent method factor as recommended by Podsakoff et al.⁷³ For the controlled model with latent common method factor, χ²/df = 1.325, CFI = 0.970, TLI = 0.964, RMSEA = 0.032, SRMR = 0.042. Compared with the five-factor model in this study, the changes of fit indices were not significant. Results of the two methods above both indicated that common method bias had a limited impact on this study.

Descriptive Statistics and Correlation Analysis

The mean, standard deviation, and correlation coefficients of the variables are shown in Table 2. Authoritarian-benevolent leadership is positively associated with LMX ambivalence (r = 0.584, p < 0.001), and LMX ambivalence is negatively associated with work engagement (r = −0.414, p < 0.001). Trait mindfulness is positively associated with work engagement (r = 0.317, p < 0.001).

Table 2 Descriptive Statistics

Hypotheses Testing

This study took use of SPSS 26.0 software for a hierarchical regression analysis. Hypothesis 1 proposed that authoritarian-benevolent leadership was positively associated with subordinates’ LMX ambivalence. As shown in Table 3, the results of Model 2 indicated that authoritarian-benevolent leadership was positively associated with LMX ambivalence (β = 0.570, p < 0.001), which supported Hypothesis 1. Hypothesis 2 proposed that LMX ambivalence mediated the relationship between authoritarian-benevolent leadership and subordinates’ work engagement. To test the indirect effect, Model 4 of the SPSS PROCESS macro was used,⁷² and the results showed that LMX ambivalence partially mediated the association between authoritarian-benevolent leadership and work engagement (b = −0.010, SE = 0.005, 95% CI [−0.020, −0.002], excluding zero). Therefore, Hypothesis 2 was supported.

Table 3 Hierarchical Regression Estimations

Hypothesis 3 posited that at lower levels of trait mindfulness, the negative relationship between LMX ambivalence and work engagement would be stronger. From Model 7 in Table 3, the regression coefficients of the interaction of LMX ambivalence and trait mindfulness were significant (β = 0.159, p < 0.01, ΔR² = 0.024). Figure 2 reveals that the relationship between LMX ambivalence and work engagement was stronger at lower levels of subordinates’ trait mindfulness, which supported Hypothesis 3.

Figure 2 Interaction Effect of LMX Ambivalence and Trait Mindfulness.

Model 14 of the SPSS PROCESS macro was used to test the moderated mediation effect of Hypothesis 4,⁷² which proposed that subordinates’ trait mindfulness moderated the indirect effect of LMX ambivalence on the relationship between authoritarian-benevolent leadership and work engagement. The results showed that the moderated mediating effect was significant (b = 0.024, SE = 0.009, 95% CI = [0.006, 0.040], excluding zero), which indicated that the indirect effect of authoritarian-benevolent leadership on subordinates’ work engagement through LMX ambivalence was moderated by subordinates’ trait mindfulness. Specifically, for subordinates with low trait mindfulness, the indirect effect of authoritarian-benevolent leadership on subordinates’ work engagement via LMX ambivalence was −0.028 (SE = 0.007, 95% CI [−0.042, −0.014], excluding zero). For subordinates with high trait mindfulness, the indirect effect was −0.003 (SE = 0.005, 95% CI [−0.014, 0.008], including zero). The result suggests that with an increase of subordinates’ trait mindfulness, the negative indirect effect of authoritarian-benevolent leadership on work engagement decreases. Therefore, Hypothesis 4 was supported.

Discussion

Based on social information processing theory, this study explored the effect of authoritarian-benevolent leadership on work engagement in constructing a moderated mediation model. The hypotheses proposed were all supported. Firstly, the result of Hypothesis 1 showed that authoritarian-benevolent leadership was negatively associated with LMX ambivalence, which was consistent with previous studies suggesting that subordinates hardly balancing and recognizing leadership inconsistency or leaders’ conflicting behaviors may lead to subordinates’ cognitive dissonance and stress.^13,74 The result provided support for the downsides of ambidextrous leadership and indicated that the inconsistent social information received from authoritarian-benevolent leaders could do harm to subordinates’ emotional and behavioral outcomes as bringing about LMX ambivalence and impairing work engagement. Secondly, the results also shed light on the mediating effect of LMX ambivalence on the relationship between authoritarian-benevolent leadership and work engagement. Existing studies have pointed that the social information from leaders engenders a socially supportive climate for employees, which affects employee outcomes,⁴⁰ such as employee task performance and career advancement.⁷⁵ In line with these findings, focusing on an important employee outcome of work engagement, the result of Hypothesis 2 suggested that LMX ambivalence mediated the effect of authoritarian-benevolent leadership on work engagement. Finally, subordinates’ trait mindfulness was proved to be a valuable moderator, which moderated the effect of LMX ambivalence on work engagement and the indirect effect of authoritarian-benevolent leadership on work engagement through LMX ambivalence. The moderating role of trait mindfulness in organizational research is unclear,⁶⁶ and few empirical studies have been conducted to investigate the moderating role of trait mindfulness on the relationship between authoritarian-benevolent leadership and work engagement. However, this study sheds light on the moderating effect of trait mindfulness, demonstrating its importance in the workplace. Specifically, for subordinates with high levels of trait mindfulness, both the effect of LMX ambivalence on work engagement and the indirect effect of authoritarian-benevolent leadership on work engagement were weaker than when subordinates’ trait mindfulness levels were low.

Theoretical Implications

The findings of this study contribute to the literature in several ways. First, this study analyses the mechanism by which authoritarian-benevolent leadership affects subordinates’ LMX ambivalence and work engagement, which deepens the understanding of authoritarian-benevolent leadership. On the one hand, this study provides a new insight into the dark side of ambidextrous leadership styles, and responds to the call for the study of authoritarian-benevolent leadership style.¹¹ Previous studies have mainly focused on the positive effects of ambidextrous leadership, which are proved to be positive in improving employees’ trust⁷⁶ and innovative behavior,¹¹ but research on the negative effects of authoritarian-benevolent leadership on subordinates’ status and behaviors is still insufficient. Based on social information processing theory, this study proves that authoritarian-benevolent leadership positively influences subordinates’ LMX ambivalence and attenuates subordinates’ work engagement, which enriches the literature of ambidextrous leadership that considers the negative side of authoritarian-benevolent leadership, thus offering a deeper understanding of this research topic. On the other hand, previous research has respectively explored the influence of authoritarian leadership and benevolent leadership on work engagement. The results showed that authoritarian leadership damaged employees’ work engagement since it was taken as a hindrance job demand,^16,77 and benevolent leadership may either improve work engagement when considered as a job resource,⁷⁷ or impair employees’ work engagement since its nature of showing consideration may signal that one could get rid of penalties and reap profits even if they loaf on the job.⁷⁸ However, the impact of authoritarian-benevolent leadership style on followers is less clear.²⁰ This study focused on the unique feature of inconsistency possessed by authoritarian-benevolent leadership style and analyzed this effect based on social information processing theory, which provides a new theoretical basis for future research to explore the impacts of ambidextrous leadership.

Secondly, this research contributes to the existing knowledge by opening the black box of the indirect mechanism of authoritarian-benevolent leadership on work engagement through an important theoretical lens provided by social information processing theory. Specifically, we confirmed that LMX ambivalence, defined as a subjective ambivalent experience of judging both positively and negatively towards leader–follower relationship,²³ mediates the relationship between authoritarian-benevolent leadership and work engagement. Because the relationship between authoritarian-benevolent leadership and work engagement is still not clear, this study provides initial support for LMX ambivalence as a mediator that can explain this link. Meanwhile, the concept of ambivalence has been paid more and more attention in the field of management research,⁷⁹ and the research on the effect of LMX ambivalence is still weak.²⁸ LMX ambivalence is a variable which has raised the attention of scholars in recent years and few research has been done to investigate its antecedents and its influences on employees’ attitudes and behaviors.⁵³ Paradoxes in organizations have always been an important research topic in the academic field of management, and employees experiencing both in a positive and negative way has a strong research significance.⁵⁹ Previous studies have focused on the negative effects of LMX ambivalence on employees’ task performance,²³ pro-organizational behaviors,¹³ and career commitment.⁵³ This study explores the relationship between authoritarian-benevolent leadership, LMX ambivalence and work engagement, and enriches the antecedent and outcome research on LMX ambivalence.

Thirdly, this study uncovered a unique individual character as trait mindfulness in explaining the effect of authoritarian-benevolent leadership on work engagement. Few empirical studies discussed the moderators between leader-follower ambivalent relationships and outcomes.^45,80 And it is not clear that why mindfulness might matter for organizations.⁶⁶ It is important to explore how subordinates with different personal traits vary in their psychological and behavioral reactions to the authoritarian-benevolent leadership style.⁸¹ Based on the social information processing theory, this study demonstrated that mindfulness as an individual trait could moderate the indirect effect between authoritarian-benevolent leadership and work engagement. The findings pointed out that subordinates with higher levels of trait mindfulness might suffer less from the detrimental effect of authoritarian-benevolent leadership and LMX ambivalence. It extends the existing research findings on the boundary conditions for how authoritarian-benevolent leaders exert influence on subordinates and provides some theoretical support for the way to restrain leaders in exerting negative influence on subordinates.

Practical Implications

This study also provides some important practical implications. First, the findings reveal the risk of authoritarian-benevolent leadership damaging subordinates’ work engagement. This conclusion can make organizations more clearly understand the potential damage of this leadership style. Leadership training programs or professional workshops should be arranged, where leaders could enhance their paradoxical mindsets, be encouraged to avoid the random switch between authoritarian and benevolent leadership behaviors, and master more leadership skills. It helps subordinates seize the law of superiors’ managing strategies and make reasonable anticipations. Meanwhile, leaders should be asked to get involved in self‑reflection or feedback seeking from their subordinates to reflect on their leader behaviors and styles. They need to keep in mind the potential harmful impacts of their leadership behaviors on employees’ emotional and behavioral outcomes and take effort to develop high-quality leader-subordinate relationships.

Secondly, it is important for leaders to pay attention to the potential existence of LMX ambivalence. LMX ambivalence is a widely held cognitive experience for employees in the workplace,⁵³ and a useful way to avoid its harmful effect is to make subordinates clarify the intention of leaders.²³ Leaders need to increase their communications with subordinates to make sure the subordinates’ full understanding of the instructions, pay more attention to the psychological changes of subordinates, and timely appease their ambivalent emotions, so as to minimize subordinates’ LMX ambivalence. Organizations could create opportunities for leaders and employees to get along with each other in forms of team-building activities and training processes, so that they have channels to share their thoughts and attitudes to improve the qualities of LMX. In the meanwhile, employees need to be trained to understand that the existence of LMX ambivalence is a common experience,²³ and be provided with stress management programs to maintain in good emotions.

Thirdly, given the detrimental effect of authoritarian-benevolent leadership and LMX ambivalence on subordinates’ work engagement, effective measures are needed to improve employees’ mindfulness. Previous study has proved that a high level of team mindfulness promotes team performance.⁸² For example, organizations could select employees with higher level of trait mindfulness in recruiting by filling out mindfulness questionnaires, or provide employees with mindfulness training programs in forms of breathing, yoga and meditative exercises. These measures help employees maintain a higher level of concentration at work, and reduce the negative impact from the external environment.⁵⁵ In the meanwhile, organizations are suggested to provide employees with suitable places for mindfulness training. For example, they can set a mindfulness moment in the activity rooms, and encourage employees to do mindfulness exercises collectively. Thereby employees can easily practice mindfulness programs to maintain in good working conditions, and effectively avoid interference from the external environment, focus on the tasks, and enhance their work engagement.

Limitations and Future Research

Our research has several limitations and indicates some suggestions for future research. First of all, as cross-sectional designed research, it is hard to avoid common method bias. Although this study did a time-lagged survey and collected data at two different time intervals, variables of LMX ambivalence and work engagement were collected at the same time point, and data were self-reported, which made it difficult to clarify the causality between the variables and to avoid common method bias. Future research may design three-time lag studies, adopt longitude studies from different sources or do lab experiments to mitigate the effects of common method bias. In the meanwhile, the sample coverage of this study is limited. For greater statistical power, future research can enlarge the sample size, and replicate similar studies in different contexts and cultural settings to increase the external validity of the research findings. Secondly, demographic variables including gender, age, education and tenure were controlled following the previous studies.^13,70 While future research may consider other potential control variables such as LMX quality which may influence the subjective ambivalence engendered by authoritarian-benevolent leadership.²³ Thirdly, previous research has shown that paradox is a variable with dual character. The experience of ambivalence is a state of experiencing tensions, and it can influence individuals both in a positive and negative way. In this study, we identified the negative effect of LMX ambivalence on work engagement, while the potential positive outcomes are not discussed, such that scholars often associate experiencing tensions with innovation.⁸³ Future research can consider the “double-edged sword” effect of LMX ambivalence and explore the mechanism of this variable from an integrated perspective. Finally, this study only investigated the moderating role of individual trait mindfulness in the research framework. Future research can explore more individual difference variables as potential boundary conditions. For example, cognitive thinking modes and personalities such as Zhongyong mindset and cynicism might have significant impact on the model. As a traditional Chinese thinking pattern, Zhongyong mindset emphasizes individuals’ thinking in a comprehensive way and emphasizing the interests of the whole, which help ease subordinates’ tensions.⁸⁴ And cynicism is discovered to help employees mitigate the negative experience from LMX ambivalence in the workplace.⁸¹ Examining how these factors may enhance or inhibit the relationship between authoritarian-benevolent leadership and work engagement would provide valuable insights into the existing leadership research field.

Conclusion

Guided by social information processing theory, this study explored how authoritarian-benevolent leadership exerted detrimental effects on subordinates’ work engagement. The results showed that: (1) authoritarian-benevolent leadership had a significantly negative effect on work engagement through leader-member exchange ambivalence; (2) subordinates’ trait mindfulness moderated the effect of leader-member exchange ambivalence on work engagement; (3) subordinates’ trait mindfulness moderated the indirect effect of authoritarian-benevolent leadership on subordinates’ work engagement through leader-member exchange ambivalence. In doing this, the current study advances the literature by discovering the underlying mechanism of how authoritarian-benevolent leadership impairs subordinates’ work outcomes, revealing the boundary conditions that can reduce this negative effect in improving subordinates’ trait mindfulness, and providing empirical evidence for social information processing theory. In practice, the findings provide suggestions for organizations to pay attention to the dark side of authoritarian-benevolent leadership and the improvement of employees’ trait mindfulness. This research will inspire future research in providing a deeper understanding of the relationship between authoritarian-benevolent leadership and subordinates’ psychological and behavioral reactions.