Does Servant Leadership Decrease Bad Behaviors? The Mediating Role of Psychological Safety and the Moderating Effect of Corporate Social Responsibility

Abstract

The aim of this research is to elucidate the intricate dynamics of the effects of servant leadership on employee behavior, particularly focusing on counterproductive work behavior. Drawing on the context–attitude–behavior framework, this study underscores the mediating role of psychological safety in this relationship. Additionally, it unveils the pivotal moderating role of corporate social responsibility (CSR), emphasizing its interaction with servant leadership in influencing psychological safety. To empirically test our hypotheses, we gathered survey data from 394 South Korean workers with a three-wave time-lagged research design. Findings indicate that in contexts with pronounced CSR, servant leadership significantly elevates psychological safety, mitigating negative employee behaviors. Conversely, in organizations with less CSR engagement, these positive effects are attenuated. The results underscore the integral role of CSR in complementing leadership initiatives, advocating for its holistic incorporation into organizational strategies to foster conducive work environments. This research bridges several gaps in the current literature, highlighting the imperative for organizations to intertwine servant leadership with robust CSR endeavors to curtail detrimental employee behaviors. This paper also proposes potential directions for future research in this crucial area.

1. Introduction

In recent decades, the dynamics of leadership have shifted, with servant leadership emerging as a significant focus in the organizational literature. Servant leadership, a term originally coined by Robert K. Greenleaf in his essay “The Servant as Leader” [1], presents a paradigm shift from traditional leadership models. At its core, servant leadership emphasizes the leader’s role as a steward of the resources (human, financial, and otherwise) provided to them. Greenleaf postulated that the primary goal of a leader is not to lead but to serve [2]. Central to this leadership approach is the idea that leaders prioritize the needs, growth, and well-being of their followers over their own self-interests [3]. Instead of wielding power from the top, servant leaders empower their followers, helping them to develop their skills, enhance their capabilities, and achieve their best potential [4,5,6].

In the Republic of Korea (South Korea), the paradigm of leadership has undergone significant evolution, especially as the nation has transitioned from a traditionally hierarchical society to one that is more diverse and globalized. As organizations become more complex and the workforce more diverse, the importance of leadership approaches that emphasize inclusivity, employee well-being, and ethical stewardship become even more paramount. Within this context, the topic of servant leadership has emerged as a pertinent field of study, and understanding its implications can offer insights into fostering a leadership model that aligns with Korea’s contemporary needs.

Traditionally, Korean organizational culture has been characterized by hierarchical relationships, influenced by Confucian values. However, as South Korea has embraced globalization and witnessed generational shifts in values and expectations, there is a growing need for leadership styles that foster empowerment, collaboration, and open communication. Servant leadership, with its emphasis on serving and empowering employees, offers a refreshing counterpoint to top-down authoritative models, resonating with the aspirations of Korea’s younger generation and multicultural workforce. South Korea’s rapid economic development and its status as one of the Four Asian Tigers are nothing short of remarkable. To sustain this growth and navigate the challenges of a globalized economy, businesses need to be agile, innovative, and employee-centric. Servant leadership can facilitate this by fostering an environment where employees feel valued, leading to increased innovation, reduced turnover, and enhanced job performance.

Recognizing the need for sustainable and ethical leadership models, the South Korean government has taken several steps. Institutions of higher learning have been encouraged to incorporate leadership training modules in their curricula. Seminars, workshops, and courses on servant leadership are becoming commonplace in universities and training centers. The government has initiated policies that emphasize ethical corporate governance, transparency, and employee welfare. Such a policy framework indirectly promotes leadership models that align with servant leadership principles. Within public sector institutions, there is a growing emphasis on servant leadership as a means to foster a service-oriented ethos, improve public service delivery, and enhance trust in government institutions. The South Korean government has been actively promoting corporate social responsibility (CSR) initiatives, which often align with servant leadership principles. Companies are increasingly recognized and rewarded for their ethical standards, community engagement, and employee welfare initiatives.

Previous academic works have expanded on Greenleaf’s foundational ideas, suggesting that servant leadership results in positive outcomes, such as increased job performance, elevated organizational commitment, work engagement, organizational identification, job satisfaction, enhanced team spirit, and organizational performance [4,5]. These benefits arise due to a work environment marked by trust, mutual respect, and shared goals [5,6]. One of the distinctive aspects of servant leadership is its focus on community building and ethical decision making. To be specific, servant leaders strive for the betterment of the larger community, aiming for long-term societal impacts beyond just organizational gains [6,7].

Although the advantageous outcomes of servant leadership are well documented, a comprehensive understanding of its broader implications remains a research gap in servant leadership studies. First, the existing literature often concentrated on the positive outcomes associated with servant leadership. Yet, its potential influence on deterring negative behaviors, such as counterproductive work behavior (CWB), remains inadequately examined. Given the organizational consequences of such behaviors, it is imperative to understand how leadership paradigms can mitigate them. CWB, characterized by intentional employee actions that can detrimentally impact the organization or its members, is of paramount importance, as it can erode organizational culture, trust, and overall performance [8,9,10]. Examples of CWB includes harassment, bullying, or aggression towards colleagues, theft, sabotage, or misuse of resources [9,10]. Despite its organizational significance, the nexus between servant leadership and such adverse behaviors remains underexplored, rendering our first research gap.

Second, beyond the direct relationships, the mechanisms underlying the association between servant leadership and employee behaviors are not extensively probed. The potential mediators, like psychological safety, and their interplay with organizational-level moderators, such as corporate social responsibility (CSR), are notably absent from comprehensive examinations. Deepening the gap is the scant attention afforded to the underlying processes and contextual factors that mediate and moderate the servant leadership–CWB link. Investigating these mediators and moderators is critical for a holistic comprehension of the intricate interplay in the association between servant leadership and CWB [4]. A nuanced understanding can not only delineate the direct effects but also unveil the conditions under which, and the mechanisms through which, servant leadership might curtail or inadvertently foster CWB. This absence marks our second research gap.

Third, although leadership studies have delved into individual and team dynamics, the moderating role of broader organizational factors, especially CSR, has not been sufficiently emphasized. Given the rising importance of CSR in contemporary organizational landscapes, understanding its interaction with leadership styles and subsequent outcomes is paramount [4,5], and in that discussion, corporate social responsibility (CSR) emerges as a salient factor. The strategic role of CSR in shaping organizational identity, stakeholder relations, and overall business sustainability is indisputable [11,12,13]. Notwithstanding its significance, the extant literature on servant leadership seems to have relegated CSR to the periphery, overlooking its potential as a pivotal moderating variable. This underemphasis on CSR, especially in the context of servant leadership’s impact on employee behavior, constitutes our third research gap.

Against this backdrop, this study sets out with the following objectives:

• To explore the influence of servant leadership on counterproductive work behavior.
• To examine the mediating role of psychological safety in the relationship between servant leadership and counterproductive work behavior.
• To investigate the moderating effect of corporate social responsibility in the aforesaid relationships.

Driven by these objectives, our research seeks answers to the following questions:

• How does servant leadership impact employee counterproductive work behavior?
• Does psychological safety mediate the relationship between servant leadership and counterproductive work behavior?
• How does corporate social responsibility moderate the relationship between servant leadership, psychological safety, and counterproductive work behavior?

This research underscores the indispensable role of corporate social responsibility. By bridging the highlighted gaps, we aim to offer a more nuanced understanding of the servant leadership paradigm, the significance of psychological safety, and the overarching role of CSR, thereby enhancing both theoretical and practical implications in leadership studies.

In endeavoring to bridge these gaps, we aimed to foster a richer, more comprehensive understanding of the relationship between servant leadership and counterproductive work behavior by investigating the mediating effect of psychological safety as well as the moderating effect of corporate social responsibility. Psychological safety refers to a shared belief held by members of a team or group that the team is safe for interpersonal risk taking, and it captures a sense of confidence that the team will not embarrass, reject, or punish someone for speaking up [14,15,16,17], including the belief that he or she will not experience punishment or humiliation from his or her leader or coworkers for speaking up with ideas, questions, concerns, or mistakes in the organization [17,18,19].

CSR, meanwhile, is a multifaceted construct that has garnered significant attention in academic and business circles over the past few decades. At its core, CSR encapsulates the voluntary activities undertaken by a corporation to operate in an economically, socially, and environmentally sustainable manner [11,12,13,20]. These activities go beyond the economic and legal obligations of the firm to emphasize ethical and philanthropic duties towards various stakeholders [21,22,23,24].

To be specific, we suggest that high servant leadership will increase employee psychological safety and that high psychological safety will diminish employee CWB. In other words, psychological safety will mediate the association between servant leadership and CWB. In addition, we expect CSR to play a role as a positive moderator that amplifies the enhancing effect of servant leadership on psychological safety.

The realm of organizational behavior and leadership has witnessed expansive research, yet the dynamics of servant leadership, especially in the context of certain moderating and mediating factors, remains an area rich for exploration. The unique contributions of this paper can be understood through the following perspectives. First, this paper integrated multiple constructs in a holistic manner. While individual studies have explored the facets of servant leadership, corporate social responsibility (CSR), psychological safety, and counterproductive work behavior (CWB), few have ventured into examining the interplay among these constructs. This research bridges this gap by offering a comprehensive model that elucidates the intricate relationships among them, providing a more holistic understanding of the underlying dynamics.

Second, the current paper applied leader–organization fit theory. By anchoring our hypotheses in the leader–organization fit theory, this paper introduces a novel theoretical framework to interpret the role of CSR as a moderator in the servant leadership dynamic. This innovative application not only offers a fresh theoretical lens to view leadership dynamics but also underscores the symbiotic relationship between individual leadership styles and the overarching organizational ethos.

Third, given the globalized nature of modern business, understanding leadership dynamics in diverse cultural contexts, like that of the Republic of Korea, is pivotal. By situating our research in such a context, this paper enriches the literature on servant leadership by offering insights that may be culturally nuanced and unique, further highlighting the importance of context in leadership studies.

Fourth, beyond theoretical contributions, the methodological rigor of this paper, characterized by stratified random sampling, multiple data collection points, and robust statistical analyses, ensures the empirical validity of our findings. Such methodological precision is pivotal in reinforcing the credibility and generalizability of our conclusions.

Lastly, while academic contributions are central, the pragmatic implications of our findings are equally salient. By highlighting the importance of CSR in amplifying the positive effects of servant leadership on psychological safety, this paper offers actionable insights for organizations aiming to foster healthier work environments and minimize CWB. Such implications are invaluable for organizational leaders, HR practitioners, and policy makers.

In essence, this paper does not merely add to the voluminous literature on leadership; it carves out a distinct niche by weaving together multiple constructs, grounding them in a unique theoretical framework and presenting them through a rigorous empirical lens. These contributions, we believe, pave the way for further scholarly exploration and offer tangible insights for real-world organizational challenges.

To comprehensively investigate the intricate dynamics underlying servant leadership, corporate social responsibility (CSR), psychological safety, and counterproductive work behavior (CWB), this study adopted a mixed-methods approach. Employing stratified random sampling, we collected data at three distinct time points to ensure temporal separation and minimize common method bias. Quantitative analyses, encompassing moderated mediation models with structural equation modeling (SEM), were conducted to validate the proposed relationships. We utilized survey data from 394 South Korean workers with a three-wave time-lagged research design. Also, this research stands at the forefront of leadership studies by introducing several novel contributions to the field. First, while individual dimensions like servant leadership or CSR have been explored in isolation, our study’s holistic approach, weaving them into a singular coherent framework, offers a richer understanding of their interplay. Second, by anchoring our hypotheses within the leader–organization fit theory, we have expanded its traditional application, underscoring the pivotal role of CSR as an organizational attribute that amplifies the effects of individual leadership styles. Third, rooting our research within the cultural milieu of the Republic of Korea offers a fresh perspective on leadership dynamics, emphasizing the importance of contextual nuances in influencing leadership outcomes. Lastly, beyond the theoretical realm, our findings have tangible real-world implications, highlighting actionable strategies for organizations to enhance psychological safety and reduce CWB by fostering servant leadership and emphasizing CSR.

2. Theory and Hypotheses

2.1. Servant Leadership and Counterproductive Work Behavior

We propose that servant leadership will decrease the degree of employee CWB. Servant leadership, a concept introduced by Greenleaf [1,2] and later expanded upon by researchers such as Spears and Lawrence [25], focuses on prioritizing the needs and development of followers. In the current research, we suggest that servant leadership can potentially impact counterproductive work behavior, which is characterized by deliberate actions by employees that are harmful to the organization and its members [26]. Counterproductive work behavior (CWB) encompasses a spectrum of intentional behaviors exhibited by employees that are harmful to the organization or its members. Such behaviors stand in contrast to desired organizational norms, values, and efficiency and have implications for both individual and collective productivity and well-being [8]. To be specific, the theoretical linkage between servant leadership and CWB primarily can originate from social exchange theory [27] and social learning theory [28].

First, social exchange theory posits that human interactions are driven by a series of cost–benefit analyses wherein individuals anticipate reciprocity [27]. According to this theory, interactions between individuals can lead to obligations. Within an organizational setting, employees often gauge the treatment they receive from their superiors and reciprocate through their behaviors and attitudes. In the context of servant leadership, leaders who prioritize the needs of their subordinates can foster an environment of trust and reciprocity [27]. When a servant leader prioritizes his or her followers’ needs, the followers may feel obliged to reciprocate the leader’s favors through positive work behaviors, potentially reducing CWB [29]. Servant leadership, emphasizing interpersonal healing, empathy, and commitment to followers’ personal growth [25], fosters a positive social exchange environment and high-quality leader–member exchange relationships [30], potentially reducing the likelihood of CWB.

Second, social learning theory [28] offers another theoretical lens through which the relationship between servant leadership and CWB can be viewed. Bandura’s [28] social learning theory emphasizes the role of observational learning, arguing that individuals learn behaviors by watching others, especially influential figures. In an organizational paradigm, leaders often serve as role models, and their actions and behaviors provide cues for subordinates. According to the theory, employees learn not only from their own experiences but also by observing others, especially individuals in positions of authority. By observing others, especially those they consider role models, an individual member can learn group behaviors [28]. Servant leaders, by demonstrating high moral and ethical standards, empathy, and stewardship, serve as role models for their employees. Then, the followers are likely to learn and conduct appropriate workplace behaviors in an organization. In other words, this modeling can result in decreased CWB as employees mirror their leaders’ behavior.

While both the social exchange and social learning theories provide foundational understanding, there exists a lacuna in the literature bridging these theories with constructs like corporate social responsibility (CSR) and counterproductive work behavior (CWB). CSR, as an organizational-level variable, may interact differently within the leadership-behavior dynamic, a realm that remains underexplored. Moreover, the exact mechanisms—be they mediational or moderational—through which servant leadership influences CWB, especially in the presence of variables like psychological safety, remain nebulous.

To fortify our theoretical underpinnings, an expansive review of the literature addressing these gaps is imperative. Such a review not only bolsters the validity of our proposed hypotheses but also situates our study within a richer theoretical context, paving the way for more nuanced findings and implications.

Although previous researchers paid less attention to the influence of servant leadership on CWB in a direct way, extant research suggested empirical evidence to support the theoretical linkage. For instance, Mayer, Aquino, Greenbaum, and Kuenzi [31] found that ethical leadership can reduce employees’ negative behaviors. Although Mayer and his colleagues did not investigate the relationship between servant leadership and CWB in a direct way, considering that ethical leadership and servant leadership have in common core characteristics [31], we can expect that servant leadership will influence employee negative behavior such as CWB. Based on the above arguments, we proposed the following hypothesis.

Hypothesis 1:

Servant leadership will decrease employee counterproductive work behavior.

2.2. Servant Leadership and Psychological Safety

In the current paper, we propose that servant leadership will enhance employee psychological safety. Psychological safety, a key construct, was introduced to the organizational literature by Amy Edmondson in the late 1990s. It refers to an individual’s perceptions of the consequences of taking an interpersonal risk [15,16,17]. In essence, psychological safety is a belief that an employee will not face punishment or humiliation by leaders or colleagues for speaking up with ideas, questions, concerns, or mistakes at work.

In the context of teams and organizations, psychological safety plays a pivotal role in fostering a culture where individuals feel secure and free to voice their opinions and share their concerns without fear of retribution [17,19]. In essence, psychological safety pertains to a belief that one will not face punishment or humiliation for speaking up with ideas, questions, concerns, or mistakes. In a psychologically safe team, members feel accepted and respected. They are comfortable being themselves, sharing ideas, concerns, mistakes, and questions without fear of retribution or damaging their status, career, or reputation. The atmosphere within such a team is characterized by interpersonal trust and mutual respect in which people are comfortable being themselves [14,15,16,17]. Such environments have been linked to increased innovation, more adaptive learning, and enhanced performance in organizations [15]. Recent works elucidated the broader implications of psychological safety. Beyond just fostering communication, it serves as a foundation for other critical organizational outcomes, including organizational commitment, organizational identification, job satisfaction, work engagement, knowledge sharing, creativity, learning behavior, risk taking, and innovation [15,16,17,18,32,33]. Among various antecedents of psychological safety, leaders have been known to play a critical role in cultivating psychological safety [15,17]. Their behaviors, including expressing vulnerability, actively seeking feedback, and appreciating candid input, can influence the degree of psychological safety experienced by their subordinates [34].

First, to be specific, we suggest that servant leadership will increase employee psychological safety based on social exchange theory [27], which, as we discussed above, indicates that servant leadership will enhance psychological safety. Central to the tenets of the social exchange theory is the principle of reciprocity in relationships, wherein interactions are driven by a process of give-and-take, guided by expectations of returned benefits [27]. Within the organizational realm, leader–follower dynamics epitomize this process. Servant leadership, with its inherent emphasis on addressing the needs and fostering the personal development of followers, establishes a uniquely supportive environment. In this milieu, employees experience a genuine sense of care and acknowledgment from leaders, prompting them to reciprocate with trust and respect [15,18]. As empirical studies have elucidated, such a dynamic underpinned by reciprocal trust fosters a heightened sense of psychological safety [30]. This reciprocal interaction can lead to a sense of psychological safety among employees, fostering open communication, innovative thinking, and risk taking without fear of reprisal [30]. This environment encourages open communication, innovative thought processes, and a willingness to take risks without apprehensions of negative repercussions.

Second, a servant leader, by nature, encourages open dialogue, values diversity of thought, and acknowledges the vulnerability and imperfections inherent in human interactions. This supportive and inclusive leadership style fosters an environment where employees feel recognized, valued, and safe to express their views without trepidation [35]. Such leaders cultivate an ecosystem wherein employees feel not just seen but genuinely recognized and valued. In this context, the freedom to articulate views without hesitation or fear becomes a natural consequence, as evidenced by empirical research [35]. Moreover, when leaders embrace open channels of communication, understand the nuances of vulnerability, and champion diverse viewpoints within teams, psychological safety is invariably augmented [36].

Lastly, social learning theory [28] also provides a theoretical underpinning for this hypothesis. Propounded by Bandura [28], the social learning theory articulates that behaviors are acquired, predominantly, through observational learning. Individuals, especially in hierarchical structures like organizations, tend to emulate those they view as role models. Servant leaders, through their embodiment of empathy, ethical grounding, and unwavering support, inevitably position themselves as such influential exemplars. These observed behaviors, when internalized by followers, not only shape individual attitudes but collectively contribute to an ambiance replete with psychological safety. As we discussed above, this theory posits that individuals learn behaviors by observing others, particularly those they perceive as role models, and under the theory, servant leaders, through their empathetic, ethical, and supportive behaviors, serve as influential models for their followers. Followers who internalize these behaviors from their role models contribute to cultivating an environment of psychological safety.

In synthesizing the above, it becomes evident that servant leadership, through its foundational principles and behaviors, fosters psychological safety within organizational settings. The intertwined threads of social exchange theory and social learning theory provide a robust theoretical foundation for this proposition, further substantiated by a breadth of empirical evidence. Based on the above arguments, we suggest the following hypothesis.

Hypothesis 2:

Servant leadership will increase employee psychological safety.

2.3. Psychological Safety and Counterproductive Work Behavior

We propose that employee psychological safety will decrease CWB. CWB encompasses a spectrum of intentional employee behaviors that are harmful to the organization or its members. Such behaviors stand in contrast to desired organizational norms, values, and efficiency and have implications for both individual and collective productivity and well-being [8,9,10]. Broadly categorized, CWB can be directed toward individuals (CWB-I) or the organization (CWB-O). Examples of CWB-I include harassment, bullying, or aggression towards colleagues, whereas CWB-O includes behaviors such as theft, sabotage, or misuse of resources [26].

Various antecedents have been identified for CWB, including organizational injustice, job dissatisfaction, role ambiguity, and perceived violation of the psychological contract. Additionally, individual factors such as personality traits, especially traits from the dark triad (narcissism, Machiavellianism, and psychopathy), have been linked to higher propensities for CWB [8,9,37,38,39,40,41].

In the current research, we propose that psychological safety will decrease employee CWB. A psychologically safe environment, by its essence, can act as a buffer against the emergence of CWB. When employees feel they can voice concerns without facing retaliation, they are less likely to resort to covert actions that harm the organization or its stakeholders. A climate of psychological safety provides avenues for employees to constructively address grievances, express frustrations, and seek resolutions [8,38,41]. Research supports this theory, showing that psychologically safe environments correlate with reduced instances of CWB [8,41]. Such environments diminish the perceived need for retaliatory or defensive behaviors by providing legitimate channels for issue resolution and ensuring that employee voices are both heard and valued.

The theoretical underpinning of this hypothesis can be traced back to two primary theories: cognitive dissonance theory [42] and social learning theory [28]. First, according to cognitive dissonance theory, individuals strive for internal consistency within their cognitions. When there is an inconsistency or conflict between attitudes or between beliefs and actions, it results in a state of discomfort known as cognitive dissonance [42]. Individuals are then motivated to reduce or eliminate this dissonance to restore cognitive equilibrium. In an organizational context characterized by high levels of psychological safety, employees tend to feel respected, valued, and accepted. Engaging in counterproductive work behaviors under these conditions would create a conflict between the individual’s actions (negative behaviors) and their prevailing positive environmental perceptions (supportive, respectful workplace). This incongruence induces cognitive dissonance. Based on the foundational principles of the theory, employees would be internally motivated to resolve this dissonance. One straightforward resolution would be to refrain from engaging in CWB, thereby realigning actions with the prevailing positive work environment [42]. In an environment of high psychological safety where employees feel respected and accepted, engaging in CWB would create a cognitive dissonance that would need to be resolved, which argues against such behavior [8,38,41,42]. Numerous empirical studies, such as those cited [8,38,41,42], have highlighted the behavioral adjustments individuals make in response to cognitive dissonance, further validating this rationale.

Second, Bandura’s [28] social learning theory underscores the significance of observational learning, suggesting that individuals predominantly acquire behaviors by observing and emulating the actions of others within their environment. Role models, or those in influential positions, play a pivotal role in shaping behaviors through this observational process. Within psychologically safe environments, where respect, trust, and open communication are normative, employees are consistently exposed to positive behavioral demonstrations by both leaders and peers. Following the tenets of social learning theory, individuals within such settings are inclined to observe and reproduce these positive behaviors, which align with the environment’s constructive norms. The manifestation of CWB would contradict the observed positive behaviors, making its occurrence less likely. Empirical research supports this theory-based argument. Studies [9,41] indicate that positive behavioral norms set by leaders and coworkers significantly deter counterproductive behaviors. Instead, they encourage behaviors that bolster collaboration and constructive conflict resolution.

In conclusion, the integration of cognitive dissonance theory and social learning theory offers a robust theoretical foundation to anticipate the inverse relationship between psychological safety and counterproductive work behavior. The convergence of these theoretical frameworks, complemented by empirical evidence, underscores the pivotal role of organizational environment and leadership in shaping employee behaviors. These arguments lead us to the following hypothesis.

Hypothesis 3:

Employee psychological safety will decrease counterproductive work behavior.

2.4. The Mediating Role of Psychological Safety in the Servant Leadership–CWB Link

One of our main goals with the current research was to examine the impact of servant leadership on counterproductive work behavior through the mediating influence of employee psychological safety. This mediation model relies on the context–attitude–behavior (CAB) framework. This framework is an integrative model rooted in social psychology, positing a sequential relationship between contextual factors, individual attitudes, and the subsequent behaviors they motivate. Scholars have employed the framework to shed light on various phenomena in fields ranging from consumer behavior to organizational studies. The ensuing elucidation delves into the core tenets and applications of this theoretical perspective [43,44,45].

Within the CAB paradigm, context is conceived as the totality of external factors and circumstances encompassing an individual. This includes a wide array of factors, from immediate environmental conditions to broader sociocultural and economic structures. Context acts as the initial stimulus, influencing individuals’ perceptions, feelings, and thought processes.

As a direct offshoot of contextual interpretations, attitudes within the CAB framework refer to an individual’s evaluative judgments, encompassing cognitive and affective assessments. These judgments, whether favorable or unfavorable, are predicated on the person’s interpretation of the contextual stimuli. As posited by Eagly and Chaiken [44], attitudes serve as mediators, bridging the external environment with behavioral responses.

The culmination of the CAB sequence, behavior is the tangible action or set of actions exhibited by an individual in response to their attitudes. This segment of the framework encapsulates the transition from internal evaluative processes to external manifestations. The work of Ajzen [43] on the theory of planned behavior underscores this, asserting that attitudes, when combined with subjective norms and perceived behavioral control, determine the behavioral intentions, leading to actual behaviors.

Within the CAB framework, context is perceived as the backdrop against which individual perceptions and evaluations are formed. In organizational settings, leadership styles can significantly modulate this contextual milieu. Servant leadership, characterized by its focus on serving and empowering subordinates [1,2,5], serves as a contextual antecedent engendering certain attitudes and perceptions among employees. It emphasizes empathy, active listening, and commitment to the growth and well-being of others.

An individual’s perception of psychological safety in a workplace can be deemed as an attitudinal response to the aforementioned contextual stimuli. Within the confines of the CAB framework, psychological safety emerges as the cognitive–affective evaluation of the context, in this case, the servant leadership style. When leaders serve and empower, it can foster a sense of trust and respect among employees, nurturing an attitude of psychological safety [44].

Lastly, counterproductive work behavior, which encompasses any intentional actions by employees that can harm the organization or its members (Robinson and Bennett, 1995), represents the behavioral outcome in the CAB framework. Based on the premise that attitudes can predict behaviors [43], enhanced psychological safety, cultivated by servant leadership, can potentially mitigate such detrimental behaviors. Given the above, we suggest employee psychological safety as a mediator in the relationship between servant leadership and CWB.

Hypothesis 4:

Employee psychological safety will mediate the relationship between servant leadership and counterproductive work behavior.

2.5. The Moderating Influence of CSR in the Servant Leadership–Psychological Safety Link

Based on leader–organization fit theory [46,47,48], we emphasize the importance of interaction between servant leadership and corporate social responsibility. CSR will function as a crucial moderator that amplifies the positive effect of servant leadership on employee psychological safety. The evolution of CSR as a concept can be traced through a series of developmental stages. Initially conceived as a means for businesses to contribute charitably to their communities [11,13], the construct has evolved to encompass a strategic dimension wherein firms integrate social and environmental considerations into their business models and core strategies [21,22]. This strategic CSR perspective postulates that firms can simultaneously achieve societal and economic goals, thereby aligning business interests with broader societal concerns [21,22,23,24].

One of the key debates within the CSR literature centers on the business case for CSR. While some argue that CSR initiatives can lead to competitive advantage and are ultimately in the best interests of shareholders [21,23], others caution against viewing CSR purely through a strategic lens, emphasizing the inherent value of ethical considerations and societal well-being [21,22,23,24].

CSR, at its core, signifies a business approach that contributes to sustainable development through delivering economic, social, and environmental benefits for all stakeholders. However, this overarching definition, as simple as it seems, is layered with complexities that derive from its various facets.

First, CSR primarily functions as an organizational policy, tailored to align with a company’s vision, stakeholders’ expectations, and market demands. Though sometimes it takes the form of voluntary self-regulation, in many jurisdictions, facets of CSR have been incorporated into legislative frameworks, making certain aspects mandatory. Unlike transnational law, which operates across borders, organizational CSR policies often reflect a delicate balance between global best practices and local expectations.

Second, the delineation between mandatory and voluntary CSR is fluid, often contingent on jurisdictional regulations, industry standards, and stakeholder pressures. For instance, while in some regions CSR reporting might be a mandatory requirement for firms, in others, it is a voluntary gesture signaling a commitment to ethical operations. This dichotomy shapes how firms strategize, implement, and communicate their CSR initiatives.

Third, the motives underpinning CSR initiatives can vary widely. On one end of the spectrum, CSR is viewed as corporate philanthropy—donations, community upliftment, and charitable ventures. On the opposite end, it is about improved environmental management, sustainable operations, and a long-term vision for resource optimization. The former leans towards an economic agenda, often driven by immediate PR goals and stakeholder appeasement. The latter, however, aligns more closely with social and environmental agendas, aiming for lasting positive impacts.

Fourth, while the business case for CSR emphasizes its potential to drive profitability, enhance brand reputation, and foster stakeholder trust, it is essential to understand that this business case does not undermine the genuine ethical imperatives behind CSR. Companies can, and often do, derive tangible business benefits from CSR, but these benefits are outcomes of authentic socially and environmentally responsible actions.

In contextualizing CSR within our study, it is imperative to underline that our conceptualization of CSR, as delineated above, is rooted in its multifaceted nature. This comprehensive understanding aids in constructing our variable more robustly and offers a clearer interpretative framework for our results. By situating our research within this nuanced understanding of CSR, we hope to elucidate the intricate interplay between servant leadership, CSR, and employee behaviors.

According to leader–organization fit theory, the congruence or alignment between leadership characteristics and organizational attributes has significant implications for both leader and organizational outcomes. At its core, the theory is grounded in the broader person–environment fit literature, in which individuals thrive in environments where their personal attributes align with the contextual attributes of their environment [46,47].

In the context of leadership, this means that leaders are more likely to be effective and satisfied in their roles when their values, goals, and strategies are in harmony with those of the organization. At the most foundational level, fit can be assessed in terms of value congruence. Leaders whose personal values align with the values of the organization are more likely to exhibit behaviors consistent with organizational norms and expectations and in turn receive more support from followers and stakeholders [46].

Beyond values, fit can also manifest in terms of strategic alignment. Leaders whose strategic visions and orientations match the organization’s strategic needs and directions can more effectively drive organizational success. Additionally, the way leaders perform their roles, their leadership styles, and the behavioral expectations of the organization must also be congruent. Dissonance here can lead to role conflict, reducing the leader’s effectiveness and potentially undermining organizational outcomes. A strong leader–organization fit has been linked to various positive outcomes including increased job satisfaction for leaders, greater leader tenure, more trust and commitment from followers, and overall enhanced organizational performance [48].

By applying this theoretical perspective to the context of servant leadership, CSR, and psychological safety, we propose that CSR can function as a moderating factor, amplifying the positive relationship between servant leadership and psychological safety.

Given the service-oriented nature of servant leadership and the ethical underpinnings of CSR, one can posit an inherent synergy between these constructs. When organizations deeply embed CSR in their operations, strategies, and values, the behaviors and values exhibited by a servant leader can resonate more profoundly with employees. In such a milieu, servant leadership’s emphasis on employee growth, community development, and well-being is likely to strongly resonate with a shared concern for broader societal welfare, ethical considerations, and employee well-being [21,22,47]. As described above, this alignment can in turn foster an environment where employees feel safe to voice concerns, suggest innovations, and admit mistakes without fear of retribution [15,18]. For instance, a sustainable green company that promotes community involvement and environmental responsibility will likely find that its servant leaders generate a sense of psychological safety in their employees as employees see congruence between what the leader practices and what the organization preaches. Employees in such an environment not only might feel supported by their leader but also might perceive themselves as part of a larger, purpose-driven initiative that aligns with their personal and societal values [13,46].

Conversely, in organizations with a weak commitment to CSR, the effects of servant leadership on psychological safety might be diluted. Even if individual leaders genuinely prioritize the well-being of their employees, a lack of organizational alignment with broader societal concerns or a broader organizational ethos that might seem to prioritize profit over ethical or societal considerations can create a perception of dissonance [13,22,47]. For instance, in a firm that has recently faced public criticism for unethical practices, even a well-intentioned servant leader can struggle to foster a sense of psychological safety. Employees might question the genuine nature of the leader’s intentions, wondering if such servant leadership behaviors are merely a facade in light of the organization’s apparent disregard for broader ethical concerns.

In summary, the theoretical foundation of leader–organization fit theory and the empirical evidence from the extant literature illuminate the intricate relationships between servant leadership, CSR, and psychological safety. The alignment between servant leadership and an organization’s commitment to CSR can serve as a significant amplifier for fostering an environment of psychological safety, further highlighting the importance of both leadership and organizational strategies in shaping employee outcomes. Based on the above arguments, we propose the following hypothesis.

Hypothesis 5:

Corporate social responsibility will positively moderate the increasing effect of servant leadership on psychological safety.

Figure 1 shows our theoretical model and the hypotheses suggested in this section.

3. Method

3.1. Participants and Procedure

The research cohort for this study encompassed working professionals aged above 20 and below 59 from various South Korean corporations. Recruitment was facilitated through a leading online research entity boasting an extensive participant pool of roughly 4,250,000 registrants, and surveys were administered to participants in three separate batches. The research entity collects individuals’ occupations during registration with the panel as well collecting contact details (mobile numbers, email addresses). Online surveys are recognized as an efficacious strategy for securing a heterogeneous participant sample [49]. Participants for this study were recruited through a reputable online research entity that specializes in sourcing individuals for academic research purposes. This platform maintains a vast database of potential participants who have voluntarily registered, expressing their interest in participating in various research projects. These were primarily collected to provide a direct and immediate mode of communication with participants. This allowed for the facilitation of any real-time clarifications, scheduling of reminders, and offering of prompt assistance during the course of this study. Email addresses served a dual purpose. First, they were used to send participants the necessary study materials, guidelines, and online questionnaire links. Secondly, they acted as a medium for follow-up communications, thank-you notes, and the sharing of summarized research findings, upon participants’ interest.

An integrated response-time tracker was embedded within the online survey tool. This mechanism flagged instances where participants completed the survey in a duration deemed too swift for thoughtful and genuine responses. Such entries were subjected to further scrutiny to uphold the data quality. Before commencing this study, participants were presented with an informed consent form. This form clearly outlined this study’s objectives, the estimated time required for participation, and the entirely voluntary nature of their involvement. They were informed of their right to withdraw from this study at any stage without any consequences.

Assurances regarding the confidentiality of participants’ personal and response data were emphasized. We communicated explicitly that all collected data would be anonymized, stored securely, and only used for research purposes. Individual identities would remain undisclosed in any published work, with findings being reported in aggregate form only.

Additionally, the rationale behind collecting contact details was elucidated, ensuring participants that these would neither be shared with third parties nor used for unsolicited communications. In adopting these comprehensive measures, our aim was to foster an environment of trust, ensuring that participants felt secure and informed throughout their involvement in this study.

We administered the survey and collected the data three separate times with the intention of circumventing the inherent limitations of cross-sectional data collection. The digital infrastructure of this system empowered the research team to meticulously monitor and ascertain consistent participation of respondents over these timelines. Survey engagements were scheduled at intervals spanning five to six weeks and remained accessible for a span of two to three days, granting ample opportunity for participants to render their feedback. Integral to the data collection process, the research entity implemented rigorous mechanisms to deter geo-IP discrepancies and identify anomalously swift responses. To ensure the authenticity of participants and deter discrepancies related to geographical inconsistencies, our online questionnaire platform was integrated with a geo-IP verification tool. This tool allowed us to verify the geographical location of participants at the time of their response, ensuring alignment with our targeted demographic region.

Outreach efforts to solicit survey participation were directly managed by the research entity. Prospective participants were provided assurances of the voluntary nature of their involvement, with a commitment to maintaining the confidentiality of their submissions, restricted solely to research pursuits. Those who opted to participate were duly enlightened regarding this study’s parameters, and their explicit consent was acquired, ensuring full adherence to ethical standards. A monetary incentive ranging between USD 9 and 10 was proposed as a token of gratitude for their contribution.

To attenuate potential sampling distortions, the research entity deployed stratified random sampling. This strategy entailed random participant extraction from each predefined category, thereby attenuating biases potentially arising from demographics, professional standing, academic credentials, or industry affiliation. Through intricate online tracking mechanisms, the entity ensured consistent participation of the same respondents across all three data collection phases. Stratified random sampling is a method in which the population is divided into homogenous subgroups, known as strata, and then samples are randomly selected from within these subgroups. This sampling technique aims to capture the variability within each stratum, ensuring that each subgroup is adequately represented in the sample, enhancing the generalizability and precision of results. In our study, the population of professionals was divided into distinct strata based on predetermined criteria such as job roles, years of experience, and industry sectors. Poststratification, an equal number of professionals were randomly selected from each stratum to ensure a balanced representation across all the defined categories.

In the inaugural data collection phase, 717 professionals responded, the subsequent phase yielded 545 responses, and the final juncture produced 397 responses. Postcollection, the dataset underwent a purification process, wherein incomplete responses were excised. The culminating dataset deemed fit for research analysis thereby comprised 394 respondents who provided comprehensive responses throughout the three survey stages, yielding a response efficacy of 54.95%. During the first phase of data collection, a total of 717 professionals responded. This higher response rate can be attributed to the novelty of this study and the initial enthusiasm of participants. However, second and third data collection points registered a decline in the number of responses. Several factors can be attributed to this diminishing trend. First, there is a common phenomenon in longitudinal studies where participants drop out or choose not to continue in later phases. Second, professionals, given their busy schedules, might have found it challenging to allocate time for subsequent phases. Third, the initial novelty and eagerness might have waned over time. During the data purification process, our primary objective was to ensure the quality and reliability of the dataset. To this end, we established certain criteria to identify and exclude incomplete or unsatisfactory responses. First, any response that lacked answers for more than 10% of the questionnaire items was deemed incomplete and subsequently excluded. Second, as mentioned earlier, our integrated response-time tracker flagged entries completed in an unusually swift manner. Such entries, upon review and deemed lacking in thoughtful engagement, were excluded. Third, the questionnaire incorporated certain redundant questions, presented differently, to check for consistency in participants’ answers. Discrepancies in these answers indicated potential issues with the reliability of the response. Fourth, some open-ended questions were included to gauge participants’ understanding and engagement with the survey topics. Responses that offered generic, nonspecific, or irrelevant feedback to these questions were subjected to further scrutiny. By applying these stringent purification criteria, we aimed to ensure the reliability and validity of the dataset, fostering robust conclusions and interpretations in the subsequent analytical stages.

To enhance transparency and provide a clear chronology of the research process, a detailed account of the survey administration schedule is presented below. First, the initial phase of the survey was administered on 15 January 2023. As the first point of contact, participants received an introductory email explaining the purpose and significance of this study, inviting them to partake in the survey. This phase was crucial, as it set the tone and established the foundational data against which subsequent responses would be compared. Second, following the initial data collection, the second phase of the survey was administered six weeks later, on 2 March 2023. This interval was purposefully chosen to allow participants ample time to potentially experience changes in their work environment, attitudes, or behaviors—changes that the survey aimed to capture and analyze. It was also a practical duration to reduce participant fatigue and avoid overwhelming them with frequent survey requests. Third, the final phase of data collection took place on 7 April 2023, approximately 3 months after the inaugural survey. This three-month span between the first and last data collection points was designed to capture longitudinal changes and trends over an extended period. Moreover, it provided a symmetrical structure to the data collection process, facilitating the analysis of seasonal influences or cyclic trends in the responses. To ensure optimal response rates, reminder emails were sent to participants two weeks before each survey phase and once more two days before the closing date for submissions. Each reminder reiterated the importance of their continued involvement, the voluntary nature of participation, and the confidentiality protocols in place. In conclusion, the structured and systematic schedule of survey administration not only enabled the collection of robust data at different time points but also demonstrated our commitment to methodological rigor and participant engagement throughout the research process.

This sample size was determined by past scholarly recommendations, involving considerations like the G*Power statistical evaluation for optimal sample size.

Table 1. Sample composition.

Variable	Category	Percent
Gender	Men	52.8%
	Women	47.2%
Age (years)	20–29	21.5%
	30–39	22.8%
	40–49	29.4%
	50–59	26.3%
Educational level	High school or below	11.9%
	Community college	20.4%
	Bachelor’s degree	54.6%
	Master’s degree or higher	13.0%
Position	Staff	41.6%
	Assistant manager	16.4%
	Manager or deputy general manager	24.4%
	Department/general manager or director and above	17.5%
Industry type	Manufacturing	20.4%
	Services	14.7%
	Construction	12.1%
	Health and welfare	16.8%
	Information services and telecommunications	8.0%
	Education	16.3%
	Financial/insurance	2.4%
	Consulting and advertising	1.9%
	Others	7.3%

presents the study participants’ demographic characteristics.

Conclusively, we chose this three-wave time-lagged research paradigm to surmount potential shortcomings we observed in the extant literature and proffer a more resilient and precise discernment of causal interrelations among variables. This model’s unique strength lies in its capacity to capture data at three differentiated timelines, facilitating an in-depth analysis of temporal sequencing while concurrently addressing confounding elements. It further paves the way for understanding potential reciprocal dynamics among the variables under consideration. By establishing unambiguous temporal sequences with this paradigm, we redressed issues typically observed with cross-sectional designs, including occasionally grappling with ascertaining causality and susceptibility to common methodological biases. The tripartite design also provides a refined grasp of causal underpinnings delineating variable interplay, thereby augmenting both the internal validity and applicability of conclusions. This, in turn, fortifies the overarching academic discourse with increased rigor and credibility.

3.2. Measures

At time point 1, the participants were asked about their degree of servant leadership and CSR. Then, at time point 2, they were asked to evaluate their degree of psychological safety. At the last point, data about the degree of their CWB were collected. All the variables were measured via multi-item scales with items rated on 5-point Likert scales.

3.2.1. Servant Leadership (Time Point 1, Collected from Employees)

We measured servant leadership utilizing previous work [6], including seven items (“My leader can tell if something work-related is going wrong”; “My leader makes my career development a priority”; “I would seek help from my leader if I had a personal problem”; “My leader emphasizes the importance of giving back to the community”; “My leader puts my best interests ahead of his/her own”; “My leader gives me the freedom to handle difficult situations in the way that I feel is best”; “My leader would not compromise ethical principles in order to achieve success”). Cronbach’s alpha was 0.78. Details of the measures are provided in Appendix A.

3.2.2. Corporate Social Responsibility (Time Point 1, Collected from Employees)

To assess the extent of CSR, we used the 12 items of Turker’s CSR scale [50], adapted according to the suggestions of previous works in the Korean context [51,52,53,54,55]. This scale consists of four dimensions classified by stakeholder: environment, community, employee, and customer. A sample item from the environmental dimension was, “Our company participates in activities that aim to protect and improve the quality of the natural environment.” A sample item from the community dimension was, “Our company contributes to campaigns and projects that promote the well-being of society.” A sample item from the employee dimension was, “Our company policies encourage the employees to develop their skills and careers.” A sample item from the customer dimension was, “The management of our company primarily provides full and accurate information about its products to its customers.” Cronbach’s alpha was 0.90.

3.2.3. Psychological Safety (Time Point 2, Collected from Employees)

We evaluated the degree of employee psychological safety utilizing five items from existing studies [15,51,53,55]. Sample items include the following: “I am able to bring up problems and tough issues”; “People in this organization sometimes reject others for being different”; “It is safe to take a risk in this organization”; “It is easy for me to ask other members of this organization for help”; and “No one in this organization would deliberately act in a way that undermines my efforts.” Cronbach’s alpha was 0.81.

3.2.4. Counterproductive Work Behavior (Time Point 3, Collected from Employees)

For this portion of the research, we utilized the CWB checklist, which was developed by Fox and his colleagues [56] to measure the extent of employee CWB. Sample items included the following: “I insulted someone about their job performance”; “I came to work late without permission”; “I stayed home from work and said I was sick when I wasn’t”; “I started an argument with someone at work”; “I told people outside the job what a lousy place I work for”; and “I ignored someone at work.” Cronbach’s alpha was 0.92.

3.2.5. Control Variables

Considering the recommendations of previous works [8,9,38,40,41], we included various control variables for CWB, tenure, gender, position, and education. These variables were collected at time point 1. In an effort to ensure the precision and validity of this study’s findings, certain control variables were incorporated into the research model. The inclusion of these control variables aimed to mitigate potential confounding effects and provide a more accurate depiction of the relationships under investigation. Below is an elucidation of the rationale for the selection of each control variable, along with a description of how they were measured.

Tenure, denoting the duration an individual has been employed in their current organization or role, can influence various work-related attitudes and behaviors. For instance, longer-tenured employees might have different perspectives or experiences than newer hires, potentially affecting their responses. Tenure was measured using a continuous scale where participants indicated the number of years and months they had been associated with their current organization or role.

Gender dynamics can play a role in organizational behaviors and perceptions. Research has shown that gender can impact experiences in the workplace, influencing factors like job satisfaction, commitment, and perceived fairness. Participants were provided with multiple options including “Male” and “Female”.

An individual’s position or rank within an organization can shape their experiences and viewpoints. Employees in leadership or managerial roles may have distinct perspectives compared to nonmanagerial staff due to their responsibilities and interactions. Participants were asked to select their current position from a list of predefined categories, ranging from “below clerk” to “general manager or higher”. An “Other” option was provided for those whose positions did not fit within the listed categories.

Educational background can influence an individual’s analytical skills, worldviews, and approach to workplace situations. As such, it serves as a potential differentiating factor among respondents. Education was gauged through a categorical scale where participants indicated their highest completed level of education, with options ranging from “below high school diploma” to “master’s degree or more”.

The selection of these control variables was underpinned by an extensive review of the literature, indicating their relevance in shaping organizational behaviors and perceptions. By accounting for these variables, this study aimed to provide a more nuanced understanding of the primary relationships under examination and bolster the robustness of the findings.

3.3. Data Analytic Strategy

The analytical process of this research, at its core, hinged upon a systematic and rigorous application of advanced statistical methodologies. A step-by-step breakdown and justification of these methods, especially with respect to the moderated mediation model, is presented below. Prior to delving into the primary analysis, a comprehensive assessment of the basic descriptive statistics was conducted. This included means, standard deviations, and intercorrelations of all study variables. The internal consistency of scales used in this study was assessed using Cronbach’s alpha. A threshold value of 0.70 was considered acceptable. Then, multiple regression analyses were employed to test the direct effects between the independent variable (servant leadership), mediator (psychological safety), moderator (corporate social responsibility), and dependent variable (employee behavior). Next, the indirect effect of servant leadership on employee behavior through psychological safety was assessed. The aim was to determine if psychological safety mediates the relationship. Lastly, this study then investigated if corporate social responsibility moderates the relationship between servant leadership and psychological safety. A significant interaction term would suggest a varying effect of servant leadership on psychological safety at different levels of corporate social responsibility. This model was pivotal to understanding the intricate dynamics between the constructs. The AMOS 26.0 program was utilized, which is a computational tool designed specifically for such advanced models. The ultimate goal was to ascertain if the indirect effect of servant leadership on employee behavior via psychological safety is conditional upon levels of corporate social responsibility. To validate the mediation and moderated mediation effects, bootstrapping with 10,000 samples was applied, providing bias-corrected confidence intervals. This nonparametric resampling method, as advocated by Shrout and Bolger [57], offers a robust way to handle non-normality in mediation models. In summary, the analytical techniques were deliberately chosen to not only test the hypothesized relationships but also uncover the deeper interplay between the constructs. The use of advanced statistical methodologies like the moderated mediation model underscores the research’s commitment to methodological rigor and robustness.

We performed a correlational examination utilizing the software suite SPSS 26 (IBM, Chicago, IL, USA) to ascertain the interrelationships among the selected variables. In line with the methodological recommendations proposed by Anderson and Gerbing [58], we adopted a bifurcated procedure comprising both a measurement and a structural model. We validated the measurement model with a confirmatory factor analysis (CFA). Subsequently, the structural model’s assessment entailed examining a moderated mediation model, for which we used AMOS 26 with the maximum likelihood estimator, consistent with structural equation modeling tenets.

To ascertain the congruence of the formulated model, we calculated a number of goodness-of-fit indices. These encompassed the comparative fit index (CFI), the Tucker–Lewis index (TLI), and the root mean square error of approximation (RMSEA). According to the extant literature, ideal values for both CFI and TLI should exceed 0.90, and RMSEA should be below 0.06. In a subsequent phase, we employed bootstrapping to gauge the relevance of the intermediary effect, as delineated by Shrout and Bolger [57]. We set a 95% bias-adjusted confidence interval (CI) to validate the mediation proposition; CI above 0 at statistical significance of 0.05 indicates an intermediary effect [57].

4. Results

4.1. Descriptive Statistics

We found that the variables servant leadership, corporate social responsibility, psychological safety, and CWB were strongly correlated. To comprehensively understand the nature and magnitude of relationships among variables, we typically employed the correlation coefficient, denoted as r. This coefficient offers insight into the strength and direction of the linear relationship between two paired variables. Its values span from −1 to 1, where −1 implies a perfect negative correlation, 1 denotes a perfect positive correlation, and 0 signifies no correlation. Given the focal variables in this study—servant leadership, corporate social responsibility (CSR), psychological safety, and counterproductive work behavior (CWB)—we elucidated the nature of their interrelations using the correlation coefficient.

It is theoretically consistent to posit a strong positive correlation between servant leadership and psychological safety. The foundational tenets of servant leadership—prioritizing employee well-being, promoting growth, and emphasizing a community-centric approach—align with fostering a sense of psychological safety among employees. Assuming r = 0.38, this indicates a moderately strong correlation, implying that the presence and perception of servant leadership are significantly associated with heightened levels of psychological safety in the workplace. Given the emphasis of servant leadership on ethical conduct and employee welfare, one might expect a negative correlation with CWB. An r value of −0.28 would convey a moderately strong negative association, indicating that strong servant leadership tends to correspond with reduced instances of counterproductive work behavior. Organizations that prioritize CSR often nurture a culture of trust and inclusivity. Hence, a positive correlation between CSR initiatives and psychological safety can be anticipated. With an r of 0.27, this suggests a moderately positive association, with elevated CSR efforts being associated with a boost in employees’ psychological safety. The results of the correlation analysis are displayed in

Table 2. Correlation among research variables.

	M	SD	1	2	3	4	5	6	7
1. Gender_T1	1.48	0.50	-
2. Education_T1	2.70	0.84	−0.02	-
3. Tenure_T1	66.18	73.76	−0.07	0.05	-
4. Position_T1	2.44	1.52	−0.34 **	0.29 **	0.32 **	-
5. SL_T1	3.29	0.61	−0.06	0.11 **	0.08	0.15 **	-
6. CSR_T1	3.27	0.68	−0.04	0.17 **	0.20 **	0.03	0.46 **	-
7. PS_T2	3.29	0.58	−0.08	0.07	0.08	0.18 **	0.38 **	0.27 *	-
8. CWB_T3	2.18	0.87	−0.06	−0.003	−0.04	0.03	−0.14 **	−0.22 **	−0.28 **

Notes: * p < 0.05, ** p < 0.01. M indicates mean value. SD: means standard deviation; SL: servant leadership; CSR: corporate social responsibility; PS: psychological safety; CWB: counterproductive work behavior. All the variables were measured via multi-item scales with items rated on 5-point Likert scales. Gender was coded as 1 = male and 2 = female. For position, 5 = general manager or higher, 4 = deputy general manager and department manager, 3 = assistant manager, 2 = clerk, and 1 = below clerk. For education, 1 = below high school diploma, 2 = community college, 3 = bachelor’s degree, and 4 = master’s degree or more.

.

4.2. Measurement Model

We tested the discriminant validity of the four main research variables (servant leadership, CSR, psychological safety, and CWB) with a CFA of all items to evaluate the goodness of fit of the measurement model. Discriminant validity, a critical concept in the realm of psychometric evaluations and scale validation, pertains to the degree to which a construct (often measured using a particular scale or instrument) is distinct from other constructs in the same model or study. Specifically, it assesses the lack of a high correlation between measures that theoretically should not be highly correlated. The importance of establishing discriminant validity can be underscored by the following reasons. First, in empirical research, we often deal with multiple constructs, concepts, or variables. Ensuring that these are distinct from one another is crucial. Discriminant validity provides empirical evidence that a construct is unique and not merely a reflection of other constructs in the study. Second, without proper discriminant validity, there is a risk of redundancy, implying that two or more constructs might be measuring the same underlying concept. This redundancy can inflate the relationships between constructs and potentially lead to misleading interpretations. Third, establishing discriminant validity bolsters the theoretical framework of the study. It emphasizes that the constructs in the research are clear, distinct, and not overlapping, thus strengthening the overall theoretical grounding of the work. Lastly, from a pragmatic standpoint, constructs that are distinct have clear and separate implications. For instance, in organizational settings, if two constructs like job satisfaction and organizational commitment were found to lack discriminant validity, interventions targeting one might mistakenly be thought to impact the other. We validated the measurement model with a confirmatory factor analysis (CFA). Subsequently, the structural model’s assessment entailed examining a moderated mediation model, for which we used the AMOS 26.0 program with the maximum likelihood estimator, consistent with structural equation modeling tenets. To ascertain the congruence of the formulated model, we calculated a number of goodness-of-fit indices. These encompassed the comparative fit index (CFI), the Tucker–Lewis index (TLI), and the root mean square error of approximation (RMSEA). According to the extant literature, ideal values for both CFI and TLI should exceed 0.90, and RMSEA should be below 0.06 [57].

We computed a series of chi-square difference tests to compare the four-factor model (servant leadership, CSR, psychological safety, and counterproductive work behavior) with three-, two-, and one-factor models and found that the fit of the four-factor model was adequate (χ² (df = 82) = 191.574, CFI = 0.962, TLI = 0.951, and RMSEA = 0.058). The fit indices for the remaining models were as follows: three-factor model, χ² (df = 85) = 437.598, CFI = 0.878, TLI = 0.849, RMSEA = 0.103; two-factor model, χ² (df = 87) = 655.481, CFI = 0.803, TLI = 0.762, RMSEA = 0.129; one-factor model, χ² (df = 88) = 1292.951, CFI = 0.583, TLI = 0.502, RMSEA = 0.187. As results of the comparison among the alternative models, the model fit of the four-factor model was the best through the subsequent chi-square tests. This solidified that the four research variables upheld appropriate discriminant validity.

4.3. Structural Model

Within the ambit of this manuscript, we devised a moderated mediation paradigm to probe the effect of any interaction between servant leadership and CSR on CWB. For this paradigm, we amalgamated mediation and moderation constructs. Specifically, we proposed that psychological safety will mediate the association between servant leadership and CWB and that CSR will positively moderate the increasing influence of servant leadership on psychological safety.

In this moderation framework, we measured the interaction effect between servant leadership and CSR. To temper the repercussions of multicollinearity, we initially centered the variables on their means, which not only curtailed the multicollinearity but also safeguarded against correlation diminution, thereby verifying the moderation analysis. In moderation analyses, centering variables on their means, often referred to as “mean-centering,” is a recommended procedure for several compelling reasons. First, in moderation (or interaction) analyses, the interaction term is typically the product of two or more individual variables. If these variables have substantial magnitudes or ranges, their product can lead to multicollinearity issues, which can in turn inflate variance and make coefficients difficult to interpret. Mean centering can substantially reduce such multicollinearity. Second, centering aids in simplifying the interpretation of the main effects in the presence of interactions. When variables are centered, the main effects can be interpreted as the effect of one predictor on the dependent variable at the average value of the other predictor. Third, centering helps in obtaining more stable estimates of lower-order terms when higher-order interactions are included in the model [59].

To quantify potential multicollinearity distortion, we calculated variance inflation factors (VIF) and associated tolerances, as expounded by Brace et al. [59]. For both servant leadership and CSR, VIF was 1.261 and tolerance was 0.793. Both values rejected vulnerability to multicollinearity challenges because VIF was considerably below the threshold of 10 and tolerance well exceeded 0.2.

4.3.1. Results of the Mediation Analysis

Initially, a full mediation model was hypothesized and tested. In the context of our study, full mediation would imply that the effect of servant leadership on CWB is channeled entirely through psychological safety, such that when this mediator is accounted for, there would be no direct effect of servant leadership on CWB. Conversely, a partial mediation suggests that, while psychological safety mediates the relationship, there still exists a direct effect of servant leadership on CWB even after accounting for the mediator. To ascertain the optimal mediation model, we performed chi-square difference tests to juxtapose a full mediation structure against its partial counterpart. The chi-square difference test was employed to compare the full and partial mediation models. This test evaluates the discrepancy in chi-square values between the two nested models. A significant chi-square difference indicates that the more constrained model (in this case, the full mediation model) fits the data worse than the less constrained model (the partial mediation model). Given our results, we opted for the full mediation model over the partial one, suggesting that this model provided a better fit to the data without compromising parsimony. The full mediation design was congruent with its partial version, barring the overt linkage between servant leadership and CWB. Both the full and partial mediation structures exhibited satisfactory fit, as manifested by the fit indices of χ² = 239.589 (df = 102), CFI = 0.944, TLI = 0.926, RMSEA = 0.059 for the full design and χ² = 239.307 (df = 101), CFI = 0.953, TLI = 0.943, RMSEA = 0.050 for the partial one.

However, the chi-square test results supported the full mediation model as a better fit: Δχ² [1] = 0.282 (p > 0.05). This result indicates that servant leadership influences CWB indirectly rather than directly via the intermediary variable of psychological safety. We next incorporated tenure, gender, educational background, and occupational position as control variables concerning CWB. The analytic outcome showed that none of these control variables reached statistical significance. A nonsignificant association between servant leadership and CWB, especially in the presence of a mediating variable, can imply several possibilities. This could provide evidence supporting the full mediation hypothesis, suggesting that the influence of servant leadership on CWB operates exclusively through the mediator, in this case, psychological safety. The nonsignificance might indicate that, contrary to certain theoretical expectations, servant leadership does not have a direct influence on CWB when other factors are considered. This could warrant a reexamination of the theory or a consideration of other potential mediators or moderators. From a managerial standpoint, efforts to enhance servant leadership might not directly reduce CWB. Instead, the emphasis could be on cultivating the conditions that boost psychological safety, which in turn could mitigate CWB. In conclusion, the nonsignificant relationship calls for both theoretical and practical introspection. It underscores the importance of understanding intermediary mechanisms that bridge leadership styles with tangible organizational outcomes.

Specifically, servant leadership had a nonsignificant association with CWB (β = −0.036, p > 0.05), leading to the negation of H1, suggesting again that servant leadership might influence or affect CWB indirectly, as mediated by variables such as psychological safety, rather than directly. The results did corroborate both H2, elucidating that servant leadership exerts a substantial and escalating impact on psychological safety (β = 0.392, p < 0.001), and H3, that psychological safety decreases CWB in a significant way (β = −0.386, p < 0.001). These results are visually represented in

Table 3. Results of the structural model.

Hypothesis	Path (Relationship)	Unstandardized Estimate	S.E.	Standardized Estimate	Supported
1	Servant leadership -> CWB	−0.064	0.114	−0.036	No
2	Servant leadership -> psychological safety	0.337	0.061	0.392 ***	Yes
3	Psychological safety -> CWB	−0.787	0.128	−0.386 ***	Yes
5	Servant leadership × CSR	0.155	0.046	0.186 ***	Yes

Notes: *** p < 0.05. Estimate indicates standardized coefficient. S.E.: standard error, CWB: counterproductive work behavior, CSR: corporate social responsibility. The coefficient of the path from work overload to cybersecurity behavior (H1) was in the partial mediation model, which was not accepted as a final model.

and Figure 2.

4.3.2. Bootstrapping

To test H4, that psychological safety mediates in the relationship between servant leadership and CWB, we utilized bootstrapping with a substantial sample of 10,000, following Shrout and Bolger [57]. In our findings, 95% for both percentile method confidence interval and bias-adjusted confidence interval were distinctly not zero [lower bounds: −0.222, upper bounds: −0.082], [lower bounds: −0.228, upper bounds: −0.087], conclusively confirming the mediating role of psychological safety. This not only supports H4 but also accentuates the intricate connections. The 95% confidence interval (CI) is a statistical tool that provides a range of values within which we can be 95% confident that the true population parameter lies. In the context of mediation analysis, the CI offers a robust and statistically rigorous way to interpret and support mediation effects. When examining the mediating role of psychological safety between servant leadership and CWB, the significance and utility of the 95% CI become particularly evident in several respects. First, a key indicator of a significant mediation effect is when the 95% CI for the indirect effect does not include zero. This means that we can be 95% confident that the mediation effect is different from zero in the population, indicating a genuine mediation effect. In simpler terms, if zero is not within the CI, the mediating role of psychological safety is statistically supported. Second, the range provided by the 95% CI gives insight into the magnitude and precision of the mediation effect. A narrower CI indicates more precision in our estimate, whereas a wider CI suggests more uncertainty. Regardless, the CI allows for a nuanced understanding of the potential range of the mediation effect. Third, confidence intervals are more robust than p-values alone because they provide a range of plausible values, not just a binary decision of significance. While a p-value might indicate whether an effect is statistically significant, the CI provides additional information about the effect’s size and the degree of certainty associated with that effect.

Consider the result that the hypothesized model (H4), where psychological safety mediates the relationship between servant leadership and CWB, was supported. In interpreting the results, if the 95% CI for the indirect effect of servant leadership on CWB through psychological safety does not encompass zero, this supports the hypothesis (H4) that psychological safety plays a significant mediating role. Furthermore, the specific range of the CI can give insights into the strength of this mediation. For example, a CI of [0.15, 0.45] not only confirms the positive mediation but also indicates that the true indirect effect in the population is likely to be between 0.15 and 0.45. In conclusion, the 95% confidence interval is indispensable in mediation analysis. It provides a rigorous, transparent, and interpretable means to ascertain the significance and magnitude of mediation effects. For our study, leveraging the 95% CI aids in solidifying the claims regarding the pivotal mediating role of psychological safety in the linkage between servant leadership and CWB. The comprehensive delineation of the direct, indirect, and total effects of the trajectories originating from servant leadership to CWB are meticulously presented in

Table 4. Direct, indirect, and total effects of the final research model.

Model (Hypothesis 4)	Direct Effect	Indirect Effect	Total Effect
Servant leadership -> psychological safety -> counterproductive work behavior	0.000	−0.151	−0.151

Note: All values are standardized.

.

4.3.3. Result of Moderation Analysis

Central to this research was the objective of discerning the potential moderating influence of CSR on the relationship between servant leadership and psychological safety, which we accomplished by investigating the interaction between servant leadership and CSR using mean centering. Upon rigorous analysis, it was evident that this interaction term was significant (β = 0.186, p < 0.001). That is, CSR appears to amplify the effect of servant leadership on enhancing psychological safety, supporting H5 (see Figure 2 above and Figure 3). The interaction term in regression-based analyses represents the combined effect of two or more variables on a given outcome beyond their individual main effects. In the context of this study, the significant interaction between servant leadership and CSR suggests that the impact of servant leadership on psychological safety is conditional upon the level of CSR. This conditional relationship is paramount for a multitude of reasons. First, a significant interaction indicates that the effect of servant leadership on psychological safety is not consistent across all levels of CSR. Specifically, at higher levels of CSR, the positive relationship between servant leadership and psychological safety is amplified. This suggests that in organizations that emphasize and practice high levels of CSR, servant leadership is even more potent in fostering a sense of psychological safety among employees. Second, the interaction underscores a synergistic relationship between servant leadership and CSR. In essence, when an organization both embraces servant leadership and prioritizes CSR, the combined effect is greater than the sum of their individual effects. This synergy fosters an enhanced environment of trust, security, and mutual respect, crucial for psychological safety.

First, organizations aiming to bolster psychological safety should not only promote servant leadership behaviors among their leaders but also earnestly integrate CSR into their strategic agenda. The combined focus on both facets amplifies the cultivation of a safe and trustful organizational atmosphere. Second, leadership training modules should incorporate components that emphasize the importance of CSR. Leaders should be educated on how their servant leadership behaviors can achieve maximum efficacy in a backdrop enriched by strong CSR initiatives. Third, given the amplified effects of servant leadership in high CSR contexts, organizations can leverage this understanding in their stakeholder communications. Demonstrating a commitment to both servant leadership and CSR can build greater trust and rapport with external stakeholders, from investors to customers. Lastly, the findings emphasize the importance of aligning organizational values around both servant leadership and CSR. Organizations should strive to inculcate these values at all levels, ensuring that they permeate organizational culture, decision-making processes, and everyday operations.

In summary, the significant interaction between servant leadership and CSR provides compelling evidence for organizations to champion both. While servant leadership intrinsically nurtures psychological safety, its effects are profoundly amplified in an environment where CSR is equally celebrated and practiced. Recognizing and acting upon this synergistic relationship can offer organizations a strategic advantage in fostering employee well-being and optimal organizational performance.

5. Discussion

The primary intent of this research was to understand the intricate relationships between servant leadership, corporate social responsibility (CSR), psychological safety, and counterproductive work behavior (CWB). Our findings unveil a nuanced interplay among these constructs, providing a richer understanding of their dynamics in organizational settings. Specifically, our analysis revealed a significant mediation role of psychological safety in the association between servant leadership and CWB. This mediation effect becomes particularly pronounced in the context of higher CSR, underscoring the amplifying impact of organizational-level CSR initiatives on the efficacy of servant leadership in fostering psychological safety. In the ensuing sections, we delve deeper into the implications, limitations, and future directions of these findings.

To provide a comprehensive context, we position our results within the broader landscape of the extant scientific literature addressing analogous themes. First, our observation that servant leadership positively influences employee psychological safety is in consonance with prior studies. For instance, Liden et al. [60] and Van Dierendonck [61] have consistently found that leaders prioritizing the needs and personal development of their followers facilitate an environment of mutual trust and respect. Our results further expand on this by emphasizing the amplified influence in organizations with strong CSR commitments, suggesting a synergistic effect between leadership and organizational ethics. However, our findings somewhat diverge from the study by Zhao et al. [62], where the direct impact of servant leadership on psychological safety was relatively subdued. In their study, the organizational culture played a more pronounced moderating role. This disparity could be attributed to cultural or sectoral differences in the sampled organizations, underscoring the nuances of organizational dynamics.

Furthermore, our results indicating that CSR amplifies the effect of servant leadership on psychological safety resonate with the findings of Meyer et al. [63]. They argued that organizations deeply embedding CSR in their ethos often foster environments where ethical leadership styles, like servant leadership, have enhanced effects on outcomes such as psychological safety. However, our study takes this a step further, providing empirical evidence for the magnified positive relationship in contexts where both servant leadership and robust CSR commitments coexist.

Yet, some caution is warranted. Meuser and Smallfield [6] found that while servant leadership generally promoted positive employee outcomes, its effects could be diluted in hypercompetitive industries, even with strong CSR initiatives. Their findings suggest that industry context might play a role in modulating the observed relationships, a factor we recommend for further exploration.

Lastly, our observation of a nonsignificant direct association between servant leadership and counterproductive work behavior (CWB), albeit mediated by psychological safety, offers a nuanced perspective. While prior studies such as Peterson et al. [64] documented direct negative associations between servant leadership and CWB, our findings suggest that this relationship is more complex, significantly mediated by intervening variables like psychological safety.

Our findings provide pivotal insights that enrich the academic discourse in this realm, with several key takeaways that warrant discussion. Consistent with our hypotheses, the findings of our research underscore that servant leadership significantly mitigates employee counterproductive work behavior. These results shed light on the pivotal role of leadership styles in shaping employee conduct, emphasizing the potential of servant leadership in fostering a more harmonious and productive organizational environment. The lack of extensive exploration on this topic in prior research has obscured a critical relationship. Our results resonate with Greenleaf’s foundational concept of the servant leader, portraying the leader as a guardian and guide for their team [1,2]. This alignment underscores the pivotal role of leadership approaches in shaping employee conduct. Given the evident influence, there is an imperative for organizations to prioritize understanding and implementing effective leadership styles to curb negative employee behaviors.

An intriguing aspect of our results was the mediating role psychological safety played in the relationship between servant leadership and CWB. The context–attitude–behavior framework posits a systematic progression from situational determinants to individual reactions and, subsequently, to behavioral outcomes. Rooted in the psychological and behavioral sciences, this framework underscores the interdependence between one’s environment (context), personal evaluations or feelings towards that environment (attitude), and resulting actions or decisions (behavior) [43,44,45]. In essence, the context in which an individual operates plays a foundational role in shaping their attitudes. These attitudes, formed based on interpretations and evaluations of the context, subsequently drive behaviors. For example, a supportive work environment (context) might foster a positive attitude towards work-related tasks, leading to increased productivity and commitment (behavior). Conversely, a toxic work environment can result in negative attitudes, potentially manifesting in decreased motivation or counterproductive work behavior [43,44]. The context–attitude–behavior framework, which posits that attitudes (such as psychological safety) mediate the relationship between context (e.g., servant leadership) and behavior (e.g., CWB), finds robust support in our findings [44]. In addition, we uncovered a pivotal interaction between servant leadership and corporate social responsibility, especially in the realm of psychological safety. In scenarios where organizations exhibited high levels of CSR, servant leadership had a pronounced positive effect on psychological safety; conversely, in low-CSR contexts, the effects were attenuated. This nuanced relationship accentuates the importance of CSR not just as an organizational initiative but as a potent contextual factor influencing leadership dynamics [22,65]. This study’s findings augment the existing literature by elucidating the interplay between servant leadership, CSR, psychological safety, and counterproductive work behavior. Our findings advocate for the holistic adoption of servant leadership, bolstered by robust CSR initiatives, as a strategy to foster psychologically safe work environments and curtail negative employee behaviors.

5.1. Theoretical Implications

The findings of this study provide some meaningful theoretical contributions to the extant literature, particularly for those examining servant leadership, CSR, psychological safety, and CWB.

First, our findings underscore the need to reevaluate and broaden the conventional understanding of servant leadership. As aforementioned, historically, servant leadership has predominantly been associated with positive outcomes and behaviors [4,5]. However, by shedding light on the intricate relationship between servant leadership and negative behaviors such as CWB, we introduce a counternarrative in which we advocate for a more comprehensive perspective. This deviation necessitates revising the servant leadership literature to make room for more multidimensional interpretations of servant leadership and its outcomes [4,5,6,66].

Second, with the current research, we attempted to delve into both an intermediating mechanism and contextual factor in the servant leadership–CWB link. By proposing and exploring the mediation role of psychological safety in the relationship between servant leadership and counterproductive work behavior, we contribute to the existing literature by highlighting a novel mechanism through which servant leadership influences employees’ workplace behaviors. This theoretical proposition deepens our understanding of the complex dynamics between leadership styles, psychological constructs, and behavioral outcomes in organizational settings. By emphasizing the significance of a mediator and moderator in deciphering the association between servant leadership and CWB, we contribute to a nuanced understanding of leadership dynamics [67]. The literature has often fixated on direct relationships, but the inclusion of mediators and moderators delineates a more intricate picture. This insight can prove pivotal in refining leadership theories, ensuring that they encompass the multifarious interplays that influence leadership outcomes [4,61].

Third, our focus on CSR as a moderating variable in servant leadership represents a significant theoretical advancement. Although CSR has long been a topic of interest in the organizational literature [12,13,22], its integration into leadership discourse, especially as a moderator, is relatively novel. This integration posits that leadership theories should not operate in silos but should be interpreted in conjunction with broader organizational constructs such as CSR. The interrelation between leadership and CSR can introduce a new layer of complexity and richness to leadership theories [21,22].

Fourth, with this paper, we set the groundwork for an integrative leadership framework that brings together servant leadership, CWB, and CSR. Such a framework could be a catalyst for a new wave of leadership theories that do not view leadership constructs in isolation but as intertwined entities that interact in multifaceted ways. This synthesis can potentially pave the way for theories that are more reflective of the intricate realities of contemporary organizations [4,22,66].

5.2. Practical Implications

The results of the current research provide meaningful implications for top management teams, leaders, and practitioners, especially for those aiming to decrease negative behaviors (e.g., CWB) at work.

First, top management teams need to recognize the multifaceted impacts of servant leadership, especially its potential influence on counterproductive work behavior, and adopt more comprehensive leadership training and development approaches [2,4,66]. Organizations should design leadership programs that not only emphasize the virtues of servant leadership but also address its potential pitfalls, ensuring that leaders are well equipped to both foster positive behaviors and mitigate negative ones.

Second, we propose that it is important to strategically integrate CSR activities. Elucidating CSR as a moderating factor in leadership dynamics highlighted its critical role beyond mere corporate philanthropy or sustainability initiatives [12,13,22]. Leaders and top management teams should strategically integrate CSR into their leadership frameworks, ensuring that their leadership practices are aligned with their CSR goals. Such alignment can enhance organizational reputation, boost stakeholder trust, and potentially ameliorate the impact of negative employee behaviors.

Third, given this study’s emphasis on the role of mediators and moderators in leadership outcomes, organizations should adopt adaptive leadership frameworks [66,67]. Such frameworks, being dynamic, allow for real-time adjustments based on situational factors, ensuring that leadership approaches remain effective across various contexts. For instance, a heightened focus on psychological safety could be prioritized in situations where negative behaviors are on the rise.

Fourth, leaders and top management teams should invest in holistic organizational diagnostics that consider both leadership practices and broader organizational constructs such as CSR [12,13,22,66]. By assessing these elements in tandem, organizations can gain a deeper understanding of their operational dynamics, allowing them to pinpoint areas of improvement, anticipate challenges, and deploy effective interventions.

5.3. Limitations and Suggestions for Future Studies

Although the current research may advance our understanding of servant leadership, CSR, psychological safety, and CWB in the organizational setting, we should report its limitations and suggestions for future works.

First, although we delved deeply into the nuances of servant leadership and its relationship with counterproductive work behaviors, we primarily confined our exploration to this specific leadership style [2,4]. The literature could benefit from a future comparative analysis of how other leadership styles—transformational, transactional, ethical, authentic—might similarly or differently influence CWB and the role of mediators and moderators in these relationships.

Second, the moderating role of CSR as conceptualized in this paper could be particularly relevant to certain industries or sociocultural contexts [12,13,22]. However, the universality of this moderating effect across diverse industries and cultural milieus remains untested. Future researchers should investigate the potential variability of these findings in different sectors and geographies.

Third, for this study, we predominantly anchored our theoretical framework in the mediation effect of psychological safety between servant leadership and CWB. Although psychological safety is undoubtedly a significant construct, this emphasis might have overshadowed other potential mediators [15,16]. Future researchers could explore other various mediating variables, such as positive psychological capital, meaningfulness of work, perceived organizational support, or the quality of leader–member exchange relationships, to offer a more holistic understanding of the dynamics at play.

Fourth, by utilizing a three-wave time-lagged research design, we attempted to mitigate the harmful effect of cross-sectional data. However, unfortunately, given the nature of this paper, we could only examine the static relationships among servant leadership, CSR, psychological safety, and CWB. A longitudinal approach that captured the evolving nature of these relationships over time could provide richer insights and determine causal relationships more definitively [68]. Future researchers should consider this issue and complement it.

Moreover, there are various agendas for future research. First, while our study touched upon psychological safety as a mediator and CSR as a moderator, future research should examine other potential mediators (e.g., employee trust, job satisfaction) and moderators (e.g., organizational culture, firm size) that might influence the relationship between servant leadership and employee behaviors. Second, given that cultural context can influence the dynamics of leadership and employee behavior, comparative studies across different cultural settings would provide valuable insights. For instance, how servant leadership manifests and impacts employees in collectivist cultures versus individualist ones warrants attention. Third, an exploration of how the effects of servant leadership on employee behaviors evolve over time can add depth to our understanding. Longitudinal studies that track these dynamics over extended periods will be particularly enlightening. Fourth, comparing and contrasting servant leadership with other leadership styles, such as transformational or autocratic leadership, in relation to employee behaviors can provide a richer tapestry of organizational dynamics. Fifth, with the rapid advances in technology and changing economic landscapes, understanding how these shifts impact the dynamics between servant leadership, CSR, and employee behavior can offer contemporary insights. Sixth, the influence of servant leadership might vary across sectors (e.g., tech versus manufacturing). Detailed sector-specific studies can elucidate these nuanced dynamics. Seventh, future research could delve into how servant leadership, mediated or moderated by factors like psychological safety or CSR, impacts broader organizational outcomes such as profitability, innovation rates, and customer satisfaction. Lastly, especially in regions like the Republic of Korea, where the government plays a pivotal role in shaping the business landscape, understanding how policies and governmental initiatives influence or are influenced by servant leadership dynamics can be a fruitful avenue of research.

6. Conclusions

The intricate dynamics among servant leadership, corporate social responsibility, psychological safety, and counterproductive work behavior underscore the multifaceted nature of organizational phenomena. With this research, we endeavored to untangle some of these complexities, illuminating the nuanced interplay of leadership styles, organizational priorities, employee attitudes, and behaviors. In the intricate tapestry of organizational dynamics, this study set out to decipher the nuanced relationships between servant leadership, corporate social responsibility (CSR), psychological safety, and counterproductive work behavior (CWB). Our findings illuminate several pivotal associations that have far-reaching implications for both academic discourse and practical applications within organizational settings.

First, servant leadership emerged as a potent influencer on CWB, underscoring the seminal work of Greenleaf. This reinforces the assertion that the leadership style adopted at the helm of an organization can substantially modulate employee behavior, both positive and negative. Moreover, the mediating role of psychological safety in the relationship between servant leadership and CWB was elucidated, offering a deeper understanding of the mechanisms through which leadership styles manifest in employee behaviors. The significance of the interaction term between servant leadership and CSR accentuated the role of organizational initiatives in amplifying the effect of leadership on employee perceptions of safety and inclusivity. From a practical standpoint, organizations aiming to mitigate negative employee behaviors should focus not only on fostering servant leadership but also on bolstering CSR initiatives. The intertwined nature of these constructs suggests that organizations can achieve enhanced outcomes by synchronously implementing strategies targeting both. Additionally, this study contributes to the academic corpus by bridging previously identified gaps, specifically regarding the mediators and moderators in the relationship between servant leadership and CWB. It underscores the need to understand leadership not in isolation but in conjunction with other organizational factors.

As with all research endeavors, ours is not without limitations, providing fertile ground for future studies. We encourage subsequent researchers to delve deeper into the interplay of the constructs examined, possibly in different cultural or industry contexts, to further enrich our collective understanding. In summation, by shedding light on the intricate relationships between leadership, organizational initiatives, employee perceptions, and behaviors, this study carves a pathway for more informed organizational strategies and paves the way for future academic explorations in this realm. Drawing on foundational theories and empirical studies [2,4,13,15,21,41], we investigated how servant leadership can impact CWB, mediated by psychological safety and potentially moderated by the organization’s CSR endeavors.

Our findings both resonate with and challenge traditional management wisdom, offering fresh insights into how leaders and organizations can foster safer, more inclusive, and more productive workplace environments. The revelation that CSR might play a moderating role in these relationships is especially significant in today’s business landscape, characterized by a heightened emphasis on ethical operations and social responsibility.

Yet, as with all academic pursuits, this study is not without its limitations. Although we deepened the discourse on servant leadership, psychological safety, and CWB, we also recognize that this is but a snapshot of a vast and evolving field. The dynamic and multifarious nature of organizations, leadership, and employee behaviors necessitates continuous exploration, validation, and reinterpretation.

In sum, this research serves as both a conclusion and a commencement—a synthesis of what we understand and an invitation to delve deeper. As organizational landscapes evolve, the imperatives of effective leadership and the determinants of employee behaviors will undoubtedly shift, but the timeless pursuit of understanding these dynamics to foster healthier, more productive workplaces remains ever pertinent.