1. Introduction
Physical and psychological well-being, together with the collective future of humanity, is dependent on preserving the environment. Environmental issues pose a great challenge for organizations and require the mobilization of considerable resources and competencies to achieve a transition towards sustainability [
1]. Corporate greening is a paramount challenge that the organizational word is facing today [
1,
2] and employees’ play a vital role in overcoming these challenges [
3]. Employees’ inertia pertaining to the environmental issues has been a major concern for managers that has led them to explore a wide range of research avenues in the hope of gaining a better understanding of the factors related to environmentally responsible behaviors. In recent years, there has been a burgeoning scholarly interest in examining employees’ extra role behaviors towards the environment [
4] and this has offered valuable insights that could shape employees’ green behavior in the workplace. Despite progress on academic and business fronts, organizations continue to face substantial human resource challenges in realizing environmental sustainability initiatives [
5,
6,
7,
8]. Ones and Dilchert [
9] highlighted that “What organizations do is a function of decisions, behaviors, and performance of their members. Organizational initiatives stem from employees. Therefore, understanding, promoting, influencing, and changing environmental behaviors of employees are keys to environmental sustainability of organizations”. Norton et al. [
10] defined employees’ voluntary green behavior that involves personal initiative exceeding organizational expectations, including prioritizing environmental interests, initiating environmental programs and policies, lobbying and activism, and encouraging others. Nurturing employees’ voluntary green behavior not only serves corporate greening objectives, but can also prevent further environmental degradation by positively affecting environmental change.
Leadership is considered to be at the heart of nurturing individual, group and organizational outcomes [
11]. Leaders can influence a wide range of diversified organizational outcomes, including environmental considerations [
12]. However, the literature has reported the influence of different leadership styles like transformational leadership [
13,
14], transactional leadership [
15], ethical leadership [
16], spiritual leadership [
17], and responsible leadership [
18] on employees’ extra role behaviors towards the environment. Noticeably, the missing link from the researchers’ attention has been the effect of servant leadership and the mechanism through which it operates to influence employees’ voluntary green behavior (EVGB). While comparing servant leadership with other emerging forms of positive leadership, Hoch et al. [
19], in their series of meta-analysis, asserted that servant leadership showed its distinctiveness and ability to better explain the variety of outcomes over and above the other forms of leadership. Researchers have reasoned that servant leaders go beyond other types of leaders, primarily in two spheres: focusing on the needs of their followers, and recognizing their own social responsibility [
20]. The essence of servant leadership stands on the principle that it should develop their followers, in such a way that they would themselves emerge as servant leaders [
21]. Servant leaders consider it their moral responsibility to safeguard the good of all the stakeholders, including employees, customers, and community [
22], and environment is no exception to this. The characteristics of servant leadership, such as stewardship, creating value for the community, servanthood, and altruistic calling are implied in the concept of servant leadership [
23]. These characteristics highlight that servant leaders act selflessly and are likely to create a sense among their subordinates of caring for the good of wider society. SLs pay too much attention to community service, while EVGB are closely aligned with safeguarding the community through environmental concern. Hence, investigating how servant leadership influences EVGB is genuinely meaningful.
Besides examining the direct association of servant leadership with EVGB, a comprehensive understanding of the mechanisms through which servant leadership leads to such behavior is also under consideration in this research. Scholars maintain that psychological empowerment could potentially be the underlying mechanism, which may describe the relationship of servant leadership with employees’ behavioral outcomes [
21,
24]. When employees realize that their leadership believes in creating value for society and stands up for their initiatives to protect and care for the environment, they are likely to feel a sense of meaning, competence, self-determination and impact [
25]. The concept of empowerment is present in almost all the definitions of servant leadership. Employees’ sense of psychological empowerment gives them the feeling of competence and control in performing jobs. This inner confidence of empowerment is essential in realizing EVGB. Psychological empowerment, being at the heart of servant leadership characteristics and an antecedent of EVGB, is examined as an intervening variable through which servant leadership potentially exerts its influence towards EVGB. Furthermore, self-determination theory (SDT) stresses the vitality of autonomous motivation in shaping employee behaviors [
26]. We operationalize autonomous motivation for the environment (AME) as being involved in those behaviors which align with one’s intrinsic goals and originates from the inner-self [
27]. We relied on SDT [
26] in proposing AME as another underlying mechanism between servant leadership and EVGB. AME is predicted to be essential in engaging an employee in voluntary green behavior. In doing so, we enhanced the explanatory power of the research model and better understanding of the path(s) through which servant leadership will lead to EVGB. The inclusion of AME, based on SDT, enabled us to establish a sequential mediation path that transmits the effect of servant leadership on EVGB through the underlying mechanism of psychological empowerment and AME.
Precisely, we aimed to offer four substantial contributions to the literature of servant leadership and EVGB. At first, despite a call to study the influence of various leadership styles on EVGB [
10], as we understand, no empirical research has studied the direct impact of servant leadership on EVGB. Second, understanding the mechanism through which a leadership style exerts its influence towards employees’ outcomes is of vital importance for academicians and practitioners [
28]. This research is an attempt, where psychological empowerment and AME are explored as independent and sequential mediators between the relationship of servant leadership and EVGB. Third, in line with the argument to develop a multilevel approach to enrich the understanding of environmental sustainability in organizational contexts [
10,
29], this research examined pragmatic predictors within a holistic model that included contextual (servant leadership) and personal level (psychological empowerment and AME) antecedents of EVGB. Finally, this research was conducted in Pakistan, a developing country where environmental laws are not up to standard. Particularly, Pakistan is among the top in the list of badly affected countries, due to environmental change and global warming [
30]. Thus, the context of this research is highly meaningful for the practitioners.
We organized this article as follows: The coming
Section 2 presents the hypotheses based on theoretical underpinnings and empirical evidences.
Section 3 offers research methods encompassing the context, sampling and procedure, and measures adopted.
Section 4 deals with the analysis, while in
Section 5, discussion of the results is featured, along with theoretical and managerial implications. The last section,
Section 6, includes conclusions, limitations, and future research avenues.
4. Results and Analysis
The analysis was carried out by employing variance-based SEM, using partial least squares [
74,
75] through the Smart-PLS 3.2.8 (Boenningstedt, Germany) software [
76]. In recent years, partial least squares structural equation modeling (PLS-SEM) has shown its footprints over diversified disciplines including marketing, accounting, human resource management, and many others [
77]. Rather than following the crowd, authors have preferred PSL-SEM for several reasons. At first, PLS-SEM allows for the analysis of complex models with multiple constructs, indicators, and relationships [
78]. Second, the latest guidelines on the use of PLS-SEM proved its superiority over other techniques in the assessment of mediation analysis [
79]. Third, psychological empowerment and AME were designed as second-order constructs, and PSL-SEM is a better choice for dealing with the models that have higher order constructs [
78,
79]. Fourth, PLS-SEM offers better ‘statistical power’ [
80]. Lastly, it is considered to be equally efficient for exploratory- and prediction-oriented research [
81]. A two-stage approach, namely involving (i) measurement model evaluation, and (ii) structural model evaluation, was employed to analyze the results of PLS-SEM [
82].
4.1. Measurement Model Evaluation
In this section, the distinction between ‘reflective’ and ‘formatively’ designed constructs needed to be considered at the outset. All the constructs designed are ‘reflective’ in nature. Besides, psychological empowerment and AME were designed as second-order reflective-reflective constructs. We followed the two-step approach for the measurement model evaluation of second-order constructs. The latent variable scores of the lower order reflective constructs were used as manifest variables of the higher order constructs.
4.1.1. Reliability
At first, the individual indicator’s reliability was ensured by standardized factor loadings and it was established when an indicator had a standardized factor loading of ≥0.70 on its associated construct [
83]. The second step requires examining the internal consistency-reliability of the constructs. The latest guidelines on reporting the results of PLS-SEM recommended the use of Dijkstra and Henseler’s [
84] ‘
ρA’ as an approximately precise measure of reliability over traditional Cronbach’s alpha and composite reliability.
Table 2 presets values of factor loadings,
ρA, Cronbach’s alpha, and ‘CR’ that are between 0.7 to 0.9 [
81], and unanimously confirms internal consistency/reliability of the first and the second order reflective constructs.
4.1.2. Convergent Validity
The 3rd stage of the measurement model evaluation requires the establishment of the convergent validity of the constructs, and for this purpose, average variance extracted (AVE) is a widely used metric [
83]. An AVE value of 0.5 or greater means that the construct explains more than half of the variance of the indicators which constitute that construct [
85]. The findings presented in
Table 2 confirm that the AVE values of all the constructs are more than the recommended value of the threshold.
4.1.3. Discriminant Validity
The latest guidelines on the evaluation of the PLS-SEM measurement model advocate the use of the heterotrait–monotrait (HTMT) ratio of correlations, over the traditional Fornell and Larkers approach [
74], to establish the discriminant validity of the constructs. When the constructs have a higher conceptual distinction, a HTMT threshold of up to 0.85 was recommended [
83]. The findings in
Table 3 present the mean and standard deviation, along with ensuring the discriminant validity.
4.2. Structural Model Evaluation
Relationships among the dependent and independent variables were assessed, while analyzing the structural model through the size and direction of path coefficients, values of the coefficient of determination, and
t-values [
77,
86]. Following recent studies, we treated psychological empowerment and AME as higher order (single factor) constructs [
12,
63,
66]. Before we proceeded further, assessment of the collinearity was carried out with the variance inflation factor (VIF) technique and all the values accorded below the threshold of 3 [
81], stressing that the model was not contaminated due to common method bias [
87].
4.2.1. Path Coefficient (β)
A path weighting scheme was run with default settings of the Smart-PLS 3.2.8 software. The bootstrapping procedure was run with 5000 subsamples, and no sign change option was chosen. Significance levels for one-tailed testing, along with percentile bootstrap were the other options selected to run the procedure. For the statistical significance of hypotheses, the value of path coefficients should be established through the percentile bootstrap confidence interval [
88] and its direction should be consistent with the respective hypotheses.
Table 4 represents the detail of the hypotheses, where the findings are confirmed through
t-values and the percentile bootstrap confidence interval. We observed 1.96 as the cutoff criterion for t-statistics [
66,
89]. The results reveal that all hypothesized relationships are significantly supported, except H1. Besides, from the control variables, the influence of Education on EVGB is significant, while other control variables are not significant.
4.2.2. Coefficient of Determination (R2)
The R
2 value of the endogenous construct represents the within-sample predictive power of the structural model [
78]. As a rough estimate, the R
2 value of 0.25, 0.50, and 075 represents the ‘weak’, ‘moderate’, and ‘strong’ traits, respectively [
78]. The R
2 value for the key target construct of employees’ voluntary green behavior is 72.2%, which means that all the antecedents explained had substantial variance in EVGB. The R
2 values of all the endogenous constructs are listed in
Table 4.
4.2.3. Blindfolding (Q2)
PLS-SEM contains a supplementary method to evaluate the predictive ability of the structure through blindfolding approach denoted by Q
2 [
81,
90]. To find the values of Q
2 in Smart-PLS 3.2.8, a blindfolding process using a cross-validated redundancy approach was employed. More than zero values for all the endogenous constructs were obtained, which ensures the predictive accuracy of the model [
81,
90].
Table 4 presents the Q
2 values of the endogenous constructs.
4.2.4. Out-of-Sample Predictive Quality-PLSpredict
Lastly, the out-of-sample predictive quality of the model was assessed through the Smart-PLS option of PLS predict, by following the procedure proposed by Shmueli [
91]. The Q
2 predict values for all the indicators of the key target construct of employees’ voluntary green behavior were found to be positive (see
Table 5). Moreover, for all the indicators of EVGB, the PLS-SEM results have a smaller prediction error compared to the linear model benchmark. Therefore, the model established a high predictive power [
92].
4.3. Mediation Analysis
To establish mediation, the significance of the direct, indirect effect and total effect needs to be assessed [
93,
94]. In this research, direct effect is the value of path coefficient (
β) from servant leadership to EVGB. Indirect effect is the product of the direct effect from servant leadership to mediating variable (psychological empowerment and/or AME) and the direct effect from those mediating variable(s) to EVGB. Then, the total effect is the sum of the direct effect and indirect effect. To obtain results, 5000 subsamples were bootstrapped with 315 observations per subsample. No sign change option to determine the significance of the path coefficients with
p less than 0.05 two-tailed was the other setting. The significance of the hypotheses was established through a confidence interval, such that it should not contain a ‘0’ value [
78]. The results of the mediation analysis are shown in
Table 6 and
Figure 2.
5. Discussion and Implications
5.1. Discussion
This research intended to examine the impact of servant leadership on employees’ voluntary green behavior, first directly and then through independent and sequential mediation of psychological empowerment, and autonomous motivation for the environment. Findings revealed that the direct impact of servant leadership on EVGB was not supported. This result is inconsistent with the literature, where the direct or indirect effect of servant leadership was investigated on employees’ organizational citizenship behavior for the environment [
40,
41,
42], a construct similar to EVGB. The reason behind this finding is that both psychological empowerment and autonomous motivation for the environment fully mediates the influence of servant leadership on EVGB. It is pertinent to state that without introducing both the mediators, i.e., psychological empower and AME, servant leadership has a positive significant effect on EVGB. However, after introducing multiple mediations, the effect of servant leadership on EVGB becomes insignificant.
The finding that servant leadership has a positive influence on autonomous motivation for the environment is congruent with the existing literature, where servant leadership has shown its positive impact on intrinsic motivation [
48], a component of autonomous motivation. The more a leader demonstrates the servant leadership style, the more the followers feel autonomously motivated. Then, the positive impact of AME on EVGB can be explained through the lens of SDT, which asserts that employees’ voluntary behaviors can be channelized through their enhanced autonomous motivation. This finding is in accordance with the recent studies, where the autonomous motivation for the environment leads to employees’ extra role behavior towards the environment [
12,
17]. Furthermore, the results confirm the mediating role of autonomous motivation for the environment between servant leadership and the EVGB relationship. Servant leaders consider the benefit of all the stakeholders, within and outside the organization, and their support for the environment provides the necessary nutrition of AME, which channelizes EVGB. This means that AME is one of the mechanisms that servant leadership opts to enhance; EVGB. This finding is partially consistent with the existing empirical research, where AME served the intervening role in predicting employees’ extra role behavior towards the environment [
12,
49,
51], and this strongly corroborated with the application of self-determination theory.
Next, the servant leadership is hypothesized to have a positive impact on employees’ psychological empowerment. Results strongly supported this postulation and validated the earlier research, where servant leadership was found to be a significant predictor of employees’ psychological empowerment [
64]. The empowerment of followers is considered as one of the vital characteristics of servant leaders. Furthermore, this study supported the positive influence of psychological empowerment on EVGB. The dimensions of psychological empowerment, including meaning, competence, self-determination and impact, enable employees to take initiatives in performing extra-role behaviors like EVGB. Furthermore, the hypothesis regarding the mediating role of employees’ psychological empowerment between the relationship of servant leadership and EVGB has also been supported. This result is partially consistent with the earlier studies, where psychological empowerment was investigated as the moderator between servant leadership and employees’ outcomes [
61] other than EVGB.
The introduction of multiple mediators sequentially offers a deeper insight into the influencing mechanism of servant leadership towards EVGB. The findings showed that psychological empowerment and AME sequentially mediate the relationship between servant leadership and EVGB. This is a unique empirical contribution of this research and draws its support by integrating self-determination and psychological empowerment theories. Interestingly, after the introduction of multiple mediators in the structural model, the impact of servant leadership on EVGB becomes non-significant, which proves that psychological empowerment and AME fully mediate the impact of the servant leadership on EVGB. Alternately, whatever the positive influence servant leadership exerted on EVGB was through the mediating mechanisms of psychological empowerment and AME. As far as the relative importance of psychological empowerment and autonomous motivation for the environment is concerned, psychological empowerment offers more significant intervening mechanisms. We explored another interesting finding while analyzing sequential mediation of psychological empowerment and AME. The path from servant leadership to AME remains marginally significant, which means that servant leadership largely influences AME through the intervening role of psychological empowerment. Thus, this study highlights the critical value of employees’ psychological empowerment to servant leaders, while shaping AME and EVGB.
5.2. Theoretical Implications
This research contributes in various ways to the literature of servant leadership and employees’ voluntary green behavior. At first, Norton, Parker, Zacher and Ashkanasy [
10] invited researchers to explore the influence of different leadership styles on EVGB, and there is a paucity of research that has examined the influence of servant leadership style on employees’ voluntary green behavior. Prior research has investigated only the moderating role of servant leadership in the relationship of CSR and OCBE [
40]. Therefore, this research is an attempt to advance the knowledge stream of servant leadership by investigating its direct relationship with EVGB. When it comes to protecting the environment, servant leaders are considered to care for the environment over their own or the organization’s financial gain and promote pro-environmental values among the stakeholders, including employees [
40,
42].
Secondly, this research enriches the literature on the influencing mechanism of servant leadership in predicting employees’ outcomes. Understanding the influencing mechanism of servant leadership has already been desired by the researchers [
28]. We proposed psychological empowerment and AME as the underlying mechanisms through which servant leadership would enhance EVGB. Besides, we offered a sequential mediation mechanism where servant leadership transmits its influence in the form of employees’ psychological empowerment, which then lifts their autonomous motivation for the environment, and finally, they demonstrate higher levels of voluntary green behaviors. These mechanisms are fully backed by the concepts of self-determination theory (SDT) and are highly valuable for the research fraternity, due to their comprehensive depiction of the underlying working of servant leadership. Furthermore, we advanced the literature on SDT by establishing the influence of servant leadership with autonomous motivation for the environment. None of the existing research has examined this relationship empirically.
Third, we advanced the knowledge of EVGB by investigating its antecedents at a contextual and personal level in a single conceptual model. Although limited empirical studies have already investigated psychological empowerment and AME as independent antecedents of EVGB, conceptualization of these antecedents in the form of sequential mediation is a unique theoretical insight for EVGB literature. This research specifically enhances the nomological network of EVGB, as well as servant leadership constructs.
5.3. Practical Implications
We offered organizational practitioners a roadmap to ‘grow’ their employees into ‘environmental activists’ for the sustainable development of their organization. First, this research reinforced the importance of servant leadership. Findings revealed that servant leadership positively influences employees’ psychological empowerment and AME, that in turn leads to their enhanced voluntary green behavior. The top management of organizations, while selecting managers, should give priority to those individuals who can demonstrate servant leadership in organizations. Besides, organizations may offer its managers training and development opportunities that can enhance their level of servant leadership. The concept of the servant leader is based on the belief in creating value for the community, serving others, delegating power, and empowering followers to grow and succeed in their professional and personal lives. Such leaders, through regular interaction with subordinates, teach them the importance of giving back to the community and enhance EVGB. It can lift a manager’s level, as being a servant leader improves interaction with employees, and this can enhance EVGB. Secondly, the importance of employees’ psychological empowerment for their autonomous motivation for the environment and engagement in voluntary green behavior in the workplace has also been highlighted. Organizational managers should realize their subordinates’ about their capabilities, competencies and should afford them freedom in doing their jobs. Third, organizations should have a keen focus on the employees’ autonomous motivation for the environment and continuously take initiatives that can enhance or at least keep this motivation alive.
6. Limitations, Future Research Directions, and Conclusions
Despite numerous theoretical and practical offerings, this research is not without limitations. First, the cross-sectional nature of the data confines this research’s ability to establish any inference about causality. We suggest longitudinal investigations of this study’s model to overcome this limitation. Second, the data for this research were collected from employees and their immediate supervisors employed in the power sector organizations of Pakistan. Though the context of this research is highly valuable for practitioners, it is different from the context of other sector/industries and countries. Thus, the findings of this research need care while generalizing. For enhanced generalizability, similar studies in different sectors/industries and countries are proposed. Third, we employed general servant leadership in our conceptual model, due to the non-availability of a valid scale to measure environment specific servant leadership. We invite scholars to prepare and validate an environment specific servant leadership scale, so that future researchers can employ environment specific servant leadership in their research models. Fourth, this research included psychological empowerment and AME as mediators; aspirant researchers recommend the examination of other factors at the organizational, team, and individual level as mediating and/or moderating variables. Fifth, we gathered data from a single source and employed the same technique, which may have caused some degree of bias. Lastly, we recommend that aspirant researchers model AME and psychological empowerment as lower order constructs, dimensions level, in their theoretical models.
In the burgeoning research area of employees’ green behavior at the workplace, we tried to identify the influencing mechanisms of servant leadership towards employees’ voluntary green behavior. A multi-theory perspective comprising self-determination and psychological empowerment theories could be employed to develop a multi-level conceptual model of this research, including servant leadership, psychological empowerment, and AME as antecedents of EVGB. Servant leadership was found to be a substantial predictor of psychological empowerment, autonomous motivation for the environment, and VEGB. In addition, psychological empowerment and AME was validated as being a predictor of EVGB.