Examining Social Sustainability in Organizations

Abstract

Social sustainability in organizations has been externally focused and, given its socio-ecological importance and potential, it requires further attention. This study examines social sustainability in business organizations (a) as an antecedent of environmental and economically related constructs, and (b) as a component of a multiplicity of relationships among social, environmental, and economically associated constructs. We examine a diverse set of 41 model configurations of 10 first-order constructs, and a diverse set of second-order constructs following a gradient ranging from a relatively moderate degree of order to quasi-random construct arrangements ending with model configurations fully determined at random. These sets reflect variability and commonality among respondents from various organizations, industries, and regions of the United States. Constructs may be antecedents, mediators, or outcomes. Analyses were conducted using PLS-SEM software. Results show (a) that social constructs are at the core of organizational life, given their frequent moderate to strong positive effects on other social constructs, as well as on environmental and economically related organizational constructs; and (b) construct variability within sustainability dimensions. Contributions include examining both constructs’ gradient ordering and constructs’ variability effects contingent on both construct nature and position. Theoretical and practical implications, as well as limitations and future research, are discussed.

1. Introduction

The generation of sustainability problems, and their solutions, stem from social relationships. A key set of social relationships takes place in the workplace. At a fundamental level, social relationships are required for any activity to generate economic value [1]. Progress toward social sustainability is by and large a function of the quality of people’s relationships. Therefore, the quality of relationships in organizations determines the degree of organizational social (in) sustainability. We define organizational social sustainability as the firm’s ability to continuously evolve successfully by appropriately integrating social, economic, and environmental dimensions.

Despite the significance of workplace social relationships, organizational sustainability studies have given a greater degree of attention to the biophysical and economic dimensions than to the social ones [2,3,4]. Owing to its external focus, social sustainability has paid insufficient attention to employees. For instance, corporate social responsibility addresses corporate behavior mostly in terms of broad relationships with communities and society; and/or confined mainly to compliance with labor rights or concerned with integrating social sustainability in organizations to serve economic performance [5,6]. This occurs despite (a) many firms touting that people are their “best/most important resource” [7], (b) increased competition for talent in some industry sectors [8], and (c) rising job precariousness [9,10]. Disinclination to address social aspects in organizations may relate to an unwillingness to deal openly with power, politics, and substantive democracy in the workplace [11,12,13] and fear that a greater degree of attention to the social may negatively impinge the firm’s economic performance [14]. Similarly, very little happens if “business as usual” continues (e.g., widespread greenwashing [15]). In other words, power asymmetries, as expressed in inappropriate attention to the social, seem to preclude a fair distribution of value-added in organizations.

Failure to address the social in organizations curtails the effective “use” of people and developing employee potential. Similarly, not creating new collaborative forms to generate a more egalitarian distribution of value-added exacerbates socio-ecological organizational problems. The relative lack of attention to social sustainability in organizations contrasts with its high socio-ecological importance and potential. In contrast, favorable social dynamics in organizations support positive environmental and economic synergies [16,17]. In other words, the social influences the environmental and economic in organizations in fundamental ways [18].

Social sustainability studies are numerous and increasing [19,20,21,22,23,24,25]; see also the journal’s Special Issue on social sustainability edited by [26] and by [27]. Sustainability research may involve various scales, diverse levels, and multiple methodologies, perspectives, and focuses [16,28,29]. Complex and varied sets of antecedents, mediators, moderators, and outcomes potentially produce multiple synergies and trade-offs among constructs. Although sustainability research is abundant, there is much yet to be researched, particularly regarding social sustainability in organizations.

Since social sustainability in organizations has been externally focused and has received less attention than the biophysical and economic dimensions, there is the need to further consider the human side of enterprise [30,31,32] and to examine the social organization, as it relates as a system, to organizations’ environmental and economic dimensions.

Constructs both within and between organizations may be differentially positioned, have different interactions in type and intensity, and receive different degrees of attention. Different degrees of construct coupling occur because of differences in path dependencies, evolutions, perspectives, and operationalizations in teams and other units within and between organizations. This is manifested in various degrees of construct ordering, ranging from relatively tight ordering (e.g., established construct sequences such as motivation-engagement-performance) to weak, very weak, or nonexistent ordering (e.g., constructs’ weak or no correlation due to constructs’ nature and/or distances). In addition, there is a variability of construct ordering within the social dimension itself. For example, comprehensive constructs such as perceived organizational support (herein POS) may include a large and diverse set of both strong and weak relationships, whereas less-encompassing constructs such as task-defined employee performance may have a smaller, and a less diverse, set of significant couplings. Most research focuses on relatively strong relationships. Likewise, research chiefly refers to highly circumscribed conditions (e.g., one or few models with few constructs). This study attempts to partly address these shortcomings.

The purpose of this study is to relate environmental, social, and economically associated dimensions of sustainability in organizations while highlighting the social dimension. Our approach examines social sustainability (a) as an antecedent of environmental and economically related constructs, and (b) as a component of a multiplicity of relationships among social, environmental, and economically associated constructs. First, using a range of model configurations (Figure 1), we examine, using first-order constructs, the effects of the social in organizations, studying the effects of green transformational leadership, transactional leadership, POS, autonomous motivation, job engagement, and job satisfaction on both (a) attitudes toward the natural environment, environmental commitment, and pro-environmental behaviors (environmental constructs), and (b) employee performance. Except for green transformational leadership and transactional leadership, all constructs refer to employees. In addition, we examine alternative configurations of first- and second-order constructs of social, environmental, and economically related constructs. Second, to investigate direct and indirect effects, many of which have not yet been examined, we study 29 model configurations. These configurations were generated both quasi-randomly and completely at random (for details, see Section 3). This set of construct combinations is in accordance with various histories, philosophies/perspectives, resources, and purposes of the firms included in our sample (respondents from a range of organizations, industries, and regions of the United States). The complex set of key relationships examined (conceptual models in Figure 1, Figure 2, Figure 3 and Figure 4) is just a subset of multiple potential sets of constructs associating social, environmental, and economically related dimensions in organizations. To the best of our knowledge, the construct sets examined in this study have not been studied previously. In a few words, our study is an attempt to further examine the social dimension in organizations by assessing a multiplicity of effects among organizational social, environmental, and economically related constructs.

This paper makes the following theoretical contributions. First, in examining social sustainability as an antecedent of environmental and economically related constructs, the model configurations among green transformational leadership, transactional leadership, POS, autonomous motivation, job engagement, and job satisfaction show moderate to strong effects both within the social itself and in interactions with environmental (e.g., attitudes toward the natural environment, environmental commitment, and pro-environmental behaviors) and economically related constructs (e.g., employee performance). The diverse set of first- and second-order model configurations validates the importance of the social in organizations. In other words, social constructs are at the core of organizational life. Second, the diverse degrees of constructs’ ordering studied affirm the centrality of the social in organizations and show the importance of paying attention to variability, as well as to the need to study in detail and integrate the social with environmental and economic dimensions. Findings suggest that construct variability within the social dimension should be further investigated, as well as its resulting interactions and effects as social variability increasingly integrates with variability in the environmental and economic dimensions. Furthermore, findings intimate that, in discussing organizational sustainability at the country level, generalization may jeopardize specificity.

In sum, social sustainability is fundamental for organizational life because it antecedes and drives the economic, and environmental dimensions. In addition, progress in social sustainability in organizations will contribute to employees’ well-being and develop employees’ potential as well as firms’ productivity, innovation, and competitiveness.

2. Literature Review and Hypotheses

The theoretical basis of our research includes spillover theory [33,34], social exchange theory [35], role theory [36], green transformational leadership theory [37], LMX theory [38], POS theory [39], and conservation of resources theory [40]. Spillover theory posits that effects on a given realm may manifest in another. It justifies multiple social, economic, and environmental interactions, particularly indirect effects. Social exchange theory puts forward that employee constructs’ effects result from a cost–benefit analysis of the employee in their exchanges with the organization and the employee’s environment. Role theory postulates that a role consists of certain expectations of what to do and what not to do. Green transformational leadership theory predicates that transformational leaders align employees’ interests with those of the organization while protecting the natural environment. LMX theory proposes that the quality of the employee-leader relationship may be expressed in respect, well-being, and performance. POS theory puts forward that employees will perform according to the degree to which they feel appreciated by their organization. Conservation of resource theory submits that maintaining existing, and pursuing new, resources manifest in employees’ social, economic, and environmental constructs. The overall rationale of these theories helps explain, from different perspectives, and with some degree of overlap, the correlative effects of social, environmental, and economically related constructs in organizations.

The section below provides justification for both the paths of our conceptual set of models and our hypotheses (see Figure 1, Figure 2, Figure 3 and Figure 4). It may be noted that the hypotheses are of a “higher order”. They do not refer just to single construct relationships. Due to space limitations, the literature review focuses on some crucial sets of construct relationships and refers to relationships that are not explicitly mentioned in the hypotheses.

2.1. Green Transformational and Transactional Leadership Relationships with POS, Autonomous Motivation, Job Engagement, and Job Satisfaction

Green transformational leadership gives a greater degree of attention to pro-environmental values and pro-environmental behaviors than both traditional transformational leadership [41,42] and transactional leadership. However, green transformational leadership studies tend to consider mostly green antecedents, green mediators, green moderators, and green outcomes [43,44,45]. Our green transformation and transactional leadership constructs combine environmentally related items and “traditional” transformational leadership items, reflecting the varying degrees of attention given to green issues by employees and leaders in a diverse population of firms.

Green transformational leaders play a crucial role in enhancing followers’ POS [43,46,47,48,49]. Employees who work for transactional leaders have limited perceptions of organizational support and feel less indebted to their leaders and organizations [50]. Similarly, green transformational leaders enhance autonomous motivation via supportive psychological states, such as competency and attachment with followers [43,51]. Furthermore, green transformational leaders promote subordinates’ innovation [52,53], nurture self-initiation, clearly transmit the firm’s vision, infuse pride, and achieve followers’ confidence, which induces individuals’ autonomous motivation [54,55,56]. Transactional leaders motivate followers through rewards and promises, which decreases employee autonomy because corrective measures or disciplinary action is needed if there is negative feedback [57,58]. Thus, followers’ motivation is lower for transactional than for transformational leaders [59]. Likewise, green transformational leaders stimulate followers’ job engagement [60,61]. Job engagement is positively associated with eco attitudes and initiatives, as well as green behaviors by employees [62,63]. Although transactional leaders may positively influence employee engagement and performance [64,65], they have a weaker influence on work engagement than transformational leaders [64,66].

Green transformational leaders enhance employee job satisfaction by shaping attitudes and behaviors toward the natural environment [67,68,69,70,71]. Similarly, transformational leaders positively impact employee job satisfaction [72,73,74,75,76]. In contrast, the relationship between transactional leaders and job satisfaction may be positive [77,78], non-significant [79], or positive in the short run but negative in the long run [80,81]. Such a relationship may depend on the rewards or punishment features [82,83]. Thus, we hypothesize:

Hypothesis 1 (H1).

Green transformational leadership positive associations with POS, autonomous motivation, job engagement, and job satisfaction will be higher than those for transactional leadership.

2.2. POS’ Mediating Effects

A large set of constructs depicting social relationships is positively or negatively associated with POS [48,84,85,86,87], suggesting POS as the antecedent, constitutive, or outcome, of such constructs. For example, POS has been positively associated with the leader’s concern with employee well-being [88,89,90], leader–member exchange [91,92], value congruence with organization [93,94], organizational justice [95,96], job security [97,98,99], flexible work programs [100,101], work–family balance [102,103], development opportunities [104,105], enriching job characteristics [85,106], work autonomy [107,108], participation in decision-making [109,110], trust in the organization [111,112], trust in coworkers [113], coworker support [114,115], supervisor support [116,117], team support [118,119], identification with the organization [120,121], self-esteem [122,123], and organizational citizenship behaviors [124,125].

Mirroring POS’ positive associations [110], consulted employees concerning their understanding and experience of POS and found multiple forms of lack of organizational support (e.g., interpersonal mistreatment by superiors, failure to provide proper resources, failure to provide feedback or supervision, ignoring employee input, failure to fairly reward performance, unfairly punishing employees, failure to provide on-the-job help, assigning unreasonable job demands, failure to give proper recognition/credit, failure to provide training or instruction, engaging in employee favoritism, treating employees with suspicion, failure to understand employee health issues, distrust of employee decision making, reduced benefits, failure to understand employee family or personal issues, failure to accommodate employee lifestyle, restricted work autonomy, forcing employee to use ineffective methods, meager base pay, failure to stop mistreatment of employees, few or no benefits, unfair performance review, inhibition of career development, and prohibition of socializing). Such relationships may reflect organizational philosophy, resources, and policies, as well as past and present leadership effects. Consequently, POS may be viewed as partly mirroring, in various degrees and combinations, the constructs referred to above. In other words, due to its comprehensiveness, POS may be considered a proxy for many social indicators within organizations.

POS positively relates to trust in leaders, attitudes, commitment, behaviors, and employee and organizational performance [48,86,87,126,127]. Likewise, POS is positively associated with job engagement, employee motivation, and job satisfaction [38,77,100,128,129]. Based on the above literature, we hypothesize:

Hypothesis 2 (H2).

POS mediates the relationships between green transformational leadership and transactional leadership with attitudes toward the natural environment, environmental commitment, pro-environmental behaviors, autonomous motivation, job engagement, job satisfaction, and employee performance.

2.3. Relationships Attitudes toward the Natural Environment-Environmental Commitment-Pro-Environmental Behaviors

Attitudes toward the natural environment involve caring for environmentally friendly activities and issues [130,131]. Environmental commitment entails a combination of pro-ecological values and beliefs [132]. Pro-environmental behaviors encompass learning and developing ideas that are operationalized in actions, processes, and offerings seeking to minimize environmental damage [133].

Prior research shows a positive association between attitudes toward the natural environment and environmental commitment [63,131,134,135,136,137]. Similarly, both attitudes toward the natural environment and environmental commitment are positively related to pro-environmental actions [71,138,139,140,141,142]. Likewise, attitudes toward the natural environment are crucial for organizational effectiveness [143,144]. Based on the above, we hypothesize:

Hypothesis 3 (H3).

Attitudes toward the natural environment, environmental commitment, and pro-environmental behaviors are positively interrelated.

2.4. Relationships Autonomous Motivation-Job Engagement-Job Satisfaction

Autonomous motivation entails competence, autonomy, and relatedness where the employees work innovatively, willingly, and proactively [57,145,146]. It is also influenced by socio-environmental features, job environment, and POS [147].

Autonomously motivated employees are actively involved in their actions with interest, pleasure [146], and autonomy [148]. Thus, the motivational process significantly stimulates individuals’ job engagement, autonomy, affiliation, and expertise [149,150,151,152,153,154]. Job satisfaction implies the pleasurable and positive emotional state resulting from the perceived relationship between what one wants from a job and what one perceives is offered [155]. Autonomous motivation enhances job satisfaction [51,57,147,156,157] through enthusiasm, enjoyment, efficacy, and encouragement in the workplace [158].

Job engagement entails undertaking job activities with enthusiasm, vigor, dedication, and absorption [159,160,161]. Research shows job engagement is positively related to job satisfaction [161,162,163,164,165,166]. Therefore, we hypothesize:

Hypothesis 4 (H4).

Autonomous motivation, job engagement, and job satisfaction are positively interrelated.

2.5. Model Antecedents of Employee Performance

Transformational leaders influence followers through creating self-interest and high standards and support with innovative, creative, and challenging tasks to achieve performance [75,167,168], whereas transactional leaders improve employee performance through rewards [169,170,171].

POS enhances job performance through social exchange [48,110,172]. Positive POS encourages employee proactiveness, confidence, and the development of employee capabilities [173]. POS also improves job engagement, perceived value of performance [172,174,175], and employee outcomes [172].

Autonomous motivation enhances employee performance [176,177,178,179]. Similarly, job engagement entails vigor to perform job duties and responsibilities thereby positively relating to employee performance [179,180]. Furthermore, research has shown a positive relationship between job satisfaction and employee performance [181,182].

Job enlargement, environmental initiatives, and management systems may enhance positive attitudes to protect the natural environment as well as job performance [131,183,184]. Likewise, environmental knowledge, awareness, and concern about the natural environment may enhance pro-environmental attitudes, environmental commitment, and pro-environmental behaviors [131] as well as employee performance [185,186,187,188,189,190]). Based on the above, we hypothesize:

Hypothesis 5 (H5).

Green transformational leadership, transactional leadership, POS, autonomous motivation, job engagement, job satisfaction, attitudes toward the natural environment, environmental commitment, and pro-environmental behaviors are positively associated with employee performance.

3. Method

This research examines a range of model configurations considering social sustainability (a) as an antecedent of environmental and economically related constructs, and (b) as a multiplicity of relationships among social, environmental, and economically related constructs (Figure 1, Figure 2, Figure 3 and Figure 4). This set of model configurations follows a gradient from a relatively moderate degree of order (e.g., including sequences such as attitudes toward the natural environment–environmental commitment–pro-environmental behavior, or motivation–job engagement–job satisfaction) to quasi-random construct arrangements, ending with model configurations fully determined at random. The ordering gradient stems not only from the diversity of constructs’ arrangements but also from the diversity in the social dimension, environmental and economic dimensions, and the resulting interactions. However, we focus on the social dimension. The rationale for examining this ordering gradient is that diverse ordering is found both within and between organizations. Since such an ordering gradient is ubiquitous, it is necessary to study it. Below, we explain the components of the studied construct orderings.

3.1. Moderate Construct Ordering: Social Sustainability as an Antecedent of Environmental and Economically Related Constructs

Based on the complexity of the social in organizations [191], we examine it (a) as the effects of factors external to the employee, in terms of green transformational and transactional leadership, and the effects of various configurations of constructs internal to the employee, including, among others, the following relationship sets: autonomous motivation–job engagement–job satisfaction; POS–autonomous motivation–job engagement–job satisfaction; green transformational leadership-POS–autonomous motivation–job engagement–job satisfaction, and (b) as alternative model configurations of first- and second-order latent variables. In the latter, the constructs examined were: SOC1: autonomous motivation, job engagement, and job satisfaction; SOC2: SOC1, and POS; SOC3: autonomous motivation, perceived organizational support, green transformational leadership, and transactional leadership; SOC4: autonomous motivation, job engagement, job satisfaction, perceived organizational support, and green transformational leadership; SOC5: autonomous motivation, job engagement, job satisfaction, perceived organizational support, green transformational leadership, and transactional leadership, and POS. ENV1: attitudes toward the natural environment; ENV2: ENV1, and environmental commitment; ENV3: ENV2, and pro-environmental behaviors. ECO1: employee performance; ECO2: ECO1, pro-environmental behaviors, job engagement, and job satisfaction; ECO3: ECO2, and autonomous motivation (Figure 2). The scope of second-order constructs was based on theory (e.g., Ref. [192] considering employee performance as degrees of employee value-adding activities) and on focus (e.g., the researchers’ interest expressed in degrees of construct comprehensiveness).

Table 1. Acronyms and meanings of first- and second-order constructs.

GreenTFL	Green Transformational Leadership
TRL	Transactional leadership
POS	Perceived organizational support
ATE	Attitudes toward the natural environment
EC	Environmental commitment
PB	Pro-environmental behaviors
AM	Autonomous motivation
JE	Job engagement
JS	Job satisfaction
EP	Employee performance
SOC1	Autonomous motivation, job engagement, and job satisfaction
SOC2	Autonomous motivation, job engagement, job satisfaction, and perceived organizational support
SOC3	Autonomous motivation, perceived organizational support, green transformational leadership, and transactional leadership
SOC4	Autonomous motivation, job engagement, job satisfaction, perceived organizational support, and green transformational leadership
SOC5	Autonomous motivation, job engagement, job satisfaction, perceived organizational support, green transformational leadership, and transactional leadership
ENV1	Attitudes toward the natural environment
ENV2	Attitudes toward the natural environment, and environmental commitment
ENV3	Attitudes toward the natural environment, environmental commitment, and pro-environmental behaviors
ECO1	Employee performance
ECO2	Employee performance, pro-environmental behaviors, job engagement, and job satisfaction
ECO3	Employee performance, pro-environmental behaviors, autonomous motivation, job engagement, and job satisfaction

shows the meanings of constructs’ acronyms.

3.2. Construct Ordering Less Than “Moderate”: Sustainability in Organizations as a Component of a Multiplicity of Relationships among Social, Environmental, and Economically Related Constructs

The social need not be an antecedent of the economic nor the environmental. In this section, we examine the social using both quasi-random and random construct configurations.

3.2.1. Quasi-Random Model Configurations

Employees’ and firms’ variability may be reflected in many construct combinations, which may be partly circumscribed and facilitated by a certain degree of ordering and looseness determined by employees’ and firms’ multiple identities and the number, type, and strength of construct relationships, as well as by the degrees of freedom originating both within and among firms. We observe boundary constraints and looseness resembling a loosely coupled system [193] or organized anarchy [194]. This seems to be the case for our sample, which includes respondents from a range of organizations, industries, and regions in the United States. Therefore, the dataset may be considered as one quasi-random instantiation of “the U.S. business organization”. For illustrative purposes, and in a simplified version of a loosely coupled system [193], some of the potential expressions of “the U.S. business organization” were operationalized in 14 model configurations in a 3 × 3 × 3 × 1 arrangement (Figure 3) in which the location of a specific construct in a given model configuration was assigned at random except for the construct located at the top left corner of a given model and the (final) outcome. The last two constructs were selected because they were considered key indicators of social, environmental, or economically related dimensions. Figure 3 shows only one version of the two selected constructs. In the other seven model configurations (not shown), the final outcome becomes the top left antecedent and what was the left top hand antecedent becomes the final outcome.

3.2.2. Random Model Configurations

The higher the magnitude of social and physical spaces in which employees interact, the higher the probability of randomness in such relationships.

The last model set includes 15 model configurations determined completely at random (Figure 4). In this set, a given model configuration includes three “social” constructs, selected at random from the total of social constructs, three “environmental” constructs, and employee performance. Our “U.S. business organization” data from a range of respondents, organizations, industries, and regions of the United States, at a specific time, make plausible relationships that have been less frequently studied, such as employee performance-job satisfaction, pro-environmental behaviors- attitudes toward the natural environment, and environmental commitment-green transformational leadership. In contrast to the model set described in the previous section, these model sets do not necessarily consider social sustainability constructs as antecedents, nor are they necessarily driven by economically related constructs as final outcomes.

3.3. Data Collection Procedure

Data were obtained using Amazon Mechanical Turk. The conceptual models and the questionnaire were formulated by the authors drawing from the extant literature. The survey respondents were followers. They received financial compensation for their participation. Items were randomized. Initially, we collected data from 50 respondents, reviewed such data, and determined that the questionnaire had no major issues. Thereafter, we collected sets of 250, 300, and 300 respondents. Respondents’ geographical distribution was diverse because they were from 34 out of 50 states. The states with more respondents were New York, Pennsylvania, Texas, Kansas, Maryland, California, Florida, Illinois, Washington, and New Hampshire. For each set, we verified whether or not the responses were acceptable. Only respondents producing questionnaires deemed acceptable were paid. We used three check questions (e.g., please answer Strongly Agree) to verify that respondents were reading and paying attention to the questions. Similarly, we considered only questionnaires in which the respondent spent an average response time of six seconds or more per question. We analyzed only fully answered questionnaires. Out of 900 respondents, 608 complied with the data “cleaning” criteria.

Amazon Mechanical Turk participants were demographically diverse and work for multiple organizations in various industries and regions. Thus, they may be more representative of a given country, as well as more reliable, than traditional data collection methods because of qualified participants and validation of the survey data [195,196].

3.4. Common Method Variance

Harman’s one-factor test [197] explained 32% of the variance. Using the correlation marker technique [198], that is, adjusting the original correlations by the lowest correlation, that is, between job satisfaction and age (0.0006) showed no statistical differences (p > 0.05) between the original and the corrected correlations.

In sum, considering the design questionnaire steps, the measures taken to clean the data, Harman’s one-factor test, and the correlation marker technique results suggest that common method variance was not an issue in this study.

3.5. Model Assessment

Items loadings were higher than 0.5 [199] and item cross-loadings were minor.

Table 2. Constructs’ reliability.

Constructs	Cronbach’s Alpha	Dijkstra/Henseler RhoA	Composite Reliability
GreenTFL	0.914	0.916	0.927
TRL	0.852	0.864	0.887
POS	0.897	0.899	0.916
AM	0.883	0.888	0.904
ATE	0.814	0.817	0.878
EC	0.888	0.890	0.912
PB	0.915	0.919	0.929
JE	0.830	0.833	0.873
JS	0.792	0.794	0.858
EP	0.739	0.754	0.836

Notes: N = 608. See Table 1 for acronyms’ meanings.

shows Cronbach’s alpha, rhoA, and composite reliability values. Values higher than 0.70 are considered acceptable for Cronbach alpha [200], rhoA [201], and composite reliability [202].

Adjusted determination coefficients for most constructs in models appearing in

Table A1. Moderate ordering: First-order constructs’ model configurations results.

	Model 1.1		Model 1.2		Model 1.3		Model 1.4
Paths	Path Coeff. (95% Bias Corrected CI)	Total Effects	Path Coeff. (95% Bias Corrected CI)	Total Effects	Path Coeff. (95% Bias Corrected CI)	Total Effects	Path Coeff. (95% Bias Corrected CI)	Total Effects
GreenTFL ⟶ POS	0.801 *** (0.738 to 0.855)	0.801 ***	0.802 *** (0.743 to 0.855)	0.802 ***	0.802 *** (0.741 to 0.855)	0.802 ***	0.802 *** (0.741 to 0.855)	0.802 ***
GreenTFL ⟶ AM		0.627 ***	0.478 *** (0.355 to 0.593)	0.738 ***	0.482 *** (0.353 to 0.604)	0.744 ***	0.482 *** (0.361 to 0.606)	0.745 ***
GreenTFL ⟶ ATE		0.578 ***	0.298 *** (0.161 to 0.421)	0.654 ***	0.216 ** (0.042 to 0.390)	0.656 ***	0.254 ** (0.085 to 0.409)	0.648 ***
GreenTFL ⟶ EC		0.545 ***	0.396 *** (0.251 to 0.539)	0.675 ***	0.318 *** (0.176 to 0.469)	0.677 ***	0.272 *** (0.143 to 0.416)	0.677 ***
GreenTFL ⟶ JE		0.635 ***	0.214 ** (0.078 to 0.357)	0.742 ***	0.388 *** (0.258 to 0.510)	0.75 ***	0.187 ** (0.037 to 0.340)	0.742 ***
GreenTFL ⟶ JS		0.623 ***	0.416 *** (0.313 to 0.520)	0.795 ***	0.572 *** (0.472 to 0.679)	0.802 ***	0.397 *** (0.283 to 0.524)	0.796 ***
GreenTFL ⟶ EP		0.553 ***	0.192 * (0.005 to 0.372)	0.604 ***	0.140 (−0.071 to 0.331)	0.605 ***	0.157 (−0.053 to 0.355)	0.601 ***
GreenTFL ⟶ PB		0.507 ***	0.145 ** (0.029 to 0.253)	0.572 ***	0.066 (−0.073 to 0.189)	0.575 ***	0.066 (−0.057 to 0.184)	0.577 ***
TRL ⟶ POS	0.038 (−0.024 to 0.107)	0.038	0.032 (−0.028 to 0.102)	0.032	0.032 (−0.030 to 0.101)	0.032	0.032 (−0.031 to 0.100)	0.032
TRL ⟶ AM		0.030	0.196 *** (0.130 to 0.269)	0.207 ***	0.206 *** (0.139 to 0.271)	0.217 ***	0.204 *** (0.136 to 0.266)	0.215 ***
TRL ⟶ ATE		0.027	0.104 ** (0.033 to 0.183)	0.118 **	0.069 (−0.008 to 0.153)	0.118 **	0.084 ** (0.008 to 0.168)	0.118 **
TRL ⟶ EC		0.026	0.137 *** (0.073 to 0.207)	0.182 ***	0.082 * (0.011 to 0.155)	0.184 ***	0.126 *** (0.051 to 0.202)	0.223 ***
TRL ⟶ JE		0.030	0.023 (−0.040 to 0.086)	0.104 **	−0.003 (−0.071 to 0.055)	0.102 ***	0.008 (−0.051 to 0.073)	0.105 **
TRL ⟶ JS		0.030	−0.032 (−0.084 to 0.025)	−0.005	−0.038 (−0.089 to 0.014)	−0.006	−0.058 (−0.121 to 0.011)	−0.007
TRL ⟶ EP		0.026	0.064 (−0.020 to 0.160)	0.08	0.055 (−0.048 to 0.164)	0.08	0.069 (−0.028 to 0.180)	0.066
TRL ⟶ PB		0.024	0.214 *** (0.138 to 0.294)	0.32 ***	0.186 *** (0.115 to 0.267)	0.322 ***	0.199 *** (0.124 to 0.280)	0.363 ***
POS ⟶ ATE	0.721 *** (0.653 to 0.778)	0.721 ***	0.444 *** (0.314 to 0.560)	0.444 ***	0.342 *** (0.224 to 0.459)	0.394 ***	0.403 *** (0.271 to 0.526)	0.434 ***
POS ⟶ EC	0.321 *** (0.222 to 0.439)	0.681 ***	0.042 (−0.102 to 0.183)	0.208 **	0.007 (−0.111 to 0.128)	0.234 ***	−0.027 (−0.170 to 0.118)	0.213 **
POS ⟶ PB	0.139 ** (0.041 to 0.240)	0.633 ***	0.048 (−0.056 to 0.152)	0.168 **	−0.018 (−0.126 to 0.079)	0.165 **	−0.003 (−0.110 to 0.097)	0.175 **
POS ⟶ AM	0.783 *** (0.735 to 0.819)	0.783 ***	0.324 *** (0.214 to 0.440)	0.324 ***	0.327 *** (0.210 to 0.448)	0.327 ***	0.328 *** (0.206 to 0.445)	0.328 ***
POS ⟶ JE	0.429 *** (0.320 to 0.538)	0.793 ***	0.350 *** (0.220 to 0.481)	0.458 ***		0.159 ***	0.280 *** (0.151 to 0.418)	0.456 ***
POS ⟶ JS	0.540 *** (0.435 to 0.645)	0.778 ***	0.333 *** (0.232 to 0.444)	0.402 ***		0.049 **	0.340*** (0.234 to 0.458)	0.403 ***
POS ⟶ EP	0.691 *** (0.610 to 0.753)	0.691 ***	0.513*** (0.355 to 0.674)	0.513 ***	0.484 *** (0.282 to 0.666)	0.484 ***	0.391 *** (0.174 to 0.600)	0.514 ***
ATE ⟶ EC	0.499 *** (0.388 to 0.601)	0.499 ***	0.375 *** (0.254 to 0.492)	0.375 ***	0.362 *** (0.241 to 0.485)	0.362 ***	0.351 *** (0.219 to 0.477)	0.351 ***
ATE ⟶ PB		0.362 ***		0.216 ***		0.197 ***	…	0.197 ***
ATE ⟶ JE							0.179 *** (0.086 to 0.276)	0.161 ***
ATE ⟶ JS							−0.003 (−0.098 to 0.097)	−0.040
ATE ⟶ EP					0.322 *** (0.201 to 0.462)	0.267 ***	0.314 *** (0.197 to 0.452)	0.258 ***
EC ⟶ PB	0.726 *** (0.630 to 0.816)	0.726 ***	0.576 *** (0.456 to 0.695)	0.576 ***	0.544 *** (0.412 to 0.669)	0.544 ***	0.562 *** (0.432 to 0.680)	0.562 ***
EC ⟶ JE							−0.071 (−0.203 to 0.050)	−0.051
EC ⟶ JS							−0.125 (−0.252 to 0.001)	−0.106
EC ⟶ EP					−0.074 (−0.240 to 0.078)	−0.152	−0.077 (−0.244 to 0.075)	−0.159 *
PB ⟶ JE							0.035 (−0.086 to 0.138)	0.035
PB ⟶ JS							0.034 (−0.086 to 0.147)	0.034
PB ⟶ EP					−0.145 (−0.301 to 0.001)	−0.145	−0.147 (−0.305 to 0.002)	−0.147
AM ⟶ JE	0.465 *** (0.348 to 0.575)	0.465 ***	0.335 *** (0.222 to 0.456)	0.335 ***	0.486 *** (0.363 to 0.625)	0.486 ***	0.326 *** (0.209 to 0.446)	0.333 ***
AM ⟶ ATE					0.043 (−0.125 to 0.210)	0.161	0.094 (−0.072 to 0.266)	0.094
AM ⟶ EC					0.289 *** (0.153 to 0.445)	0.316 ***	0.268 *** (0.136 to 0.412)	0.301***
AM ⟶ JS		0.140 ***		0.051 **		0.149 ***	0.260 *** (0.144 to 0.373)	0.234 ***
AM ⟶ EP					0.031 (−0.176 to 0.249)	0.055	0.086 (−0.110 to 0.277)	0.042
AM ⟶ PB					0.149 * (0.021 to 0.282)	0.342 ***	0.178 *** (0.062 to 0.302)	0.348 ***
JE ⟶ JS	0.301 *** (0.181 to 0.417)	0.301 ***	0.151 (0.055 to 0.256)	0.151 **	0.306 *** (0.186 to 0.414)	0.306 ***
JE ⟶ ATE					0.27 *** (0.145 to 0.394)	0.244 ***
JE ⟶ EC					−0.039 (−0.147 to 0.075)	0.022
JE ⟶ EP					0.033 (−0.141 to 0.187)	0.049
JE ⟶ PB					0.036 (−0.061 to 0.123)	0.055
JS ⟶ EP					0.054 (−0.099 to 0.183)	0.054
JS ⟶ ATE					−0.086 (−0.202 to 0.027)	−0.086
JS ⟶ EC					−0.088 (−0.203 to 0.008)	−0.119 *
JS ⟶ PB					0.022 (−0.060 to 0.103)	−0.043

Notes: N = 608; *** p < 0.001; ** p < 0.01; * p < 0.05; 95% bias corrected confidence intervals appear in parentheses. See Table 1, in main text, for acronyms’ meanings.

(see Appendix A) varied slightly and were, on average, higher than 0.700, except for employee performance (ECO1), which was on average about 0.500. Configurations of second-order latent variables (see

Table A2. Moderate ordering: First and second order constructs’ model configurations results.

	Model 2.1	Model 2.2	Model 2.3	Model 2.4	Model 2.5	Model 2.6	Model 2.7	Model 2.8
Paths	Path Coeff. (95% Bias Corrected CI)	Path Coeff. (95% Bias Corrected CI)	Path Coeff. (95% Bias Corrected CI)	Path Coeff. (95% Bias Corrected CI)	Path Coeff. (95% Bias Corrected CI)	Path Coeff. (95% Bias Corrected CI)	Path Coeff. (95% Bias Corrected CI)	Path Coeff. (95% Bias Corrected CI)
GreenTFL ⟶ ECO1	0.128 (−0.098 to 0.344)				0.108 (−0.126 to 0.340)			0.108 (−0.131 to 0.333)
GreenTFL ⟶ ECO3		0.343 *** (0.265 to 0.418)				0.343 *** (0.269 to 0.419)
GreenTFL ⟶ ENV2		0.455 *** (0.317 to 0.572)				0.455 *** (0.320 to 0.569)
GreenTFL ⟶ ENV3	0.305 *** (0.196 to 0.435)				0.313 *** (0.201 to 0.446)			0.313 *** (0.193 to 0.443)
GreenTFL ⟶ POS	0.802 *** (0.736 to 0.852)	0.803 *** (0.741 to 0.854)				0.803 *** (0.738 to 0.855)
GreenTFL ⟶ SOC1	0.483 *** (0.355 to 0.604)
GreenTFL ⟶ SOC2					0.839 *** (0.792 to 0.875)			0.829 *** (0.783 to 0.868)
TRL ⟶ ECO1	0.036 (−0.060 to 0.144)		0.004 (−0.090 to 0.113)		0.011 (−0.085 to 0.119)			0.011 (−0.084 to 0.124)
TRL ⟶ ECO3		0.152 *** (0.101 to 0.206)				0.152 *** (0.099 to 0.205)
TRL ⟶ ENV2		0.157 *** (0.083 to 0.239)				0.157 *** (0.083 to 0.242)
TRL ⟶ ENV3	0.240 *** (0.158 to 0.330)		0.235 *** (0.157 to 0.323)		0.208 *** (0.135 to 0.290)			0.208 *** (0.133 to 0.288)
TRL ⟶ POS	0.034 (−0.026 to 0.106)	0.032 (−0.029 to 0.101)				0.032 (−0.032 to 0.102)
TRL ⟶ SOC1	0.127 *** (0.080 to 0.173)
TRL ⟶ SOC2					0.109 *** (0.063 to 0.164)			0.109 *** (0.056 to 0.167)
TRL ⟶ SOC4			0.496 *** (0.399 to 0.580)
POS ⟶ ECO1	0.467 *** (0.217 to 0.693)
POS ⟶ ECO3		0.277 *** (0.192 to 0.362)				0.277 *** (0.196 to 0.362)
POS ⟶ ENV2		0.318 *** (0.196 to 0.445)				0.318 *** (0.196 to 0.441)
POS ⟶ ENV3	0.122 (−0.010 to 0.253)
POS ⟶ SOC1	0.417 *** (0.305 to 0.539)
SOC1 ⟶ ECO1	0.084 (−0.186 to 0.323)
SOC2 ⟶ ECO1					0.568 *** (0.353 to 0.774)			0.568 *** (0.356 to 0.786)
SOC4 ⟶ ECO1			0.662 *** (0.450 to 0.832)
SOC3 ⟶ ECO2							0.671*** (0.612 to 0.731)
SOC5 ⟶ ECO1				0.684 *** (0.468 to 0.858)
SOC1 ⟶ ENV3	0.327 *** (0.168 to 0.478)
SOC2 ⟶ ENV3					0.428 *** (0.277 to 0.561)			0.428 *** (0.273 to 0.567)
SOC4 ⟶ ENV3			0.724 *** (0.637 to 0.800)
SOC5 ⟶ ENV3				0.856 *** (0.805 to 0.894)
SOC3 ⟶ ENV2							0.843*** (0.796 to 0.880)
ENV2 ⟶ ECO3		0.299 *** (0.227 to 0.369)				0.299 *** (0.231 to 0.369)
ENV2 ⟶ ECO2							0.301*** (0.234 to 0.364)
ENV3 ⟶ ECO1	0.042 (−0.154 to 0.248)		0.026 (−0.169 to 0.233)	0.004 (−0.182 to 0.220)	0.032 (−0.158 to 0.243)			0.032 (−0.159 to 0.245)

Notes: N = 608; *** p < 0.001; 95% bias corrected confidence intervals appear in parentheses. See Table 1, in main text, for acronyms’ meanings.

in Appendix A) show that the more encompassing a given construct, the higher the adjusted determination coefficient (e.g., 0.471 for ECO1 in Model 2.3 and Model 2.4 to 0.899 for ECO3 in Model 2.2 and Model 2.6). However, the location of a construct in a particular model may significantly influence its determination coefficient (e.g., ECO1, a construct just referring to job tasks, has a higher determination coefficient, 0.471 in Model 2.3 and Model 2.4 than does SOC4, a far more encompassing construct, 0.245 in Model 2.3). These results confirm that the coefficient of determination is a function of both the nature/content of the construct, that is, whether or not the construct is encompassing and its position in a given model owing to the number and strength of direct and indirect effects. While this is obvious, it is usually not given the importance that it deserves partly because it requires making valid model comparisons across studies. Overall, adjusted coefficients of determination were moderate to high. Coefficients of determination for the complete set of model configurations studied will be provided upon request.

Variance inflation factors were below 10.0 [203]. Thus, multicollinearity was not an issue. The Heterotrait–Monotrait (HTMT) ratio criterion, utilizing 5000 bootstrapping samples, was used to evaluate discriminant validity. HTMT values of all binary construct combinations were lower than 1, which is acceptable [204]. Mardia’s multivariate test (https://webpower.psychstat.org/models/kurtosis, accessed on 22 February 2022) was statistically significant for both skewness and kurtosis (p < 0.001). Therefore, path coefficients and 95% bias-corrected confidence intervals were generated using 5000 bootstrapping samples.

The effects of social desirability, age, job experience, race, gender, marital status, educational level, industry, and job tenure were controlled for in all analyses.

3.6. Demographics

Forty-two percent of the respondents were female and 58% were male; 90% were married and 7% were single (the rest were either divorced, separated, or widowed). In terms of education, 22% had graduate degrees, 68% had bachelor’s degrees, 3% had associate degrees, and the rest had high school. The respondents worked in banking and financial services (22%); education, professional and scientific services (16%); manufacturing, mining, and quarrying (14%); health care and pharmaceuticals (9%); accounting and consulting (12%); retail and wholesale trade (6%); or other (21%). Regarding job experience, 2% had less than 1 year of experience, 33% had 1 to 4 years, 36% had 4 to 8 years, 20% had 8 to 12 years, and 9% had more than 12 years of experience.

3.7. Measures

In selecting the constructs’ measures we relied, whenever possible, on scales with proven psychometric properties. What follows defines the constructs, refers to the scales used to measure them, and provides an example item. The questionnaire used in this study will be provided by the first author upon request.

Green transformational leadership refers to “leadership behaviors that motivate followers to achieve environmental goals and inspire followers to perform beyond expected levels of environmental performance” [43], p. 109. Green transformational leadership was measured with 8 items from a scale developed by [205] and 5 items from a scale developed by [133]. We modified some items to reflect the green nature of green transformational leadership. An example item is, “My supervisor speaks convincingly about his vision of the natural environment”. Transactional leadership entails “a leader exchanging something of economic, political, or psychological value with a follower…these exchanges are based on the leader identifying performance requirements and clarifying the conditions under which rewards are available for meeting these requirements” [206], p. 1861. Transactional leadership was measured with 8 items from a scale developed by [205]. An example item is, “My supervisor provides assistance in exchange for my effort”.

POS refers to “the extent to which employees perceive that their contributions are valued by their organization and that the firm cares about their well-being” [207], p. 501. POS was measured with 9 items from a scale developed by [207]. An example item is, “The organization values my contribution to its well-being”. Autonomous motivation is defined as “active engagement with tasks that people find interesting and that, in turn, promote growth; it presupposes that human are naturally active and that there are natural tendencies toward development that require nutriments to function effectively” [57], p. 233. Autonomous motivation was measured with 11 items from a scale developed by [208]. An example item is, “I work because my work is my life, and I don’t want to fail”. Job engagement refers to “a positive, fulfilling, work-related state of mind that is characterized by vigor, dedication, and absorption; it is not a momentary and specific state; rather, it is a more persistent and pervasive affective-cognitive state that is not focused on any particular object, event, individual, or behavior.” [160], p. 74. Job engagement was measured with 7 items from a scale developed by [209]. An example item is, “At work, I feel bursting with energy”. Job satisfaction refers to “all characteristics of the job itself and the work environment which individuals find rewarding, fulfilling, and satisfying, or frustrating and unsatisfying” [210], p. 255. Job satisfaction was measured with 5 items from a scale developed by [211]. An example item is, “Overall, I am satisfied with my job”.

Attitudes toward the natural environment refer to “the collection of beliefs, affect, and behavioral intentions a person holds regarding environmentally related activities or issues” [130], p. 31. We measure attitudes toward the natural environment with 4 items from a scale developed by [212]. An example item is, “I have strong positive feelings about eco-friendly behaviors”. Environmental commitment refers to “environmental behaviors entailing a wide variety of behaviors such as involvement in specific environmental practices, thinking about environmental improvements, and offering suggestions to be more environmentally friendly” [213], p. 37. Environmental commitment was measured with 6 items from a scale developed by [214]. An example item is, “I feel a sense of duty to support the environmental efforts of my company”. Pro-environmental behaviors were conceptualized as “a broad set of environmentally-responsible activities such as learning more about the environment, developing and applying ideas for reducing the company’s environmental impact, developing green processes and products, recycling and reusing, and questioning practices that hurt the environment” [133], p. 81. Pro-environmental behaviors were measured with 11 items from a scale developed by [133]. An example item is, “At work, I apply new ideas for reducing our impact on the natural environment”.

Employee performance refers to “the total expected value to the organization of the discrete behavioral episodes that an individual carries out over a standard period” [192], p. 82. Employee performance, as a first order construct, was measured with 4 items from a scale developed by [208]. An example item is, “I always meet all the formal performance requirements of my job”.

Social desirability was defined as “participants’ tendency to give ‘desirable’ answers in response to attitudinal questionnaires to put forward a more socially acceptable self-image” [215], p. 2. The social desirability scale was measured with 5 items from a scale developed by [215]. An example item is “Would you smile at people every time you meet them?”

4. Results and Discussion

Table 3. Means and correlations.

Variables	Mean	Std. Dev.	(GreenTFL)	(TRL)	(POS)	(AM)	(JE)	(JS)	(ATE)	(EC)	(PB)	(EP)
GreenTFL	5.475	0.791	1.000
TRL	5.158	0.979	0.453 ***	1.000
POS	5.499	0.832	0.816 ***	0.395 ***	1.000
AM	5.448	0.797	0.831 ***	0.574 ***	0.787 ***	1.000
JE	5.503	0.818	0.793 ***	0.449 ***	0.796 ***	0.799 ***	1.000
JS	5.587	0.813	0.796 ***	0.359 ***	0.791 ***	0.756 ***	0.736 ***	1.000
ATE	5.545	0.858	0.702 ***	0.410 ***	0.719 ***	0.670 ***	0.702 ***	0.601 ***	1.000
EC	5.431	0.952	0.776 ***	0.530 ***	0.711 ***	0.779 ***	0.701 ***	0.635 ***	0.757 ***	1.000
PB	5.334	0.929	0.738 ***	0.624 ***	0.668 ***	0.783 ***	0.685 ***	0.623 ***	0.683 ***	0.852 ***	1.000
EP	5.594	0.802	0.619 ***	0.302 ***	0.671 ***	0.580 ***	0.578 ***	0.579 ***	0.602 ***	0.506 ***	0.456 ***	1.000

Notes: N = 608; *** p < 0.001. See Table 1 for acronyms’ meanings.

shows means and constructs’ correlations.

4.1. Moderate Ordering: Social Sustainability Constructs as Antecedents of Environmental and Economically Related Constructs

4.1.1. Relationships among First Order Constructs

Table A1 (see Appendix A) shows path coefficients, 95% bias-corrected confidence intervals, and total effects for Models 1.1, 1.2, 1.3, and 1.4 (see Figure 1). Green transformational leadership relationships with POS (β = 0.801, 0.802, and 0.802 for Models 1.1, 1.2, and 1.3, respectively) and job satisfaction (β = 0.416, 0.572 for Models 2.2 and 2.3, respectively) were strong; its relationships with environmental commitment (β = 0.396 and 0.318 for Models 1.2 and 1.3, respectively) and autonomous motivation (β = 0.478, and 0.482 for Models 1.2 and 1.3, respectively) were moderate; relationships with job engagement (β = 0.214 and 0.388 for Models 1.2 and 1.3, respectively), attitudes toward the natural environment (β = 0.298 and 0.296 for Models 1.2 and 1.3, respectively), and pro-environmental behaviors (β = 0.145 for Model 1.2 and nonsignificant for Models 1.3., and 1.4, respectively) were relatively weak. Green transformational leadership was not directly related to employee performance. However, its total effects on employee performance were highly significant (p < 0.001). All green transformational leadership’s total effects but one (with environmental commitment) were moderate to high, ranging from 0.575 to 0.802. In contrast, relationships between transactional leadership and other constructs were nonsignificant or weak. Similarly, transactional leadership’s total effects on POS, job satisfaction, pro-environmental behaviors, and employee performance were not significant, whereas those on autonomous motivation, job engagement, attitudes toward the natural environment, and environmental commitment were weak (e.g., β ranging from 0.023 to 0.217).

As expected, green transformational leadership had more frequent and stronger positive relationships than transactional leadership with both other social constructs and with environmentally and economically related constructs [43,216,217,218]. The nature of “traditional” transformational leadership implies enhancing quality social relationships and adapting to change and, since the “greening” of businesses has been extant for several years, it may be expected that effective “traditional” transformational leaders will increasingly consider environmental matters. In other words, “traditional” transformational leaders may have become, or are becoming, green transformational leaders. In contrast, the narrow, task-focused, quid pro quo character of transactional leadership frequently resulted in nonsignificant relationships with other social, environmental, and economically related organizational constructs. Thus, it may be expected that, as the greening of business proceeds, transactional leaders will be negatively selected. These results support H1.

POS’ path coefficients with attitudes toward the natural environment (β = 0.721, 0.444, and 0.342 for Models 1.1, 1.2, and 1.3, respectively), autonomous motivation (β = 0.783, 0.324, and 0.327 for Models 1.1, 1.2, and 1.3, respectively), job engagement (β = 0.429, and 0.350 for Models 1.1, and 1.2, respectively), job satisfaction (β = 0.540, and 0.333 for Models 1.1 and 1.2, respectively), and employee performance (β = 0.583, 0.486, and 0.466 for Models 1.1, 1.2, and 1.3, respectively) were moderate to strong. The relationships of POS with environmental commitment (β = 0.321 for Model 1.1) and pro-environmental behaviors (β = 0.139 for Model 1.1) were relatively weak, and those relationships were not significant for Models 1.2, and 1.3, although the total effects were significant. The relationships of POS with environmental commitment and pro-environmental behaviors were fully mediated in Models 1.2 and 1.3.

In concordance with POS’s broad encompassing nature reflected in being constitutive or positively or negatively associated with a large set of constructs in organizations (see Section 2.2), results indicate that POS strongly mediates green transformational leadership effects with, and has positive moderate to large total effects on, autonomous motivation, job engagement, job satisfaction, attitudes toward the natural environment, environmental commitment, pro-environmental behaviors, and employee performance. However, the total effects of POS decrease when considering green transformational leadership’s direct relationships with outcome constructs, whereas adjusted coefficients of determination of the latter increase (see Table A1 in Appendix A). These results reaffirm green transformational leadership’s impacts, alluded to above. As previously noted, green transformational leadership related positively and strongly (β = 0.801, 0.802, and 0.802 for Models 1.1, 1.2, and 1.3, respectively) to POS. These findings seem to be in line with prior research [219,220,221,222].

POS is a central social construct, which shows moderate relationships with other “traditional” social constructs (e.g., job engagement, job satisfaction) and relatively weak relationships with the environmental constructs examined. Overall, findings suggest that internalization of environmental responsibilities by employees requires further development. Overall, these results support H2.

Path coefficients of relationships attitudes toward the natural environment and environmental commitment (β = 0.499,.375, and 0.362 for Models 1.1, 1.2, and 1.3, respectively) and environmental commitment and pro-environmental behaviors (β = 0.726, 0.576, and 0.544 for Models 1.1, 1.2, and 1.3, respectively) were positive moderate to strong. These results partly agree with extant research [217,223,224]. Path coefficients and total effects decreased, as the number of relationships in the models increased, whereas, in most cases, adjusted determination coefficients increased. The latter reflects mostly the indirect effects of green transformational leadership and POS. Most relationships between environmentally related constructs with job engagement, job satisfaction, and employee performance (see Table A1, Models 1.3 and 1.4 in Appendix A) were not significant. Environmentally related constructs were weakly related to otherwise strong social constructs, such as POS. This suggests that to increase employees’ contributions to organizational sustainability, it requires employees’ environmental concerns and behaviors to relate to “traditional” social constructs more strongly. Overall, the relationships attitudes toward the natural environment–environmental commitment–pro-environmental behaviors support H3.

The level of concern for the environment in the United States is lower than in most European, Latin American, and African countries [225]. Environmental issues are not considered a high priority [226]. Opinions, support, and behaviors about the environment are substantially polarized ideologically [225,227] and geographically [228]. Ideological and geographical variability may generate canceling-out effects, making it more difficult to find strong relationships at the national level. Citizens’ stable and relatively low level of concern for environmental matters may be an obstacle to internalizing the natural environment in organizational processes. However, organizations that wish to consider environmental matters seriously may play a key role in the future by bridging, and positively changing, the individual level (e.g., denial of direct effects of environmental changes on the individual) and the societal level (e.g., individuals not personally involved in solutions).

Autonomous motivation related positively moderately (β = 0.465, 0.335, and 0.486 for Models 1.1, 1.2, and 1.3, respectively) to job engagement, and job engagement related positively moderately low (β = 0.301, 0.151, and 0.306 for Models 1.1, 1.2, and 1.3, respectively) to job satisfaction. These findings appear to agree with past studies [148,151,229,230]. The decrease in path coefficient from Models 1.1 and 1.2 is explained by the direct effects, in Model 1.2, of green transformational leadership on both job engagement and job satisfaction. Thus, the (social) relationships among autonomous motivation–job engagement–job satisfaction support H4.

Most constructs were not associated with employee performance. Only the relationships between green transformational leadership–employee performance (β = 0.145 in Model 1.2), POS-employee performance (β = 0.691, 0.513, 0.484, and 0.391 for Models 1.1, 1.2, 1.3, and 1.4, respectively), and attitudes toward the natural environment-employee performance (β = 0.322, and 0.314 for Model 1.3 and 1.4, respectively) were significant. These findings resemble prior research [41,49,172]. Employee performance’s poor or absent relationships may stem from its narrow focus on job task compliance. Overall, H5 was not supported by the data. Employee performance’s alternative formulations appear below.

In sum, the social in organizations seems differentiated in its effects. Green transformational leadership and POS showed multiple, and at times strong, relationships. Attitudes toward the natural environment, environmental commitment, and pro-environmental behaviors, as well as motivation, job engagement, and job satisfaction, had moderate relationships, in general, although those relationships were stronger among construct sets (e.g., environmental, social).

4.1.2. Relationships among First and Second Order Constructs: Alternative Formulations of Social, Environmental, and Economically Related Constructs

Table A2 (see Appendix A) displays the results of the various configurations involving second-order latent variables for the social, environmental, and economically related constructs that were studied. Overall, in most relationships, the more encompassing a certain second-order latent variable (e.g., ECO1 to ECO2 to ECO3 or SOC1 to SOC2 to SOC3 to SOC4 to SOC5 or ENV1 to ENV2 to ENV3), the higher the path coefficient. For example, this is observed in the relationships between transactional leadership and ECO1, ECO2, and ECO3; between green transformational leadership and SOC1 and SOC2; between SOC1, SOC2, and SOC3 and ECO1, and between SOC1, SOC2, SOC3, and SOC4 and ENV3. As discussed above, this is in line with the comprehensiveness of the social construct, as compared to that of the environmental and economically related constructs, in that the more content of the second-order latent variable, the higher the likelihood of a stronger relationship. However, there are a few exceptions. For example, the path coefficient in POS-ECO1 (β = 0.467) in Model 2.1 decreases in the relationship POS-ECO2 (β = 0.277) in Model 2.2, which may be due to a more diverse set encompassed in ECO2. Similarly, the path coefficient in POS-ENV2 (β = 0.318) in Model 2.2 decreases in the relationship POS-ENV3 (β = 0.122) in Model 2.1 possibly because part of the effect of POS, in Model 2.1, goes to SOC1, whereas SOC1 is not present in Model 2.2. Various operationalizations of the social, environmental, and economically related constructs resulted in varying relationship strengths. Nonetheless, these results reaffirm the centrality of the social dimension in organizations

4.2. Construct Ordering Less Than “Moderate”: Social Sustainability in Organizations as a Multiplicity of Relationships among Social, Environmental, and Economically Related Constructs

This section refers, and discusses, results for both quasi-random and random model configurations.

4.2.1. Quasi-Random Model Configurations

Table A3. Quasi-random model configurations results.

Model 3.1					Model 3.2
Paths	Path Coeff.	p Values	Total Effects	p Values	Paths	Path Coeff.	p Values	Total Effects	p Values
POS ⟶ AM	0.203	0.001	0.676	0.000	PB ⟶ AM	0.166	0.004	0.689	0.000
POS ⟶ ATE	0.720	0.000	0.720	0.000	PB ⟶ ATE	0.686	0.000	0.686	0.000
POS ⟶ EC	0.065	0.390	0.658	0.000	PB ⟶ EC	0.532	0.000	0.838	0.000
POS ⟶ JE	0.250	0.000	0.709	0.000	PB ⟶ JE	0.026	0.645	0.595	0.000
POS ⟶ JS	…	…	0.433	0.000	PB ⟶ JS	…	…	0.413	0.000
POS ⟶ EP	0.403	0.000	0.633	0.000	PB ⟶ EP	−0.075	0.253	0.380	0.000
POS ⟶ PB	−0.005	0.915	0.590	0.000	PB ⟶ GreenTFL	0.306	0.000	0.638	0.000
POS ⟶ GreenTFL	0.338	0.000	0.677	0.000	PB ⟶ POS	0.010	0.841	0.565	0.000
POS ⟶ TRL	−0.032	0.686	0.359	0.000	PB ⟶ TRL	0.660	0.000	0.636	0.000
ATE ⟶ AM	−0.034	0.432	0.395	0.000	ATE ⟶ AM	−0.006	0.890	0.392	0.000
ATE ⟶ EC	0.396	0.000	0.588	0.000	ATE ⟶ EC	0.243	0.000	0.343	0.000
ATE ⟶ JE	0.175	0.001	0.407	0.000	ATE ⟶ JE	0.218	0.000	0.567	0.000
ATE ⟶ JS	0.602	0.000	0.602	0.000	ATE ⟶ JS	0.602	0.000	0.602	0.000
ATE ⟶ EP	0.252	0.001	0.320	0.000	ATE ⟶ EP	0.446	0.000	0.664	0.000
ATE ⟶ PB	0.066	0.158	0.526	0.000	ATE ⟶ GreenTFL	0.143	0.002	0.498	0.000
ATE ⟶ GreenTFL	0.197	0.000	0.444	0.000	ATE ⟶ POS	0.208	0.000	0.628	0.000
ATE ⟶ TRL	−0.003	0.960	0.316	0.000	ATE ⟶ TRL	−0.048	0.418	−0.048	0.505
JS ⟶ AM	0.183	0.000	0.349	0.000	JS ⟶ AM	0.235	0.000	0.429	0.000
JS ⟶ EC	−0.030	0.575	0.154	0.002	JS ⟶ EC	−0.036	0.439	0.083	0.024
JS ⟶ JE	0.104	0.027	0.268	0.000	JS ⟶ JE	0.171	0.001	0.432	0.000
JS ⟶ EP	0.113	0.108	0.113	0.108	JS ⟶ EP	0.362	0.000	0.362	0.000
JS ⟶ PB	0.026	0.542	0.207	0.000	JS ⟶ GreenTFL	0.442	0.000	0.491	0.000
JS ⟶ GreenTFL	0.385	0.000	0.390	0.000	JS ⟶ POS	0.297	0.000	0.515	0.000
JS ⟶ TRL	−0.029	0.673	0.096	0.085	JS ⟶ TRL	−0.086	0.184	−0.054	0.365
EP ⟶ AM	0.028	0.528	0.053	0.291	EP ⟶ AM	0.077	0.052	0.133	0.002
EP ⟶ EC	…	…	0.022	0.345	EP ⟶ EC	…	…	0.033	0.037
EP ⟶ JE	−0.013	0.764	0.011	0.836	EP ⟶ JE	0.032	0.426	0.105	0.034
EP ⟶ PB	−0.063	0.063	−0.032	0.441	EP ⟶ GreenTFL	0.135	0.004	0.135	0.004
EP ⟶ GreenTFL	0.047	0.336	0.047	0.336	EP ⟶ POS	…	…	0.057	0.015
EP ⟶ TRL	0.021	0.743	0.038	0.542	EP ⟶ TRL	0.055	0.333	0.059	0.273
GreenTFL ⟶ AM	0.290	0.000	0.452	0.000	GreenTFL ⟶ AM	0.328	0.000	0.365	0.000
GreenTFL ⟶ EC	0.473	0.000	0.473	0.000	GreenTFL ⟶ EC	0.243	0.000	0.243	0.000
GreenTFL ⟶ JE	0.153	0.049	0.280	0.000	GreenTFL ⟶ JE	0.196	0.010	0.318	0.000
GreenTFL ⟶ PB	0.074	0.253	0.460	0.000	GreenTFL⟶ POS	…	…	0.169	0.000
GreenTFL ⟶ TRL	0.133	0.143	0.351	0.000	GreenTFL ⟶ TRL	0.024	0.782	0.026	0.752
EC ⟶ AM	0.217	0.000	0.295	0.000	EC ⟶ AM	0.136	0.014	0.137	0.014
EC ⟶ JE	−0.033	0.584	0.060	0.335	EC ⟶ JE	−0.048	0.490	0.002	0.977
EC ⟶ PB	0.509	0.000	0.645	0.000	EC ⟶ POS	…	…	0.035	0.240
EC ⟶ TRL	0.460	0.000	0.460	0.000	EC ⟶ TRL	0.010	0.913	0.010	0.913
TRL ⟶ AM	0.169	0.000	0.169	0.000	TRL ⟶ AM	0.119	0.000	0.119	0.000
TRL ⟶ JE	…	…	0.053	0.000	TRL⟶JE	…	…	0.044	0.001
TRL ⟶ PB	0.189	0.000	0.217	0.000	TRL ⟶ POS	−0.050	0.170	−0.009	0.781
AM ⟶ JE	0.316	0.000	0.316	0.000	AM ⟶ JE	0.364	0.000	0.364	0.000
AM ⟶ PB	0.160	0.013	0.167	0.008	AM ⟶ POS	0.252	0.000	0.343	0.000
JE ⟶ PB	0.022	0.635	0.022	0.635	JE ⟶ POS	0.248	0.000	0.248	0.000
Model 3.3					Model 3.4
Paths	Path Coeff.	p Values	Total Effects	p Values	Paths	Path Coeff.	p Values	Total Effects	p Values
JS ⟶ AM	0.770	0.000	0.770	0.000	PB ⟶ AM	0.782	0.000	0.782	0.000
JS ⟶ ATE	−0.057	0.318	0.607	0.000	PB ⟶ ATE	0.353	0.000	0.699	0.000
JS ⟶ EC	−0.103	0.064	0.648	0.000	PB ⟶ EC	0.58	0.000	0.864	0.000
JS ⟶ JE	0.298	0.000	0.746	0.000	PB ⟶ JE	0.164	0.006	0.698	0.000
JS ⟶ EP	0.237	0.002	0.582	0.000	PB ⟶ JS	0.065	0.294	0.644	0.000
JS ⟶ PB	0.026	0.539	0.633	0.000	PB ⟶ EP	−0.051	0.532	0.465	0.000
JS ⟶ GreenTFL	0.332	0.000	0.800	0.000	PB ⟶ GreenTFL	0.257	0.000	0.759	0.000
JS ⟶ POS	0.218	0.000	0.787	0.000	PB ⟶ POS	0.015	0.790	0.696	0.000
JS ⟶ TRL	…	…	0.426	0.000	PB ⟶ TRL	…	…	0.433	0.000
AM ⟶ ATE	0.070	0.342	0.520	0.000	AM ⟶ ATE	−0.068	0.369	0.341	0.000
AM ⟶ EC	0.307	0.000	0.699	0.000	AM ⟶ EC	0.099	0.093	0.278	0.000
AM ⟶ JE	0.572	0.000	0.582	0.000	AM ⟶ JE	0.715	0.000	0.683	0.000
AM ⟶ EP	0.227	0.020	0.356	0.000	AM ⟶ JS	0.406	0.000	0.670	0.000
AM ⟶ PB	0.156	0.014	0.682	0.000	AM ⟶ EP	0.377	0.000	0.591	0.000
AM ⟶ GreenTFL	0.503	0.000	0.558	0.000	AM ⟶ GreenTFL	0.577	0.000	0.642	0.000
AM ⟶ POS	0.168	0.007	0.482	0.000	AM ⟶ POS	0.222	0.000	0.666	0.000
AM ⟶ TRL	0.553	0.000	0.553	0.000	AM ⟶ TRL	0.554	0.000	0.554	0.000
TRL ⟶ ATE	0.064	0.105	0.072	0.123	TRL ⟶ ATE	−0.031	0.406	−0.069	0.165
TRL ⟶ EC	0.146	0.002	0.151	0.003	TRL ⟶ EC	−0.013	0.679	−0.040	0.290
TRL ⟶ JE	0.019	0.566	0.019	0.566	TRL ⟶ JE	−0.058	0.143	−0.058	0.143
TRL ⟶ EP	−0.015	0.785	−0.011	0.849	TRL ⟶ JS	−0.07	0.046	−0.123	0.007
TRL ⟶ PB	0.189	0.000	0.274	0.000	TRL ⟶ EP	−0.02	0.744	−0.039	0.554
TRL ⟶ GreenTFL	0.021	0.520	0.019	0.552	TRL ⟶ GreenTFL	−0.08	0.033	−0.088	0.030
TRL ⟶ POS	−0.043	0.139	−0.028	0.404	TRL ⟶ POS	−0.064	0.051	−0.116	0.021
JE ⟶ ATE	0.320	0.000	0.384	0.000	JE ⟶ ATE	0.279	0.000	0.372	0.000
JE ⟶ EC	0.055	0.360	0.103	0.080	JE ⟶ EC	0.005	0.924	0.046	0.360
JE ⟶ EP	0.237	0.001	0.237	0.001	JE ⟶ JS	…	…	0.175	0.000
JE ⟶ PB	0.023	0.618	0.086	0.107	JE ⟶ EP	0.331	0.000	0.331	0.000
JE ⟶ GreenTFL	…	…	0.029	0.061	JE ⟶ GreenTFL	…	…	0.061	0.012
JE ⟶ POS	0.208	0.000	0.302	0.000	JE ⟶ POS	0.253	0.000	0.371	0.000
EP ⟶ ATE	0.236	0.000	0.268	0.000	EP ⟶ ATE	0.256	0.000	0.280	0.000
EP ⟶ EC	−0.037	0.465	0.042	0.383	EP ⟶ EC	0.009	0.804	0.065	0.062
EP ⟶ PB	−0.065	0.060	−0.018	0.687	EP ⟶ JS	…	…	0.119	0.000
EP ⟶ GreenTFL	0.122	0.005	0.122	0.005	EP ⟶ GreenTFL	0.186	0.000	0.186	0.000
EP ⟶ POS	0.149	0.008	0.206	0.000	EP ⟶ POS	0.172	0.004	0.254	0.000
GreenTFL ⟶ ATE	0.258	0.002	0.258	0.002	GreenTFL ⟶ ATE	0.13	0.074	0.130	0.074
GreenTFL ⟶ EC	0.395	0.000	0.425	0.000	GreenTFL ⟶ EC	0.206	0.009	0.224	0.004
GreenTFL ⟶ PB	0.074	0.260	0.308	0.000	GreenTFL ⟶ JS	…	…	0.118	0.016
GreenTFL ⟶ POS	0.166	0.017	0.201	0.003	GreenTFL ⟶ POS	0.254	0.002	0.270	0.001
ATE ⟶ EC	…	…	0.020	0.135	ATE ⟶ EC	…	…	0.009	0.276
ATE ⟶ PB	0.067	0.151	0.076	0.109	ATE ⟶ JS	0.009	0.866	0.066	0.241
ATE ⟶ POS	0.138	0.000	0.138	0.000	ATE ⟶ POS	0.122	0.002	0.122	0.002
POS ⟶ EC	0.148	0.039	0.148	0.039	POS ⟶ EC	0.07	0.169	0.070	0.169
POS ⟶ PB	−0.005	0.923	0.071	0.291	POS ⟶ JS	0.47	0.000	0.467	0.000
EC ⟶ PB	0.511	0.000	0.511	0.000	EC ⟶ JS	−0.047	0.471	−0.047	0.471
Model 3.5					Model 3.6
Paths	Path Coeff.	p Values	Total Effects	p Values	Paths	Path Coeff.	p Values	Total Effects	p Values
PB ⟶ AM	0.393	0.000	0.674	0.000	GreenTFL ⟶ AM	0.575	0.000	0.768	0.000
PB ⟶ ATE	0.686	0.000	0.686	0.000	GreenTFL ⟶ ATE	0.702	0.000	0.702	0.000
PB ⟶ EC	…	…	0.519	0.000	GreenTFL ⟶ EC	…	…	0.531	0.000
PB ⟶ JE	0.028	0.677	0.625	0.000	GreenTFL ⟶ JE	0.305	0.000	0.778	0.000
PB ⟶ JS	0.120	0.058	0.553	0.000	GreenTFL ⟶ JS	0.575	0.000	0.800	0.000
PB ⟶ EP	−0.148	0.042	0.426	0.000	GreenTFL ⟶ EP	0.161	0.139	0.662	0.000
PB ⟶ GreenTFL	0.065	0.311	0.621	0.000	GreenTFL ⟶ PB	0.072	0.270	0.568	0.000
PB ⟶ POS	−0.059	0.231	0.587	0.000	GreenTFL ⟶ POS	0.193	0.005	0.804	0.000
PB ⟶ TRL	0.565	0.000	0.623	0.000	GreenTFL ⟶ TRL	−0.080	0.417	0.336	0.000
ATE ⟶ AM	0.168	0.002	0.410	0.000	ATE ⟶ AM	0.049	0.249	0.275	0.000
ATE ⟶ EC	0.757	0.000	0.757	0.000	ATE ⟶ EC	0.757	0.000	0.757	0.000
ATE ⟶ JE	0.286	0.000	0.527	0.000	ATE ⟶ JE	0.241	0.000	0.318	0.000
ATE ⟶ JS	0.041	0.506	0.420	0.000	ATE ⟶ JS	0.004	0.945	0.080	0.175
ATE ⟶ EP	0.325	0.000	0.605	0.000	ATE ⟶ EP	0.316	0.000	0.293	0.001
ATE ⟶ GreenTFL	−0.009	0.820	0.526	0.000	ATE ⟶ PB	0.065	0.165	0.550	0.000
ATE ⟶ POS	0.195	0.000	0.599	0.000	ATE ⟶ POS	0.183	0.000	0.307	0.000
ATE ⟶ TRL	−0.046	0.463	−0.031	0.698	ATE ⟶ TRL	−0.008	0.891	0.345	0.000
EC ⟶ AM	0.319	0.000	0.319	0.000	EC ⟶ AM	0.298	0.000	0.298	0.000
EC ⟶ JE	−0.007	0.924	0.185	0.021	EC ⟶ JE	−0.056	0.400	0.073	0.363
EC ⟶ JS	0.115	0.124	0.217	0.016	EC ⟶ JS	−0.027	0.674	−0.005	0.949
EC ⟶ EP	−0.081	0.347	0.091	0.350	EC ⟶ EP	−0.206	0.012	−0.165	0.069
EC ⟶ GreenTFL	0.212	0.001	0.401	0.000	EC ⟶ PB	0.511	0.000	0.657	0.000
EC ⟶ POS	0.069	0.216	0.253	0.001	EC ⟶ POS	−0.009	0.877	0.051	0.514
EC ⟶ TRL	−0.058	0.521	0.007	0.935	EC ⟶ TRL	0.297	0.001	0.459	0.000
AM ⟶ JE	0.601	0.000	0.601	0.000	AM ⟶ JE	0.434	0.000	0.434	0.000
AM ⟶ JS	…	…	0.332	0.000	AM ⟶ JS	…	…	0.132	0.000
AM ⟶ EP	0.165	0.094	0.381	0.000	AM ⟶ EP	0.080	0.424	0.182	0.044
AM ⟶ GreenTFL	0.231	0.000	0.460	0.000	AM ⟶ PB	0.156	0.015	0.253	0.000
AM ⟶ POS	0.223	0.000	0.475	0.000	AM ⟶ POS	0.150	0.009	0.278	0.000
AM ⟶ TRL	0.342	0.000	0.273	0.001	AM ⟶ TRL	0.528	0.000	0.506	0.000
JE ⟶ JS	0.553	0.000	0.553	0.000	JE ⟶ JS	0.304	0.000	0.304	0.000
JE ⟶ EP	−0.001	0.987	0.216	0.011	JE ⟶ EP	−0.022	0.797	0.105	0.163
JE ⟶ GreenTFL	0.111	0.037	0.321	0.000	JE ⟶ PB	0.023	0.628	0.015	0.767
JE ⟶ POS	0.249	0.000	0.419	0.000	JE ⟶ POS	0.218	0.000	0.295	0.000
JE ⟶ TRL	−0.018	0.784	−0.121	0.106	JE ⟶ TRL	0.010	0.892	−0.043	0.604
JS ⟶ EP	0.100	0.151	0.220	0.001	JS ⟶ EP	0.057	0.409	0.151	0.037
JS ⟶ GreenTFL	0.245	0.000	0.305	0.000	JS ⟶ PB	0.029	0.501	−0.014	0.749
JS ⟶ POS	0.308	0.000	0.308	0.000	JS ⟶ POS	0.253	0.000	0.253	0.000
JS ⟶ TRL	−0.193	0.002	−0.190	0.002	JS ⟶ TRL	−0.167	0.016	−0.170	0.014
POS ⟶ EP	0.387	0.001	0.387	0.001	POS ⟶ EP	0.370	0.002	0.370	0.002
POS ⟶ GreenTFL	0.142	0.020	0.164	0.004	POS ⟶ PB	−0.005	0.923	−0.030	0.546
POS ⟶ TRL	…	…	0.006	0.816	POS ⟶ TRL	…	…	−0.009	0.707
EP ⟶ GreenTFL	0.055	0.185	0.055	0.190	EP ⟶ PB	−0.064	0.060	−0.068	0.061
EP ⟶ TRL	0.015	0.808	0.015	0.808	EP ⟶ TRL	−0.024	0.712	−0.024	0.712
TRL ⟶ GreenTFL	−0.025	0.448	−0.025	0.448	TRL ⟶ PB	−0.190	0.000	0.190	0.000
Model 3.7					Model 3.8
Paths	Path Coeff.	p Values	Total Effects	p Values	Paths	Path Coeff.	p Values	Total Effects	p Values
EP ⟶ AM	0.069	0.063	0.544	0.000	PB ⟶ AM	0.233	0.000	0.750	0.000
EP ⟶ ATE	0.260	0.000	0.601	0.000	PB ⟶ ATE	0.341	0.000	0.678	0.000
EP ⟶ EC	0.091	0.095	0.475	0.000	PB ⟶ EC	0.689	0.000	0.847	0.000
EP ⟶ JE	0.590	0.000	0.590	0.000	PB ⟶ JE	0.688	0.000	0.688	0.000
EP ⟶ JS	…	…	0.440	0.000	PB ⟶ JS	…	…	0.512	0.000
EP ⟶ PB	−0.051	0.138	0.422	0.000	PB ⟶ EP	−0.072	0.363	0.460	0.000
EP ⟶ GreenTFL	0.084	0.040	0.561	0.000	PB ⟶ GreenTFL	0.033	0.594	0.694	0.000
EP ⟶ POS	0.156	0.005	0.624	0.000	PB ⟶ POS	−0.062	0.200	0.628	0.000
EP ⟶ TRL	−0.028	0.663	0.290	0.000	PB ⟶ TRL	0.560	0.000	0.635	0.000
JE ⟶ AM	0.330	0.000	0.751	0.000	JE ⟶ AM	0.321	0.000	0.561	0.000
JE ⟶ ATE	0.374	0.000	0.541	0.000	JE ⟶ ATE	0.381	0.000	0.458	0.000
JE ⟶ EC	0.483	0.000	0.650	0.000	JE ⟶ EC	0.171	0.000	0.230	0.000
JE ⟶ JS	0.745	0.000	0.745	0.000	JE ⟶ JS	0.745	0.000	0.745	0.000
JE ⟶ PB	0.038	0.407	0.658	0.000	JE ⟶ EP	0.041	0.636	0.534	0.000
JE ⟶ GreenTFL	0.145	0.003	0.713	0.000	JE ⟶ GreenTFL	0.154	0.002	0.604	0.000
JE ⟶ POS	0.208	0.000	0.670	0.000	JE ⟶ POS	0.218	0.000	0.709	0.000
JE ⟶ TRL	0.001	0.991	0.415	0.000	JE ⟶ TRL	−0.013	0.840	0.009	0.895
JS ⟶ AM	0.260	0.000	0.338	0.000	JS ⟶ AM	0.264	0.000	0.278	0.000
JS ⟶ ATE	0.004	0.938	0.078	0.129	JS ⟶ ATE	0.060	0.322	0.076	0.185
JS ⟶ EC	0.225	0.000	0.225	0.000	JS ⟶ EC	0.078	0.054	0.078	0.054
JS ⟶ PB	0.025	0.548	0.240	0.000	JS ⟶ EP	0.055	0.402	0.307	0.000
JS ⟶ GreenTFL	0.270	0.000	0.418	0.000	JS ⟶ GreenTFL	0.289	0.000	0.392	0.000
JS ⟶ POS	0.229	0.000	0.353	0.000	JS ⟶ POS	0.249	0.000	0.390	0.000
JS ⟶ TRL	−0.186	0.004	0.046	0.384	JS ⟶ TRL	−0.187	0.003	−0.098	0.085
EC ⟶ AM	0.351	0.000	0.351	0.000	EC ⟶ AM	0.189	0.001	0.189	0.001
EC ⟶ ATE	…	…	0.077	0.002	EC ⟶ ATE	…	…	0.011	0.526
EC ⟶ PB	0.542	0.000	0.706	0.000	EC ⟶ EP	…	…	0.100	0.017
EC ⟶ GreenTFL	0.240	0.000	0.327	0.000	EC ⟶ GreenTFL	0.203	0.002	0.258	0.000
EC ⟶ POS	0.025	0.627	0.131	0.001	EC ⟶ POS	0.026	0.645	0.111	0.043
EC ⟶ TRL	0.277	0.001	0.453	0.000	EC ⟶ TRL	−0.061	0.507	0.004	0.968
AM ⟶ ATE	0.218	0.001	0.218	0.001	AM ⟶ ATE	0.058	0.473	0.058	0.473
AM ⟶ PB	0.149	0.026	0.266	0.000	ATE ⟶ EP	…	…	0.190	0.000
AM ⟶ GreenTFL	0.273	0.000	0.268	0.000	AM ⟶ GreenTFL	0.287	0.000	0.280	0.000
AM ⟶ POS	0.128	0.025	0.198	0.000	AM ⟶ POS	0.169	0.004	0.235	0.000
AM ⟶ TRL	0.506	0.000	0.503	0.000	AM ⟶ TRL	0.345	0.000	0.343	0.000
ATE ⟶ PB	…	…	0.001	0.967	ATE ⟶ EP	…	…	0.104	0.005
ATE ⟶ GreenTFL	0.027	0.484	0.027	0.475	ATE ⟶ GreenTFL	0.060	0.168	0.062	0.150
ATE ⟶ POS	0.123	0.001	0.128	0.000	ATE ⟶ POS	0.185	0.000	0.198	0.000
ATE ⟶ TRL	−0.013	0.819	−0.013	0.819	ATE ⟶ TRL	−0.040	0.528	−0.040	0.528
TRL ⟶ PB	0.190	0.000	0.189	0.000	TRL ⟶ EP	0.076	0.134	0.067	0.187
TRL ⟶ GreenTFL	−0.023	0.417	−0.023	0.417	TRL ⟶ GreenTFL	−0.033	0.295	−0.033	0.295
TRL ⟶ POS			−0.004	0.503	TRL ⟶ POS	…	…	−0.006	0.373
GreenTFL ⟶ PB	0.073	0.260	0.074	0.242	GreenTFL ⟶ EP	0.179	0.066	0.274	0.003
GreenTFL ⟶ POS	0.158	0.030	0.158	0.030	GreenTFL ⟶ POS	0.198	0.005	0.198	0.005
POS ⟶ EP	0.484	0.000	0.484	0.000	POS ⟶ EP	0.484	0.000	0.484	0.000
Model 3.9					Model 3.10
Paths	Path Coeff.	p Values	Total Effects	p Values	Paths	Path Coeff.	p Values	Total Effects	p Values
EC ⟶ AM	0.201	0.000	0.539	0.000	POS ⟶ AM	0.286	0.000	0.637	0.000
EC ⟶ ATE	0.503	0.000	0.683	0.000	POS ⟶ ATE	0.32	0.000	0.587	0.000
EC ⟶ JE	0.001	0.992	0.521	0.000	POS ⟶ EC	0.031	0.499	0.460	0.000
EC ⟶ JS	0.231	0.003	0.468	0.000	POS ⟶ JE	0.311	0.000	0.707	0.000
EC ⟶ EP	0.507	0.000	0.507	0.000	POS ⟶ JS	0.608	0.000	0.726	0.000
EC ⟶ PB	…	…	0.232	0.000	POS ⟶ EP	0.674	0.000	0.674	0.000
EC ⟶ GreenTFL	0.280	0.000	0.620	0.000	POS ⟶ PB	…	…	0.308	0.000
EC ⟶ POS	0.032	0.514	0.584	0.000	POS ⟶ GreenTFL	0.272	0.000	0.689	0.000
EC ⟶ TRL	0.006	0.943	0.135	0.099	POS ⟶ TRL	−0.071	0.302	0.157	0.019
EP ⟶ AM	0.089	0.015	0.433	0.000	EP ⟶ AM	0.07	0.093	0.295	0.000
EP ⟶ ATE	0.303	0.000	0.357	0.000	EP ⟶ ATE	0.228	0.000	0.396	0.000
EP ⟶ JE	0.058	0.160	0.440	0.000	EP ⟶ EC	−0.024	0.443	0.365	0.000
EP ⟶ JS	0.342	0.000	0.465	0.000	EP ⟶ JE	0.002	0.974	0.206	0.000
EP ⟶ PB	0.458	0.000	0.458	0.000	EP ⟶ JS	0.094	0.060	0.171	0.001
EP ⟶ GreenTFL	0.135	0.003	0.414	0.000	EP ⟶ PB	0.457	0.000	0.457	0.000
EP ⟶ POS	0.167	0.003	0.517	0.000	EP ⟶ GreenTFL	0.076	0.096	0.286	0.000
EP ⟶ TRL	0.039	0.473	0.312	0.000	EP ⟶ TRL	0.062	0.247	0.354	0.000
PB ⟶ AM	0.171	0.002	0.391	0.000	PB ⟶ AM	0.307	0.000	0.447	0.000
PB ⟶ ATE	0.117	0.057	0.117	0.057	PB ⟶ ATE	0.366	0.000	0.366	0.000
PB ⟶ JE	0.090	0.117	0.292	0.000	PB ⟶ EC	0.47	0.000	0.663	0.000
PB ⟶ JS	0.310	0.000	0.272	0.000	PB ⟶ JE	0.069	0.216	0.268	0.000
PB ⟶ GreenTFL	0.125	0.057	0.253	0.000	PB ⟶ JS	0.195	0.001	0.173	0.000
PB ⟶ POS	−0.008	0.869	0.195	0.003	PB ⟶ GreenTFL	0.255	0.000	0.351	0.000
PB ⟶ TRL	0.640	0.000	0.634	0.000	PB ⟶ TRL	0.668	0.000	0.655	0.000
ATE ⟶ AM	−0.058	0.186	0.008	0.865	ATE ⟶ AM	−0.05	0.242	0.001	0.985
ATE ⟶ JE	0.215	0.000	0.234	0.000	ATE ⟶ EC	0.257	0.000	0.268	0.000
ATE ⟶ JS	…	…	0.003	0.523	ATE ⟶ JE	0.157	0.001	0.174	0.001
ATE ⟶ GreenTFL	0.063	0.123	0.064	0.113	ATE ⟶ JS	…	…	0.001	0.711
ATE ⟶ POS	0.120	0.001	0.184	0.000	ATE ⟶ GreenTFL	0.083	0.077	0.083	0.077
ATE ⟶ TRL	−0.057	0.335	−0.057	0.335	ATE ⟶ TRL	−0.034	0.571	−0.034	0.571
TRL ⟶ AM	0.108	0.001	0.100	0.014	TRL ⟶ AM	0.113	0.001	0.128	0.001
TRL ⟶ JE	0.043	0.161	0.022	0.595	TRL ⟶ EC	−0.025	0.405	−0.005	0.866
TRL ⟶ JS	−0.060	0.194	−0.060	0.194	TRL ⟶ JE	0.051	0.092	0.047	0.186
TRL ⟶ GreenTFL	0.008	0.798	−0.018	0.649	TRL ⟶ JS	−0.033	0.328	−0.033	0.328
TRL ⟶ POS	−0.051	0.106	−0.047	0.228	TRL ⟶ GreenTFL	0.015	0.649	0.004	0.912
JS ⟶ AM	0.255	0.000	0.373	0.000	JS ⟶ AM	…	…	0.072	0.005
JS ⟶ JE	0.253	0.000	0.388	0.000	JS ⟶ EC	−0.077	0.068	−0.007	0.857
JS ⟶ GreenTFL	0.422	0.000	0.422	0.000	JS ⟶ JE	0.157	0.002	0.232	0.000
JS ⟶ POS	0.216	0.000	0.423	0.000	JS ⟶ GreenTFL	0.325	0.000	0.325	0.000
GreenTFL ⟶ AM	…	…	0.097	0.001	GreenTFL ⟶ AM	…	…	0.072	0.027
GreenTFL ⟶ JE	0.320	0.000	0.320	0.000	GreenTFL ⟶ EC	0.212	0.004	0.212	0.002
GreenTFL ⟶ POS	0.159	0.019	0.241	0.001	GreenTFL ⟶ JE	0.231	0.006	0.231	0.006
JE ⟶ AM	0.304	0.000	0.304	0.000	JE ⟶ AM	0.312	0.000	0.312	0.000
JE ⟶ POS	0.208	0.000	0.257	0.000	JE ⟶ EC	−0.043	0.362	0.003	0.950
AM ⟶ POS	0.160	0.006	0.160	0.006	AM ⟶ EC	0.148	0.008	0.148	0.008
Model 3.11					Model 3.12
Paths	Path Coeff.	p Values	Total Effects	p Values	Paths	Path Coeff.	p Values	Total Effects	p Values
POS ⟶ AM	0.180	0.002	0.779	0.000	ATE ⟶ AM	−0.016	0.725	0.662	0.000
POS ⟶ ATE	0.178	0.003	0.728	0.000	ATE ⟶ EC	0.756	0.000	0.756	0.000
POS ⟶ EC	0.713	0.000	0.713	0.000	ATE ⟶ JE	0.216	0.000	0.698	0.000
POS ⟶ JE	0.352	0.000	0.794	0.000	ATE ⟶ JS	0.256	0.000	0.592	0.000
POS ⟶ JS	0.662	0.000	0.780	0.000	ATE ⟶ EP	0.442	0.000	0.612	0.000
POS ⟶ EP	0.555	0.000	0.678	0.000	ATE ⟶ PB	…	…	0.647	0.000
POS ⟶ PB	…	…	0.610	0.000	ATE ⟶ GreenTFL	0.099	0.033	0.692	0.000
POS ⟶ GreenTFL	0.437	0.000	0.807	0.000	ATE ⟶ POS	0.119	0.001	0.710	0.000
POS ⟶ TRL	−0.129	0.117	0.352	0.000	ATE ⟶ TRL	−0.041	0.481	0.386	0.000
EC ⟶ AM	0.105	0.053	0.463	0.000	EC ⟶ AM	0.138	0.014	0.659	0.000
EC ⟶ ATE	0.432	0.000	0.518	0.000	EC ⟶ JE	0.001	0.987	0.409	0.000
EC ⟶ JE	0.049	0.422	0.278	0.000	EC ⟶ JS	0.215	0.016	0.444	0.000
EC ⟶ JS	0.039	0.557	0.165	0.003	EC ⟶ EP	0.014	0.876	0.102	0.213
EC ⟶ EP	0.096	0.250	0.047	0.524	EC ⟶ PB	0.856	0.000	0.856	0.000
EC ⟶ PB	0.856	0.000	0.856	0.000	EC ⟶ GreenTFL	0.362	0.000	0.597	0.000
EC ⟶ GreenTFL	0.266	0.000	0.415	0.000	EC ⟶ POS	0.033	0.504	0.394	0.000
EC ⟶ TRL	−0.048	0.587	0.567	0.000	EC ⟶ TRL	−0.044	0.638	0.556	0.000
PB ⟶ AM	0.251	0.000	0.316	0.000	PB ⟶ AM	0.256	0.000	0.396	0.000
PB ⟶ ATE	0.105	0.167	0.077	0.256	PB ⟶ JE	0.086	0.123	0.261	0.000
PB ⟶ JE	0.090	0.116	0.178	0.003	PB ⟶ JS	0.267	0.001	0.267	0.001
PB ⟶ JS	0.148	0.011	0.148	0.011	PB ⟶ EP	−0.084	0.267	0.012	0.882
PB ⟶ EP	−0.083	0.287	−0.063	0.423	PB ⟶ GreenTFL	0.242	0.001	0.245	0.001
PB ⟶ GreenTFL	0.168	0.005	0.160	0.008	PB ⟶ POS	−0.007	0.884	0.174	0.016
PB ⟶ TRL	0.559	0.000	0.654	0.000	PB ⟶ TRL	0.565	0.000	0.645	0.000
JS ⟶ AM	0.156	0.000	0.165	0.000	JS ⟶ AM	0.218	0.000	0.280	0.000
JS ⟶ ATE	−0.039	0.462	0.004	0.938	JS ⟶ JE	0.257	0.000	0.306	0.000
JS ⟶ JE	0.150	0.006	0.156	0.003	JS ⟶ EP	0.360	0.000	0.360	0.000
JS ⟶ EP	0.135	0.048	0.135	0.048	JS ⟶ GreenTFL	…	…	0.096	0.006
EP ⟶ GreenTFL	…	…	0.016	0.270	JS ⟶ POS	0.218	0.000	0.405	0.000
JS ⟶ TRL	−0.137	0.033	−0.065	0.308	JS ⟶ TRL	−0.168	0.015	−0.064	0.334
EP ⟶ AM	0.038	0.380	0.070	0.109	EP ⟶ AM	0.082	0.034	0.170	0.000
EP ⟶ ATE	0.207	0.000	0.211	0.000	EP ⟶ JE	0.059	0.164	0.147	0.009
EP ⟶ JE	0.039	0.417	0.069	0.157	EP ⟶ GreenTFL	0.267	0.000	0.267	0.000
EP ⟶ GreenTFL	0.115	0.024	0.115	0.024	EP ⟶ POS	0.167	0.003	0.264	0.000
EP ⟶ TRL	0.038	0.541	0.057	0.293	EP ⟶ TRL	0.024	0.694	0.061	0.261
GreenTFL ⟶ AM	0.277	0.000	0.277	0.000	GreenTFL ⟶ AM	0.329	0.000	0.329	0.000
GreenTFL ⟶ ATE	−0.015	0.837	0.005	0.939	GreenTFL ⟶ JE	0.318	0.000	0.319	0.000
GreenTFL ⟶ JE	0.240	0.008	0.242	0.007	GreenTFL ⟶ POS	0.159	0.021	0.276	0.000
GreenTFL ⟶ TRL	−0.080	0.407	0.034	0.711	GreenTFL ⟶ TRL	−0.101	0.285	0.023	0.798
AM ⟶ ATE	−0.129	0.099	−0.129	0.084	AM ⟶ JE	…	…	0.016	0.226
AM ⟶ JE	…	…	0.019	0.190	AM ⟶ POS	0.153	0.010	0.137	0.014
AM ⟶ TRL	0.411	0.000	0.411	0.000	AM ⟶ TRL	0.376	0.000	0.376	0.000
TRL ⟶ ATE	−0.011	0.757	0.000	0.991	TRL ⟶ JE	0.043	0.154	0.043	0.154
TRL ⟶ JE	0.045	0.126	0.045	0.126	TRL ⟶ POS	−0.052	0.112	−0.043	0.187
JE ⟶ ATE	0.232	0.000	0.232	0.000	JE ⟶ POS	0.212	0.000	0.212	0.000
Model 3.13					Model 3.14
Paths	Path Coeff.	p Values	Total Effects	p Values	Paths	Path Coeff.	p Values	Total Effects	p Values
GreenTFL ⟶ AM	0.304	0.000	0.636	0.000	EC ⟶ AM	0.188	0.000	0.537	0.000
GreenTFL ⟶ ATE	0.161	0.025	0.425	0.000	EC ⟶ ATE	0.450	0.000	0.571	0.000
GreenTFL ⟶ EC	0.213	0.005	0.636	0.000	EC ⟶ JE	0.039	0.531	0.396	0.000
GreenTFL ⟶ JE	0.230	0.006	0.505	0.000	EC ⟶ JS	0.053	0.448	0.296	0.000
GreenTFL ⟶ JS	0.453	0.000	0.536	0.000	EC ⟶ EP	0.039	0.601	0.187	0.005
GreenTFL ⟶ EP	0.210	0.036	0.303	0.001	EC ⟶ PB	0.649	0.000	0.796	0.000
GreenTFL ⟶ PB	0.453	0.000	0.631	0.000	EC ⟶ GreenTFL	0.212	0.000	0.526	0.000
GreenTFL ⟶ POS	…	…	0.180	0.000	EC ⟶ POS	…	…	0.210	0.000
GreenTFL ⟶ TRL	0.455	0.000	0.455	0.000	EC ⟶ TRL	0.531	0.000	0.531	0.000
TRL ⟶ AM	0.115	0.001	0.338	0.000	TRL ⟶ AM	0.108	0.002	0.374	0.000
TRL ⟶ ATE	−0.011	0.780	0.275	0.000	TRL ⟶ ATE	−0.011	0.746	0.150	0.003
TRL ⟶ EC	−0.025	0.409	0.279	0.000	TRL ⟶ JE	0.053	0.107	0.324	0.000
TRL ⟶ JE	0.051	0.100	0.270	0.000	TRL ⟶ JS	−0.034	0.316	0.279	0.000
TRL ⟶ JS	−0.035	0.273	0.144	0.001	TRL ⟶ EP	0.029	0.531	0.280	0.000
TRL ⟶ EP	0.008	0.871	0.205	0.003	TRL ⟶ PB	0.235	0.000	0.281	0.000
TRL ⟶ PB	0.357	0.000	0.425	0.000	TRL ⟶ GreenTFL	−0.025	0.460	0.254	0.000
TRL ⟶ POS	0.396	0.000	0.396	0.000	TRL ⟶ POS	0.396	0.000	0.396	0.000
POS ⟶ AM	0.183	0.002	0.323	0.000	POS ⟶ AM	0.267	0.000	0.490	0.000
POS ⟶ ATE	0.380	0.000	0.435	0.000	POS ⟶ ATE	0.361	0.000	0.368	0.000
POS ⟶ EC	0.030	0.511	0.208	0.006	POS ⟶ JE	0.369	0.000	0.614	0.000
POS ⟶ JE	0.312	0.000	0.458	0.000	POS ⟶ JS	0.612	0.000	0.681	0.000
POS ⟶ JS	0.406	0.000	0.413	0.000	POS ⟶ EP	0.634	0.000	0.634	0.000
POS ⟶ EP	0.499	0.000	0.499	0.000	POS ⟶ PB	0.144	0.002	0.119	0.017
POS ⟶ PB	0.211	0.003	0.174	0.018	POS ⟶ GreenTFL	0.141	0.016	0.530	0.000
EP ⟶ AM	0.040	0.321	0.025	0.610	EP ⟶ AM	0.075	0.061	0.080	0.104
EP ⟶ ATE			−0.023	0.163	EP ⟶ ATE	…	…	−0.003	0.537
EP ⟶ EC	−0.023	0.487	−0.064	0.203	EP ⟶ JE	0.021	0.678	0.039	0.470
EP ⟶ JE	0.002	0.969	0.001	0.988	EP ⟶ JS	0.104	0.042	0.097	0.059
EP ⟶ JS	0.052	0.234	0.049	0.262	EP ⟶ PB	−0.039	0.258	−0.039	0.258
EP ⟶ PB	−0.074	0.153	−0.074	0.153	EP ⟶ GreenTFL	0.056	0.168	0.100	0.045
PB ⟶ AM	0.223	0.000	0.234	0.000	PB ⟶ AM	0.194	0.000	0.235	0.000
PB ⟶ ATE	0.318	0.000	0.318	0.000	PB ⟶ ATE	0.066	0.386	0.066	0.386
PB ⟶ EC	0.470	0.000	0.577	0.000	PB ⟶ JE	0.104	0.072	0.155	0.014
PB ⟶ JE	0.070	0.222	0.126	0.020	PB ⟶ JS	0.177	0.004	0.174	0.004
PB ⟶ JS	0.062	0.257	0.040	0.416	PB ⟶ GreenTFL	0.065	0.297	0.178	0.010
ATE ⟶ AM	−0.058	0.153	−0.024	0.581	ATE ⟶ AM	−0.096	0.028	−0.048	0.307
ATE ⟶ EC	0.256	0.000	0.252	0.000	ATE ⟶ JE	0.167	0.002	0.156	0.004
ATE ⟶ JE	0.157	0.001	0.146	0.003	ATE ⟶ JS	−0.049	0.363	−0.049	0.363
ATE ⟶ JS	−0.068	0.159	−0.068	0.159	ATE ⟶ GreenTFL	−0.009	0.828	−0.015	0.748
JS ⟶ AM	…	…	0.037	0.015	JS ⟶ AM	…	…	0.071	0.003
JS ⟶ EC	−0.077	0.065	−0.079	0.067	JS ⟶ JE	0.231	0.000	0.231	0.000
JS ⟶ JE	0.157	0.002	0.157	0.002	JS ⟶ GreenTFL	0.245	0.000	0.287	0.000
JE ⟶ AM	0.234	0.000	0.234	0.000	JE ⟶ AM	0.306	0.000	0.306	0.000
JE ⟶ EC	−0.043	0.363	−0.009	0.854	JE ⟶ GreenTFL	0.112	0.032	0.182	0.001
AM ⟶ EC	0.146	0.007	0.146	0.007	AM ⟶ GreenTFL	0.230	0.000	0.230	0.000

Notes: N = 608; 95% bias corrected confidence intervals appear in parentheses. See Table 1, in main text, for acronyms’ meanings.

(see Appendix A) shows the results for the 14 model configurations designed to report the effects of diverse quasi-random arrangements of environmental, social, and economically related constructs. Considering our sample as a loosely coupled system, findings show three results. First, key social constructs such as POS has 42 of 48 (87.5%) significant positive relationships with other constructs; green transformational leadership has 37 of 50 (75%) significant positive relationships with other constructs; and job satisfaction has 46 of 66 (69.7%) significant positive relationships with other constructs. Second, employee attitudes toward the natural environment have 38 of 68 (55.88%) significant positive relationships with other constructs; environmental commitment has 37 of 69 (56.06%) significant positive relationships with other constructs; and pro-environmental behaviors has 42 of 69 (60.87%) positive significant relationships with other constructs. Third, transactional leadership has 17 of 54 (31.48%) significant positive relationships with other constructs, and employee performance has 30 of 68 (44.12%) significant positive relationships with other constructs. Thus, the more encompassing the construct, the larger the number of significant relationships with other constructs. These effects remain after adjusting for the number of constructs in the sets of the environmental, social, and economically related constructs.

4.2.2. Completely at Random Model Configurations

Table A4. Random model configurations results.

Model 4.1		Model 4.2		Model 4.3		Model 4.4		Model 4.5
Paths	Path Coeff.	Paths	Path Coeff.	Paths	Path Coeff.	Paths	Path Coeff.	Paths	Path Coeff.
ATE ⟶ EC	0.250 ***	ATE ⟶ EC	0.259 ***	ATE ⟶ EP	0.392 ***	ATE ⟶ EC	0.263 ***	EC ⟶ ATE	0.423 ***
EP ⟶ ATE	0.222 ***	ATE ⟶ EP	0.337 ***	ATE ⟶ GreenTFL	−0.007	ATE ⟶ EP	0.316 ***	EC ⟶ EP	0.027
EP ⟶ EC	−0.022	ATE ⟶ PB	0.404 ***	ATE ⟶ JE	0.292 ***	ATE ⟶ JE	0.170 **	EC ⟶ JE	0.101
EP ⟶ GreenTFL	0.096 *	ATE ⟶ TRL	−0.019	EC ⟶ ATE	0.559 ***	EC ⟶ JE	0.046	EC ⟶ PB	0.709 ***
EP ⟶ JS	0.093	EC ⟶ EP	−0.059	EC ⟶ EP	−0.010	EP ⟶ EC	−0.020	EC ⟶ POS	0.116
GreenTFL ⟶ ATE	0.143	EC ⟶ TRL	−0.013	EC ⟶ GreenTFL	0.280 ***	EP ⟶ JE	0.049	EP ⟶ ATE	0.209 ***
GreenTFL ⟶ EC	0.247 ***	EP ⟶ TRL	0.098	EC ⟶ JE	0.083	PB ⟶ ATE	0.374 ***	GreenTFL ⟶ ATE	−0.068
JS ⟶ ATE	−0.084	JS ⟶ ATE	0.088	EC ⟶ JS	0.380 ***	PB ⟶ EC	0.589 ***	GreenTFL ⟶ EC	0.779 ***
JS ⟶ EC	−0.060	JS ⟶ EC	0.025	EC ⟶ PB	0.856 ***	PB ⟶ EP	−0.143 *	GreenTFL ⟶ EP	0.208
JS ⟶ GreenTFL	0.320 ***	JS ⟶ EP	0.141 *	EP ⟶ GreenTFL	0.098 *	PB ⟶ JE	0.127 *	GreenTFL ⟶ JE	0.615 ***
PB ⟶ ATE	0.332 ***	JS ⟶ PB	0.217 ***	JE ⟶ EP	0.205 **	PB ⟶ POS	0.673 ***	GreenTFL ⟶ PB	0.190 ***
PB ⟶ EC	0.494 ***	JS ⟶ TRL	−0.041	JE ⟶ GreenTFL	0.211 ***	PB ⟶ TRL	0.657 ***	GreenTFL ⟶ POS	0.406 ***
PB ⟶ EP	0.006	PB ⟶ EC	0.587 ***	JS ⟶ ATE	0.189 **	POS ⟶ ATE	0.472 ***	JE ⟶ ATE	0.194 **
PB ⟶ GreenTFL	0.297 ***	PB ⟶ EP	−0.102	JS ⟶ EP	0.271 ***	POS ⟶ EC	0.138 **	JE ⟶ EP	0.125
PB ⟶ JS	0.173 ***	PB ⟶ TRL	0.608	JS ⟶ GreenTFL	0.350 ***	POS ⟶ EP	0.523 ***	JE ⟶ POS	0.406 ***
POS ⟶ ATE	0.297 ***	POS ⟶ ATE	0.651 ***	JS ⟶ JE	0.442 ***	POS ⟶ JE	0.513 ***	PB ⟶ ATE	0.077
POS ⟶ EC	0.048	POS ⟶ EC	0.112 *	PB ⟶ ATE	0.089	POS ⟶ TRL	−0.048	PB ⟶ EP	−0.117
POS ⟶ EP	0.670 ***	POS ⟶ EP	0.432 ***	PB ⟶ EP	−0.111	TRL ⟶ ATE	−0.008	PB ⟶ JE	0.130
POS ⟶ GreenTFL	0.299 ***	POS ⟶ JS	0.790 ***	PB ⟶ GreenTFL	0.115	TRL ⟶ EC	0.006	PB ⟶ POS	−0.001
POS ⟶ JS	0.611 ***	POS ⟶ PB	0.210 **	PB ⟶ JE	0.126 *	TRL ⟶ EP	0.051	POS ⟶ ATE	0.139
POS ⟶ PB	0.673 ***	POS ⟶ RL	−0.055	PB ⟶ JS	0.301	TRL ⟶ JE	0.039	POS ⟶ EP	0.464
Model 4.6		Model 4.7		Model 4.8		Model 4.9		Model 4.10
Paths	Path Coeff.	Paths	Path Coeff.	Paths	Path Coeff.	Paths	Path Coeff.	Paths	Path Coeff.
ATE ⟶ EC	0.293 ***	EC ⟶ ATE	0.466 ***	ATE ⟶ EC	0.648 ***	ATE ⟶ EC	0.400 ***	ATE ⟶ EC	0.264 ***
ATE ⟶ EP	0.571 ***	EP ⟶ ATE	0.269 ***	ATE ⟶ EP	0.393 ***	ATE ⟶ EP	0.382 ***	ATE ⟶ EP	0.311 ***
ATE ⟶ JS	0.138 *	EP ⟶ EC	0.044	ATE ⟶ JE	0.405 ***	ATE ⟶ PB	0.020	ATE ⟶ JE	0.176 ***
ATE ⟶ PB	0.516 ***	EP ⟶ JS	0.145 **	ATE ⟶ JS	0.025	EC ⟶ EP	−0.045	ATE ⟶ JS	−0.036
ATE ⟶ POS	0.176 ***	GreenTFL ⟶ ATE	0.070	ATE ⟶ PB	0.084	EC ⟶ PB	0.682 ***	ATE ⟶ PB	0.419 ***
EC ⟶ POS	0.140 **	GreenTFL ⟶ EC	0.313 ***	EC ⟶ EP	0.023	GreenTFL ⟶ ATE	0.393 ***	EC ⟶ JE	0.037
EP ⟶ EC	0.002	GreenTFL ⟶ EP	0.628 ***	EC ⟶ JE	0.180 ***	GreenTFL ⟶ EC	0.473 ***	EP ⟶ EC	−0.023
EP ⟶ JS	0.324 ***	GreenTFL ⟶ JS	0.654 ***	EC ⟶ JS	0.116	GreenTFL ⟶ EP	0.236 *	EP ⟶ JE	0.033
EP ⟶ POS	0.222 ***	GreenTFL ⟶ PB	0.579 ***	EC ⟶ PB	0.668 ***	GreenTFL ⟶ JE	0.791 ***	EP ⟶ JS	0.100
JS ⟶ EC	0.082 *	GreenTFL ⟶ TRL	0.455 ***	JE ⟶ EP	0.168 *	GreenTFL ⟶ JS	0.553 ***	JS ⟶ EC	0.030
JS ⟶ POS	0.426 ***	JS ⟶ ATE	0.046	JE ⟶ JS	0.564 ***	GreenTFL ⟶ PB	0.109	JS ⟶ JE	0.234 ***
PB ⟶ EC	0.599 ***	JS ⟶ EC	−0.028	JS ⟶ EP	0.263 ***	JE ⟶ ATE	0.408 ***	PB ⟶ EC	0.587 ***
PB ⟶ EP	0.041	PB ⟶ ATE	0.120	PB ⟶ EP	−0.122	JE ⟶ EC	0.051	PB ⟶ EP	−0.108
PB ⟶ JS	0.421 ***	PB ⟶ EC	0.623 ***	PB ⟶ JE	0.231 *	JE ⟶ EP	0.116	PB ⟶ JE	0.113 ***
PB ⟶ POS	0.073	PB ⟶ EP	−0.029	PB ⟶ JS	0.166 ***	JE ⟶ JS	0.311 ***	PB ⟶ JS	0.187 ***
TRL ⟶ ATE	0.412 ***	PB ⟶ JS	0.105	TRL ⟶ ATE	0.412 ***	JE ⟶ PB	0.072	POS ⟶ ATE	0.720 ***
TRL ⟶ EC	0.007	TRL ⟶ ATE	−0.026	TRL ⟶ EC	0.264 ***	JS ⟶ ATE	−0.019	POS ⟶ EC	0.119 *
TRL ⟶ EP	0.038	TRL ⟶ EC	−0.003	TRL ⟶ EP	0.071	JS ⟶ EC	−0.020	POS ⟶ EP	0.523 ***
TRL ⟶ JS	−0.059	TRL ⟶ EP	0.031	TRL ⟶ JE	0.016	JS ⟶ EP	0.172 *	POS ⟶ JE	0.367 ***
TRL ⟶ PB	0.412 ***	TRL ⟶ JS	−0.050	TRL ⟶ JS	−0.058	JS ⟶ PB	0.038	POS ⟶ JS	0.622 ***
TRL ⟶ POS	−0.020	TRL ⟶ PB	0.361 ***	TRL ⟶ PB	0.235 ***	PB ⟶ EP	−0.102	POS ⟶ PB	0.372 ***
Model 4.11		Model 4.12		Model 4.13		Model 4.14		Model 4.15
Paths	Path Coeff.	Paths	Path Coeff.	Paths	Path Coeff.	Paths	Path Coeff.	Paths	Path Coeff.
ATE ⟶ EC	0.495 ***	ATE ⟶ EC	0.514 ***	EC ⟶ ATE	0.424 ***	ATE ⟶ EC	0.252 ***	ATE ⟶ EC	0.260 ***
ATE ⟶ EP	0.314 ***	ATE ⟶ EP	0.448 ***	EC ⟶ EP	0.080	ATE ⟶ EP	0.405 ***	ATE ⟶ EP	0.310 ***
ATE⟶ GreenTFL	0.024	ATE ⟶ JE	0.249 ***	EC ⟶ PB	0.690 ***	ATE ⟶ PB	0.264 ***	ATE ⟶ PB	0.358 ***
ATE ⟶ PB	0.043	ATE ⟶ PB	0.061	EC ⟶ POS	0.089	ATE ⟶ POS	0.186 ***	EC ⟶ EP	−0.025
EC ⟶ EP	−0.064	EC ⟶ EP	−0.069	EP ⟶ ATE	0.203 ***	EP ⟶ EC	−0.025	JE ⟶ ATE	0.365 ***
EC ⟶ GreenTFL	0.262 ***	EC ⟶ JE	0.159 ***	EP ⟶ POS	0.219 ***	EP ⟶ POS	0.230 ***	JE ⟶ EC	0.024
EC ⟶ PB	0.728 ***	EC ⟶ PB	0.588 ***	GreenTFL ⟶ ATE	−0.067	GreenTFL⟶ ATE	0.649 ***	JE ⟶ EP	0.052
GreenTFL ⟶ EP	0.159	EP ⟶ JE	0.103 *	GreenTFL ⟶ EC	0.617 ***	GreenTFL ⟶ EC	0.220 ***	JE ⟶ JS	0.321 ***
JS ⟶ ATE	0.088	EP ⟶ PB	−0.048	GreenTFL ⟶ EP	0.400 ***	GreenTFL ⟶ EP	0.454 ***	JE ⟶ PB	0.195 ***
JS ⟶ EC	0.154 *	JE ⟶ PB	0.075	GreenTFL ⟶ JE	0.791 ***	GreenTFL ⟶ PB	0.407 ***	JS ⟶ ATE	−0.021
JS ⟶ EP	0.070	JS ⟶ ATE	0.521 ***	GreenTFL ⟶ PB	0.131 ***	GreenTFL ⟶ POS	0.488 ***	JS ⟶ EC	0.017
JS ⟶ GreenTFL	0.337 ***	JS ⟶ EC	0.245 ***	GreenTFL ⟶ POS	0.329 ***	GreenTFL ⟶ TRL	0.455 ***	JS ⟶ EP	0.111
JS ⟶ PB	0.104 *	JS ⟶ EP	0.347 ***	JE ⟶ ATE	0.193 ***	PB ⟶ EC	0.497 ***	JS ⟶ PB	0.163 ***
PB ⟶ EP	−0.089	JS ⟶ JE	0.407 ***	JE ⟶ EC	0.205 ***	PB ⟶ EP	−0.182 *	PB ⟶ EC	0.562 ***
PB ⟶ GreenTFL	0.121 *	JS ⟶ PB	0.099 **	JE ⟶ EP	0.309 ***	PB ⟶ POS	0.087	PB ⟶ EP	−0.075
POS ⟶ ATE	0.650 ***	TRL ⟶ ATE	0.225 ***	JE ⟶ PB	0.092	POS ⟶ EC	0.025	POS ⟶ ATE	0.442 ***
POS ⟶ EC	0.235 **	TRL ⟶ EC	0.232 ***	JE ⟶ POS	0.335 ***	TRL ⟶ ATE	0.116 **	POS ⟶ EC	0.105 *
POS ⟶ EP	0.381 ***	TRL ⟶ EP	0.026	PB ⟶ ATE	0.076	TRL ⟶ EC	−0.007	POS ⟶ EP	0.406 ***
POS ⟶ GreenTFL	0.276 ***	TRL ⟶ JE	0.067 *	PB ⟶ EP	−0.116	TRL ⟶ EP	0.040	POS ⟶ JE	0.806 ***
POS ⟶ JS	0.790 ***	TRL ⟶ JS	0.358 ***	PB ⟶ POS	0.022	TRL ⟶ PB	0.331 ***	POS ⟶ JS	0.531 ***
POS ⟶ PB	0.041	TRL ⟶ PB	0.216 ***	POS ⟶ ATE	0.144 *	TRL ⟶ POS	−0.036	POS ⟶ PB	0.130

Notes: N = 608; *** p < 0.001; ** p < 0.01; * p < 0.05; 95% bias corrected confidence intervals appear in parentheses. See Table 1, in main text, for acronyms’ meanings.

(see Appendix A) shows the results for 15 model configurations determined completely at random. Results show that, of a total of 315 relationships, 111 (35.23%) were nonsignificant, and of those 111 nonsignificant relationships, 81 (72.97%) involved at least one environmental construct. These findings reaffirm the importance of the effects of social constructs. Similarly, these results suggest that environmental constructs are less strongly relational and relatively more isolated than social constructs. This is in agreement with research showing that environmental issues are not a priority for the average U.S. citizen [225,226].

Given the multi-organizational, multi-industry, and multi-regional nature of this study sample, this model set illustrates only part of the diversity of model construct interactions existing in U.S. organizations. Model components, starting points, interactions, and outcomes are multiple. Thus, the study of a large and diverse set of construct interactions is necessary to situate meaningful knowledge generalizations.

In sum, a diverse set of 41 model configurations suggest that social sustainability constructs in organizations, by being broader in scope, tend to relate more frequently and more strongly to other constructs than environmental and economically related constructs. These results apply to relationships at moderate ordering as well as to weaker orderings (e.g., quasi-random and random model configurations). The effects of environmentally related constructs may be constrained because such constructs refer only to the environmental dimension. However, we may optimistically expect that, since environmental considerations are increasingly constituting organizational processes, environmental constructs will be further embedded in organizations and have future larger effects on organizations.

5. Theoretical and Practical Implications

The purpose of this research was to examine environmental, social, and economically related constructs of sustainability in organizations (a) considering social sustainability as an antecedent of environmental and economically associated constructs, and (b) viewing sustainability as a multiplicity of relationships among social, environmental, and economically related constructs. This study makes two major contributions. First, in examining social sustainability as an antecedent of environmental and economically related constructs, the model configurations among green transformational leadership, transactional leadership, POS, autonomous motivation, job engagement, and job satisfaction show moderate to strong effects both within the social itself and in interactions with environmental (e.g., attitudes toward the natural environment, commitment toward the natural environment, and pro-environmental behaviors) and economically related constructs (e.g., employee performance). The diverse set of first- and second-order model configurations validates the importance of the social in organizations. Second, the diverse degrees of constructs’ ordering are studied to validate the centrality of the social in organizations and show the importance of paying attention to variability, as well as to the need to study in detail and integrate the social with environmental and economic dimensions. Findings suggest that construct variability within the social dimension should be further investigated, as well as its resulting interactions and effects as social variability increasingly integrates with variability in the environmental and economic dimensions.

For example, the broad scope and potential reach of some social sustainability constructs such as POS and green transformational leadership contrasts with the narrow/limited impact of constructs such as transactional leadership and task-defined employee performance. Since organizational life is relational [231], and therefore, social, all constructs may be deemed to be social. For instance, attitudes toward the natural environment, environmental commitment, and pro-environmental behaviors are generated by and/or are associated with or impact organizational (and societal) relational constructs [232]. However, internalization of the environmental and enactment of the corresponding behaviors by employees needs improvement. Similarly, since environmental considerations may be increasingly required as constitutive of organizational processes, every construct may (in the future) be deemed “green”. At the same time, we need to acknowledge that, at present, not everything in organizations is really “green”. Consequently, enhancing and properly balancing the environmental, social, and economic components in organizations is complex, challenging, and interesting. Given the potential diversity involved in integrating sustainability dimensions in organizations, it is necessary to specify constructs, their relationships, and boundary conditions, as well as meanings assigned to constructs that may be equally or similarly named but have different contents, items, and/or meanings. In addition, broadly encompassing constructs may be too vague and less useful for practitioners than specific constructs. Furthermore, findings intimate that, in discussing organizational sustainability at the country level, generalization may jeopardize specificity. Variability within and between organizations leads to a certain degree of both structuring and looseness in construct relationships, which may decrease the generalization of research findings. At the same time, since the “average” organization does not exist, generalizable knowledge may underestimate constructs’ effects or/and may not fully apply to specific firm conditions. This is an instantiation of Thorngate’s impostulate [193]. Thus, we must properly balance the emic and the etic in organizations. While looseness stemming from within and between employee and organizational variability may provide room to maneuver for practitioners, it is, at the same time, an obstacle in closing the gap between theory and practice. Clarity about interplays between ordering and looseness within and between organizations offers a more realistic perspective than just focusing on (strong) ordering, the dominant approach so far, which may overestimate the degree of control that may be exercised, as well as what can be achieved in organizations. As a result, more comprehensive models and multiple configurations of such models may add contextuality, and realism, to research results.

6. Limitations and Future Research

This study has several limitations. The study utilized cross-sectional data. Constructs were examined at the employee level. Similar studies of the social at the team, organizational, and other collective forms seem necessary. Although our sample included respondents from 34 out of 50 states pertaining to various organizations, industries, and regions, we do not claim country representativeness. Future studies may include larger diverse samples. We acknowledge that the set of possibilities is narrowly represented by the study of a set of 10 constructs and some of their relationships, which were used only for illustrative purposes. Given organizational complexity, more comprehensive formulations of the social, environmental, and economically related constructs, their interactions, and their dynamics are needed. At the same time, a balance between construct comprehensiveness and specificity is needed. Since the social may be a diverse set, it does not necessarily reflect employee interests. Future research may focus on social constructs that explicitly reflect employee interests. These indications for future research may enhance the generalizability and robustness of research findings regarding the diversity of interactions among social, environmental, and economically related constructs in organizations.

7. Conclusions

A large and diverse set of model configurations, depicting contrasting degrees of ordering and diverse positions among social, environmental, and economically related constructs suggest larger effects of social constructs on multiple employee outcomes than did environmentally and economically related constructs. Social sustainability in organizations is complex, complicated, and challenging. Given its importance, potential, and multiple manifestations, it requires substantial conceptual and empirical efforts. Nonetheless, since social sustainability is fundamental for sustainability in general future research will help to improve employee well-being and development, organizational productivity, and innovation as well as the future.