Impact of Compassion on Voice Behavior: A Moderated Mediation Model

Employee voice is a constructive and change-oriented communication that aims to improve a situation. In line with conservation of resource theory, our research proposed a moderated mediation model by examining the indirect effect of compassion on voice behavior through the mediating effect of affective commitment, and also examined the conditional effect of managerial support in the mediated relationship of compassion and voice behavior. Data were obtained from employees and their immediate supervisor in the public sector in three times at regular intervals of one week within a 2-month span of time. By using PROCESS macro on an actual sample of employees (300) and supervisors (19), our study found that compassion is positively associated with affective commitment that, in turn, is positively associated with voice behavior. Our study also found that affective commitment mediates the relationship between compassion and voice behavior. Furthermore, managerial support negatively moderates the relationship between affective commitment and voice behavior as well as mediating effect of affective commitment between compassion and voice behavior. The study finding adds to the deeper understanding of the pivotal construct, i.e., voice behavior. In addition to recommendations for more empirical research on voice behavior, theoretical and practical implications are given.


Introduction
Employee voice is a communication that is discretionary for the manifestation of ideas, opinions, and concerns about work-related issues to make a positive contribution towards work or the organization [1]. Voice behavior is appreciated by organizations as it assists the organization to identify problems [2] that might be ignored otherwise; it also helps in making innovations and adjustments to the rapidly evolving competitive business world [3,4]. Employees can contribute to the organization's successful functioning by integrating voice behavior and, therefore, they can provide recommendations about how to improve the organization's current situation [5]. Adopting voice is not only important for organizations but also for the employees as it helps in improving organizational innovation [6], employee motivation [7], job satisfaction, leader member exchange (LXM) [8], innovative work behavior [9], career growth [10], creative work involvement [11], team performance [12], and perceived organizational status [13] and also enhances employees' capacity for meeting different demands on the job [14]. Furthermore, voice behavior reduces employee turnover intention [15,16]. Despite the fact that voice behavior generates different advantages, indulging in voice behavior is not free from risks [17]. However, as there are different advantages of voice behavior, scholars are giving a lot of attention to identifying factors that facilitate this behavior in the workplace [18,19]. Prior studies have examined many individual and situational factors that can prompt employees' voice behavior, i.e., thriving at work, ethical leadership [20], moral efficacy [21], perceived nature of work, and person-organization fit [22]. Therefore, our study aims to extend the extant literature of voice behavior by theorizing affective commitment as a new predictor of employee voice behavior. Few studies have examined whether affective commitment is a predictor of employee voice behavior [23]. By favoring this empirical assumption, our study proposed that affective commitment encourages employee voice behavior. Only those employees exhibit voice behavior who are highly committed to the organization, its goals, and values [24,25]. Employees with increased organizational commitment are extra work oriented as compared to those who are less committed [26]. Affective commitment is characterized as the desire of the employee to remain a member of the organization, an intention to make an effort for the organization, and a belief in the organization's norms and values [27].
Furthermore, suffering is an unavoidable part of the human condition [28] and organizational life [29], which appears in all workplaces in one way or another. It includes a broad category of individuals who experience both physical and emotional distress, psychological stress, and detached feelings that may be precipitated by specific events and situations [28]. Individuals take the pain from their own experiences to work. Apart from this, it can also occur within the organization due to bad communications with bosses, colleagues, or consumers [30]. This situation has signified the importance of compassion in the workplace to alleviate the pain in times of suffering [31]. Compassion is an altruistic behavior and virtue that contains an urge to minimize others' anguish [32]. However, organizational scholars and practitioners have recently begun to pay more attention to workplace compassion by recognizing its positive effect on employees' attitude and behaviors [29]. Prior research explains that compassion at work creates many beneficial outcomes for employees [30], i.e., collective self-esteem, work-related identity [33], sense of work engagement, organizational citizenship behavior, service-oriented performance, knowledge sharing, and lower burnout [34]. However, the influence of compassion on the employee affective commitment was empirically examined by few studies [29,30,35]. Therefore, our study aims to explore how experiencing compassion influenced the attitude or behavior of employees, i.e., affective commitment in the organizational setting. In addition, our study examined the intervening role of affective commitment between compassion and employee voice behavior. This is the first study to our knowledge that explores the indirect effect of compassion on voice behavior through affective commitment. Next, we assume that the moderating effect of managerial support between affective commitment and voice behavior. We also anticipate that managerial support has the potential to influence the positive mediated association between compassion and voice. Managerial support is one construct to promote commitment towards the organization [36,37], however, despite its numerous benefits, managers may take this behavior as intimidating and become defensive and may use different ways to discourage it [17]. Therefore, managerial support is introduced as a moderator in the affective commitments and voice behavior relationship and the chances of voice behavior taking place decrease. Employees' higher perception of support from management lessens the positive association between compassion and voice behavior through affective commitment. In this, we used conservation of resource theory to support our theoretical model. Grounded upon conservation of resources (COR) theory; [38] compassion helps people to deal effectively with psychological or physiological stress [39] by providing social support, moral support, money, time, and gifts in times of suffering. When employees experienced compassion, in the workplace, they feel like they are respected which encourages them to form a positive image of the organization and stay committed with organization [35,40], that enable them to raise their voice. The COR perspective [38] maintains that social support (e.g., understanding by employers) helps in increasing and maintaining resources, thus protecting individuals from losing any resources. These conserved resources help employees to maintain commitment [41] by making an individual effort towards job raises their affective commitment that can further increase voice behavior. So, in order to fill the gap, the main goals of our study are: (1) to examine the direct impact of compassion on voice behavior which in turn influences employee voice behavior, (2) to explore the intervening role of affective commitment between compassion and voice behavior, (3) to investigate the moderation effect of managerial support on the mediating effect of affective commitment between compassion and voice behavior.
This research provides many theoretical contributions. First, this study expands the COR literature by investigating how compassion generates voice behavior via affective commitment and also whether the moderating impact of managerial support between affective commitment and voice behavior raise the chances of voice behavior or not. Second, a model is developed that explains the mechanism of compassion and voice behavior through the involvement of affective commitment, as well as the affective commitment and voice behavior relationship through the conditional effect of managerial support. Thus, by testing those relationships, a novel path is established. Third, this research has important insights for organizations as well as their managers by reflecting on ways of improving affective commitment and compassion that enables the employees to show specific positive behavior, i.e., voice behavior. Lastly, this study has extended the voice behavior literature by exploring how voice behavior can be influenced by various factors. This study makes more of an effort to increase the generalizability of voice behavior in the Asian organizational context, as prior studies on voice behavior were carried out in various Western countries. Figure 1 shows all the proposed relationships.
raises their affective commitment that can further increase voice behavior. So, in order to fill the gap, the main goals of our study are: (1) to examine the direct impact of compassion on voice behavior which in turn influences employee voice behavior, (2) to explore the intervening role of affective commitment between compassion and voice behavior, (3) to investigate the moderation effect of managerial support on the mediating effect of affective commitment between compassion and voice behavior.
This research provides many theoretical contributions. First, this study expands the COR literature by investigating how compassion generates voice behavior via affective commitment and also whether the moderating impact of managerial support between affective commitment and voice behavior raise the chances of voice behavior or not. Second, a model is developed that explains the mechanism of compassion and voice behavior through the involvement of affective commitment, as well as the affective commitment and voice behavior relationship through the conditional effect of managerial support. Thus, by testing those relationships, a novel path is established. Third, this research has important insights for organizations as well as their managers by reflecting on ways of improving affective commitment and compassion that enables the employees to show specific positive behavior, i.e., voice behavior. Lastly, this study has extended the voice behavior literature by exploring how voice behavior can be influenced by various factors. This study makes more of an effort to increase the generalizability of voice behavior in the Asian organizational context, as prior studies on voice behavior were carried out in various Western countries. Figure 1 shows all the proposed relationships.

Compassion and Affective Commitment
Compassion in the workplace has been defined as the response by the heart to the suffering of other people that includes observing the pain of members of an organization, experiencing feelings of empathy for their suffering, and trying to act in a way to demonstrate love, kindness, and warmth to relieve their affliction [31,42]. Compassion is composed of a relational process that includes giving, receiving, or observing the display of empathy among organizational employees, and consoling those who are in pain [32]. Recovery from personal loss becomes easy for employees when their colleagues display compassion, which plays a major role in acknowledging their suffering and putting effort to relieve the pain [43]. Employees display compassion towards others through behavior and interactions within an organization [44]. Compassion in the working context is a social behavior that helps employees deal with daunting situations in their personal and professional lives [44] as it inspires and enables other people to work together, flourish, and thrive [45,46].
Affective commitment is the important determinant of loyalty and dedication. Prior research explained that when employees experience thriving at work [47], pleasant job experiences [48], have a desire to help other people, perception of organizational support [49], achieve high-performance work practices [50], and receive workplace civility and family motivation [51] as a result of it, they feel higher emotional attachment towards their organization. Therefore, we theorized that when employees experience compassion in the workplace, they form the perception that their organization

Compassion and Affective Commitment
Compassion in the workplace has been defined as the response by the heart to the suffering of other people that includes observing the pain of members of an organization, experiencing feelings of empathy for their suffering, and trying to act in a way to demonstrate love, kindness, and warmth to relieve their affliction [31,42]. Compassion is composed of a relational process that includes giving, receiving, or observing the display of empathy among organizational employees, and consoling those who are in pain [32]. Recovery from personal loss becomes easy for employees when their colleagues display compassion, which plays a major role in acknowledging their suffering and putting effort to relieve the pain [43]. Employees display compassion towards others through behavior and interactions within an organization [44]. Compassion in the working context is a social behavior that helps employees deal with daunting situations in their personal and professional lives [44] as it inspires and enables other people to work together, flourish, and thrive [45,46].
Affective commitment is the important determinant of loyalty and dedication. Prior research explained that when employees experience thriving at work [47], pleasant job experiences [48], have a desire to help other people, perception of organizational support [49], achieve high-performance work practices [50], and receive workplace civility and family motivation [51] as a result of it, they feel higher emotional attachment towards their organization. Therefore, we theorized that when employees experience compassion in the workplace, they form the perception that their organization takes care of them, hence, increasing their affective commitment [40]. Compassion also helps employees to strike a balance between their work and personal lives, reducing interpersonal disagreements and, helping them in raising their affective commitment [52]. Besides, subordinates whose supervisors show compassion towards them feel high levels of positive emotions which increase their affective commitment towards the organization [29]. Likewise, employees show more emotional attachment to their organization when they experience compassion for people in their workplace [35]. Therefore, based on the conservation of resources [38], our study theorized that compassion will positively impact employee affective commitment. According to this theory, when people have resources, they handle stressful incidents better. This theory expresses the role of personal, material, and energy resources. Therefore, we stated that individuals who better handle and move on from their adverse situation also continue their work by lessening their unbearable fear and anxiety [30] when they experience compassion in the form of social, moral, and verbal support, polite listening, meals, hugs, capital, gifts, time, and pliability [32] from coworkers and supervisors in times of suffering. These finite amounts of resources (material, social, and psychological) in the organizational context in times of suffering help them to show positive behaviors [38], i.e., affective commitment. It is also believed that compassion at the workplace helps employees to deal effectively with psychological and physiological stress (e.g., in case of a natural disaster) [39]. We anticipate that compassion within an organization provides relief to those employees who are suffering and, hence, this results in a positive attitude, i.e., affective commitment. This leads to our first hypothesis:

Affective Commitment and Voice Behavior
Affective commitment is constructed as the emotional attachment a person shows towards their organization [53]. It describes the extent of emotional attachment, involvement, and identification of employees with their organization [54]. Employees who rank highly in affective commitment tend to be more energetic at work, so the functioning of the organization could be improved even when faced with complications [55]. Affective commitment leads to a number of positive outcomes such as reduced withdrawal behavior (absenteeism and lateness) [56], turnover intention [57], workplace deviant behavior [58], improved performance, involvement in organizational citizenship behaviors [59], enhanced physical, as well as psychological, well-being [60], and work engagement [61]. Affective commitment helps employees to indulge in different extra-role behaviors [62]. Therefore, employees align with the organization's goals and beliefs by displaying voice behavior when they possess a higher affective commitment [24]. Employees are more likely to exhibit voice behavior when they have a special attachment to their organization [63]. Prior studies found that affective commitment is the strongest indicator of employee voice behavior [23,25]. Therefore, in line with these prior studies and conservation of resource theory, we anticipate that employees will be involved in extra-role behavior (i.e., voice behavior) when they are emotionally attached to their employing organization. According to COR theory, commitment is a personal resource, therefore employees can increase commitment and also increase the possibility of resource preservation by providing personal involvement in the job that they are doing [41,64]. Therefore, we hypothesized this as: Hypothesis 2 (H2). Affective commitment is positively related to voice behavior.

Mediating Effect of Affective Commitment
In this study, we theorized that compassionate behavior may lead to affective commitment towards an organization, due to the attachment and identification of employees with their organization. Affective commitment helps in building up the identification of employees with their workplace [65], so it can enhance the possibility of extra-role behaviors, e.g., voice behavior [25]. Studies noted that highly effective employee have a stronger sense of identity and consider the interest of organization as their own. When problems arise, these employees are less likely to give up and are more likely to voluntarily share new ideas, give warnings, or promote positive improvements [24]. In addition, employees with high levels of affective involvement put in more effort to enhance organizational functioning for the betterment of their organizations by facing many difficulties in guaranteeing improvement as well as challenging the status quo [23]. Employees with higher rates of compassionate behavior have high chances of providing emotional support to their colleagues as it not only makes them understand their organizational mission, but also makes them align with the purpose of their work [66]. Thus, a sense of emotional attachment towards the organization is developed among employees [67,68] when people are able to support their actions and have power over their work [69,70] and believe their organization is honest [71]. In that instance, there is a high possibility of voice behavior being exhibited. Previously, it was seen that affective commitment with an organization, organizational justice, and autonomy are linked with voice behavior in a positive manner [72]. By building on the given argument and based on the COR theory, our study hypothesized that affective commitment plays the role of a mediator in the compassion and voice behavior relationship. Experienced compassion in the workplace provides the many resources to employees that encourage their affective commitment that, in turn, lead them to voice behavior. So, we hypothesized this as:

Hypothesis 3 (H3).
Affective commitment mediates the relationship between compassion and voice behavior.

Moderating Role of Managerial Support
Managerial support is defined as the extent to which overall impressions are formed by employees about their managers about that whether they value the contributions made by their subordinates, support them, and show concern for their welfare [73]. Supervisor support is also considered to be an informal kind of support at the organizational level [74]. The construct of managerial support has attracted significant attention from researchers in organizational research, as it is linked with many beneficial individual and organizational outcomes, i.e., retention, performance, and productivity [75,76]. Employees feel more efficacious and engaged in their work when they receive support from their managers [77]. Prior research has highlighted that managerial support is one of the major contextual factors that impact the commitment of employees [36,78]. Eisenberger et al. [79] highlighted that there are high chances for employees to be effectively committed to their organization when they feel that their organization supports them. Employees' commitment towards their organization is also greatly affected by supervisor support [80]. However, research has established that there are many benefits of managers welcoming ideas and opinions of employees at different levels of hierarchy in the organization. These benefits might lead one to believe that managers are open to encouraging the voice of employees [81]. Constructive voice provides different benefits to managers and to the units led by them [15]. However, deeper investigation of managers' action in organizations shows that many managers indulge in behaviors that discourage voice behavior [17]. As raising voice by employees for the purpose of giving ideas and suggestions is considered to be a voluntary behavior, so managers may perceive it as subordinates indulging in extra-role behaviors thus, threatening the status quo [81]. Moreover, as per Hobfoll [38], the resources of a person can not only be increased, but could also be reinforced or replaced (in case any resources are lacking) through the utilization of social support. In order to clarify the resource's nature, Hobfoll [38] empirically tested a list of 74 resources in which understanding (compassion) received from employer and co-workers' support were included in social resources that are work related and helps in future resource preservation. Therefore, in this study, we posit that managerial support is the critical phenomenon that can change the positive effect of affective commitment and employee voice behavior. Affectively committed employees will feel hesitant to raise their voice in the workplace, especially when they receive the support from their supervisors. When they perceive that their manager fulfills their needs and are more concerned about their well-being as a result of, they do not also raise their voice in the workplace for constructive changes and better organizational effectiveness. So, we hypothesized this as:

Hypothesis 4 (H4).
Managerial support moderates the relationship between affective commitment and voice behavior, for example, when managerial support increases, the frequency of raising voice decreases.
In addition, we anticipate that managerial support weakens the indirect effects of compassion on voice behavior by taking affective commitment as a mediator, as presented in Figure 1 (moderated mediation model). Hence, we propose the moderated mediation hypothesis:

Hypothesis 5 (H5).
Managerial support moderates the mediated relationship between compassion and voice behavior. This mediated relationship increases with the decreasing frequency of managerial support rather than high support.

Procedures and Participants
A public sector organization of South Asia was chosen. The reason behind choosing this organization was to check the frequency of voice in a public sector institution, because employees often hesitant to raise their voice due to rigid and strict rules and regulations and complex hierarchies of government institutions. Though the organization has offices in different cities, for this study, two major offices of the institution in two big cities were targeted because they had the highest number of employees and getting access to them was relatively easy. Questionnaires were distributed among employees ranging from Grade 07 to 19. Around 30 supervisors at various levels of the organization were contacted for the survey. Each supervisor had a staff of almost 50 to 60 employees under them. Out of the 30 contacted, only 19 agreed to participate in the survey, after getting approval of their staff members. In the first office, 200 questionnaires were distributed as it had a big unit, whereas in the second office, 150 questionnaires were distributed among various full-time employees of the organization. The data collection through the questionnaires was done at three different time intervals (time lagged data structure) in both the offices independently. Time lagged data were used for receiving more accurate (bias-free) data by analyzing whether events that took place in the workplace during the different data collection intervals had any effect on employee responses or not. There was a gap of one week between each questionnaire distributed at time 1, time 2, and time 3.
The survey was self-report as well as multi-source to strengthen the design of the study. This way, responses were collected from employees as well as their immediate supervisors, thus increasing the authenticity of the responses received. The participants were assured that their anonymity would be maintained. We approached both the offices and the participants through research associates. These research associates were fully trained to properly explain the questions to the respondents so that no confusion occurred in understanding the questions. Upon the distribution of questionnaires, participants were asked to think about themselves and others' behavior over the last month and then answer the questions. Similarly, the supervisors were also asked to rate employees' voice behavior after responses from employees were received.
At time period 1, data were collected from all employees that were part of the sample in the form of a self-report survey. Questionnaires were handed over to different supervisors of Grade 17, 18, and 19 who then distributed them among their staff members. The staff members were required to answer various questions regarding managerial support. It took almost a week to get the filled questionnaires back. Three hundred and fifty questionnaires were distributed out of which 310 completely filled questionnaires were returned. After 1 week, the participants who took part in the survey conducted at time period 1 and gave complete responses were again contacted by their supervisors for a second survey at time period 2. In this survey, the participants responded regarding compassion and affective commitment. Three hundred and ten questionnaires were distributed, out of which a final of 300 competely filled questionnaires were received that were used in the final study.
Finally, at time period 3, a total of 300 participants who gave complete responses for data collected at time period 1 and 2 were rated by their supervisors for their voice behavior. The supervisors were given enclosed questionnaires regarding employees' voice and were asked to fill them. So, the data at time period 1 and time period 2 were collected from employees, whereas those at time period 3 were collected from the supervisors of the same employees, thus making the data self-reported as well as multi-source. The total responses in the study were 300. Two hundred and seventy-three (91.0%) respondents were male and 27 (9.0%) were female. In terms of marital status, 250 (83.3%) respondents were married and 50 (16.66%) were single. The age of the respondents was between 15 and 60 years. The mean age was 39 years. The education of respondents was between 10-19 years (Mean = 15 years) and tenure was between 1-40 years (11 years).

Managerial Support (α = 0.78)
Managerial support was assessed at time 1 by using a six-item scale of Anderson et al. [82]. The sample items were "My supervisor is supportive when I have a work problem" and "My supervisor is fair and doesn't show favouritism in responding to employee's personal or family needs".

Compassion (α = 0.80)
Compassion was measured at time 2 by using a three-item scale of Lilius et al. [29]. The sample items include "I frequently experienced compassion from supervisor" and "I frequently experienced compassion from co-workers".

Affective Commitment (α = 0.87)
Affective commitment was measured at time 2 using a three-item scale developed by Mowday et al. [83]. The sample items were "I am proud to tell people who I work for" and "I share many of the values of the organization".

Voice Behavior (α = 0.96)
Voice behavior was computed at time 3 using a six-item scale used by Van Dyne and Lee Pine [84]. The sample items were "This particular subordinate develops and makes recommendations concerning issues that affect this work group" and "This particular subordinate speaks up and encourages other in this group to get involved in issues that affect the group".
All the measures were adopted for their proven psychometric properties: reliability and construct validity (convergent, discriminant, and predictive). All constructs were measured on a five-point Likert type scale ranging from strongly disagree (1) to strongly agree (5).

Data Analysis Approach
The empirical data were statistically analyzed using a two-step procedure. First, descriptive analysis, correlational analysis, reliability, and validity of the measure were examined by using SPSS and AMOS (24). Second, our proposed moderated mediation model (hypotheses) was tested using Hayes's [85] techniques. PROCESS Macros is the best and recommended technique to test the indirect and conditional effect [47].

Measurement Validation
A measurement model was evaluated prior to testing the framed hypotheses by computing confirmatory factor analysis in AMOS 20 to confirm the fit indices of the various study variables. According to the statistical analysis carried out by the authors' proposed (four-factor) model composed of compassion, affective commitment, managerial support, and voice behavior, as considered suitable to data, significant findings were revealed: χ 2 = 262.179, p < df = 129, Comparative Fit Index (CFI) = 0.97, Turkish Lower Index (TLI) = 0.96, Root Mean Square Error of Approximation (RMSEA) = 0.05. As all these values are in the acceptable range, i.e., IFI, CFI, TLI < 0.95, and RMSEA > 0.08 [86], as compared to fit indices of three factors model (χ 2 = 608.79, p < df = 132, CFI = 0.89, TLI = 0.88, RMSEA = 0.11) that combined compassion, managerial support, and affective commitment into one factor, two factor mdoel (χ 2 = 1335.17, p < df = 134, CFI = 0.74, TLI = 0.70, RMSEA = 0.17) that combined compassion and affective commitment into one factor, and combined voice behavior and managerial support into a one factor, also one-factor model (χ 2 = 2297.41, p < df = 135,CFI = 0.53, TLI = 0.47, RMSEA = 0.23) that combined all factors into one (Table 1). Furthermore, the validity (convergent and discriminant) of measures was estimated in line with Fornell and Larcker's [87] recommendations. The researchers [87] stated that the acceptable limit of composite reliability (CR) is <0.80 and average variance extract (AVE) is <0.50 for ensuring validity. All variables have convergent validity as CR and AVE of all factors were greater than 0.50 and 0.7, respectively, in this study. Similarly, the criteria of discriminant validity were attained as the diagonal √ AVE of all factors were larger than the inter-construct correlation values (see Table 2).

Descriptive Analysis and Intercorrelations
In line with the hypothesized relationship, the correlation analysis showed that compassion is positively linked with affective commitment (r = 0.44, p < 0.01). So, H1 is supported. The relationship between affective commitment and voice behavior is positive, thereby (r = 0.16, p < 0.01) supporting H2. The results showed that the managerial support and affective commitment relationship is significant and positive (r = 0.28, p < 0.01) and the managerial support and voice behavior positive relationship is also significant (r = 0.42, p < 0.01), providing support for H4 that managerial support has a moderator relationship between affective commitment and voice behavior, consequently affecting the frequency of voice behavior (Table 3).

Hypotheses Testing Results
Moderated mediation analyses (Model 14 PROCESS) with 1000 bootstrapping [85] were conducted to test whether compassion indirectly influences voice behavior through the intervening role of affective commitment and whether managerial support moderates the affective commitment and voice behavior relationship. Table 4 depicts the results of a simple mediation model. In line with Hypothesis 1, the results demonstrate that the direct effect of compassion on affective commitment is significant and positive (β = 0.38, t = 8.64, p < 0.01). Favoring the assumption of Hypothesis 2, the results showed that affective commitment is related with voice behavior in a positive way (β = 0.14, t = 1.97, p < 0.01). These results provide the support for mediation paths and are consistent with Hypothesis 3. The results showed that commitment mediates the compassion and voice behavior relationship. This effect is positive and significant as the Sobel test and normal theory support each other (β = 0.05, p < 0.05, LLCI = 0.009, ULCI = 0.116, z = 1.91).  and +1 SD (4.63 < high mean value). The indirect effect was significantly different from zero among low mean value (β = 0.04, LLCI = 0.000, ULCI = 0.110), mean value (β = 0.00 LLCI = −0.041, ULCI = 0.400), and high mean value (β = −0.04, LLCI = −0.108, ULCI = −0.005) for managerial support (see Table 5).

Discussion
Based on the conservation of resource theory, our research tested a moderated mediation model by examining the indirect effect of compassion on voice behavior through the mediating effect of affective commitment, as well as examined the conditional effect of managerial support on the mediated relationship of compassion and voice behavior in a public sector organization. The study findings showed that compassion is positively related to affective commitment. These findings are consistent with prior studies which demonstrated that experiencing compassion in the work setting enabled the employees to show a positive attitude or behavior, i.e., affective commitment [29,30,35]. This study showed that affective commitment raises the chances of exhibiting voice behavior in line with the studies of [23][24][25] that observed that affective commitment is the strongest contributor to voice behavior. Moreover, the study findings showed that affective commitment intervenes in the compassion and voice behavior relationship by creating a connection between the two. This showed that there is a psychological and emotional link between an organization and employee in line with the study of Cao et al. [88] that examined affective commitment as a mediating variable in the relationship of work family conflict and job performance. The results also showed that managerial support acts as a moderator among the path between affective commitment and voice behavior which decreases the possibility of voice behavior taking place. Lastly, the results showed that managerial support negatively moderates the indirect effect of compassion on voice behavior via the incorporation of affective commitment as a mediator. These study findings are different from the studies of Mukanzi and Senaji [89] and Mazzetti et al. [90], who found a positive moderation effect of managerial support, while our study found a negative conditional effect.

Conclusions
This research used PROCESS Macros (Model 14) to empirically test the indirect effect of compassion on employee voice behavior through the mediating effect of affective commitment, as well as the moderation of managerial support between affective commitment and voice behavior also the conditional effect of managerial support on the mediated effect of affective commitment between compassion and voice behavior. In line with our proposed moderated mediation model and conservation of resource theory, the study shows that compassion has a positive effect on employees' affective commitment, which in turn is linked with voice behavior. Compassionate behavior creates more emotional attachment among employees towards their organization that encourages the employees to become involved in extra-role behavior, i.e., voice behavior. The study findings give evidence for the mediating effect of affective commitment in the compassion and voice behavior relationship, and managerial support moderates the affective commitment and voice behavior association by decreasing voice frequency. Employees will not raise their voice in an organization with higher managerial support. This study also finds that managerial support negatively moderates the indirect influence of compassion on voice behavior via the incorporation of affective commitment as a mediator. This study not only makes an addition to the voice behavior literature by giving enough attention to direct and indirect paths which influence the relationship of compassion and voice behavior, but it also describes an understanding of the mediating role of affective commitment between compassion and voice behavior, which might help organizations to better regulate compassion and encourage their employees' level of emotional attachment. This study also helps managers to know how they can use various approaches that will encourage employees to indulge in voice behavior, which could ultimately affect organizational success and productivity.

Theoretical Contributions
The study findings contributed to the literature in the following ways. First, this research extends the voice behavior literature by emphasizing factors that can influence voice behavior, i.e., affective commitment. Prior studies have investigated the many personal and situational predictors of employee voice behavior [20][21][22]. However, the influence of affective commitment on employee voice behavior has been less explored in the literature. Therefore, our study empirically explored the relationship of affective commitment with voice behavior in public sector organizations. Second, our study contributed to the compassion literature in organizational research by examining its positive effect in an organizational setting. Previous studies examined the role of compassion in the creation of positive outcomes in the area of sociology and medicine [46] and this topic has been overlooked by researchers in organizational research. In addition, by replicating findings in the Asian context, the study results showed that compassion is the key contributor to affective commitment. The effect of compassion on employees' affective commitment in various cultures shows that compassion at work is a universally essential factor in the achievement of efficiency and sustainability in complex and unpredictable environments. Third, the beneficial influence of compassion on voice behavior was explored in this study by analyzing the mediating effect of affective commitment, a precedent of voice behavior. By investigating why and how affective commitment affects the relationship between compassion and employee voice behavior, this finding complements the current body of literature of affective commitment in the field of organizational behavior. Past studies explored the mediating role of affective commitment [25,88], while the mediating effect between compassion and voice behavior was examined for first time, to our knowledge. Fourth, this study investigates the conditional effect of managerial support on voice behavior by impacting the affective commitment and voice behavior path. As few studies have examined the moderation of managerial support in the literature, therefore, this research is among the first of those studies that have examined the conditional effect of managerial support on affective commitment and employee voice behavior. Furthermore, our study adds to the managerial support literature by exploring its dark side. By drawing relationship on the basis of COR theory, our study has provided an insight into the effectiveness of this theory in the Asian context. COR theory explains the different actions that may be taken by people (or their supervisors) for increasing resources [91,92]. However, compassion, managerial support, and affective commitment are those factors that obviously secure, sustain, and improve the resources of employees.

Practical Implications
This study provides many practical implications for practitioners and managers who are in favor of promoting compassion, affective commitment, and voice behavior in the workplace. First, the study findings showed that compassion is important for increasing the affective commitment of employees; therefore, our study suggested that to set up a compassionate culture in an organization, managers should give attention to the implementation of a systematic compassionate climate that is part of the organizational goals and strategies, a reliable and trustworthy way of providing support to employees who go through suffering, instilling an increased level of uprightness and integrity in the organizational members, developing reciprocal relations based upon being concerned and helpful, and tolerating mistakes and encouraging forgiveness across the organization [30]. This research also indicated that managers should consider developing employee assistance programs (EAPs) and programs for employee support [93] so that workers gain emotional support through these programs in emergency situations. Organizations can also develop a compassionate culture by cultivating a collaborative environment, acknowledging positive attribute and strengths in front of others, and fostering an open dialogue culture in the workplace. Second, as affective commitment prompts voice behavior, therefore, organizations can increase employees' affective commitment by using various strategies that may include implementing an effective employee recruitment system, initiating programs that support employee involvement [93], and providing them different opportunities for career management [94]. Organizations can create emotional attachment among employees by giving positive feedback, developing trust, and creating a positive culture in the workplace to promote open innovation [95,96]. Open innovations [97] encourage inter-organizational collaborations among employees of different organizations for idea generation [98,99] and prompt employees' internal and external knowledge integration for innovation [100]. As innovation is deemed as an essential prerequisite for the strategic and construction maintenance of organizations that could improve companies' competitive advantage and ensure prosperity and long-term sustainability [101]. Third, voice behavior can play a vital role in organizational functioning and effectiveness. Therefore, a study suggested that an organization can enhance employees' voice behavior through a variety of ways, i.e., a system for distributing rewards should be incorporated in organizations that encourages employee contribution and encourages them to exhibit voice behavior [102]. There are various studies that have highlighted the importance of good superior and subordinate relationships for enhancing voice behavior [71,103]. Therefore, it is important for supervisors to encourage voice behavior of employees as it not only benefits the organization or work units of employees but also helps employee development. Hence, such behavior should be dealt with in a positive way and should not be punished by superiors. By stressing freedom of speech in the workplace, managers proactively embrace and promote voice behavior from employees [23]. Organizations should routinely offer constructive feedback on the suggestions that employees give, thereby promoting voice behavior by enhancing the sense of self-efficacy among employees. The positive behaviors of managers can play a vital role in the creation of a voice environment [104].

Limitations and Future Directions
This study is not free from certain limitations, like earlier studies. First, this study was conducted in the context of a public organization which could be a possible limitation when it comes to the generalization of results to private sector organizations. Future research should be carried out to look into the predictors and outcomes of employee voice behavior among other sectors, i.e., banking, telecommunication, education, and manufacturing organizations. Second, this study was cross-sectional, inhibiting the risk of causality rolling out [105]. Therefore, it would be better if future studies could investigate the hypotheses given in this study by using longitudinal studies or experimental studies to create a causal relation in field studies. Third, in this study, data were gathered from employees and their supervisors at different points in time by using a time lagged structure with a temporal gap of one week. There is a high chance of artificial inflation of the demonstrated relationships due to the presence of common method bias. To cater to this, future studies may collect the data with appropriate temporal gaps. Fourth, this study investigates the indirect effect of compassion on voice behavior through affective commitment, drawing on COR theory. Future studies can further expand the theoretical perspective of the relationship between compassion and voice behavior by examining it from the perspective of social exchange theory and incorporating other mediating variables, i.e., thriving at work, etc. Fifth, there are various forms of voice behavior, i.e., destructive, constructive, defensive, and supportive [106]. Future researchers can investigate the effect of compassion, affective commitment, and managerial support on these specific types of voice behavior, or use different directions from compassion and managerial support that can lead to these voice behaviors. In addition, this study addresses the frequency of voice behavior, thereby overlooking the quality of voice behavior. Therefore, future researchers can use mixed-method research designs to establish voice behavior frequency and quality [105]. Sixth, in this study, we examined the association among compassion, affective commitment and voice behavior on an individual level. Future studies may examine these associations on organizational and team levels. In addition, future studies can continue to investigate the antecedents and outcomes of compassion in the existing organizational literature, as it is still limited. Future research can also explore the negative moderation of managerial support on employees' attitudes and behaviors.