Examining organizational justice, emotional exhaustion and workplace bullying as antecedents of organizational commitment: A developing country’s perspective

This study examines the effects of Emotional Exhaustion, Emotional Labor and Workplace Bullying on Organizational Commitment and moderating role of Organizational Justice on these relationships. For this quantitative study of crosssectional nature, data have been collected from sample of the sales representatives and service employees working in cellular companies of Pakistan. Of 700 questionnaires that were distributed, 335 completely filled questionnaires were returned. Data have been analyzed using SPSS 20 & PROCESS and results suggest a positive relation between Organizational Commitment and Emotional Labor, while Emotional Exhaustion and Workplace Bullying are found negatively correlated to Organizational Commitment. Moreover, Organizational Justice shows moderating effect on Emotional ExhaustionOrganizational Commitment relationship only.


Introduction
Despite significant efforts, neither the modern day organizations have succeeded in fully attaining organizational commitment of their employees, nor could the scholars fully explain the dynamics surrounding this concept. A large survey of 50 Fortune 1000 firms (please note that most of them are best in the world) involving 1.2 million employees shockingly reveals that in 85% of the firms, employees turn less committed within first six months at the work floors (Chamorro-Premuzic & Garrad, 2017). Lack of Organizational commitment has serious negative repercussions for organizations. Data from world's most sophisticated and well managed telecom and tech organizations of US reveal a loss of $ 650 billion in 2018 alone, when less committed employees exposed 2.8 billion records of consumer data-containing personally identifiable information (Goren, 2019). The concept of organizational commitment has, therefore, redrawn immense managerial as well as scholarly attention (Yu, Yen, Barnes & Huang, 2019). In organizational contexts that are rapidly evolving at the back of sprouting technology, there is a need to have a fresh scholarly look at the antecedents of organizational commitment.
The concept of organizational commitment finds its roots in Becker's (1960) work. Thenceforth, researchers have extensively examined the effects, determinants and theoretical basis of organizational commitment. This construct is considered as a key factor that determines employees' loyalty, productivity, hard work, absenteeism, retention, turnover, prosocial and citizenship behaviour, volunteering for extra tasks and participation in organizational matters etc. (Allen & Meyer, 1990;Allen & Shanock, 2013;Yip, Muhammad & Segaran, 2019). This concept has also been rigorously analysed in terms of its determinants (Paul & Anantharaman, 2004;Yu et al., 2019;Chen, Zhou & Klyver, 2019). Observed antecedents of organizational commitment are categorized as a) personal-e.g., attitudes and work ethics, b) job related-e.g., significance assigned to a task, autonomy at work, complexity of job etc. and c) organizational-e.g., human resource practices and leaders behaviors etc. (Sharma, Mohapatra & Rai, 2013).
However, there are still few unexplored aspects that may build or destroy organizational commitment. For example, employees are expected to be more committed to their organizations, when work floor is not emotionally shattering and insolent. A work floor that is less emotionally exhausting, less emotional-labor-intensive and protects employees from bullying through well executed organizational justice system shall facilitate continued commitment of the employees. Notwithstanding the crucial role that emotional exhaustion, emotional labor, bullying and organizational justice may play in organizational commitment of the employees, studies that integrate these variables and examine their influence on organizational commitment are scarce. More importantly, research that examines the effects of these variables on organizational commitment of employees in service sector is virtually non-existent. Front line employees in services sector are more prone to face emotional exhaustion and labor owing to the fact that being representative of their organizations, the front-line service personnel have direct interaction with the customers (Adil & Kamal, 2013: Cho, Rutherford & Park, 2013Sardeshmukh, Goldsby & Smith, 2020). Furthermore, fewer jobs and poor enforcement of laws to protect employees in the context of developing country (e.g., Pakistan) make matter worst. Therefore, as recommended by previous researchers (Bhatti, Alkahtani, Hassan & Sulaiman, 2015;Cho et al., 2013;Vignoli, Guglielmi, Bonfiglioli & Violante, 2015;Thompson, Carlson, Kacmar & Vogel, 2019), this study examines if emotional exhaustion, emotional labor, workplace bullying and organizational justice explain organizational commitment in a developing country's context. This study is organized in the following format: First, the variables of the study are discussed very briefly in the literature review section leading to hypotheses building. Then the methodology, data analysis and results are presented that are followed by discussion and conclusion.

Organizational commitment
Since its inception in 1960s, Organizational commitment has been defined in many ways. Organizational commitment is commonly viewed as an employee's emotional attachment to the organization. Further, it has been described as the "the relative strength of an individual's identification with and involvement in to a particular organization" (Mowday, Steers & Porter, 1979, p. 226). Organizational commitment signifies a psychological state which describes employee's identification with an organization as well as his/her involvement in that organization (O'Reilly & Chatman, 1986;Kanta & Srivalli, 2019).
In Human Resource Management & Organizational Behavior literature, organizational commitment is one of the most studied factors and has an undisputable place due to its strong association with employees' retention and their job performance (Kuriakose, 2020). High organizational commitment among employees ensures their efficiency (Meyer & Allen, 1997) and loyalty to their organizations (Hunt, Chonko & Wood, 1986). The outcomes of organizational commitment have been studied in literature on individual selfrealization, absenteeism, turnover rate, job role and effort, levels of responsibility and autonomy, performance and many others (e.g. Jung & Yoon, 2012;Al-Jabari & Ghazzawi, 2019).

Emotional labor
Emotional labor is defined as the "the management of feeling to create a publicly observable facial and bodily display" for a wage (Hochschild, 1983, p. 7). Mostly literature on emotional labor conceptualizes this concept in two strategies: deep acting and surface acting. Deep acting refers to the inner feelings of an individual, which do not meet the performance requirement of an organization. An individual having a deep psychological processing (thinking, memory and imagination) surpasses negative emotions or stimulates positive emotions, so as to experience the emotions in line with the emotional performance of the organizational requirements (Grandey, 2000). Deep acting is about conforming actual emotions to the expected or desired emotions, for which an employee makes effort to align his/her actual feelings to the desired emotions as per situation (Hennig-Thurau, Groth, Paul & Gremler, 2006). While on the other hand, surface acting is related to the visual part in which individual adjusts emotions with respect to the required organizational emotions, when the emotions actually perceived by the individual do not conform to the performance rules required by the organization (Hochschild, 1983). Moreover, surface acting is masking out of real emotions and the emotions are expressed without actually feeling them. It involves suppressing of emotions or hiding of actual emotions by the fake ones (Cho et al., 2013).
Employees directly interacting with customers are directed to display organizationally desired emotions that in turn cause emotional labor among them (Julian, 2008). It is a condition in which a person manages and alters his/her feelings and behaviors to conform to an ideal (Barry, Olekalns & Rees, 2019). Emotional labor indicates expression of those rules that are desirable to an organization through deep acting as well as surface acting on the part of the employees (Cho et al., 2013;Kim, 2020).

Emotional exhaustion
Emotional exhaustion is defined as the feelings of being emotionally exhausted and overextended by one's work (Maslach & Jackson, 1981). Emotional exhaustion is actually a state in which an employee's energy is sapped out due to the excess of emotional demands (Chen, Chang & Wang, 2019). It is a state or feeling of extreme emotional fatigue and lethargy (Peng, Zhang et al., 2014;Sardeshmukh et al., 2020) that drains an employee physically, psychologically and emotionally (Lu, Sun & Du, 2015). Thus, emotional exhaustion has negative impacts on employees' behaviors (Cropanzano, Rupp & Byrne, 2003;Thompson et al., 2019).
Emotional exhaustion occurs when employees perform "people-work" of some kind. Service employees are prone to emotional labor as organizations often require them to show firm desired emotions, while interacting with clients. Professionals working in the service industry are required to promote the welfare of their clients, which imposes numerous emotional, cognitive and behavioral demands (De Jonge & Dormann, 2003). If mandated emotions do not reflect the true feelings of employees, then employees feel less committed towards their organization, thus, resulting into job alienation (Adelmann, 1996;Yang & Chang, 2008). Ongoing hassles and excessive demands of customers can lead towards the emotional and physical depletion, particularly among service industries employees, as they repeatedly encounter the verbal aggression and negative customer reactions which make them emotionally exhaust (Cordes & Dougherty, 1993). Therefore, service industries employees are considered more prone to be emotionally exhausted.

Workplace bullying
Bullying or belittling at workplace refers to the insulting, neglecting, niggling, negatively disturbing and maneuvering someone's work assignments. Bullying is a disturbing procedure in which people encounter belittling acts, where their ego is hurt due to orderly harmful social deeds (Einarsen, Hoel & Notelaers, 2009). Workplace bullying is a condition in which an employee experiences negative, annoying and aggressive behaviors/acts of other employees repeatedly and consistently (Cooper-Thomas, Gardner et al., 2013) creating an effect of humiliation, intimidation, fright and punishment (Einarsen, Hoel, Zapf & Cooper, 2020;Rai & Agarwal, 2018). Moreover, bullying at workplace is comprised of organized harmful behaviors performed by the perpetrators over a period of time. These harmful conducts or behaviors plunge victims into a situation to cause harm to the target immensely and the victims recognize that they are insecure and helpless to defend themselves (Einarsen et al., 2020). It is evident from the past literature that bullying at workplace occurs in many organizations throughout the world (Einarsen et al., 2020) particularly in the Asian firms, whereby its intensity is relatively high and it adversely affects employees' commitment (Nielsen & Einarsen, 2012;Cooper-Thomas et al., 2013;Sheehan, McCabe & Garavan, 2018).

Organizational justice
Organizational justice shows the employees' perception that the treatment and outcomes they receive in their organizations are fair, just and aligned with the expected moral and ethical standards (Cropanzano, Bowen & Gilliland, 2007;Cole, Bernerth, Walter & Holt, 2010). Past studies describe four dimensions of organizational justice based on distinct assessment of fairness in the decision making of an organization. These dimensions include distributive justice, procedural justice, interpersonal justice and informational justice.
Distributive justice refers to the perceived fairness of outcomes based on an employee comparing the ratio of his/her inputs and outcomes to the inputs and outcomes of referent others (Adams, 1965). Similarly, procedural justice proposes that individuals evaluate fairness not only on outcomes, but also on fairness in the decision-making process and the ability to have voice in this process (Levanthal, 1980). Likewise, interpersonal justice refers to the fairness perceptions of interpersonal treatment, while informational justice reflects fairness perceptions of the adequacy and truthfulness of explanations (Colquitt, 2001). An organization with fair procedural, distributive and interactional justice ensures commitment, loyalty and satisfaction of its employees (Ayobami & Eugene, 2013).

Organizational commitment and emotional labor
Previous research offers mixed results on relationship between organizational commitment and emotional labor. For example a part of research shows significant positive impacts of emotional labor on employees' commitment to the organization (Ghalandari, Jogh, Imani & Nia, 2012;Cho et al., 2013). Notably, specific emotional labor strategies may have a significant impact on organizational commitment (Cho et al., 2013). Several studies have reported negative implications of emotional labor on employees' organizational commitment (Yang & Chang, 2008;Grandey & Sayre, 2019). This scholarly disagreement is primarily due to the contexts and industries, where these studies have been conducted. Therefore, there is a need to empirically re-examine this relationship. In this study, we expect that the relationship between organizational commitment and emotional labor is significant and positive. This expectation is grounded in the fact that like Cho et al. (2013) we examine the service and retail sector in a developing country's context (Cho et al., 2013 studied this relationship in South Korean context). We expect to find the same results as those of Cho et al. (2013). Therefore, we hypothesize the following: H1: Emotional labor positively affects employees' organizational commitment.

Organizational commitment and emotional exhaustion
Research on relationship between emotional exhaustion and organizational commitment reveals mixed results. For instance, a study by Cho et al. (2013) suggests that organizational commitment has a positive relationship with emotional exhaustion. While another study offers negative relationship (Lapointe, Vandenberghe & Panaccio, 2011). In this work, we empirically revisit this relationship. As emotional exhaustion increases burnout (Thompson et al., 2019), therefore, on the same lines we expect that it may lead to decreased organizational commitment of employees. Therefore, we hypothesize that:

Organizational commitment and workplace bullying
The existing literature proposes that organizational commitment has a negative relationship with workplace bullying (Nielsen & Einarsen, 2012;Sheehan et al., 2018). A meta-analytic review contends that workplace bullying increases employees' turnover intentions and decreases organizational commitment and job satisfaction (Nielsen & Einarsen, 2012). Similarly, anti-bullying initiatives taken by organizations may buffer the negative relationship between organizational commitment and workplace bullying (Cooper-Thomas et al., 2013). To investigate this relationship, we therefore, hypothesize that: H3: Workplace bullying negatively impacts employees' organizational commitment.

Moderating role of organizational justice
Recent research proposes that relationship between organizational justice and organizational commitment is positive (Dehkordi, Mohammadi & Yektayar, 2013). Furthermore, emotional exhaustion is in a negative relationship with organizational justice (Cole et al., 2010) and workplace bullying may negatively affect the employees' perception of organizational justice (Adoric & Kvartuc, 2007;Rai & Agarwal, 2018). A conceptual study by Bhatti et al. (2015) proposes that organizational justice moderates the effect for workplace deviance which is a concept close to workplace bullying. Although, direct relationships of organizational justice to organizational commitment and workplace bullying have been studied in literature, we suppose that the relationships between our proposed variables are moderated by organizational justice. Therefore, we expect the following: H4: Organizational justice positively moderates emotional labor-to-organizational commitment relationship.
H5: Organizational justice positively moderates the relationship between emotional exhaustion and organizational commitment such that the negative relationship between two is weakened when organizational justice is high. H6: Organizational justice positively moderates the relationship between workplace bullying and organizational commitment such that negative relationship between two is weakened when organizational justice is high.
In nutshell, this study examines not only direct relationships of variables to organizational commitment, but also investigates the moderating role of organizational justice on these relationships (Figure 1).

Research site and sample
Pakistan being the 6 th largest labor force in the world and 2 nd largest Islamic country by population has been chosen as a research site for this study to examine the hypothesized relationship in a developing country context. Through simple random sampling technique, 700 questionnaires have been distributed among the sales representatives and service employees of telecom sector of Pakistan. Out of 700, 335 completely filled questionnaires from volunteers have been returned showing 48% response rate. A sample containing 335 observations to test overall model-fit for three predictors and one moderating variable meets the two criteria for the acceptable minimum sample size (Green, 1991). Table 1 gives details of characteristics of sample.

Instruments of the study
To measure emotional labor (EL), 7-item scale developed by Kruml and Geddes (2000) has been used. Emotional exhaustion (EL) has been measured using 7-item scale of Maslach and Jackson (1981). Workplace bullying (WB) has been measured by using 16-item scale by Einarsen et al. (2009). Organizational commitment (OC) instrument has been adopted from 8-item scale of Meyer and Allen (1997). Organizational justice (OJ) is measured by using 12-item scale by Colquitt (2001). The reliability of each scale has been measured through Cronbach's Alpha (Please see Table 2).

Results and discussion
Data have been analyzed using SPSS 20 and PROCESS by Hayes (2013). For regression based moderation analyses, data have also been evaluated for the assumptions of normality, linearity, homoscedasticity and multicollinearity etc. Preliminary analyses confirm that data meet assumptions of regression. As data is interval, therefore Cook's Distance values for all variables are below threshold of 1, indicating that there are no outliers that have unduly influenced the model. The P-P plot and scatter plots confirm normality of data and the variance of residuals is constant and meets the assumption of homoscedasticity.
In Table 4, collinearity statistics show that the independent variables are not highly correlated with one another. As per criteria, the VIF scores are found to be below 10, while tolerance scores are above 0.2. Similarly, the correlation values between independent variables (Table 3) also show that the correlations between predictor variables EE, EL and WB are less than 0.80, indicating no issue of multicollinearity (Kline, 2011).  Table 3 shows Pearson's correlation among variables. The Pearson correlation coefficient at (r=0.200, p<0.01) shows a significant positive correlation between EL and OC. While, we find significant negative correlations between EE and OC (r= -0.463, p<0.01) and p<0.01). Table 4 gives the model summary with R 2 =0.249 showing that the predictor variables account for 25% variation in the dependent variable. This shows the presence of various other variables affecting OC. The adjusted R 2 at 0.242 gives a very close value to R 2 showing small difference for the final model (i.e. 0.007 or 0.7%). In Table 4, ANOVA shows that the model significantly predicts the OC for the conditions (F(3,331)=36.54, p<0.001). The estimates of b-values support the hypotheses H1, H2 and H3.  (Ghalandari et al., 2012;Cho et al., 2013). Similarly, the findings are in line with the previous research showing negative correlation between EE and OC (Cho et al., 2013;Thompson et al., 2019) and WB and OC (Nielsen & Einarsen, 2012;Sheehan et al., 2018). Unlike the research conducted in western context, this study is of prototype significance as it explicates empirically how WB, EE and EL at workplace can affect employees' OC in a developing country context like Pakistan.

Moderation analysis
Using the PROCESS by Hayes (2013), the moderating effect of organizational justice has been calculated. We have selected mean centre for products, calculation of hetroscedasticity-consistent SEs and generate data for plotting (Model 1, 2, 3) in options. We have also included in our analysis Johnson-Neyman conditioning technique to find the percent of cases in the data set. It shows at what range of the moderator, the effect of predictor is significantly positive, non-significant or significantly negative.
Though EL (β= 0.232, p<0.000) and OJ (β= 0.148, p<0.05) both are significant predictors of OC, but their interaction term shows no moderation (Please See Table 5). Similarly, the graph in Figure 2 shows that conditional effect of EL on OC at different levels of moderator (OJ) remains unchanged, signifying no moderation effect. Moreover, Johnson-Nyman test shows that in the data, 11.34% of the cases have a value of moderator less than -0.837, whereas, 88.65% of the cases have a value of the moderator greater than 0.8370. These findings do not support H4. Extant literature shows a positive correlation between OJ and OC (e.g., Dehkordi et al., 2013) and between EL and OC (Ghalandari et al., 2012;Cho et al., 2013). Although we had proposed that OJ would positively moderate the relationship between EL and OC, but the empirical findings do not support this hypothesis.  Figure 3 also confirms the moderating effect, thus supporting H5. This is in line with the present research showing that EE is found to be negatively correlated not only to OJ (Cole et al., 2010), but also to OC (Cho et al., 2013;Thompson et al., 2019). Therefore, OJ will positively moderate the relationship between EE and OC, such that the negative relationship between them will be dampened if OJ will be high. Table 7 suggests no change in R 2 (0.0000), thus no moderation effect of OJ on WB and OC is present. The graph in Figure 4 shows that conditional effect of X (WB) on Y (OC) at different levels of moderator (OJ) remains unchanged, signifying no moderation effect.
Johnson-Neyman technique indicates no statistically significant transition points within the observed range of the moderator. This leads to a clear rejection of H6. Though extant research confirms a negative correlations between WB and OJ (Adoric & Kvartuc, 2007;Rai & Agarwal, 2018) and WB and OC (Nielsen & Einarsen, 2012;Sheehan et al., 2018), but our results do not support H6. Employees' high OJ does not weaken the negative correlation between WB and OC. Overall, according to the results obtained from data analysis, four (H1, H2, H3 and H5) out of six hypotheses have been supported.
In so far as implications of our study are concerned, we suggest that keeping employees committed is essentially necessary for organizations operating in information and telecom sectors. US organizations related to these sectors lost $ 650 billion in 2018 alone when less committed employees exposed 2.8 billion records of consumer data (Goren, 2019). Recent data illustrate that lack of OC leads to high turnover intention as of every 1000 workers, 31% reported that they have quit job within first six months of joining and managers state that  Statistics, 2020). Latest statistics also reveal that over 45% of managers consider lack of commitment and burnout as main reasons for turnover intention of their workforce. Contrary to this, those employees that are highly committed to the organization are 75% less likely to show turnover intention compared to less committed employees of the organization (U.S. Bureau of Labor Statistics, 2020). Findings of our work suggest that human resource function of telecom and other capital intensive organizations in developing countries like Pakistan have to adopt a finer and integrated view of work-force stability that is synonymous to the organizational commitment of employees. Organizations that we studied employ programs like exit interviews to find why people leave. This way these companies intend to curb the reasons of turnover. We suggest that such programs are based on carelessly conceived methods, and organizations need to go back on the turnover path and employ chronological analysis to identify anchor points that trigger loss of commitment in employees. Bullying-in the form of berating, threatening, giving snide comments and unfair criticism-and emotional exhaustion-in the form of mood swings, lack of motivation and imagination, social withdrawal, missed deadlines, increased rate of absence from work etc.-do not go unnoticed by managers. Therefore, a program that aims to identify anchor points (humans, systems, processes) causing loss of organizational commitment must be designed, executed and controlled, keeping in view the fact that prevention is much more effective than exit interviews. Such a program should be based on strong core of organizational justice, because systems that ensure perception of fairness among employees give rise to more commitment towards organizations.
Secondly, our work emphasizes that human resource teams of organizations have to take up and strengthen their role of gatekeeping to identify and bar entry of bullies in the organization. This role involves rightly assessing the negative tendencies in potential candidates before hiring to avoid future mishaps. Such a role should be fully supported by organizational recruitment policies. These policies should provide clear guidance to avoid recruitment of individuals with frequent job hopping, individuals who carry no career progression at one workplace, who show bragging and boastfulness, and show no potential to work in teams etc.
On the whole our work suggests that organizational commitment of employees can be stimulated by promoting positive work environment and discouraging all negative workplace behaviors. Importance of employees in any organizations is like a backbone. Their dedication helps organizations grow. This study can help managers by providing them basis to strengthen employees' commitment and dedication to their organizations through managing their emotions and behaviors effectively.

Conclusion
Building on the work of Cho et al. (2013), our work examines the effects of workplace bullying, emotional exhaustion and emotional labor on organizational commitment taking organizational justice as a moderator in a developing country context of Pakistan. Analysis of data collected from telecommunication sector of Pakistan suggests that emotional exhaustion and workplace bullying reduce organizational commitment of employees, while employees' emotional labor is positively correlated to the organizational commitment. Moreover, organizational justice shows moderating effect on emotional exhaustionorganizational commitment relationship only. Our study recommends that organization should employ zero tolerance towards workplace bullying and put in place policies that thwart such anchor points leading to emotional exhaustion in employees. Organizational justice is another element that is of strategic significance and managers should inculcate and improve systems to ensure perception of fairness among employees to ensure their commitment towards organizations. We also suggest that to avoid negative workplace behaviors, human resource function of these organizations should put in place system to check candidates for potential negative tendencies before hiring to avoid future mishaps. We propose that future researchers should replicate this study in other industries using combination of other variables like gender, tenure/experience, job designation and employees' commitment to their co-workers etc.