The main objective of this research is to study the effect of personality, emotional intelligence (EI), affectivity, emotional labor and emotional exhaustion on counterproductive work behavior (CWB) of frontline employees in the government sector. A questionnaire was designed and distributed to 625 frontline employees working at service counters in 25 ministries in Malaysia. We received responses from 519 employees (response rate=83%). The data was analyzed using Structural Equation Modeling (SEM). The main findings are: (1) personality factors of employees drive their EI, affectivity, emotional labor, emotional exhaustion, and CWB and (2) EI and affectivity impact emotional labor, emotional exhaustion and CWB. Through the integrated model, we have studied the indirect roles of emotional labor and emotional exhaustion. This is one of the few studies that have effectively integrated the five constructs into a single framework to study their effects on CWB.
El objetivo principal de este trabajo es estudiar el efecto de la personalidad, la inteligencia emocional (IE), la afectividad, el trabajo emocional y el agotamiento emocional en el comportamiento laboral contraproducente de los empleados de primera línea del sector público. Se diseñó un cuestionario, que se distribuyó a 625 funcionarios de primera línea destinados en ventanillas de 25 ministerios de Malasia. Recibimos respuesta de 519 funcionarios (índice de respuesta del 83%). Los datos se analizaron mediante el modelado de ecuaciones estructurales (SEM). Se obtuvieron los siguientes resultados: (1) los factores de personalidad gobiernan su inteligencia emocional, afectividad, trabajo emocional, agotamiento emocional y el comportamiento laboral contraproducente (CLC) y (2) la inteligencia emocional y la afectividad influyen en el trabajo emocional, el agotamiento emocional y el CLC. Mediante el modelo integrado hemos estudiado el papel indirecto del trabajo emocional y del agotamiento emocional. Se trata de uno de los pocos estudios que han integrado eficazmente los cinco constructos en una estructura única para estudiar sus efectos sobre el CLC.
Counterproductive work behavior (CWB) is quite common among employees in many organizations, but much of it apparently goes unnoticed, unreported, or both (Bennett & Robinson, 2000). CWB can be intentional or unintentional and can result from a wide range of underlying causes and motivations. CWB is the employees’ behavior that goes against the goals of an organization. All acts of CWB violate the legitimate interests of an organization by harming the members of the organization and/or organization as a whole (Marcus & Schuler, 2004). It involves a wide spectrum of behaviors that harm employees, customers and/or the organization. These behaviors range from severe, systematic, and abusive to milder and ambiguous episodes of workplace incivility (Fox & Spector, 2005). According to Porath, MacInnes, and Folkes (2011, p. 12), “witnessing incivility among employees is not normal, it is not rare either”. Examples of CWB are: intentionally working slow, taking long breaks, sabotage of equipment, theft of property, showing favoritism, gossiping, sexual harassment, blaming others, verbal abuse, physical abuse, receiving bribe, and being corrupt. The employees studied in this research are the frontline staff at the service counters at each of the ministries in Malaysia.
Researchers have shown that CWB of employees results in enormous economic and social costs for the organizations that can possibly run into billions of dollars (Bennett & Robinson, 2000; Galperin & Burke, 2006). According to Ogbonna and Harris (2002), “the attitudes and behaviors of frontline, customer-contact service providers are a significant factor in customer's perceptions and interpretations of service encounters” (p. 163). The extant research on frontline employees seems to assume that these employees are compliant, obedient, and constructive but evidence shows otherwise (Harris & Ogbonna, 2006). A recent study has argued that “the interaction between frontline employees and customers creates an impression of what is to come in the service experience” (Dagger, Danaher, Sweeney, & McColl-Kennedy 2013, p. 488). The evidence suggests that frontline employees deliberately behave in counterproductive ways (Bennett & Robinson, 2000). According to Harris and Ogbonna (2006), industrial sociology has explored CWB extensively and the studies have been conducted mainly in the non-service sectors. The fundamental difference between the effects of CWB in service and non-service sectors is the immediacy. In the service sector, the negative impact of CWB is immediate and the actions are likely to affect the customers’ evaluations of the organization. The negative behavior of frontline government employees can make citizens form negative opinions about the government in power and these may have severe implications. In spite of the pivotal role played by the frontline employees in the service sector, the behavior of these employees are least understood and studied (Harris & Ogbonna, 2006). There is a dearth of studies addressing CWB in the service sector and especially among the frontline employees of government agencies and ministries. In this research, we address the roles of personality factors, EI, affectivity, emotional labor, and emotional exhaustion on CWB of frontline government employees. In this research, (1) CWB has two dimensions: interpersonal (CWB-I) and organizational (CWB-O).
The contributions of this study are threefold. First, this paper has integrated the five constructs (emotional intelligence, personality factors, affectivity, emotional labor, and emotional exhaustion) into a single framework to study their effects on CWB. Earlier studies have looked at some of these constructs together. We have also looked at the inter-relationships between the five constructs. We have not only assessed the relationships at construct level but also at the dimension level of two constructs, affectivity and emotional labor. Second, we have studied the impact of demographic variables (gender, age, education, and duration of service) on all the constructs; there is a dearth of research in this area. Third, we have conducted this study in a fast developing country in South-East Asia, Malaysia. Malaysia has a population of 27 million. According to statistics provided by the Public Complaints Bureau (PCB), Prime Minister's Department, Malaysia, the total complaints received from public by various sources for the year 2011 is 13,356 (Public Complaints Bureau Annual Report, 2011). Most of the complaints (about 75%) submitted to PCB are related to people's dissatisfaction with work behaviors of government personnel, especially frontline employees. Therefore, this research is timely and we believe that this situation must be prevalent in many developing countries.
Theoretical Framework and Hypotheses DevelopmentMany researchers have defined CWB and all the definitions agree that CWBs are “characterized by a disregard for societal and organizational rules and values” (Martinko, Gundlach, & Douglas 2002, p. 37). Martinko et al. (2002) in their effort to develop an integrative theory of CWB have reviewed main theoretical perspectives of CWB, emphasizing their common elements such as individual factors, situational factors, and cognitive information-processing elements (causal attributions and perceptions of disequilibria). These perspectives indicate that CWB “is the result of a complex interaction between the person and the environment in which the individual's reasoning about the environment and expected outcomes drive the individual's behavior” (p. 41). They have argued that attribution theory provides the best explanation as to why some individuals engage in CWB when exposed to certain (negative) stimuli. In our research, personality factors, emotional intelligence, and affectivity fall under the ‘individual differences’ of the integrated framework developed by Martinko et al. (2002) and emotional labor and emotional exhaustion fall under ‘situational variables’ of the framework. These factors lead to internal and/or external attributions and these in turn lead to CWB of frontline government employees. The internal attribution results in a self-destructive form of CWB such as drug abuse, alcohol use, absenteeism, passivity, depression, dissatisfaction, and lower performance. The external attribution results in a retaliatory form of CWB such as aggression, violence, abuse, sabotage, terrorism, fraud, harassment, and being corrupt. The framework of our research is given in Figure 1.
Theoretical Framework of this Research (hypothesized relationships).
Note. PY=Personality traits, EI=Emotional intelligence, APA=Positive affectivity, ANA=Negative affectivity, ELSA=Emotional labor (surface acting), ELDA=Emotional labor (deep acting), EE=Emotional exhaustion, CWB=Counterproductive work behavior.
Direct relationship between personality factors and CWB (H1). Many researchers have shown the relationships between personality factors and CWB (Cullen & Sackett, 2003; Dalal, 2005; Ones, Viswesvaran & Schmidt 2003; Salgado, 2002). These researchers have used Big Five personality dimensions to reflect personality factors and these dimensions are: conscientiousness, extraversion, neuroticism (opposite of emotional stability), agreeableness, and openness to experience. Specifically, these researchers have shown that conscientiousness, emotional stability, and agreeableness are the strongest predictors of both the dimensions of CWB (CWB-I and CWB-O). Many researchers have argued that an individual's attributions about the causes of equitable or inequitable outcomes are a primary force behind the individual's CWB (Douglas & Martinko, 2001; Martinko et al., 2002). The employees that have a high level of conscientiousness, emotional stability, and agreeableness are less likely to involve in activities that may harm other employees and the organization because of their attributions about the causes of outcomes. Based on the theory of attributions, differences in the personality factors (individual differences) between the individuals (frontline government employees) will determine if the individuals will exhibit CWB or not. We posit the following hypothesis linking personality factors and CWB:H1 There is a negative relationship between personality factors (conscientiousness, emotional stability, agreeableness, extraversion, and openness to experience) and CWB of frontline government employees.
Indirect relationships between personality factors and CWB (H1a, H1b, H1c). At the individual level, Salovey and Mayer (1990) define EI as “the subset of social intelligence that involves the ability to monitor one's own and others’ feelings and emotions, to discriminate among them and to use this information to guide one's thinking and actions.” (p. 189). Many studies have established the relationship between personality traits and EI (Brackett & Mayer, 2003; Chamorro-Premuzic, Bennett, & Furnham 2007; Myleen, Michael, & William 2009; Petrides et al., 2010). For example, Chamorro-Premuzic et al. (2007) have shown that extraversion, agreeableness, conscientiousness, and emotional stability are positively linked to EI. Through hypothesis H2, we have argued a direct relationship between EI and CWB. Based on the earlier studies and hypothesis H2, we posit the following hypothesis linking personality factors and EI:
H1a: There is a positive relationship between personality factors (conscientiousness, emotional stability, agreeableness, extraversion, and openness to experience) and EI of frontline government employees and therefore, an indirect effect of personality factors on CWB through EI.
Affectivity is the dispositional tendency to experience a variety of positive and negative mood states (Penney & Spector, 2005; Watson & Clark, 1984). The significance of affectivity is the emotional states that are seen to lie at the center of attitude formation as well as employee behaviors in organizations (Ashkanasy, Hartel, & Daus 2002). The link between personality traits and affectivity is old and it has been shown that extraversion is linked to positive affectivity and neuroticism is linked to negative affectivity (Bruck & Allen, 2003; Hoergera & Quirk, 2010). Yik and Russell (2001) argue that the other three traits, namely, conscientiousness, agreeableness, and openness, have a small but significant effect on affectivity (positive and negative). Through hypotheses H3 and H4, we have argued the direct effect of positive and negative affectivities on CWB. Based on the above arguments and hypotheses H3 and H4, we posit the following hypotheses:H1b There is a positive relationship between personality factors (conscientiousness, emotional stability, agreeableness, extraversion, and openness to experience) and positive affectivity of frontline government employees and therefore, an indirect effect of personality factors on CWB through positive affectivity. There is a negative relationship between personality factors (conscientiousness, emotional stability, agreeableness, extraversion, and openness to experience) and negative affectivity of frontline government employees and therefore, an indirect effect of personality factors on CWB through negative affectivity.
Direct relationship between EI and CWB (H2). The link between EI and CWB has been established by many researchers. There are a few meta-analytic studies that link EI to performance outcomes, general mental ability, personality factors, health, and effective leadership (Harms & Credé, 2010; Joseph & Newman, 2010; Schutte, Malouff, Thorsteinsson, Bhullar, & Rooke 2007; Van Rooy & Viswesvaran, 2004). EI can be viewed as a personality trait or as ability (Carmeli, 2003). In this research, we view EI as a competency that is expected to increase positive attitudes towards work and drive positive behavior and improved outcomes. Van Rooy and Viswesvaran (2004) have observed that EI correlates significantly with general mental ability and personality traits. These observations can help explain CWB. Messmer-Magnus, Viswesvaran, Deshpande, and Jospeh (2010) have suggested that emotionally intelligent employees are less likely to indulge in unethical activities that include CWB. According to Mayer, Caruso, and Salovey (1999), EI plays a significant role in reducing the deviant behaviors (CWBs) related to organizational tasks. Deshpande, Joseph, & Shu (2005) have shown that employees with high levels of EI do not exhibit CWB. According to Petrides, Frederickson, and Furnham (2004), employees with high levels of EI engage less in deviant behaviors than those with low EI. A recent study by Jung and Yoon (2012) corroborate the findings of previous researchers. The conclusions derived from earlier studies support the theory of attributions that states that individuals with high levels of EI (individual differences) are less likely to involve in activities linked to CWB. Based on the above arguments, we posit the following hypothesis:H2 There is a negative relationship between EI of frontline government employees and their CWB.
Indirect relationships between EI and CWB (H2a, H2b, H2c). Since the characteristics of EI such as perception and regulation of emotion may modify an employee's emotional labor behaviors, EI is considered to be an important factor, influencing emotional labor (Lee & Ok, 2012; Mikolajczak, Menil, & Luminet 2007; Psilopanagioti, Anagnostopoulos, Mourtou, & Niakas 2012). The findings of studies linking EI and emotional labor are mixed. For example, Austin, Dore, and O’Donovan (2008) and Mikolajczak et al. (2007) have shown that the relationship between EI and emotional labor (surface acting dimension) is negative; Brotheridge (2006) has shown that there is no relationship between these two constructs; Wong and Law (2002) have argued that employees high on the appraisal of self-emotions and other's emotions recognize the need to engage in emotional labor to satisfy and contribute to positive service experience of the customer. Cote (2005) provides some initial findings that support the relationship between emotional intelligence and deep acting dimension of emotional labor. He reports that individuals with high EI are more likely to engage in deep acting during interpersonal interactions. The emotional labor strategies also are posited to have differential effects on individual well-being and performance. EI is thought to be a dynamic characteristic that enables an individual to appropriately match the strategy to the situation (Barrett & Gross, 2001). Through hypothesis H5, we have highlighted the relationship between emotional labor (surface acting) and CWB. Based on the arguments, we posit the following hypotheses:H2a There is a positive relationship between EI of frontline government employees and their emotional labor (deep acting). There is a negative relationship between EI of frontline government employees and their emotional labor (surface acting) and therefore, an indirect effect of EI on CWB through emotional labor (surface acting).
Based on the theoretical underpinnings and empirical support for the relationship between emotional intelligence and burnout (Brackett, Palomera, Moisa, Reyes, & Salovey, 2010; Chan, 2006), this study expects to find significant relationships between EI and emotional exhaustion, a component of burnout. For example, feelings of emotional depletion may cause emotional exhaustion, negative feelings toward an individual's role and performance as service provider. Therefore, the presence or absence of emotional exhaustion is contingent on the level of EI. Through hypothesis H6, we have shown that there is a direct relationship between emotional exhaustion and CWB of employees. Based on the above arguments, we posit as follows:H2c There is a negative relationship between EI of frontline government employees and the emotional exhaustion experienced by them and therefore, an indirect effect of EI on CWB through emotional exhaustion.
Indirect relationships between EI and emotional exhaustion (H2d, H2e). Literature on emotional labor indicates that surface acting is more likely to have negative effects, such as depersonalization and emotional exhaustion, while deep acting seems to be associated with positive outcomes, such as personal accomplishment and affective well-being (Goldberg & Grandey, 2007; Grandey, 2003; Montgomery, Panagopolou, de Wildt, & Meenks 2006). Johnson and Spector's (2007) findings also support this notion that surface acting is positively related to emotional exhaustion, while deep acting has a negative relationship with emotional exhaustion. Recent findings by Lee (2010) show that surface acting is positively and deep acting is negatively related to emotional exhaustion and depersonalization. Through hypotheses H2a and H2b, we have shown the relationship between EI and emotional labor (surface acting), and emotional labor (deep acting). Based on the above arguments, we posit the following hypotheses:H2d There is a negative relationship between emotional exhaustion of frontline government employees and their emotional labor (deep acting) and therefore, an indirect effect of EI on emotional exhaustion through emotional labor (deep acting). There is a positive relationship between emotional exhaustion of frontline government employees and their emotional labor (surface acting) and therefore, an indirect effect of EI on emotional exhaustion through emotional labor (surface acting).
Direct relationship between affectivity and CWB (H3, H4). Negative affectivity refers to the “dispositional tendency to experience a variety of negative mood states” (Penney & Spector, 2005, p. 781). The employees who are high on negative affectivity experience high levels of distressing emotions such as anger, frustration, fear, hostility, and anxiety when confronted with stressful conditions. This is because these employees tend to be less satisfied with themselves and their environments, interpret the conditions as disturbing and threatening, perceive disequilibria and may view negative outcomes as permanent and stable (Douglas & Martinko, 2001; Fox & Spector, 2005; Martinko et al., 2002). This attribution leads the employees high on negative affectivity to indulge in CWB activities. The individuals high on positive affectivity have a more positive outlook of the world and engage more in organizational citizenship behaviors instead of CWBs (Miles, Borman, Spector, & Fox 2002). Dalal (2005) has found that positive affectivity has a negative relationship with CWB and negative affectivity has a positive relationship with CWB. In general, the results on the role of affectivity are inconclusive. For example, a study by Fox, Spector, and Miles (2001) has analyzed the role of affectivity as a moderating variable between stressor-CWB relationships and has shown the effect to be weak. Penney and Spector (2005) and Salami (2010) have found evidence on the role of negative affectivity as a moderator between job stressors and CWB. However, in our research based on the theory of attributions we study the direct effect of affectivity on CWB. Based on the above arguments, we posit the following hypothesis:H3 Negative affectivity of frontline employees has a positive relationship with their CWB. Positive affectivity of frontline employees has a negative relationship with their CWB.
Indirect and direct relationships between affectivity and emotional exhaustion (H3a, H3b, H4a, H4b).Morris and Feldman (1996) contend that an individual's predisposition to experience positive or negative affect will influence emotional dissonance. Brotheridge and Lee (2003) have posited that affectivity corresponds to both the range and intensity of emotions displayed, and the use of surface or deep acting emotional labor. Individuals with high levels of affectivity may have greater trouble, concealing their feelings with surface acting and realigning their feelings through deep acting, than individuals with low affectivity (Brotheridge & Lee, 2003). Research has consistently found a positive relationship between negative affectivity and surface acting. On the other hand positive affectivity and deep acting are positively associated (Brotheridge & Grandey, 2002; Brotheridge & Lee, 2003; Gosserand & Diefendorff, 2005; Johnson & Spector, 2007; Kammeyer-Mueller et al., 2013).
Employees who are high on negative affectivity are easily distressed, agitated, upset, dissatisfied, and emotionally exhausted (Chui & Francesco, 1993). Abraham (1998) has argued the relationship between negative affectivity and emotional exhaustion. According to Szczygiel, Buczny and Bazinska (2012), negative affectivity is positively associated with emotional exhaustion and positive affectivity decreases emotional exhaustion among the service sector employees. Lockyer (2013) has shown the relationship between affectivity and emotional exhaustion. Based on the above arguments, we posit the following hypotheses:H3a There is a positive relationship between negative affectivity and emotional exhaustion of frontline employees. There is a positive relationship between negative affectivity of front line government employees and emotional labor (surface acting) and therefore, an indirect effect of negative affectivity on emotional exhaustion through emotional labor (surface acting). There is a negative relationship between positive affectivity and emotional exhaustion of frontline employees. There is a positive relationship between positive affectivity of front line government employees and emotional labor (deep acting) and therefore, an indirect effect of positive affectivity on emotional exhaustion through emotional labor (deep acting).
Direct relationship between emotional labor and CWB (H5): Emotional labor is defined as “managing emotions so they are consistent with organizational display rules, regardless of whether they are discrepant with the individuals’ internal feelings” (Grandey, 2000, p. 96) and this emotional labor facilitates task effectiveness and self-expression. Emotional labor is “the display of expected emotions by service agents during service encounters” (Ashforth & Humphrey, 1993, p. 88). Based on the services management literature, emotional labor has relevance to service encounters for four reasons: (1) frontline service personnel represent the organization to customers, (2) most of the service transactions involve face-to-face interactions between the personnel and customers, (3) customer participation in the service encounters makes them uncertain and dynamic, and (4) services rendered are intangible. Emotional labor can have a positive effect on task performance by “regulating interaction and excluding interpersonal problems”, or it can have a negative effect on performance by “priming expectations of good service that cannot be met” (Ashforth & Humphrey, 1993, p. 107). A review of literature shows that emotional labor has been conceptualized as an internal emotional state (Hochschild, 1983; Morris & Feldman, 1996), as a display of emotions (Ashforth & Humphrey, 1993), as situational factors related to the job (Morrison & Feldman 1996), and behaviors, such as surface acting and deep acting (Grandey, 2000; Hochschild, 1983).
Bechtoldt, Welk, Zapf, & Hartig (2007) have proposed that CWB provides the possibility to escape the state of emotional dissonance that is central to surface acting. While deep acting is often regarded as a preferred emotional labor tactic over surface acting, it is not always successful in preventing customer services (frontline) employees from feeling negative emotions, and often requires significant emotional efforts. We posit the following hypothesis:H5 There is a positive relationship between emotional labor (surface acting) and CWB of frontline government employees.
Direct relationship between emotional exhaustion and CWB (H6). Emotional exhaustion refers to feelings of being drained by tasks and duties at workplaces. Emotionally exhausted employees because of emotional and mental exhaustion may expend less effort at work and may be unwilling to help others (Mulki, Jaramillo, & Locander 2006). Banks, Whelpley, Oh, and Shin (2012) explain the mechanism by which emotional exhaustion influence CWB (individuals and organization). They use the stressor-emotion model and claim that emotionally exhausted employees have lower levels of commitment, which increases the probability of exhibiting CWB. Bolton, Harvey, Grawitch, and Berber (2012) argue that emotional exhaustion leads to depersonalization and organizational dis-identification and these results in CWB of employees. According to Ito and Brotheridge (2003), emotional exhaustion can be a major factor in effective coping of stress at the workplace. When the coping strategy becomes counterproductive, it results in CWB (Spector, 1998). Therefore, CWB results from an emotional response with the intention to attack the situation and/or to “passively and indirectly cope with the situation” (Spector & Fox, 2003, p. 274). Based on the above arguments, we posit the following hypothesis:H6 There is a positive relationship between emotional exhaustion and CWB of frontline government employees.
The population for this study consists of approximately 5,200 frontline staff employed at twenty five (25) ministries located in the Federal Territory of Kuala Lumpur and Putrajaya, Malaysia. The ministries in Kuala Lumpur and Putrajaya are selected because all the 25 ministries under Malaysian government have their main ministries’ office at either Kuala Lumpur or Putrajaya. The frontline staffs are chosen because they are the government's main representatives who interact directly with customers or the public. Therefore, they have heavy responsibilities and play very important roles in public perception of the public sector or government, in general. Since one of the researchers of this study is a government employee working in one of the ministries, access to sample elements (frontline staff) has not been an issue. The questionnaire was distributed to 625 frontline staff working at the service counters in 25 ministries and the respondents were selected randomly. These counter staff act as a bridge between the citizens and the ministries. Out of 625 questionnaires distributed, 519 (response rate=83%) were returned and 512 were deemed useful for further analysis and these had no missing values. The original English version of the questionnaire was translated into Bahasa Malaysia (local language) using the back translation method with the help of a panel of language experts from the language faculty at a local public university.
MeasuresEI. The items under this construct were adopted from the scale developed by Wong and Law (2002). This construct had four dimensions: self-emotion appraisal, other's emotion appraisal, use of emotion and regulation of emotion. The scale consisted of 16 items in a five-point Likert format where one corresponded to “strongly disagree” and five corresponded to “strongly agree”. High average scores correspond to high levels of EI.
Personality traits. The items under this construct were adopted from the scale developed by Rammstedt and John (2007). This construct had five dimensions: extraversion, agreeableness, conscientiousness, neuroticism, and openness. The scale consisted of 10 items in a five-point Likert format where 1 corresponded to strongly disagree and 5 corresponded to strongly agree.
Affectivity. The items under this construct were adopted from the scale developed by Watson, Clark, and Tellegen (1988). This construct had two dimensions: positive and negative affectivity. The scale consisted of 20 items in a five-point Likert format where 1 corresponded to not at all and 5 corresponded to extremely. High average scores correspond to high levels of affectivity.
Emotional Labor. The items under this construct were adopted from the scale developed by Grandey (2003). This construct had two dimensions: surface acting and deep acting. The scale consisted of eight items (five items under surface acting and three items under deep acting). High average scores correspond to high levels of emotional labor.
Emotional Exhaustion. The items under this construct were adopted from the scale developed by Maslach and Jackson (1986). The scale consisted of nine items. The scale employed a seven-point Likert scale that ranged from never to everyday. High mean scores on this measure suggest high levels of emotional exhaustion.
CWB. The items under this construct were adopted from the scale developed by Spector et al. (2006). This scale had two dimensions: CWB-O (organizational) and CWB-I (interpersonal). The scale consisted of 45 items. The scale employed a five-point Likert scale that ranged from never to everyday. High average scores correspond to high levels of CWB. The final questionnaire and the data can be obtained from the corresponding author.
Pilot TestingA pilot test was carried out to determine the clarity of the instructions and items in the questionnaire. According to Sekaran and Bougie (2010), a reliable instrument should avoid long complex questions, double barreled questions, culture jargon, leading questions, double meaning, and sensitive questions. The researcher conducted a pilot test with 30 participants from one selected ministry. No changes were made to the questionnaire after the pilot test. The questionnaire contained questions related to frontline staffs’ EI, personality, affectivity, emotional labor, emotional exhaustion, and CWB. Before the questionnaires were distributed to the frontline staffs, a formal approval was obtained from the related ministry. Later, the questionnaires were given to the officer in-charge to distribute to the selected frontline staffs that were identified randomly. The officer in-charge was chosen to expedite the process of distribution and collection of questionnaires. The reliability scores of the constructs ranged between .70 and .92.
ResultsReliability and ValidityThe reliability of the constructs was measured using Cronbach alpha. The reliability scores are given in Table 1 and it can be seen that the scores of the constructs are between .70 and .98. The validity of the constructs was tested based on the results from Confirmatory Factor Analysis (CFA), which was performed using Lisrel 9.1 student version. We used the following criteria for the convergent and divergent validity tests (Hair, Black, Babin, & Anderson, 2010; Podsakoff, Mackenzie, Lee, & Podsakoff, 2003): (1) factor loading of items on the latent construct must be greater than .5, (2) composite reliability (CR) of each construct must be greater than .7, (3) average variance extracted (AVE) of each construct must be greater than .5 and (4) AVE of each construct must be greater than the squared correlation of that construct with other constructs. The validity test results, CR, and AVE are given in Table 1. The AVE scores for personality, emotional labor, and affectivity are below the cutoff value of .5. However, these values are above the squared correlation with other constructs.
Descriptive Statistics and Reliability Scores.
| Construct | Items | Score -Male^ | Score-Female^ | Reliability | CR | AVE | |||
|---|---|---|---|---|---|---|---|---|---|
| Mean | SD | Mean | SD | Male | Female | ||||
| EX | 2 | 7.82* | 1.51 | 8.23* | 1.55 | .70# | .72# | .70# | .40# |
| AG | 2 | 7.65* | 1.69 | 8.07* | 1.53 | ||||
| CO | 2 | 7.59* | 1.60 | 8.17* | 1.32 | ||||
| NE | 2 | 4.78 | 1.54 | 4.56 | 1.60 | ||||
| OP | 2 | 6.12 | 1.23 | 6.12 | 1.22 | ||||
| Emotional Intelligence | 16 | 75.90 | 13.33 | 75.49 | 12.31 | .948 | .946 | .87 | .63 |
| Positive Affectivity | 10 | 37.29 | 6.21 | 37.60 | 6.57 | .820 | .869 | .86 | .44 |
| Negative Affectivity | 10 | 25.65 | 6.72 | 25.11 | 6.52 | .829 | .821 | .85 | .45 |
| Emotional Labor (SA) | 5 | 18.77* | 4.47 | 19.72* | 4.04 | .815 | .839 | .85 | .42 |
| Emotional Labor (DA) | 3 | 11.37* | 3.07 | 12.00* | 2.52 | .886 | .814 | .85 | .42 |
| Emotional Exhaustion | 9 | 14.71 | 11.61 | 14.44 | 11.21 | .951 | .948 | .95 | .67 |
| CWB | 45 | 61.41* | 24.74 | 56.10* | 20.75 | .981 | .985 | .96 | .69 |
Note. SD=standard deviation, CR=composite reliability, AVE=average variance extracted, PY=Personality, EX=Extraversion, AG=Agreeableness, CO=Conscientiousness, NE=Neuroticism, OP=Openness, SA=surface acting, DA=deep acting, CWB=counter productive work behavior.
In this research, the responses to the questionnaire items have been obtained from a single source and both the dependent and focal explanatory variables are perceptual measures derived from the same respondent. This may result in common method variance. Common method variance is the “variance that is attributable to the measurement method rather than to the constructs the measures represent” (Podsakoff et al., 2003, p. 879). We performed a one-factor Herman test by loading all the items on to a single factor. Factor analysis indicated that this single factor could explain only 29% of the total variance. Podsakoff et al. (2003) indicate that there is no agreed cut-off point but a value of less than 50% is considered to be a reasonable value to indicate the reduced effect of common method variance. Since in this research only 29% of the total variance is explained by a common factor, we can argue that effect of common method variance is limited.
Descriptive StatisticsAmong the 512 respondents, 70% were females. This is not surprising given the fact that more than 60% of the university graduates in Malaysia are females. According to World Bank report on Malaysia, 68.3% of females are employed in services. About 85% of the respondents are under 35 years of age, 90% of the respondents have less than 10 years of experience and 89% of the respondents have been as frontline employees for less than six years. The mean and standard deviation of all constructs for male and female employees are given in Table 1. Based on the descriptive statistics the following conclusions can be drawn: (1) the mean level of EI can be considered to be ‘high’ (male: mean=4.74, SD=0.83; female: mean=4.72, SD=0.77), (2) the mean extraversion score can be considered to be ‘moderate’ for male and ‘high’ for female (male: mean=3.91, SD=0.75; female: mean=4.15, SD=0.78), (3) the mean agreeableness score can be considered to be ‘moderate’ for male and ‘high’ for female’ (male: mean=3.83, SD=0.85; female: mean=4.04, SD=0.77), (4) the mean conscientiousness score can be considered to be ‘moderate’ for male and ‘high’ for female (male: mean=3.80, SD=0.80; female: mean=4.09, SD=0.66), (5) the mean neuroticism score can be considered to be ‘low’ for both male and female (male: mean=2.39, SD=0.77; female: mean=2.28, SD=0.80), (6) the mean openness score can be considered to be ‘moderate’ for male and female (male: mean=3.06, SD=0.62; female: mean=3.06, SD=0.61), (7) the level of positive affectivity is ‘moderate’ (male: mean=3.73, SD=0.62; female: mean=3.76, SD=0.66) and the level of negative affectivity is ‘low’ (male: mean=2.57, SD=0.67; female: mean=2.51, SD=0.65), (8) the level of emotional labor (surface acting) is ‘moderate’ (male: mean=3.75, SD=0.89; female: mean=3.94, SD=0.81), (9) the level of emotional labor (deep acting) is ‘moderate’ (male: mean=3.79, SD=1.02; female: mean=4.00, SD=0.84), (10) the level of emotional exhaustion can be considered to be ‘low’ (male: mean=1.63, SD=1.29; female: mean=1.60, SD=1.25), (11) the level of CWB can be considered to be ‘low’ (male: mean=1.36, SD=0.55; female: mean=1.25, SD=0.46), and (12) gender differences are significant for personality, emotional labor (surface and deep acting) and CWB; personality trait score and emotional labor are higher for females but CWB is lower.
We have managed to capture 45 different CWB activities indulged by the respondents in the last year. Table 2 gives the exhaustive list of activities and the frequency of indulgence. The frequency indicates that the problem of CWB among front-line government employees may need attention. In order to understand the differences in personality traits and CWB among the respondents in various ministries, we have computed the average levels of five dimensions of personality traits and CWB. Table 3 gives the average levels. The levels indicate that (1) levels of extraversion, agreeableness, and conscientiousness can be considered to be ‘moderate’ to ‘high’, (2) levels of neuroticism and openness can be considered to be ‘low’ to ‘moderate’ and (3) level of CWB can be considered to be ‘low’.
List of CWB Activities Indulged by the Respondents.
| Frequency (% age) | ||||
|---|---|---|---|---|
| CWB activities indulged by the respondents | Once or twice a year | Once or twice per month | Once or twice per week | |
| 1 | Purposely wasted your employer's materials supplies | 24.2 | 9.0 | 2.0 |
| 2 | Daydreamed rather than did your work | 26.0 | 7.4 | 3.5 |
| 3 | Complained about insignificant things at work | 21.3 | 6.4 | 5.3 |
| 4 | Told people outside the job what a lousy place you work for | 19.5 | 5.7 | 2.8 |
| 5 | Purposely did your work incorrectly | 15.4 | 4.5 | 3.7 |
| 6 | Came to work late without permission | 16.4 | 6.4 | 3.1 |
| 7 | Stayed home from work and said you were sick when you weren’t | 17.6 | 3.3 | 2.8 |
| 8 | Purposely damaged a piece of equipment or property | 13.5 | 3.1 | 2.4 |
| 9 | Purposely dirtied or littered your place of work | 12.5 | 3.1 | 2.4 |
| 10 | Stolen something belonging to your employer | 13.1 | 3.3 | 1.4 |
| 11 | Started or continued a damaging or harmful rumor at work | 15.6 | 3.3 | 1.4 |
| 12 | Been nasty or rude to a client or customer | 16.0 | 2.9 | 1.8 |
| 13 | Purposely worked slowly when things needed to get done | 15.6 | 2.1 | 1.8 |
| 14 | Refused to take on an assignment when asked | 21.7 | 2.7 | 2.4 |
| 15 | Purposely came late to an appointment or meeting | 22.1 | 5.9 | 2.6 |
| 16 | Failed to report a problem so it would get worse | 16.0 | 3.9 | 2.3 |
| 17 | Taken a longer break than you were allowed to take | 15.6 | 2.3 | 2.3 |
| 18 | Purposely failed to follow instructions | 16.0 | 5.3 | 1.2 |
| 19 | Left work earlier than you were allowed to | 17.0 | 6.8 | 4.5 |
| 20 | Insulted someone about their job performance | 15.6 | 3.1 | 2.4 |
| 21 | Made fun of someone's personal life | 9.8 | 2.3 | 1.2 |
| 22 | Took supplies or tools home without permission | 8.8 | 3.7 | 1.6 |
| 23 | Tried to look busy while doing nothing | 10.9 | 2.7 | 2.3 |
| 24 | Put in to be paid for more hours than you worked | 19.3 | 4.3 | 1.6 |
| 25 | Took money from your employer without permission | 15.6 | 5.1 | 1.4 |
| 26 | Ignored someone at work | 15.0 | 2.9 | 1.6 |
| 27 | Refused to help someone at work | 11.3 | 5.1 | 2.2 |
| 28 | Withheld needed information from someone at work | 12.1 | 3.3 | 1.8 |
| 29 | Purposely interfered with someone at work doing his/her job | 14.6 | 4.3 | 1.0 |
| 30 | Blamed someone at work for an error you made | 14.1 | 2.5 | 2.0 |
| 31 | Started an argument with someone at work | 15.8 | 3.3 | 1.6 |
| 32 | Stole something belonging to someone at work | 8.2 | 1.8 | 2.0 |
| 33 | Verbally abused someone at work | 10.7 | 2.0 | 3.3 |
| 34 | Made an obscene gesture (the finger) to someone at work | 12.3 | 1.8 | 1.6 |
| 35 | Threatened someone at work with violence | 10.0 | 3.1 | 1.0 |
| 36 | Threatened someone at work, but not physically | 11.9 | 0.8 | 2.0 |
| 37 | Said something obscene to someone at work to make them feel bad | 10.2 | 2.1 | 1.6 |
| 38 | Hid something so someone at work couldn’t find it | 9.2 | 1.8 | 1.4 |
| 39 | Did something to make someone at work look bad | 10.4 | 1.8 | 1.4 |
| 40 | Played a mean prank to embarrass someone at work | 10.9 | 2.9 | 1.4 |
| 41 | Destroyed property belonging to someone at work | 11.1 | 1.4 | 1.6 |
| 42 | Looked at someone at work's private mail property without permission | 10.4 | 1.8 | 1.2 |
| 43 | Hit or pushed someone at work | 8.8 | 1.2 | 2.3 |
| 44 | Insulted or made fun of someone at work | 13.5 | 1.8 | 2.0 |
| 45 | Avoid returning a phone call to someone you should at work | 14.3 | 3.1 | 2.1 |
Personality Traits and CWB Scores (average) Across Various Ministries.
| Mean (on a 5-scale) and Standard Deviation | ||||||
|---|---|---|---|---|---|---|
| Ministry | Personality Traits | CWB | ||||
| Extraversion | Agreeableness | Conscientiousness | Neuroticism | Openness | ||
| Prime Ministry's Department | 3.68 (0.15) | 3.74 (0.16) | 3.78 (0.19) | 2.84 (0.15) | 3.20 (0.11) | 1.41 (0.09) |
| Home Ministry | 3.90 (0.21) | 4.12 (0.14) | 4.08 (0.15) | 2.12 (0.17) | 2.96 (0.13) | 1.13 (0.05) |
| Ministry of Defense | 3.84 (0.23) | 3.68 (0.18) | 3.84 (0.21) | 2.25 (0.23) | 3.19 (0.22) | 1.49 (0.15) |
| Ministry of Education | 4.22 (0.12) | 4.14 (0.13) | 4.16 (0.14) | 2.34 (0.14) | 3.08 (0.07) | 1.08 (0.03) |
| Ministry of Energy | 4.30 (0.14) | 4.02 (0.19) | 4.19 (0.14) | 2.14 (0.15) | 3.10 (0.15) | 1.19 (0.07) |
| Ministry of Federal Territories | 3.88 (0.16) | 3.29 (0.26) | 3.85 (0.15) | 2.47 (0.15) | 2.88 (0.14) | 1.43 (0.07) |
| Ministry of Finance | 4.14 (0.13) | 3.98 (0.16) | 4.08 (0.15) | 2.34 (0.17) | 3.08 (0.14) | 1.36 (0.18) |
| Ministry of Foreign Affairs | 3.75 (0.19) | 3.97 (0.19) | 4.06 (0.20) | 2.28 (0.23) | 3.25 (0.16) | 1.36 (0.12) |
| Ministry of Health | 4.24 (0.10) | 4.46 (0.10) | 4.18 (0.13) | 1.98 (0.15) | 3.06 (0.10) | 1.09 (0.04) |
| Ministry of Higher Education | 4.00 (0.14) | 3.93 (0.20) | 4.00 (0.14) | 2.28 (0.14) | 2.93 (0.10) | 1.22 (0.06) |
| Ministry of Housing | 4.27 (0.15) | 4.27 (0.14) | 4.02 (0.18) | 2.18 (0.18) | 3.21 (0.08) | 1.16 (0.08) |
| Ministry of Human Resource | 4.20 (0.11) | 3.82 (0.13) | 4.06 (0.17) | 2.34 (0.13) | 3.28 (0.12) | 1.40 (0.12) |
| Ministry of Information | 3.79 (0.19) | 3.63 (0.19) | 3.87 (0.19) | 2.45 (0.15) | 3.13 (0.14) | 1.32 (0.09) |
| Ministry of International Trade | 3.73 (0.25) | 3.47 (0.21) | 3.60 (0.17) | 2.63 (0.16) | 3.00 (0.15) | 1.39 (0.05) |
| Ministry of Natural Resources | 4.21 (0.19) | 4.02 (0.22) | 4.02 (0.16) | 2.17 (0.20) | 3.14 (0.10) | 1.15 (0.06) |
| Ministry of Plantation | 4.09 (0.22) | 4.28 (0.17) | 4.22 (0.17) | 2.06 (0.18) | 3.06 (0.12) | 1.03 (0.02) |
| Ministry of Rural Development | 4.15 (0.18) | 4.05 (0.15) | 3.90 (0.20) | 2.28 (0.17) | 3.08 (0.13) | 1.31 (0.13) |
| Ministry of Science & Technology | 4.25 (0.23) | 3.79 (0.23) | 3.75 (0.20) | 2.50 (0.23) | 2.82 (0.21) | 1.44 (0.09) |
| Ministry of Tourism | 4.53 (0.13) | 4.07 (0.19) | 4.47 (0.13) | 1.83 (0.20) | 2.70 (0.27) | 1.18 (0.06) |
| Ministry of Transport | 4.13 (0.16) | 4.13 (0.15) | 3.98 (0.17) | 2.27 (0.19) | 3.02 (0.12) | 1.32 (0.10) |
| Ministry of Women | 3.86 (0.19) | 3.66 (0.20) | 3.61 (0.21) | 2.71 (0.21) | 3.21 (0.17) | 1.32 (0.13) |
| Ministry of Works | 3.75 (0.27) | 3.86 (0.18) | 3.82 (0.28) | 2.36 (0.19) | 3.04 (0.19) | 1.89 (0.29) |
| Ministry of Sports | 4.19 (0.11) | 4.17 (0.16) | 4.00 (0.17) | 2.21 (0.15) | 2.98 (0.09) | 1.15 (0.08) |
| Ministry of Agriculture | 3.82 (0.17) | 3.90 (0.15) | 3.86 (0.15) | 2.66 (0.13) | 2.84 (0.09) | 1.42 (0.15) |
| Ministry of Domestic Trade | 4.20 (0.17) | 4.32 (0.12) | 4.32 (0.16) | 2.07 (0.17) | 3.11 (0.15) | 1.14 (0.05) |
The correlations between the dimensions of all constructs for male and female are given in Table 4. Based on the correlations, it can be observed that dimensions of CWB have significant correlations with dimensions of other constructs. Similarly, dimensions of emotional labor and emotional exhaustion have significant correlations with the dimensions of personality and EI.
Correlation between Constructs.
| Construct | EX | AG | CON | NE | OP | EI | APA | ANA | ELSA | ELDA | EE | CWB |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| EX | 1.00 | .465* | .451* | -.491* | -.048 | .307* | .327* | -.222* | .302* | .172* | -.197* | -.091 |
| AG | .372* | 1.00 | .467* | -.552* | -.039 | .342* | .233* | -.250* | .185* | .289* | -.344* | -.271* |
| CON | .372* | .516* | 1.00 | -.442* | .043 | .242* | .353* | -.195* | .117 | .084 | -.211* | -.240* |
| NE | -.466* | -.530* | -.566* | 1.00 | .120 | -.376* | -.140* | .300* | -.337* | -.132* | .475* | .290* |
| OP | .015 | .039 | .055 | .017 | 1.00 | .035 | .028 | .124* | -.024 | -.103 | .164* | -.017 |
| EI | .342* | .463* | .398* | -.419* | -.014 | 1.00 | .312* | -.055 | .408* | .232* | -.227* | -.236* |
| APA | .334* | .403* | .414* | -.401* | .088 | .475* | 1.00 | -.005 | .160* | .091 | .026 | -.075 |
| ANA | -.343* | -.418* | -.371* | .490* | .074 | -.218* | .034 | 1.00 | -.043 | .018 | .201* | .157* |
| ELSA | .094 | .175* | .145 | -.237 | .007 | .480* | .213* | -.081 | 1.00 | .510* | -.241* | -.085 |
| ELDA | .131 | .200* | .145 | -.185* | -.043 | .438* | .247* | -.263* | .471 | 1.00 | -.151 | -.091 |
| EE | -.196* | -.168* | -.291* | .390* | .044 | -.272* | -.075 | .347* | -.463* | -.331 | 1.00 | .484* |
| CWB | -.231* | -.379* | -.425* | .300* | -.103 | -.346* | -.109 | -.391* | -.235* | -.227* | .524* | 1.00 |
Note. Values below the diagonal are for males and values above the diagonal are for female.
EX=Etraversion, AG=Agreeableness, CON=Conscientiousness, NE=Neuroticism, OP=Openness, EI=Emotional intelligence, APA=Positive affectivity, ANA=Negative affectivity, ELSA=Emotional labor (surface acting), ELDA=Emotional labor (deep acting), EE=Emotional exhaustion, CWB=Counterproductive work behavior.
* p < .05.
The hypotheses were tested using Lisrel 9.1 version. The model fit statistics are: root mean square error approximation (RMSEA)=.042 (acceptable level<.08); chi-square/degrees of freedom=1.826 (acceptable level<3); p-value=.895 (acceptable level<.05); normed fit index (NFI)=.968 (acceptable level > .9); comparative fit index (CFI)=.985 (acceptable level > .9); root mean residual (RMR)=.030 (acceptable level<.08); goodness-of-fit index=.977 (acceptable level > .9). The significant results of the hypotheses testing are given in Table 5. We highlight some interesting results that were contrary to our hypotheses.
Significant Hypothesized Relationships.
| Significant Relationships | β value/p-value | Hypothesis supported | |
|---|---|---|---|
| From | To | ||
| Personality traits | |||
| Extraversion | EI | .115/.015 | H1a |
| Agreeableness | .208/.000 | ||
| Neuroticism | -.217/.000 | ||
| Personality traits | |||
| Extraversion | Positive affectivity | .173/.000 | H1b |
| Conscientiousness | .257/.000 | ||
| Neuroticism | .105/.030 | ||
| Personality traits | |||
| Extraversion | Negative affectivity | .118/.017 | H1c |
| Agreeableness | -.016/.001 | ||
| Neuroticism | .243/.000 | ||
| Personality traits | |||
| Conscientiousness | CWB | -.147/.000 | H1 |
| Openness | -.085/.022 | ||
| EI | CWB | -.120/.002 | H2 |
| Negative affectivity | CWB | .078/.042 | H3 |
| Emotional exhaustion | CWB | .462/.000 | H5 |
| Negative affectivity | Emotional exhaustion | .095/.021 | H3a |
| Indirect relationship between EI and emotional exhaustion through emotional labor (surface acting) | Sobel's test t-value=4.488, p-value=.000 | H2e | |
First, the negative relationship between positive affectivity and CWB of frontline employees is not supported (β=.049, p-value=.286). Counter to results by Miles et al. (2002) and Dalal (2005), our study shows a positive relationship between positive affectivity and CWB. Second, the hypothesis that establishes the relationship between emotional labor (surface acting) and CWB is not supported (β=.061, p-value=.144). Our result is contrary to the result by Bechtoldt, Welk, Zapf, & Hartig (2007). According to them, CWB is used as the avenue by employees to escape emotional dissonance and this is central to emotional labor (surface acting). Third, the hypothesis that establishes the indirect link between EI and emotional exhaustion through emotional labor (deep acting) is not supported (β=.072, p-value=.095). Our finding contradicts the finding of Cote (2005) and according to him the individuals with high EI are more likely to engage in deep acting during interpersonal interactions. Fourth, the indirect negative relationship between EI and CWB through emotional labor (surface acting) is not supported. In fact, our test reveals that there is a strong positive relationship between EI and emotional labor (surface acting) (β=.367, p-value=.000). Our result is partially supported by Wong and Law (2002). They have argued that employees high on the appraisal of self-emotions and other's emotions recognize the need to engage in emotional labor in a positive manner. But, the authors have not looked at the individual dimensions of emotional labor. Fifth, the indirect negative relationship between EI and CWB through emotional exhaustion is not supported (β=-.058, p-value=.209). Our result does not support the findings by many researchers (For example, Brackett et al., 2010). Sixth, the relationship between emotional exhaustion and emotional labor (deep acting) is not supported (β=-.070, p-value=.114). However, there is a negative relationship between emotional exhaustion and emotional labor (surface acting) (β=-.212, p-value=.000) contrary to our assertion that there is a positive relationship. A few researchers have shown a positive relationship between emotional exhaustion and emotional labor (surface acting) (Chooi, 2012; Johnson & Spector, 2007; Lee, 2010). However, Chu (2002) and Pugh, Groth and Hennig-Thurau (2011), have argued a negative relationship between emotional labor (surface acting) and emotional exhaustion. Seventh, the hypotheses that establish the links between affectivity (positive and negative) and emotional labor (surface acting and deep acting) are not supported. Kammeyer-Mueller et al. (2013) have shown a moderate relationship between these two constructs. Eighth, the hypotheses that test the relationships between positive affectivity (β=.108, p-value=.007), negative affectivity (β=.095, p-value=.021) and emotional exhaustion are partially supported. We have hypothesized the relationship between positive affectivity and emotional exhaustion to be negative. The indirect role of emotional labor between affectivity and emotional exhaustion is not supported since the relationship between affectivity (positive and negative) and emotional labor is insignificant.
Since the percentage of women is higher in our sample (70%), we have performed independent t-test as an additional analysis to study the effect of gender on all the constructs. Based on our results, there are significant differences and these are: (1) extraversion (t-value=-2.789, p-value=.005) with women recording a higher level than men, (2) agreeableness (t-value=-2.731, p-value=.007) with women recording higher level than men, (3) conscientiousness (t-value=-3.899, p-value=.000) with women showing higher level than men, (4) levels of emotional labor (surface acting) (t-value=-2.366, p-value=.018) with women recording a higher level than men, (5) levels of emotional labor (deep acting) (t-value=-2.446, p-value=.015) with women recording a higher level than men, and (4) levels of CWB (t-value=2.328, p-value=.021) with men recoding a higher level than women.
DiscussionThis research set out to address a fundamental issue related to CWB of frontline government employees. Specifically, this research has analyzed the roles (direct and indirect) of personality traits, EI, positive and negative affectivity, emotional labor, and emotional exhaustion of frontline employees on CWB. The framework has been validated by studying the frontline government employees in Malaysia. We believe that this framework can be tested in any country (developed and developing). The framework with significant relationships is given in Figure 2. Among the constructs that have a direct impact on CWB are: personality traits (conscientiousness and openness dimensions), negative affectivity, EI, and emotional exhaustion. Of all the constructs, emotional exhaustion has the strongest link with CWB. The detailed discussion based on the direct and indirect relationships follows.
Final Framework with Significant Relationships.
Note. Pex=Extraversion, Pab=Agreeableness, Pcon=Conscientiousness, Pne=Neuroticism, Pop=Openness, EI=Emotional intelligence, APA=Positive affectivity, ANA=Negative affectivity, ELSA=Emotional labor (surface acting), ELDA=Emotional labor (deep acting), EE=Emotional exhaustion, CWB=Counterproductive work behavior
The model fit statistics are: Root mean square error approximation (RMSEA)=.042; chi-square/degrees of freedom=1.826; p-value=.895; Normed fit index (NFI)=.968; Comparative fit index (CFI)=.985; Root mean residual (RMR)=.030, Goodness of fit index (GFI)=.977.
The validated framework suggests that personality traits impact EI, affectivity (positive and negative), emotional labor (deep acting and surface acting), emotional exhaustion, and CWB of frontline employees; EI and affectivity in turn impact emotional labor, emotional exhaustion, and CWB of these employees. The descriptive statistics of these constructs suggest that levels range from ‘low’ to ‘high’. The levels of negative affectivity, emotional exhaustion, and CWB are low, which suggests that the frontline employees in Malaysian public sector experience low levels of negative emotions and emotional exhaustion and therefore, involve in lesser instances of CWB. A high level of emotional intelligence indicates that government employees with the ‘right’ skill have been appointed at frontline offices in Malaysia.
The link between personality traits and EI is not surprising, as many studies have established the relationship (for example, Brackett & Mayer, 2003; Myleen et al., 2009; Petrides et al., 2010). Our study has shown that extraversion, agreeableness, and neuroticism have impact on EI and these findings agree with the findings of earlier studies (Chamorro-Premuzic et al., 2007; Lopes, Salovey, & Strauss, 2003). The relationship between personality traits and affectivity (negative and positive) has been studied and demonstrated by a few researchers (Bruck & Allen, 2003; Hoergera & Quirk, 2010; Yik & Russell, 2001). Our research indicates that (1) extraversion, conscientiousness, and neuroticism have impact on positive affectivity with conscientiousness contributing the maximum and (2) extraversion, agreeableness, and neuroticism have impact on negative affectivity with neuroticism contributing the maximum.
Our results show that conscientiousness and openness are negatively correlated to CWB. Our research shows that it is the personality traits that drive the EI, affectivity (positive and negative), emotional labor (surface and deep acting), emotional exhaustion, and CWB of frontline employees. The managers must ensure that employees with the right attitude and skills handle the front offices.
Our results show that EI has positive relationships with positive affectivity and emotional labor (surface acting) and negative relationships with CWB. The relationship between emotional intelligence and positive affectivity is supported by earlier studies (Kong & Zhao, 2013). These researchers have argued that positive and negative tendencies of individuals are significantly influenced by the way they can understand, control, and direct their emotions.
The indirect effect of EI on emotional exhaustion is a significant result of our paper. Earlier researchers have focused on the direct effects of emotional intelligence on emotional exhaustion (Brackett et al., 2010; Chan, 2006; Moon & Hur, 2011). In our paper, the direct effect is insignificant and the indirect effect on emotional exhaustion (1) through emotional labor (surface acting) (Sobel's test t-value=4.823, p-value=.000) has been found to be ‘strong’ and (2) through positive affectivity (Sobel's test t-value=2.496, p-value=.013) has been found to be ‘moderate’. Our research suggests that irrespective of the type of affectivity, frontline employees in the government are emotionally exhausted. This may be a reflection of the nature of their job that requires handling problems of citizens who come from different backgrounds. Another plausible reason for this phenomenon can be because of high percentage of women respondents (70%). According to Rubino, Volpone, and Avery (2013), women are more emotionally exhausted than men especially if they experience work-family conflicts. In the Asian society, women are expected to play a significant role at home and this may result in higher work-family conflicts.
The testing of relationships between EI, emotional labor (surface acting) and emotional labor (deep acting) have yielded mixed results. The relationship with emotional labor (deep acting) is not significant (hypothesized relationship is positive) and with emotional labor (surface acting) is strongly positive (hypothesized relationship is negative). Earlier studies have shown that EI is negatively related to emotional labor (surface) and positively related to emotional labor (deep acting) (Johnson & Spector, 2007; Lee, 2010). Employees engaged in jobs that require greater interpersonal demands, such as government frontline employees, need to expend effort and self-control to communicate organizationally desired emotions (e.g., cheerfulness, helpfulness, and friendliness) in an effective manner during service transactions and encounters (Hochschild, 1983). Our study suggests that emotional labor (surface and deep) is used by government frontline employees in Malaysia to align their displayed emotions with organizational display rules/norms. Alignment of displayed emotions is important since government frontline employees are dealing with ordinary citizens that come from different walks of life.
In line with earlier studies (Deshpande et al., 2005; Jung & Yoon, 2012; Petrides et al., 2004), our finding indicates a significant negative relationship between EI and CWB. The government frontline employees need to possess a high level of emotional intelligence to handle the citizens. A high level of EI among the employees results in low incidence of CWB.
Our finding on the relationship between positive affectivity and emotional exhaustion is contradictory to the earlier findings. Earlier studies have shown a negative relationship (Abraham, 1998; Lockyer, 2013; Szczygieł et al., 2012) and our study has shown a positive relationship. Positive affectivity refers to individuals experiencing positive moods such as joy, interest in the job, and alertness. But this does not stop government frontline employees in Malaysia from experiencing high levels of emotional exhaustion. A plausible reason can be the nature of the job that requires deft handling of various problems of citizens. The impact of negative affectivity on emotional exhaustion and emotional exhaustion on CWB is well documented (Douglas & Martinko, 2001; Fox & Spector, 2005; Martinko et al., 2002) and our results support the findings. The frontline employees experiencing negative mood states are more likely to be emotionally exhausted and are more likely to resort to CWB.
Our finding on the relationship between emotional labor (surface acting) and emotional exhaustion is contradictory to the findings of some studies (Johnson & Spector, 2007; Lee, 2010). Our study has shown a negative relationship in line with the studies by Chu (2002) and Pugh et al. (2011). The negative effect can be due to the fact that government frontline employees by displaying their emotions (emotional labor) (Ashforth & Humphrey, 1993) may feel that their emotional exhaustion is reduced. Our result also shows a positive relationship between surface acting and deep acting emotional labor. According to Grandey (2003, p. 87), surface acting is “faking in bad faith” and deep acting is “faking in good faith”. For frontline employees, faking in good or bad faith is needed to please the citizens who approach these employees with different issues.
The positive relationship between emotional exhaustion and CWB of government frontline employees in Malaysia is very strong and supports the view that excessive emotional exhaustion results in CWB (Banks et al., 2012; Mulki et al., 2006). We also observe that emotional exhaustion indirectly impacts (Sobel's test t-vale=4.820, p-value=.000) the relationship between emotional labor (surface acting) and CWB. Our model that integrates all the constructs can explain: (1) 33.2% of the variance in CWB, (2) 26.5% of the variance in emotional exhaustion, (3) 33.4% of the variance in emotional labor (deep acting), and (4) 23.1% of the variance in emotional labor (surface acting).
Managerial ImplicationsSince the frontline employees are “barometers of the business” (Dagger et al., 2013, p. 498), the managers must ensure that they possess or are trained to have the right skills to perform their jobs (Dagger et al., 2013; Porath, MacInnes, & Folkes, 2011). The selection of frontline employees who are constantly in touch with the common citizens must be done with utmost care since the effectiveness of service delivery depends upon them. The citizens form opinions about the government institutions based on their experiences with the frontline employees. Therefore, it is imperative that these employees with the following characteristics be allowed to handle the citizens: (1) high on personality traits such as extraversion, agreeableness, openness, and conscientiousness and low on neuroticism; (2) high levels of EI, and (3) low levels of negative affectivity. These characteristics will help in managing the emotional labor, emotional exhaustion, and thereby CWB of the employees.
As suggested by Grandey (2000), environment in the form of organizational support plays an important role in managing emotions and stress of service employees. How can the managers help in avoiding unpleasant incidences that may occur when the frontline employees and common citizens are interacting? We outline some of the steps that can be carried out by the managers handling frontline employees.
First, a proper training on enhancing EI of employees needs to be provided. According to Beigi and Shirmohammadi (2011) and Cherniss and Adler (2000), training given to employees helps them handle their emotions better. Second, managers can give job autonomy and therefore, emotional autonomy to employees. Some studies have shown that job autonomy minimizes the stress due to emotion regulation process (Grandey, 2000). Third, managers must support their frontline employees. An employee's perception that he/she is supported by managers and co-workers can help him/her have positive views about the working environment and this in turn can reduce emotional labor and emotional exhaustion (Grandey, 2000; Howes, Cropanzano, Grandey, & Mohler 1999). Fourth, emotional displays (emotional labor) adequate for a particular situation are dependent on culture and situation, and hence learnable (Lazanyi, 2011). A social environment at work that is conducive can help employees identify themselves with the organization and its goals and this will lead to internalization of organizational norms relating to emotional displays in place. Fifth, when the number of women employees is significant as in the case of Malaysia, organizations can provide additional support such as a child care center in the office premises and this may reduce work-family conflict among women employees and this in turn can help reduce their emotional exhaustion and instances of CWBs.
Limitations and Directions for Future ResearchLike any other study, this study is not without limitations. First, the questionnaire was distributed to all the 25 ministries in Malaysia. Though we received the feedback from all the ministries, only seven ministries returned all the questionnaires after filling (100%) and six ministries returned less than 75%. Higher response rate from more ministries could have made generalization meaningful. Second, 70% of the frontline employees were women. In this research we did address the gender influence on the results. However, future studies can perform in-depth studies related to gender and factors influencing CWB. Additional constructs such as work-family conflict and work overload can be included to understand the complete process of employees resorting to CWB. Our study has obtained responses from one source – front-line employees. We agree that the results are more robust with multiple sources. Unfortunately, the supervisors and managers refused to participate in the study. Even though we have tested and shown that the effects of common variance bias are not high, we do recognize the limitation of not obtaining information from multiple sources.
ConclusionsThis study addressed the CWB of government frontline employees in Malaysia. The study was designed to answer a fundamental question on the effect of personality, EI, and affectivity on emotional labor, emotional exhaustion, and finally on CWB of the employees. The main finding shows that personality traits drive the EI of employees, their affectivity, their emotional labor, their emotional exhaustion, and their CWB. EI and affectivity in turn affect emotional labor, emotional exhaustion, and CWB. The interrelationship between the various constructs has made this model comprehensive and different from other studies. Our integrated model has also helped us study the indirect roles of emotional labor (between EI and emotional exhaustion) and emotional exhaustion (between emotional labor and CWB).
Conflict of InterestThe authors of this article declare no conflict of interest.
