In service organizations, though the expression of positive emotions to the customers by service employees contributes to the efficacy of the organization as a whole and is related to customers’ satisfaction, it would be unrealistic to expect that service employees will always be in a good mood (Grandey 2003; Morris and Feldman 1996). In fact, when employees cannot spontaneously display either the appropriate emotions or their genuine emotions, they have to rely on compensatory strategies, called emotional labor strategies, to regulate both their feelings and the way they are expressed, in order to meet the demands of their jobs (Diefendorff et al. 2005; Grandey 2000).

Emotional labor is the process of regulating both feelings and expressions for organizational goals and it may involve enhancing, faking, or suppressing emotions to modify the emotional expression (Grandey 2000). Emotional labor strategies take various forms. Surface acting involves the simulation of emotions that are not truly felt by the careful presentation of appropriate verbal or non-verbal signals, either through the artificial expression of positive emotions or the suppression of negative emotions (Ashforth and Humphrey 1993; Diefendroff et al. 2005; Grandey 2000). Conversely, deep acting involves the real internal modification of negative emotions and attempts to actually feel the emotion that should be displayed (Ashforth and Humphrey 1993; Diefendroff et al. 2005).

Acting Mechanisms of Emotional Labor and Emotional Exhaustion

In practice, the more employees need to alter their emotional expressions, the more likely they will need to exert efforts and invest personal resources in order to alter or suppress true emotions (Baumeister et al. 1998; Muraven and Baumeister 2000; Richard and Gross 1999). Thus, the outcomes of emotional labor will vary according to the strategy used (Ashkanasy et al. 2002; Goldberg and Grandey 2007). Indeed, if one ends out truly feeling the emotion that he or she is displaying (i.e. deep acting), outcomes are expected to be more positive (Zapf 2002). The potential negative outcomes of surface acting will then also be explained, at least in part, by the emotional toll of the resulting feelings of non-authenticity or emotional dissonance (Bulan et al. 1997; Hochschild 1983; Parkinson 1991; Sutton 1991). It is thus not surprising that many authors proposed that emotional labor strategies and similar forms of demands-resources imbalances may be involved in the emergence of burnout (Cordes and Dougherty 1993; Hobföll 1989; Leiter and Stright 2009; Wilk and Moynihan 2005).

Among the components of burnout, emotional exhaustion is one of the most-often-cited consequences of emotional labor. Emotional exhaustion is a specific stress-related reaction that refers to a state of depleted energy caused by the excessive psychological and emotional demands that occur among individuals who work with people in some capacity (Jackson et al. 1987). Kruml and Geddes (2000) suggested that the degree of exhaustion which workers experience varies according to acting types. They demonstrated that surface acting (which they conceptualized as dissonance) was more strongly related to emotional exhaustion than deep acting (conceptualized as effort). Erickson and Ritter (2001) also found a positive relationship between surface acting and emotional exhaustion in a broad range of occupations. Brotheridge and Grandey (2002) found that surface acting was positively related to emotional exhaustion while deep acting showed almost no relationship. To further support this common finding, Grandey (2003) also corroborated the finding that the more often surface acting is performed, the greater the emotional exhaustion reported. Deep acting, on the other hand, has not shown to be as detrimental to the individual performing it, and in fact is related to many positive outcomes. In an experience sampling study of call-center employees, Totterdell and Holman (2003) demonstrated that surface acting was more positively associated with emotional exhaustion than deep acting. In addition, Grandey et al. (2005) also found that surface acting was positively related to emotional exhaustion. Finally, Johnson and Spector’s (2007) findings also support this notion, that is, surface acting was positively related to emotional exhaustion, while deep acting exhibited a negative relationship with emotional exhaustion. In concordance with the aforementioned literature, the first hypothesis of the present study is as follows:

  1. Hypothesis 1:

    Surface acting would positively predict emotional exhaustion.

  2. Hypothesis 2:

    Deep acting would negatively predict emotional exhaustion.

Affectivity and Emotional Labor

In recent years, most of the research on emotional labor has been situationally based. Situational variables such as frequency, duration, variety of emotional labor (Morris and Feldman 1996, 1997; Jones 1998; Grandey 1999), presence of display rules (Morris and Feldman 1996), or other job variables such as display training, quality orientation (Kruml and Geddes 2000), and job autonomy (Morris and Feldman 1996; Jones 1998; Kruml and Geddes 2000) are frequently treated as antecedents of emotional labor. Rarely, however, is emotional labor formulated as having an endogenous source of variance, one that is reflective of the ongoing state of the person as opposed to being a product of the situation. This study argues that it is the individual dispositional factor that will determine what acting mechanism (surface acting, deep acting) individuals tend to adopt during service transactions. Among the personal resources, affectivity appears to be the most relevant personal attribute in relation to employee’s choice of emotional labor strategy. Affectivity has been defined as a general tendency to experience a particular mood or to react to objects in a particular way or with certain emotions (Abraham 1998).

According to Grandey (2000), since individual differences in affectivity determine intensity and nature of felt emotions and affect subsequent behaviors, both positive and negative affectivity (NA) should relate to emotional labor. Schaubroeck and Jones (2000) argued that employees with different affective styles may engage in different mechanisms of emotional labor. Schaubroek and Jones’ idea has been empirically supported as research has consistently found a positive relationship between negative affectivity (NA) and surface acting (Brotheridge and Grandey 2002; Brotheridge and Lee 2003; Johnson 2004; Gosserand and Diefendorff 2005). However, findings on affectivity and deep acting have been scarce and are not conclusive. Johnson (2004) found that positive affectivity (PA) and deep acting were positively associated, and while deep acting and negative affectivity (NA) were negatively related, this relationship was nonsignificant. Similarly, Gosserand and Diefendorff (2005) reported a positive relationship between positive affectivity (PA) and deep acting, and a negative, albeit non-significant, relationship between negative affectivity (NA) and deep acting. As there are so few findings about two constructs that exist within the same nomological network, it is important that further research examines the relationship between affectivity and the emotional labor strategies, especially deep acting. As per the direction of pertinent literature, the present study postulated the following hypotheses:

  1. Hypothesis 3:

    Negative affectivity (NA) would be positively related with surface acting.

  2. Hypothesis 4:

    Positive affectivity (PA) would be positively related to deep acting.

Moderating Role of Affectivity in Emotional Labor and Emotional Exhaustion

In general, most studies considering the effects of NA present a somewhat dim outlook for high-NA individuals, including the likelihood of experiencing more frequent episodes of stress and frustration and lower levels of job satisfaction (Levin and Stokes 1989). Because emotional exhaustion is defined as a specific type of stress reaction, one can expect that high-NA employees will report higher levels of job related emotional exhaustion than will low-NA employees. This has further been empirically supported by Zellars et al. (1999) and more recently by Adil and Kamal (2012) who found that negative affectivity (NA) was a positive predictor of emotional exhaustion.

In most of the studies on the relationship between affectivity and emotional exhaustion, positive affectivity (PA) has often been a negative correlate of emotional exhaustion. Zellars et al. (2000) and Zellars et al. (2004) found that it was neuroticism (which is equated with negative affectivity (NA); see Brief 1998) that predicted exhaustion and depersonalization component of emotional exhaustion whereas extraversion (equated with the positive affectivity (PA); see Brief 1998) predicted diminished accomplishment component of the same. This has further been testified by Adil and Kamal (2012) who found that positive affectivity (PA) was a negative predictor of emotional exhaustion among customer services representatives.

Albeit the significant relationship between affectivity and emotional exhaustion and between emotional labor and emotional exhaustion, few researchers have tried to explore the role of affectivity in the relationship between emotional labor and emotional exhaustion. Grandey (2002) studied the effects of surface acting and deep acting on burnout, controlling for the effects of negative affectivity (NA) and interpersonal demands (i.e., interaction expectations including frequency and duration of interactions, intensity and variety of emotional expressions, and perception of display rules from the organization). They found that individuals who reported higher levels of negative affectivity (NA) also reported higher levels of burnout, but surface acting and deep acting accounted for unique variance beyond negative affectivity (NA). Although a relationship between negative affectivity (NA) and emotional labor was not hypothesized, some interesting correlations surfaced from this study. For instance, there was a positive relationship between negative affectivity (NA) and surface acting, but no significant relationship was found between negative affectivity (NA) and deep acting. Perhaps individuals high on negative affectivity (NA) are more likely to surface act than deep act because of its response-focused nature (Gross 1998a, b). That is, it may be easier for high negative affectivity (NA) individuals, who have a tendency to react to difficult situations more negatively, to modify their expressions by faking a smile than it is for them to modify their feelings by thinking positive thoughts or cognitively reappraising the situation in a more positive light. Grandey (2002) found that surface acting was positively related to negative affectivity (NA), and negatively related to positive affectivity (PA); deep acting was not related to either affectivity dimension. Because Brotheridge and Grandey did not hypothesize the effects of negative affectivity (NA) or positive affectivity (PA) on emotional labor and Grandey only hypothesized the relationship between negative affectivity (NA) and emotional labor, more research is needed to clarify the relationships between these variables.

Despite the fact that literature provides evidence for the relationship between affectivity and emotional labor, between affectivity and emotional exhaustion, and between emotional labor and emotional exhaustion (as discussed previously), no previous research has identified how affectivity, emotional labor, and emotional exhaustion may have parsimoniously been integrated in a single theoretical framework. It can be deduced from whatever literature evidence is available that high NA employees are more likely to be engaged in surface acting whereas high PA employees are more likely to undertake deep acting. Moreover, negative affectivity (NA) is a positive whereas positive affectivity (PA) is a negative predictor of emotional exhaustion. This dynamic interplay of affectivity with both emotional labor and emotional exhaustion suggests its moderating role. Thus the present study proposes:

  1. Hypothesis 5:

    Negative affectivity (NA) would moderate the relationship between surface acting and emotional exhaustion such that employees high on negative affectivity (NA) would be more likely to suffer from emotional exhaustion as compared to their counterparts who are low on negative affectivity (NA).

  2. Hypothesis 6:

    Positive affectivity (PA) would moderate the relationship between deep acting and emotional exhaustion such that employees high on positive affectivity (PA) would be less likely to suffer from emotional exhaustion as compared to their counterparts who are low on positive affectivity (PA).

The significance of this study lies in the point that it empirically contributed to the pertinent literature by specifying the precise moderating roles of positive affectivity (PA) and negative affectivity (NA) in the relationship between acting mechanisms of emotional labor and emotional exhaustion. Thus the present study not only holds the promise of theory development but also may spawn important implications for service sector industry.

Control Variables

Among the demographic variables, the present study has incorporated sex, job sector, and job experience as the pertinent control variables in relation to emotional exhaustion. Literature on gender differences in emotional exhaustion has yielded mixed result. Various studies have reported greater emotional exhaustion (Adil and Kamal 2012; Daalen et al. 2009), work exhaustion (Thomsen et al. 1999), and stress (Ptacek et al. 1992) in women. On the contrary, Adekola (2010) and Goddard and Patton (1998) found no gender differences in emotional exhaustion whereas Burke et al. (1996) reported greater emotional exhaustion in men. In Pakistani culture where gender equality is still a dream, women appear to be more emotionally exhausted since they have to be extensively involved in household work, childcare, etc. besides shouldering the demands of their jobs. The inequality between the two sexes in the workplace may also make women employees more prone to emotional exhaustion.

In a review of job burnout, Maslach et al. (2001) reported that of all the demographic variables that have been studied, age is the one that has been most consistently related to burnout. Among younger employees the level of burnout is reported to be higher than it is among those over 30 or 40 years old. Age is confounded with work experience, so burnout appears to be more of a risk earlier in one’s career. Thus, job experience and emotional exhaustion should be inversely related. Contradictory evidence, however is also available as some researchers found that job experience is not related to emotional exhaustion (Kowalski et al. 2010; Sonnentag et al. 2010) whereas others have found that the two constructs are positively related (Lewig and Dollard 2003). Based upon the demographics of sex and job experience, our final hypotheses postulated that:

  1. Hypothesis 7(a):

    Female employees would be significantly more emotionally exhausted as compared to their male counterparts.

  2. Hypothesis 7(b):

    Job experience would be inversely related to emotional exhaustion.

In Pakistani work milieu, customer services representatives in cellular services are typically less paid, have more workloads, lacks in job security, and are exposed to more demanding affective situations as compared to their banking equivalents. Consequently, they are involved in greater amount of emotional, mental, and physical exertions which may lead them to emotional exhaustion. Little research is available on the comparison of customer services representatives of banking sector and telecommunication sector in terms of emotional exhaustion, so we did not formulate any hypothesis.

Method

Participants

The sample of the present study comprised of 232 customer services representatives from various cellular service providers, and banks of Islamabad and Rawalpindi cities of Pakistan. Equal number of employees from both the professions with the job designation of customer services representatives was purposively drawn (n = 116). Furthermore, equal number of men and women were chosen from each occupational category. The minimum academic qualification of the sample was graduation. 161 of the participants had an education of 14 years (BA/BSc/BBA) whereas the remaining 71 held post graduate degrees (MA/MSc/MBA). The age range of the sample was 19 to 57 years (M = 26.36, SD = 5.37). The job experience of the sample ranged from 1 to 12 years (M = 4.18, SD = 4.25).

The inclusion criteria in the sample were limited to full time employees of banks and cellular services who have been working in customer services departments with a minimum job experience of 1 year. Furthermore, the educational baseline of the sample was graduation. The baseline of 14 years of formal education was necessary to ensure participants’ comprehension of scales which were in English language.

Instruments

The instruments used for the measurement of various variables of the present study were validated and standard measure of their corresponding constructs. All the scales were self-report Likert type rating scales in English language. A brief description of these instruments is as follows:

Emotional Labor Scale

Deep acting and surface acting were measured with the corresponding subscales of Emotional Labor Scale (Brotheridge and Lee 2003). Both the subscales consisted of three items each which were measured on a 5-point Likert type scale (‘1’ = Never, ‘5’ = Always) where high score represents high level of focal constructs and vice versa. None of the items were revers scored. Brotheridge and Lee (2002) reported excellent reliability estimates of .89 and .86 for deep acting and surface acting respectively.

Positive and Negative Affect Schedule

Positive and negative affectivity (NA) was measured through Positive and Negative Affect Schedule (PANAS) developed by Watson et al. (1988). PANAS comprises of two subscales each consisting of 10 adjectives. Respondents were asked to report how did they feel over the past month on a 5-point Likert type response format (where 1 = “very slightly or not at all” and 5 = “extremely”). The higher the score, the higher the focal construct and vice versa. Research has evidenced the sound reliability of PANAS (i.e., Morris and Feldman 1996; Jones 1998; Wright and Cropanzano 1998; Schaubroeck and Jones 2000). For example, Cronbach’s alphas of .86 and .91 for PA and .85 and .83 for NA in successive studies (Morris 1995; Schaubroeck and Jones 2000) supported internal consistency reliability.

Emotional Exhaustion Scale

Emotional exhaustion was measured through a 7-item summated scale developed by Erickson and Ritter (2001). It was a 7-point scale ranging from 0 = “never felt this way while at work” to 6 =”felt this way every day” where respondents were asked to indicate how often they had experienced each of the enlisted situations during the previous 6 months. The score ranged from 0 to 36 where high scores corresponded higher level of burn out and vice versa. The scale had been a consistent and reliable measure of burnout with a reported alpha of .90 (Erickson and Ritter 2001).

Results

The reliability estimates, descriptive statistics, and correlations of variables of the present study have been depicted in Table 1. As evident from the Table, all the measures demonstrated satisfactory levels of internal consistency and provided a reliable operationalization of their corresponding constructs. The correlation matrix revealed that all the correlations were in the expected directions and of expected magnitude. Deep acting had a significant and positive correlation with surface acting, emotional exhaustion, and both of the interaction terms (product of negative affectivity (NA) and surface acting and product of positive affectivity (PA) and deep acting), however its correlations was non-significant with both positive and negative affectivity (NA). Surface acting was positively associated with negative affectivity (NA), emotional exhaustion and both of the interaction terms; however its negative relationship was non-significant with positive affectivity (PA).

Table 1 Descriptives, alpha reliability estimates and correlations of variables (N = 232)
Full size table

The proposed hypotheses of the present study were tested through hierarchical regression analysis. Sex, organization, and job experience were entered as the control variables in the first and the subsequent models. In the second model, deep acting and surface acting were entered, the third model also included negative and positive affectivity (PA), and the fourth model comprised of the product terms of deep acting with positive and negative affectivity (NA) as well as the product terms of surface acting with both types of affectivity (see Table 2).

Table 2 Summary of hierarchical regression analysis for variables predicting emotional exhaustion (N = 232)
Full size table

Among the control variables, sex and organization remained the significant predictors of emotional exhaustion across the four models. The first model comprising of only control variables was statistically significant {F model (3, 228) = 4.27, p < .01} and explained 5.3 % variance in emotional exhaustion {R 2 = .053}.

As expected, surface acting significantly predicted emotional exhaustion {t = 2.63 p < .01} whereas deep acting, albeit having a significant positive relationship with emotional exhaustion, could not significantly predict it (see model 2, 3, & 4 in Table 2). The model was overall significant {R 2 = .10, F (5, 226) = 5.16, p < .001} and added significantly in the explained variance of emotional exhaustion {ΔR 2 = .049, F chnage (2, 226) = 6.21, p < .01}.

In the third model, affective attributes of personality were entered. The model was statistically significant {R 2 = .18, F (7, 224) = 7.17, p < .001} and added significantly in the explained variance of emotional exhaustion {ΔR 2 = .081, F chnage (2, 224) = 11.05, p < .001}. Negative affectivity (NA) significantly predicted emotional exhaustion in positive direction {t = 4.64, p < .001}.

The final model of the current investigation comprised of product terms of modes of emotional labor and affectivity for testing the moderated impact of affectivity on the relationship between modes of emotional labor and emotional exhaustion. More specifically, it included the product terms of deep acting with both positive and negative affectivity (NA) and the product terms of surface acting with both negative and positive affectivity (PA). The model was statistically significant {R 2 = .24, F (11, 220) = 6.43, p < .001} and it explained a unique variance in emotional exhaustion {ΔR 2 = .060, F chnage (4, 220) = 4.39, p < .01}. In accordance with the hypotheses, the interaction term of negative affectivity (NA) and surface acting significantly predicted emotional exhaustion in positive direction {t = 2.37, p < .01}whereas the interaction term of positive affectivity (PA) and deep acting significantly predicted emotional exhaustion in negative direction {t = 2.81, p < .01}. This suggested the differential moderating roles of positive and negative affectivity (NA) in the relationships between modes of emotional labor and emotional exhaustion (see Table 2, Model 4).

The moderating role of negative affectivity (NA) in the relationship between surface acting and emotional exhaustion is plotted in Fig. 1, which depicts that individuals high on negative affectivity (NA) are more likely to be emotionally exhausted as a consequence of being engaged in surface acting whereas the individuals who are low on negative affectivity (NA), indulging in extensive surface acting does not expose them to greater risk of emotional exhaustion.

Fig. 1
figure 1

Interactive effect of surface acting and negative affectivity on emotional exhaustion among customer services representatives

Full size image

Figure 2 depicts the interactive effect of deep acting and positive affectivity (PA) on emotional exhaustion among customer services representatives. The buffering effect of positive affectivity (PA) on the relationship between deep acting and emotional exhaustion is evident in the figure. Individuals who are high on positive affectivity (PA) are less likely to fall a prey to emotional exhaustion as corollary to deep acting, whereas individuals with low positive affectivity (PA) are at increased risk of getting emotionally exhausted if they indulge in extensive deep acting while regulating their emotions on their jobs.

Fig. 2
figure 2

Interactive effect of deep acting and positive affectivity on emotional exhaustion among customer services representatives

Full size image

Discussion

Our findings indicate that the first hypothesis of this study suggesting surface acting as positive predictor of emotional exhaustion has been supported, however, contrary to the second hypothesis of this study which postulated deep acting as negative predictor of emotional exhaustion, deep acting in fact turned out to be non-significant predictor of emotional exhaustion. According to Grandey (2003), in surface acting, employees experience emotional dissonance because of the incongruity between expressions and inner feelings. This emotional dissonance may lead them to emotional exhaustion. On the contrary, since deep acting by its very definition lessens emotional dissonance by aligning feelings with expressions, deep acting’s relationship with emotional exhaustion should be weaker than the relationship between surface acting and emotional exhaustion and that’s what the present study has exactly found. Our results replicate the existing empirical findings which have consistently demonstrated a significant association between surface acting and emotional exhaustion and a non-significant relationship between deep acting emotional exhaustion (Brotheridge and Grandey 2002; Erickson and Ritter 2001; Grandey 2003; Monaghan 2006; Noor and Zainuddin 2011).

The third and fourth hypotheses of the present study postulated relationships between affectivity and acting mechanisms of emotional labor. These hypotheses were tested through correlations (see Table 1). A significant and direct correlation was found between negative affectivity (NA) and surface acting which supported our third hypothesis. Since individuals who are high in negative affectivity (NA) experience negative emotions more frequently, they may need to hide or fake these feelings more often during their interactions with customers. Thus such employees are more likely to be engaged in surface acting. Our findings are consistent with those of Brotheridge and Grandey (2002), Brotheridge and Lee (2003), Gosserand and Diefendorff (2005), and Monaghan (2006) who reported significant positive correlation between negative affectivity (NA) and surface acting.

Our fourth hypothesis that suggested a positive relationship between positive affectivity (PA) and deep acting, however, has not been supported. Results demonstrated a non-significant albeit positive relationship between deep acting and positive affectivity (PA). Similar findings have been reported by Diefendorff et al. (2005) who equated positive affectivity (PA) with extraversion and found that deep acting was not related to extraversion. Since individuals high in positive affectivity (PA) are predisposed to experience positive emotions such as zeal and cheerfulness more frequently, such individuals may neither need to fake their emotions (surface acting) nor alter their actual feelings (deep acting) during their interactions with customers since the display of their real feelings would mostly be congruent with the organizational display rule demands.

The fifth and sixth hypotheses of this study proposed moderating roles of affectivity in the relationship between acting mechanisms of emotional labor and emotional exhaustion. These hypotheses were tested through hierarchical regression analyses (see Table 2). The results support both of these hypotheses. Negative affectivity (NA) moderated between surface acting and emotional exhaustion and significantly increased the likelihood that high NA employees may fall prey to emotional exhaustion as a consequence of being engaged in surface acting. On the other hand, positive affectivity (PA) moderated between deep acting and emotional exhaustion and it buffered the impact of deep acting on emotional exhaustion. In other words, employees who are high on positive affectivity (PA) are significantly less likely to be emotionally exhausted as a consequence of being engaged in deep acting as compared to their counterparts who are low on PA.

When a high NA individual is asked to display positive emotions, this individual may express such emotions with a greater degree of “acting,” and will be more aware of this effort. Thus, the more the NA employee wants to adhere to the organizational display rule demands, the more emotive dissonance s/he has to experience which may make her/him more vulnerable to emotional exhaustion. Weiss and Cropanzano (1996) argued that affective dispositions of individuals influence the intensity of their emotional responses to work events. Therefore, an employee who is high on NA may respond more negatively to a negative affective event at work than someone low on NA. To this end, in customer service jobs requiring positive emotional displays, employees with high levels of NA may need to perform more emotional labor in order to display the appropriate emotions when negative affective events occur (Grandey 2000). Thus, this increased amount of emotional labor on the part of high NA employees make them more vulnerable to emotional exhaustion.

Following this same line of reasoning, when a high PA employee is asked to display positive emotions during a service transaction, this individual may perform such emotional labor with very little degree of “acting” and hardly recognize the effort of “acting cheerful.” Thus a high PA employee engaged in deep acting would experience little emotive dissonance which in turn reduces the risk of being emotionally exhausted. Furthermore, individuals high on PA may respond more positively to all situations, including ones in which negative affective events occur. Therefore, these individuals may have to perform less emotional labor to display organizationally desired emotions and that’s why they are less prone to be emotionally exhausted.

In case of control variables of the present study, sex and organization were significant whereas job experience was not significant predictor of emotional exhaustion. Women turned out to be more prone to emotional exhaustion as compared to their male counterparts. Hence, our hypothesis no. 7 (a) is supported whereas hypothesis no. 7 (b) was rejected. The practices of gender socialization in Pakistani culture hold logical explanations why women are at an increased risk of being emotionally exhausted. Our society molds women into soft spoken, non–assertive, and docile beings who are expected to be modest in the expression of their truly felt emotions. Women are never supposed to be aggressive or hostile and even if they were actually experiencing some intimidating feelings, they are not allowed to let these feelings openly. This suppression of internal feelings may in turn lead women to more intense experience of emotional exhaustion as compared to their male counterparts. Women might also become more prone to emotional exhaustion because they generally employ emotion-focused coping to regulate the corresponding emotional response (Ptacek et al. 1992) in contrast with men who are more likely to use more problem-focused methods aimed directly at the stressor (Skues and Kirby 1995).

Consistent with recent research (Kowalski, et al. 2010; Sonnentag et al. 2010), our findings indicated that job experience is not relevant to emotional exhaustion. The negative relationship between job experience/age and emotional exhaustion might have spurious on account of survival bias—i.e. those who burn out early in their careers are likely to quit their jobs, leaving behind the survivors who consequently exhibit lower levels of burnout.

Our findings suggest that employees of cellular customer services were at increased risk of being emotionally exhausted as compared to their counterparts in banking sector. Although we have not formulated any hypothesis regarding the organizational differences in emotional exhaustion, yet this finding worth explanation. In Pakistani work milieu, display rule demands, salary, and job security of customer services representatives (CSRs) in banking sector and cellular services are considerably different. Display rule demands in cellular services are much more stringent and explicit as compared to banking sector. CSRs in cellular services are evaluated in terms of their adherence to display rule demands in their formal performance evaluations whereas no such performance criterion is formalized in the majority of banks. This increased level of display rule demands may make CSRs more vulnerable to emotional exhaustion (Adil and Kamal 2012; Martínez-Iñigo et al. 2009) in cellular services. Employees in the cellular customer services have been earning much less as compared to their banking equivalents, while they have to work more extensively. Finally, the employees in telecommunication services may find their jobs less secure, with relatively few fringe benefits in comparison with the banking employees who may find their careers more secure.

Limitations and Suggestions for Future Research

The reliance on self-report, cross-sectional, perceptual measures constitutes perhaps the most serious limitation of the present study as it increases the likelihood of inflating the observed relationships spuriously on account of common method variance. However, the range of correlations in the present study is .02 to .32 (excluding the correlations of interaction terms, see Table 1) and, despite the fact that many of the correlations were significant, none were aberrantly high. Moreover the cross-sectional design of this study did not allow causal interpretation of the findings.

The present study has focused on affectivity as dispositional factor that might influence the relationship between acting mechanism of emotional labor and emotional exhaustion. There could be other potentially moderating variables both in terms of personal dispositions and organizational variables which should be investigated in future studies. Grandey’s model of emotional labor (2000) provides a useful framework in this regard and proposes emotional intelligence and emotional expressivity as relevant dispositional factors besides sex and affectivity. Similarly, organizational variables like perceived organizational support and job autonomy may also play a moderating role between emotional labor and emotional exhaustion. Future studies should therefore consider a combination of both organizational and individual variables in relation to emotional labor and emotional exhaustion.

There are certain other affective variables which are exogenous to job or personality factors. Positive and negative affective events at work are example of such factors which must be studied in relation to emotional exhaustion. It could be that positive and negative emotional events may influence the commitment to display rules and people with various affective dispositions may react differently to these events resulting in differential patterns of emotional labor and varying degrees of emotional exhaustion.

Implications

The present study elucidates certain implications in customer services. The role of affectivity in emotional exhaustion makes the construct very pertinent to recruitment and selection procedures in customer services. It follows from the findings of this study that selection devices used in customer services must incorporate assessment of affectivity, since positive affectivity (PA) may play a buffering role in the relationship between deep acting and emotional exhaustion. Thus high PA provides customer services representatives a distinct shield which may protect them from emotional exhaustion as a consequence of deep acting. On the contrary, high NA employees are in a vulnerable state to emotional exhaustion had they been engaged in surface acting as mode of emotional labor.

In any customer services job, emotions are regulated both by surface acting and deep acting as at times employees may engage in surface acting and at other times they may preferably be indulged in deep acting. Given the dynamic interplay of surface acting and deep acting as two distinct modes of emotional regulation, management of customer services should ensure that the employees selected for typical jobs of customer services should be simultaneously high on positive affectivity (PA) and low on negative affectivity (NA). All other factors being equal, such employees have a competitive advantage over their counterparts since by virtue of their dispositions, they are relatively immune to emotional exhaustion no matter whether they engage in surface acting or choose to indulge in deep acting for meeting the organizational display rule demands.

Conclusion

Overall the study has demonstrated that surface acting is a significant predictor of emotional exhaustion whereas deep acting though positively correlated with emotional exhaustion, did not predict it in regression analyses. Negative affectivity (NA) was significantly correlated with surface acting, whereas deep acting did not significantly correlate with any mode of emotional labor. Finally, negative affectivity (NA) moderated between surface acting and emotional exhaustion such that high NA employees are more likely to be emotionally exhausted. The study has provided evidence for sex differences in emotional exhaustion and suggested a more vulnerable status of female customer service representatives. It also revealed that customer services representatives of cellular services sector were more emotionally exhausted as compared to their counterparts in banking sector. Finally, this research also suggests a buffering role of positive affectivity (PA) by demonstrating that high PA employees are less prone to emotional exhaustion had they been engaged in deep acting as compared to their low PA counterparts.