DO PERSONALITY AND GENDER MITIGATE THE ADVERSE EFFECTS OF EMOTIONAL LABOR? EMPIRICAL EVIDENCE FROM HEALTHCARE DURING COVID-19 PANDEMIC

Purpose: The impact of emotional labor on organizational outcomes is contingent upon numerous factors. Moreover, after the outbreak of the COVID-19 pandemic worldwide, the implications of emotional labour practices, especially among physicians, have become more uncertain. This study aims to address these notions by examining the multigroup moderating effect of social and enterprising personalities and gender on a model of emotional labour. The aim is to determine whether emotional labour affects emotional exhaustion and well-being varies across these groups. Methodology: Serial mediation and multigroup moderation methods were employed to test the proposed hypotheses. Data was collected from 400 physicians across Pakistan. Snowball sampling method was adopted for data collection. Structural equation modelling, including serial mediation and multigroup moderation techniques were employed for data analysis. Main Findings: Physicians with highly social personalities and low enterprising personalities are more likely to experience detrimental effects of surface acting. Genuine emotions improve well-being for high social and enterprising as well as low enterprising personality physicians. No across-group differences exist for deep acting. For gender, females are found to be more susceptible to negative repercussions of emotional labor than their male counterparts. Positive effect of genuine emotions remains somewhat the same for both. Implications/Applications: This study assists healthcare organizations in making the right choices during the recruitment and selection process to choose the most suitable candidates for tasks involving emotional labor, i.e., ones whose personality and gender attributes are aligned with the affective or emotional component of work. This would aid in ensuring doctors’ high psychological and physical well-being at work. Novelty/Originality of this study: This research proposes a model of emotional labour strategies and their effects on physicians' emotional exhaustion and well-being. Multigroup moderating effects of two personality types, i.e., social and enterprising, and gender, are investigated with respect to the proposed model to see if there lie any differences between physicians belonging to various personality or gender groups.


INTRODUCTION
Emotional labor and regulation have become an important part of service interactions. To enhance customer experience, the display of certain emotions is now a formal part of jobs (Gabriel, Acosta, & Grandey, 2015). However, meeting the emotional demands of work is a tiring requirement and may even cause grave problems like depression and anxiety (Yin, Huang, & Lv, 2018). Lack of emotional stability leads to emotional exhaustion and strain among personnel (David et al., 2020). If not addressed, this may escalate to low morale and well-being and high turnover intentions ( Practising emotional labor in fields like medicine carries challenges that are far more severe relative to other occupations. Especially after the outbreak of the COVID-19 pandemic worldwide, the propensity towards emotional instability and emotional exhaustion has further increased (Sinnott, Georgiadis, Park & Dixon-Woods, 2020). This is particularly true for doctors and physicians who are not only working in high-risk environments but are also more likely to experience mental and emotional exhaustion (Shaukat, Ali, & Razzak, 2020). One must have emotional or affective resources to combat the adverse effects of emotional labor. Here, the role of a physician's own personality and gender may play a key role in determining how emotions affect them during the pandemic.

Impact of COVID-19 Pandemic on Healthcare Sector
In addition to the existing challenges associated with practising emotional labor in healthcare, the spread of the COVID-19 virus presents some major hurdles for healthcare personnel. From patient care and prolonged working hours to the high al., 2019; Yu, Kim, & Qu, 2019). The authentic display, also termed genuine emotions, helps reduce emotional exhaustion by eliminating faking and pretense from service interaction (Hulsheger & Schewe, 2011;Picard, Cossette, & Morin, 2018).

Hedonic, Eudaimonic, and Physical Well-Being
Well-being is one of the most important outcomes associated with job demands, emotional exhaustion, and turnover intentions (Heijden, Mahoney, & Xu, 2019;Rostamabadi et al., 2019;Wolter et al., 2019). Two types of well-being are widely discussed in the literature. First is hedonic well-being, which involves the presence of happiness and pleasure and the absence of anything unpleasant (Diener, Lucas, & Oishi, 2018). Second is eudaimonic well-being which includes long-term life satisfaction, contentment, meaningfulness, and purpose in life (Ryff, 2018). Eudaimonic well-being is considered longlasting relative to hedonic well-being (Joshanloo, Jovanovic, & Park, 2020). Physical well-being is also paramount to psychological well-being (Gaetano, 2016 The impact of emotional labor on well-being depends upon the EL strategy being used. Surface acting is often found to reduce well-being (Ryan & Deci, 2017 Applying the JDR and COR theories in the healthcare context, our study hypothesizes that both surface and deep acting are likely to increase emotional exhaustion and cause low well-being among physicians during the pandemic. Those who opt for displaying genuine emotions at work will experience low emotional exhaustion and high well-being.
H1a: Emotional Exhaustion mediates the relationship between Surface Acting and low Hedonic Well-being.
H1b: Emotional Exhaustion mediates the relationship between Surface Acting and low Eudaimonic Well-being.
H1c: Emotional Exhaustion mediates the relationship between Surface Acting and low Physical Well-being.
H2a: Emotional Exhaustion mediates the relationship between Deep Acting and low Hedonic Well-being.
H2b: Emotional Exhaustion mediates the relationship between Deep Acting and low Eudaimonic Well-being.
H2c: Emotional Exhaustion mediates the relationship between Deep Acting and low Physical Well-being.
H3a: Emotional Exhaustion mediates the relationship between Genuine Emotion and high Hedonic Well-being.
H3b: Emotional Exhaustion mediates the relationship between Genuine Emotion and high Eudaimonic Well-being.
H3c: Emotional Exhaustion mediates the relationship between Genuine Emotion and high Physical Well-being.

Social and Enterprising Personality
A good fit between job and personality not only produces a positive effect on job and organisation (Holland, 1997; Lee, Kim, Kim, & Kim, 2017; Phan & Rounds, 2018) but also improves job satisfaction (Törnroos, Jokela, & Hakulinen, 2019) and increases the chances of career success (Su & Nye, 2017). This study has chosen the personality typology developed by Holland (1959). Out of six, our study has chosen social and enterprising personality types since both the types are theorized to be effective in social settings, teamwork, and taking the initiative (Holland, 1959;1997). Based upon JDR and COR theories, workers have varying levels of affective resources. Simply put, not every worker would deal with the job's emotional challenges the same way because of different personalities, level of resolve, patience, and psychological capital. Thus, it is proposed that individuals with different personalities will react differently to the emotional component of the job. For some, it may be naturally easy to project the required emotions, unlike others who find it more challenging. Former may have little problem conserving the resources and balancing them to meet job demands, thereby not experiencing emotional exhaustion and low well-being. In contrast, the latter group is more likely to experience exhaustion and subsequent low well-being levels.
Henceforth, keeping in mind the individual differences concerning resilience, emotional composure, and emotional regulation, this study would test the multigroup moderating effect of social and enterprising personalities on the research model. Workers with social personalities are effective in occupations where interaction with people and relationship building are vital to the task's nature. Similarly, workers with enterprising personalities are self-confident and effective in managing and leading teams (Holland, 1959;1997). Since both these personalities engage more in social and interactive activities, it is reasonable to believe that groups of workers with high versus low degree of social and enterprising H4a: Impact of emotional labor among physicians with high social personality differs from those with a low social personality.
H4b: Impact of emotional labor among physicians with high enterprising personalities differs from those with low enterprising personalities.

Gender as a Multigroup Moderator
In addition to personality, gender would also be tested for its multigroup moderating effect on the model. Research has shown a significant differential effect of gender on the outcomes of emotional labor. For instance, male workers are less affected by the negative effects of emotional labor than females (Cottingham, Erickson, & Diefendorff, 2015). Females feel stressed and exhausted due to practising emotional labor relative to their male counterparts (Han et al., 2018). Both males and females respond differently to emotional job demands at work and tackle the implications of emotional labor differently (Simpson & Stroh, 2004). Based on this premise, this study proposes that males and female physicians would also differ when it comes to the effects of emotional labor during the ongoing pandemic, including emotional exhaustion and wellbeing.

H5:
There are significant differences between male and female physicians concerning the impact of emotional labor.

SAMPLE AND METHODS
Data was collected from 400 physicians working in large hospitals in Islamabad/Rawalpindi, Lahore, Peshawar, Karachi, Bahawalpur, and other cities where the online survey could reach. The Snowball sampling method was adopted to access maximum number of respondents. Structural equation modelling, including serial mediation and multigroup moderation, is employed for data analysis. Emotional labor (surface and deep acting) is measured using 6-item Emotional Labor Scale developed by Brotheridge and Lee (2003). Genuine emotion is measured using a 3-item scale developed by Diefendorff, Croyle, and Gosserand (2005). Scale for emotional exhaustion is adapted from Maslach Burnout Inventory (Maslach & Jackson, 1981). For measuring hedonic well-being, the scale is adapted from the hedonic well-being instrument developed by the world health organisation (Sell & Nagpal, 1992). Nine items measuring the Eudaimonic well-being scale are adapted from the instrument developed by Ryff (1989). Seven items measuring physical well-being scale are adapted from the "Perceived Well-Being Scale" developed by Recker and Wong (1984). Lastly, social and enterprising personalities are measured through Holland (1997) personality typology (6 items). AMOS version 23 is used for all statistical analysis.

Confirmatory Factor Analysis
A confirmatory factor analysis was run to examine all scales' validity and reliability in the study. According to Hair et al. (2009), factor loadings for any item must not be below 0.5. However, factor loadings for ewb2, ewb3, ewb4, ewb9, pwb3, and pwb7 were below 0.5. Therefore, they were removed from the model. The final measurement model is shown in Figure  2, with all items loading above 0.5. Fit indices in There is no multicollinearity problem in the model since covariances among all latent variables are less than 0.7. Multicollinearity can cause problems in regression analysis later; thus, it is advised to merge those variables which might be highly correlated (Wanzer et al., 2019). However, the merging should be supported by theory (Dormann et al., 2012). Since both social and enterprising personality types are highly correlated and measured using the same global scale, both are examined as one construct in CFA.

Structural Model
The structural model highlights the relationships between independent and dependent variables in the model. We have built a latent structural model in figure 3, inclusive of all observed and latent variables. Results reveal that surface and deep acting increase emotional exhaustion. Conversely, genuine emotions reduce emotional exhaustion. Unlike expected, both surface and deep acting have a significant but positive relationship with hedonic, eudaimonic, and physical well-being. Deep acting only has a significant impact on hedonic well-being. As presumed, genuine emotion is found to have a positive effect on well-being dimensions. Regression weights with significance values for all direct paths are shown in table 4. It should be noted here that the study is primarily interested in mediated effects, which are assessed later.

Serial Mediation
The study seeks to examine the sequential impact of emotional labor strategies on well-being dimensions. The mediator is emotional exhaustion. Mediating effects are calculated using the latent mediation estimand developed by James Gaskin. A bootstrap sample of 1500 is taken to calculate the indirect effects. Results shown in table 5 indicate that surface and deep acting reduce well-being if the relationship is mediated by emotional exhaustion. This satisfies our initial presumption. However, the mediation is only significant for surface acting but insignificant for deep acting. Thus, H1a, H1b, and H1c are supported, and H2a, H2b, and H2c are rejected. As hypothesized, genuine emotion increases well-being via reduced emotional exhaustion. This again confirms the stance that genuine emotional display reduces strain and improves wellbeing. This provides support for H3a, H3b, and H3c.

Moderated Mediation-Multigroup Effect of Social Personality
For measuring the multigroup effect of social personality type, firstly, a global test is run to check if the two groups, i.e., physicians with high social personality and physicians with a low social personality, differ on a model level. The test is run using the command of multigroup analysis in AMOS. Table 6 shows the moderator's global effect through a p-value smaller than 0.05 that there is a significant difference between the two groups when it comes to the impact of emotional labor on well-being. Hence the hypothesis H4a is supported. The local effects highlighted in table 7 point out that when physicians with high social personality engage in surface acting, experience low well-being via emotional exhaustion. The same group experiences low well-being via emotional exhaustion if they display genuine emotions. For physicians with a low social personality, almost all mediation paths are insignificant, signifying no relationship. No difference was found between groups with respect to the impact of deep acting, except on hedonic well-being. The mediation path is significant for the latter group, revealing that physicians with low social personality experience low well-being via emotional exhaustion if they employ deep acting.

Moderated Mediation-Multigroup Effect of Enterprising Personality
For conducting multigroup moderating effect of enterprising personality, a similar method was adopted as for social personality. Table 8 shows the moderating effect of enterprising personality on the entire model. P-value less than 0.05 concludes a significant difference between physicians with high enterprising personality and physicians with low enterprising personality, with respect to the impact of emotional labor strategies on well-being dimensions. Hence, hypothesis H4b is supported. Local effects across the two groups are depicted in Table 9. It shows that the negative impact of surface acting on hedonic and physical well-being is significant for the group with a low enterprising personality, while for a high enterprising personality, the impact is insignificant. For genuine emotions, no differences exist in hedonic well-being. Both groups experience high hedonic well-being when they employ genuine emotions. However, only the group with high enterprising personality experiences high eudaimonic well-being with genuine emotions. Similarly, only the group with a low enterprising personality experience high physical well-being with genuine emotions. Finally, there are no local-level differences between the two groups with respect to the impact of deep acting on well-being.

Moderated Mediation-Multigroup Effect of Gender
The third multigroup moderator in this study is gender. Test for the global effect of gender on the model in table 10 reveals a significant difference between male and female physicians concerning the impact of emotional labor strategies on wellbeing dimensions. This leads to the support of hypothesis H5.

DISCUSSION
This research aimed towards understanding the behaviour of physicians while practising emotional labor during the COVID-19 pandemic. Differential roles of social and enterprising personalities and gender were examined to assess whether these factors play any role in accentuating or mitigating the impact of emotional labor. Structural equation modelling technique was used to assess the hypothesized relationships. Furthermore, serial mediation analysis and multigroup moderating tests were also used to study in-depth associations among the variables and the global and local differences. Multigroup analysis further highlights the role of personality and gender, upon which emotional labor is contingent. Global tests results reveal significant differences between high versus low social personality, high versus low enterprising personality, and male versus female when it comes to the effects of emotional labor on emotional exhaustion and well-being dimensions. For further in-depth analysis, local tests were run for each moderator. Results reveal that physicians with high social personality are more at risk of experiencing low well-being if they engage in surface acting (David, Shoss, Johnson, & Witt, 2020). The same group is also more likely to experience high well-being if they employ genuine emotions. The reason for this may be those social personalities may be expected to engage in emotional acting, which may take a toll on their well-being. Especially during a pandemic when social distancing is crucial, surface emotional acting is more likely to compromise doctors' well-being. That is why they feel uplifted and experience high well-being when they get to display genuine and authentic emotions. No significant local effects of deep acting on well-being dimensions were found.
Results for the enterprising personality group were somewhat different. Physicians with low enterprising personalities are more vulnerable to low hedonic and physical well-being if they engage in surface acting (Chen, Chang, & Wang, 2019). This might be because an enterprising personality entails leadership and confidence, which this group lacks. Thus, this group would find surface acting more tiring and challenging. For doctors displaying genuine emotions, no difference was found between high versus low enterprising personality with respect to hedonic well-being. However, a high enterprising personality leads to higher eudaimonic well-being, whereas a low enterprising personality leads to higher physical wellbeing when genuine emotions are displayed. The former effect is understandable as doctors with leading skills are naturally and genuinely effective in leadership roles. Thus they need not display emotional acting and therefore experience high eudaimonic well-being. Similarly, doctors with low enterprising are not naturally prone to leadership roles. Thus displaying genuine emotions may compromise their eudaimonic well-being but improve their physical well-being. No significant local effects of deep acting on well-being dimensions were found.
Finally, the results from the moderating effect of gender show that female doctors experience low hedonic, eudaimonic and physical well-being when engaging in surface acting. Conversely, no significant effect of surface acting was found on wellbeing dimensions among male doctors. Both male and female doctors experience high well-being levels when they display genuine emotions. However, the impact of genuine emotions on eudaimonic well-being was not significant among male doctors. No significant local effects of deep acting on well-being dimensions were found concerning gender.