Emotional labor is a means for employees to manage their emotions and to express only those requested by their organizations (Hochschild, 1983). Notably, in a service-oriented sector, employers often ask employees to display certain favorable emotions. To exhibit the required emotional responses, employees generally vary on two dimensions of emotional labor – surface acting or deep acting – to regulate their emotions. Surface acting is a person’s untruthful appearance of a prescribed emotion without attempting to change their authentic feelings, eventually leading to emotional exhaustion (Martínez-Iñigo et al., 2007). Deep acting refers to demonstrating a person’s authentic emotions (Ashforth and Humphrey, 1993; Grandey, 2003; Groth et al., 2009), and is often found unrelated to employee stress or physical well-being (Grandey, 2003).

Working with people is likely to involve a significant amount of emotional labor. The characteristics and context of healthcare render emotional labor inevitable for healthcare professionals. Healthcare workers use emotional labor to create and cultivate bonds with patients or clients (Fouquereau et al., 2019), in particular when they wish to reduce patients’ anxiety or need to perform an unpleasant procedure (Martínez-Iñigo et al., 2007). Gray (2010) advocates that emotional labor is a core component of nurses’ role in making patients feel safe and comfortable. The surge of research on emotional labor has demonstrated its pervasive effects on organizational outcomes (Grandey and Sayre, 2019). Previous studies have demonstrated the direct effects of emotional labor (or emotional dissonance) on organizational outcomes (Kammeyer-Mueller et al., 2013; Deng et al., 2017; Grandey and Sayre, 2019) and employee well-being (Hintsanen et al., 2014; Wagner et al., 2014; Diestel et al., 2015; Grandey and Sayre, 2019). However, relatively few studies have investigated the mechanism between emotional labor and work outcomes (Wagner et al., 2014; Deng et al., 2017). Notably, Deng et al. (2017) explored the mechanism of ego depletion between emotional labor and coworker harmful behavior, whereas Wagner et al. (2014) examined the mechanism of anxiety between surface acting and emotional exhaustion and insomnia.

The present study focuses on the impact of emotional labor on sleep, and how sleep problems can further influence job satisfaction and job performance. In particular, we argue that emotional labor is a predictive factor of sleep, which indirectly influences work outcomes. The current literature also proposes that personality may be related to sleep (Hintsanen et al., 2014; Kim et al., 2015; Allen et al., 2016) and emotional labor (Yagil and Medler-Liraz, 2017). Drawing on these studies, this research integrates the theories of emotional labor, sleep, and personality. We posit the moderating role of personality traits and examine how they adjust the relationship between emotional labor and sleep. In doing so, we highlight the need for a more comprehensive investigation of the topic (Kammeyer-Mueller et al., 2013; Grandey and Gabriel, 2015).

Accordingly, this study aims to extend the theoretical and empirical approaches to emotional labor by delineating the degree and mediating effect of sleep problems on the relationship between emotional labor and job satisfaction and performance. We propose that some people replenish their psychological resources more quickly than others, which can be attributed to variances in personality traits. Thus, we theorize that employee personality traits (i.e., conscientiousness, neuroticism, agreeableness, extraversion, and openness) moderate the relationship between emotional labor and sleep problems. Identifying these moderators will help offer insights on the processes of emotional labor.

Many service-oriented jobs require specific displayed affects from employees, that is, the expression of positive affect (e.g., cheerfulness) and the suppression of negative affect (e.g., annoyance) (Groth et al., 2009). Negative effects have been negatively associated with falling asleep, maintaining sleep, and reinitiating sleep (Fortunato and Harsh, 2006). When faced with affective requirements, the emotional regulation strategies adopted by employees can impact their sleep patterns. For instance, Wagner et al. (2014) found surface acting conducted at work to be associated with insomnia experienced during nighttime.

On the other hand, employees experiencing emotional requirements can decide to fulfill the displayed rules by making efforts to align real and required emotions. However, excessive efforts to regulate one’s emotions can lead to ego depletion (Baumeister et al., 1998) and feelings of exhaustion, which negatively impacts sleep quality (Baglioni et al., 2010). Incongruence between felt and required emotions when surface acting creates tension and emotional dissonance (Pugh et al., 2011). Individuals who regularly endure stress associated with surface acting are more prone to depression and anxiety, which decrease job performance and contribute to burnout (Weaver et al., 2019). Such dissonance (i.e., surface acting) results in sleep problems, making it increasingly difficult to control emotions and thus, triggering a vicious cycle between emotional labor and poor sleep (Palmer and Alfano, 2017).

Sleep quality is crucial for optimal brain function (Saper et al., 2005). High sleep quality stabilizes the cerebral metabolic rate and ensures adequate resources for the prefrontal cortex (Hofmann et al., 2012). In contrast, sleep loss is likely to affect psychological well-being (Diestel et al., 2015) and organizational outcomes, including safety (Weaver et al., 2019), job satisfaction (Scott and Judge, 2006), and performance (Amirian, 2014). Exploring the relationships among insomnia, emotions, and job satisfaction for 45 employees, Scott and Judge (2006) discovered that insomnia is negatively correlated with job satisfaction. Tomasko et al. (2012), however, demonstrated no difference in the performance or learning of surgical tasks between sleep-deprived medical students and the control group, although the former reported an increase in total subjective mental workload. In a study of 171 nurses and 75 junior and senior business students, Christian and Ellis (2011) found that sleep problems reduce self-control and increase hostility, resulting in heightened workplace deviance.

Deep acting produces a natural and genuine change in emotions, and the individual engaging in it may be considered more trustworthy by customers, clients, and patients (Groth et al., 2009). Unlike surface acting, deep acting does not consume as many psychological resources (Goldberg and Grandey, 2007), given the lower levels of incongruence between felt and displayed emotions (Cheung et al., 2018; Fouquereau et al., 2019). Past evidence suggested that stronger deep acting is associated with lower cognitive exhaustion (Xanthopoulou et al., 2018). Lower cognitive exhaustion is related to better sleep (Hülsheger and Shewe, 2011) and a reduced need for recovery at the end of workdays (Xanthopoulou et al., 2018). Thus, we can conclude that deep acting is negatively associated with sleep problems. Integrating the findings of these studies, we hypothesize the following:

Individuals show various responses and employ different strategies to cope with stress, and thus, we expect work outcomes to vary by individual personality. Individuals who regulate their emotions through deep acting may remain connected to their core values and beliefs, resulting in an alignment of their felt and shown emotions. By contrast, individuals who resort to surface acting may show incongruence between felt and shown emotions. This study delineates the way emotional labor strategies combine with specific employees’ personality traits to influence their sleep quality and, in turn, job outcomes to understand the mechanism between emotional labor, sleep, and work outcomes.

Personality factors have been linked to the development of sleep problems. People with high conscientiousness live longer because they participate in more healthy activities, which include regular exercise, healthy diets, avoidance of substance abuse, and fewer risk behaviors (Bogg and Roberts, 2004). Using different measurement scales, several studies have found that higher conscientiousness is associated with improved daily sleep quality and sleep efficiency (O’Connor, 2014; Kim et al., 2015). Highly conscientious individuals practice healthy lifestyles that help regulate their sleep quality. Therefore, when individuals high in conscientiousness engage in surface acting, their conscientiousness traits have a protective effect between surface acting and sleep problems. Thus, we hypothesize the following:

Individuals with neuroticism are characterized as more anxious and stressed during interactions, and thus, are less likely to experience positive emotions (Kiffin-Petersen et al., 2011). Such individuals may need to alter their feelings to maintain positive displayed emotions by engaging in more surface acting or deep acting. Studies have shown that neurotic individuals engage in more surface acting (Diefendorff et al., 2005; Kiffin-Petersen et al., 2011). Most individuals with high neuroticism tend to exhibit fewer health-promoting behaviors (Booth-Kewley and Vickers, 1994) and are more susceptible to and more likely to report somatic complaints (Watson and Clark, 1994) possibly related to poor sleep quality (Williams and Moroz, 2009; Cellini et al., 2017). Performing affective requirements and disguising real emotions will harm individuals’ sleep activities. This could exacerbate sleep problems among individuals with high neuroticism who are already more susceptible to negative emotions than individuals with low neuroticism. Therefore, we hypothesize the following:

Extraversion is characterized by warmth, cheerfulness, or vigor and is used to express enthusiasm or interest. Extraverts are more open to social influences, suggesting that they may be more willing to engage in the emotions required by their organizations (Wilson, 1981), and experience more positive emotions (Watson and Clark, 1994; Judge et al., 2009). Bono and Vey (2007) found that individuals who score high on extraversion (rather than low) respond better to organizational demands for positive emotions through deep acting. High extraverts can easily engage in deep acting as they naturally have mood traits that align with the required emotions (Diefendorff et al., 2005). Judge et al. (2009) further demonstrated that deep acting generates more positive reactions for extraverts.

Extraversion is associated with shorter sleep latency (Williams and Moroz, 2009) and better sleep quality (Hintsanen et al., 2014; Allen et al., 2016), but not with sleep duration (Williams and Moroz, 2009). By provoking desired feelings (Chi et al., 2011) and a good-faith attempt to produce internal emotions, deep acting shows less of an influence on sleep. In combination, high extraversion and deep acting engender more positive affect, thereby promoting the expression of more naturally felt emotions – and in turn, reducing sleep problems. We therefore hypothesize that extraversion moderates the negative relationship between deep acting and sleep problems.

Agreeableness is the tendency to be compassionate, kind, and considerate toward others (John and Srivastava, 1999). Agreeableness is congruent with the positive requirement of service jobs and is positively related with deep acting (Diefendorff and Richard, 2003; Kammeyer-Mueller et al., 2013). Findings on the relationship between agreeableness and sleep problems are mixed. Several studies find no relation (Williams and Moroz, 2009; Stephan et al., 2018), while others show that lower agreeableness is related to fewer sleeping hours (Randler, 2008) and higher sleep deficiency (Hintsanen et al., 2014; Cellini et al., 2017). Employing surface acting to regulate emotions increases sleep problems. The interplay of surface acting and low agreeableness exacerbates the situation (Törnroos et al., 2012). Thus, we hypothesize the following:

Individuals with openness can be described as imaginative, curious and open-minded toward new experiences. Studies have been inconclusive about the relationship among openness, emotional labor, and sleep. Though several studies report that openness is not associated with sleep quality (Hintsanen et al., 2014; Cellini et al., 2017; Stephan et al., 2018), Allen et al. (2016) found that higher openness is related to sleep difficulties. The abovementioned studies, however, do not examine the moderating effect of openness. Openness is generally considered to be an intrapsychic trait rather than an interpersonal trait (Costa and McCrae, 2012). Individuals high on openness to experience generally practice more deep acting than surface acting and, thus, are less emotionally exhausted in stressful situations (Zapf, 2002) and do not burn out easily (Bakker et al., 2002; Divinakumar et al., 2019). We, therefore, hypothesize the following:

Figure 1 summarizes our hypotheses.

Table 1 presents the means, standard deviations, and bivariate correlations for this study’s variables. First, a confirmatory factor analysis is performed to validate each measurement model. Table 2 shows the baseline measurement model (Model 1) and that all study variables loaded on their respective factors, except sleep problems (observed variable), have a good fit [χ² (d.f. = 224) = 863.164, GFI = 0.925, CFI = 0.926, TLI = 0.909, RMSEA = 0.056, and SRMR = 0.052]. Second, the one-factor model (Model 2) wherein all the variables are loaded on one global factor report poor fit [χ² (d.f. = 252) = 5371.357, GFI = 0.641, CFI = 0.384, TLI = 0.325, RMSEA = 0.153, and SRMR = 0.134]. Finally, to evaluate the discriminant validity of the latent constructs, we compare the baseline model with the one-factor model. The baseline model has a significantly lower χ² value, yielding an excellent fit [Δχ² (d.f. = 28) = 4508.19, p < 0.001]. This evidences the discriminant validity of our measures for the latent constructs. We also computed average variance extracted (AVE) with the squared correlation and find the square root of the AVE for each latent variable exceeds the correlations between the constructs. The discriminant validities are then established (Table 1).

Our findings extend the emerging body of literature on emotional labor in several ways. First, we propose a mechanism that explains how job satisfaction and performance are influenced by emotional labor through sleep problems, which is prevalent in the working population. Our results indicate that only surface acting influences sleep problems, through which it further impacts job satisfaction and performance. This is consistent with meta analytic studies from Hülsheger and Shewe (2011) and Kammeyer-Mueller et al. (2013), surface acting is generally negatively related to impaired well-being, job attitude, and job performance, whereas deep acting has weak or no relations with well-being. Requiring employees to display emotions at work that are far from their genuine emotions may exhaust employees’ mental resources, stimulate negative emotions, and influence sleep quality. Since the psychological environment influences biological outcomes, such unpleasant emotional experiences during work are reported to have an influence on sleep (Greenberg, 2006). Supervisors who fail to recognize the influence of surface action on sleep are likely to be less prepared to tackle organizational problems (e.g., injuries, unethical behavior, and job performance) that plague sleep-deprived emotional laborers (Wagner et al., 2014).

Second, we found that the effects of surface acting on sleep problems are more pronounced among people with high (as opposed to low) neuroticism and low (as opposed to high) agreeableness for the following reasons: (a) surface acting involves suppressing, amplifying, or faking expressions (Hochschild, 1983; Grandey, 2003) which increases the probability of sleep problems, (b) individuals with high neuroticism are more likely to report fewer health-promoting behaviors (Booth-Kewley and Vickers, 1994), higher insomnia (Williams and Moroz, 2009), and more emotions (Watson and Clark, 1994), and (c) people with low agreeableness dislike being told what to do as this will pressure them into being inauthentic with their emotions, thereby developing sleep problems as a result (Hintsanen et al., 2014). Further, individuals with high neuroticism or low agreeableness develop more sleep problems when performing surface acting, whereas those with low neuroticism or high agreeableness remain relatively unaffected. Overall, the interactions between surface acting and high neuroticism or low agreeableness aggravate the adverse effects of surface acting on sleep problems.

Although some studies have shown higher extraversion to be related to better sleep quality (i.e., Gray and Watson, 2002; Williams and Moroz, 2009; Hintsanen et al., 2014; Allen et al., 2016), we did not find evidence in support of this relationship. In contrast to our hypothesis, our results suggest that extraversion acts as a moderator on the negative relationship between deep acting and sleep problems only among individuals with low extraversion. One possible explanation for such effect may be that people low on extraversion are better at conciliating gaps between the authentic and the required [self], or in this case “emotions,” as they may hold more experience doing so in their day-to-day lives to maintain a balance between socializing with the outer world and recharging via focusing on their internal thoughts and authentic feelings. Moreover, taking from sleep theory, we suspect that the effect of high extraversion on deep acting and sleep problems may be mitigated by the greater susceptibility of extraverts to experience emotional injury during social contexts, thereby creating stress and subsequent sleep problems after having engaged in social activities (Rupp et al., 2010). Overall, the effect of extraversion on sleep quality remains inconclusive, as other studies have also been unsuccessful in finding evidence to support such effect (Duggan et al., 2014; Kim et al., 2015; Huang et al., 2016; Cellini et al., 2017). Future studies are encouraged to continue exploring the moderating role of extraversion between deep acting and sleep problem.

Understanding the associations between personality traits and sleep can be useful in targeting those at an increased risk of sleep problems. Further, the findings can be particularly meaningful when recruiting personnel for jobs in which the lack of vigilance may have serious consequences. Individuals should consider their personality when selecting jobs to foster the person-job fit and person-organization fit (Holland, 1985). This may improve emotional labor and its impact on job satisfaction and performance. In sum, this study demonstrates that three out of five personality traits result in interactive effects by emotional labor on sleep. Examining personality traits and emotional labor contributes to the knowledge base necessary to map phenotypes for individual variations in sleep.

This study is an extension of the efforts to explain the mechanisms of emotional labor, job satisfaction, and job performance, a line of research that is becoming increasingly crucial given the high costs related to sleep problems and work-related outcomes in healthcare. It provides a distinct boundary condition, a mediating model, which incorporates sleep problems during work as a critical factor connecting emotional labor with job satisfaction and performance. The results suggest that being sleep deprived at work may negatively affect job performance and satisfaction. In particular, certain traits such as high neuroticism and low agreeableness aggravate the effects of surface acting on sleep problems, whereas low extraversion facilitates the relationship between deep acting and low sleep problems.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving human participants were reviewed and approved by the Institutional Review Board of the Chi-Mei Medical Center (IRB No. 10109-001). The patients/participants provided their written informed consent to participate in this study.

S-CY: conceptualization, data curation, study design, methodology, and manuscript draft. S-HC: conceptualization, data curation, and manuscript draft. K-SY: data collection and formal analysis. WC: data collection and project administration. TW: review and editing. All authors contributed to the article and approved the submitted version.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.