The Art of Emotional Labor: Balancing Authenticity and Performance in Public Sector Teaching

1. INTRODUCTION

It is crucial to understand the emotional labor involved in teaching, especially in today's complex and highly developed educational setting. Teachers frequently manage their feelings when interacting with students, parents, and colleagues, which impacts their performance and job satisfaction (Brundin & Languilaire, 2022). The emotional demands of teaching, coupled with the profession's challenging and often undervalued nature (Zhang & Zhu, 2008), examine the need to understand how emotional work affects teaching outcomes and to develop strategies to support teachers in addressing these emotional challenges.

Emotion management has become a crucial concern within human resource management, especially in today's highly competitive professional environment. The necessity of managing emotions within human resources presents significant challenges in this context (Brundin & Languilaire, 2022). Teaching requires emotional regulation during interactions with students, parents, and colleagues. Faculty members must understand, regulate, and observe their emotional responses to assess how these responses influence their behavior and how they impact their interactions. Recognizing appropriate emotional conduct in the workplace highlights the importance of effective emotional management.

The teaching profession specifically requires proficient emotional regulation to enhance academic achievement and cultivate a positive learning atmosphere. Moreover, educators are frequently overburdened and undervalued (Zhang & Zhu, 2008), with heightened emotional labor (Dreer, 2021). Larrivee (2012) asserted that "teaching is emotional labor" (p.37) because it involves a distinctive form of emotional exertion that differentiates it from other careers; it generally involves highly intense, continuous, and prolonged interactions with students. Understanding teachers' emotional states is equally vital, as they can provoke emotional responses in their students. Lawless (2018) referred to the emergence of emotional labor within academia as "academic labor."

Several scholars have extensively studied the idea of emotional labor under "emotion regulation." As Hochschild (1983) explains, this is "the management of feelings to create a publicly observable facial and bodily display" (p. 7) through various techniques, including surface acting, deep acting, and genuine emotions. Employees often hide and control their emotions, considering it a job requirement because almost every job in the service industry requires a positive display of emotions, such as being friendly or happy, while suppressing negative emotions (either verbally or through facial expressions) like sadness or anger (Brotheridge & Grandey, 2002). As a result, this emotional management becomes an essential part of their work and workplace success, heavily influenced by their emotions.

Ashforth and Humphrey (1995) found that emotions are an inseparable part of the workplace, as employees' emotional responses play a critical role in shaping their job performance. They significantly influence their job performance, highlighting the importance of emotionality over rationality. A person's emotionality encompasses their tendency to experience specific emotions with varying intensities and durations. It also covers an individual's preferences for displaying or suppressing certain emotions. Performance problems can arise for employees who perform emotional labor at their workplace (Mesmer-Magnus et al., 2012). The present study aims to understand how job performance levels can be increased or decreased through the use of emotional labor strategies.

Self-efficacy constitutes a vital factor in attaining superior job performance among employees (Muliati et al., 2022). As articulated by Bandura (1982), "it is a personal judgment of how well one can execute courses of action required to deal with prospective situations" (p. 122). Beliefs related to self-efficacy can significantly influence behavior, motivation levels, and the outcomes of tasks. The primary focus of this study is on generalized self-efficacy, which "captures differences among individuals in their tendency to view themselves as capable of meeting task demands in a broad array of contexts" (Chen et al., 2001, p. 63). These attributes are not confined to specific contexts but are instead generalized across various situations.

Similarly, self-efficacy beliefs are crucial for employees' emotional displays in the workplace. For example, Sloan (2014) found that public service workers often perform surface acting to complete tasks. When their self-efficacy levels were high, they mitigated the adverse effects of intense surface acting in their jobs. However, a clear understanding of its role remains elusive, making this study essential for grasping the mediating role of self-efficacy.

Researchers often focus on the negative impacts of emotional labor (EL) like emotional exhaustion, burnout, or stress (Sulakkana & Weerasinghe, 2020; Sloan, 2014; Grandey, 2003), neglecting its potential benefits, such as improved job performance, job satisfaction, and employee well-being. Humphrey et al. (2015) confirmed that managing the effort to display positive emotions is possible. However, the strategies to regulate positive emotions and the necessary conditions for performing EL can be detrimental. Unlike most studies on EL and job performance that rely on peer or supervisor evaluations, this research solely examines individual job performance of teachers, assessed through three dimensions: task performance, contextual performance, and counterproductive work behavior.

To the best of researchers' knowledge, studies in this area have been fragmented. No research has examined self-efficacy as a mediating factor between emotional labor and job performance. Emotions in the service sector are a central focus of contemporary research. Investigations have explored this phenomenon across various contexts, including healthcare professionals (doctors and nurses), hairstylists, bus drivers, and other service providers (Yang & Chen, 2021). However, scholars like Hagenauer and Volet (2014), Mendzheritskaya and Hansen (2019) emphasize the need to investigate emotion regulation and emotional challenges in higher education. This study focuses on the positive effects of emotional labor, examining how emotional labor strategies (deep acting, surface acting, and genuine emotions) influence the job performance of faculty in the public sector. Furthermore, the research aims to illuminate the role of self-efficacy as a mediating variable between emotional labor and job performance, addressing the limited literature on this topic. This study seeks to bridge this gap, as self-efficacy is crucial in teaching, enabling instructors to effectively convey lectures, assess student engagement, and achieve goals while performing emotional labor.

This study investigates the impact of emotional labor strategies on the performance of university teachers in Lahore, Pakistan. Emotional demands influence performance, yet this area remains understudied in higher education. Most research has focused on healthcare and customer service (Brotheridge & Grandey, 2002; Sloan, 2014), leaving classroom settings largely unexplored. This research analyzes surface acting, deep acting, and authentic emotions, with self-efficacy as a mediator. It enhances understanding and suggests ways to improve performance, advancing the discourse on emotional labor. The study extends the literature by comprehensively investigating emotional labor strategies in public-sector university teaching. Introducing self-efficacy as a moderating variable presents a new perspective on how educators' beliefs affect their emotional labor and performance. This approach deepens the understanding of emotional labor and practical implications for interventions that help teachers manage their emotional workload. Using a sample of public sector university teachers provides a unique perspective on how emotional labor impacts job performance in higher education.

Literature Review

Emotional labor, a critical component within service and people-oriented professions, influences organizational expectations. As defined by Morris and Feldman (1996), emotional labour (EL) encompasses 'the effort, planning, and control required to display emotions that conform to these expectations during interpersonal interactions' (p. 987). Employees engaged in such roles are socialized into specific feeling rules that dictate how they 'should' feel when interacting with customers, clients, or patients (Thoits, 2004). Although this process is essential, it may also result in a sense of alienation as the organization supervises their emotional expressions.

Three strategies for emotional labor encompass surface acting, deep acting, and authentic emotional expression. Surface acting transpires when 'emotions are intensely felt internally, yet the employee suppresses, amplifies, or fabricates their outward expression' (Hochschild, 1983). Conversely, deep acting entails 'employees endeavoring to conform to emotional norms and genuinely evoke the desired feelings during their interactions'. Deep acting is often more advantageous because it fosters emotional congruence, whereas surface acting results in emotional dissonance (Grandey, 2003). Employees engaged in deep acting immerse themselves in cognitions and processes that cultivate these feelings, thereby enabling genuine expression (Humphrey et al., 2015). Ashkanasy and Humphrey (2011) observed that authentic emotional display aligns with organizational display rules. For instance, a hairdresser may enjoy engaging with clients, thereby allowing their emotions to conform to organizational display norms.

Hochschild (1983) delineated three conditions requisite for jobs to be classified as involving emotional labor: (a) frequent face-to-face interactions with clients, (b) the capacity of the job to evoke emotions in others, and (c) the presence of control or regulation over those emotional exchanges. Yin and Lee (2012) observed that educators satisfy all three criteria pertinent to occupations entailing emotional labor: (a) teaching necessitates face-to-face engagement with students, (b) educators are responsible for eliciting specific emotional responses in students and others, such as fear, enthusiasm, and anxiety, among others, and (c) socio-cultural controls or professional norms are observed within the educational environment. Moreover, educators encounter more complex and problematic behaviors compared to those in other service professions, including students’ need to pay attention, chatter during lectures, make discourteous remarks, or exhibit absenteeism. Such disruptions contribute to emotional exhaustion, impede the teaching process, and exert pressure on teachers to manage these situations routinely (Wróbel, 2013). Teachers serve not only as intermediaries transmitting knowledge but also as emotional motivators, requiring them to convey passion for the subject matter to foster effective student learning. Beyond their professional responsibilities, teachers must sustain a positive disposition in their personal lives, recognizing that professional acumen alone is insufficient; they must also possess robust emotional regulation skills. Consequently, teachers are not merely physical or mental laborers but are also significant emotional workers (Ye & Chen, 2015). They bring their emotional selves into the workplace daily, an essential aspect for executing their roles proficiently (Walia & Tulsi, 2014).

Campbell (1990) defined “job performance as a means to reach a goal or set of goals within a job, role, or organization" (p.704). This includes all aspects of job performance within an organization, specifically involving the execution of the job. Job performance can be further understood through its dimensions: task performance, contextual performance, and counterproductive work behavior. Borman and Motowidlo (1997) defined task performance as 'the effectiveness with which employees carry out activities that contribute to the organization's technical core, either directly by executing aspects of its technological processes or indirectly by supplying necessary materials or services' (p. 99). This includes actions part of formal reward systems. Contextual performance refers to behaviors that do not directly contribute to organizational performance but support the organizational, social, and psychological environment (Sonnentag et al., 2008, p. 428). The third dimension is counterproductive work behavior, defined by Rotundo and Sackett (2002) as voluntary actions that harm the organization's well-being (p. 69). This behavior is classified into four categories: personal aggression, property damage, production loss, and political deviance. Different types of job performance evaluations exist, including supervisor rating scales, peer evaluations, subordinate assessments, and self-rated scales, which vary based on who completes them. 

Lent et al. (1996) defined self-efficacy as "people's judgment of their capabilities to organize and execute courses of action required to attain designated types of performance" (p. 83). Bandura (2012) noted that employees boost their self-efficacy through vicarious experiences, learning from their peers' successes. Social persuasion also enhances self-efficacy; positive feedback from supervisors increases employees' confidence and self-assurance. Coaching offers essential guidance, enabling employees to achieve their goals. Employees with low self-efficacy tend to lose dedication when suppressing their emotions, whereas those with high self-efficacy remain committed. Higher self-efficacy not only equips employees to handle threats but also enables them to receive and provide positive social support. Self-efficacy functions as “a professional shield" protecting employees’ feelings (Heuven et al., 2006). A clear understanding of its role is crucial for grasping the mediating effect of self-efficacy.

The first strategy related to emotional labor is surface acting, which involves changing outward displays. Kammeyer-Mueller et al. (2013) found that masking emotions through surface acting is associated with decreased job satisfaction, higher stress levels, and poorer job performance. Goodwin et al. (2011) suggested that surface acting causes emotion suppression, which negatively affects job performance as employees carry an extra cognitive load while engaging in surface acting, draining their self-regulatory resources. These resources also include self-efficacy, so this study suggests that:

H.1 There is a negative relationship between surface acting and job performance.

H.2 There is a negative relationship between surface acting and self-efficacy.

The second strategy, deep acting, enables employees to experience the emotions they need to display genuinely. Grandey (2003) connected 'deep acting' to antecedent-focused emotion regulation. Research indicates that while surface acting causes emotional dissonance, which can harm well-being and performance, deep acting is less mentally exhausting (Brotheridge & Grandey, 2002). Performance depends on emotional expression at work, which varies by occupation. Although some studies emphasize the positive impact of deep acting on job performance, there is still limited empirical evidence linking it to self-efficacy and overall job performance (Goodwin et al., 2011). Building on existing research, this study concludes that:

H.3 There is a positive relationship between deep acting and job performance.

H.4 There is a positive relationship between deep acting and self-efficacy.

The third emotional labor strategy is genuine emotions. While employees may modify their emotions during customer interactions, they often display authentic feelings, an "automatic" regulation strategy that requires less effort. Genuine emotions are associated with positive feelings; favorable display rules encourage the experience of more positive emotions, benefiting employees. Expressing positivity contributes to job satisfaction, ultimately enhancing their job performance. Genuine emotions do not drain resources, but rather support teachers' performance. Therefore, genuine emotions are proposed to have a positive correlation with teachers' job performance and self-efficacy.

H.5 There is a positive relationship between genuine emotions and job performance.

H.6 There is a positive relationship between genuine emotions and self-efficacy.

Without high performers, an organization cannot function; high employee job performance significantly determines overall organizational performance, and the importance of personal traits, such as high self-efficacy, cannot be overlooked in achieving elevated job performance. That implies employees with high levels of self-efficacy outperform those with low levels of self-efficacy in job performance. Therefore, this study asserts a direct relationship between self-efficacy and teachers' job performance.

H.7 There is a positive relationship between self-efficacy and job performance.

This model also helps to understand the mediating role of another variable: self-efficacy. Self-competence and self-mastery are essential components of self-efficacy that help an individual overcome job stress and endure unpleasant emotions or events at the workplace. Thus, a person with increased self-efficacy might view emotional labor as a rewarding part of their job rather than as a stressful activity and a means to improve job performance. Sloan (2014) believes that self-efficacy can serve as both a mediator and a moderator when examining the influence of emotional labor strategies. However, a clear understanding of its role still needs to be established, so this study will be valuable in clarifying the mediating role of self-efficacy. 

H.8 Self-efficacy mediates the relationship between surface acting and job performance.

H.9 Self-efficacy mediates the relationship between deep acting and job performance.

H.10 Self-efficacy mediates the relationship between genuine emotions and job performance.

Discussion, as mentioned above, leads to the following research framework, as presented in

Research Model and Study Hypotheses

Methodology

This study was conducted in Pakistan, focusing on the academic staff of public universities in the country. Lahore, the capital city of Punjab, Pakistan, was chosen as the target population because of its large number of HEIs.

This study employed a cross-sectional research design to investigate the impact of emotional labor strategies and self-efficacy on the job performance of academic staff at public universities in Lahore, Pakistan. A two-wave data collection method was used, which is particularly effective for exploring temporal relationships and the mediating role of self-efficacy between emotional labor and job performance. To reduce common method bias, data were collected at two separate points: emotional labor was measured in December 2021, and self-efficacy and job performance were assessed in March 2022, providing a clearer view of causal links The decision to adopt a quantitative research approach aligns with the study's goal of measuring the impact of emotional labor strategies on job performance. Validated measures of emotional labor (Diefendorff et al., 2005), job performance (Koopmans et al., 2014; Spector et al., 2006; Williams & Anderson, 1991), and self-efficacy (Sherer & Adams, 1983) were utilized. Stratified sampling was employed to ensure the sample was drawn from key universities and departments, enhancing the study's external validity.

Public universities were identified through the Higher Education Commission (HEC) site. Solvin's formula is widely used to decide the sample size from the targeted population. It helps determine the accuracy of the results while giving the researcher an idea of the sample size.

Sample size =  =  = 331

Where N is the size of the population and e is the margin of error. The sample size of the current study is 331. For this study, proportionate stratified sampling was employed, meaning that the population size of each stratum is maintained in proportion to the sample size of the respective stratum. Three strata were drawn from the population of interest, namely, the University of the Punjab, Government College University, and Lahore College for Women University, Lahore, using the following formula.

Proportionate stratified sampling formula: nh = nh/N*n

Where, nh Sample size for hth stratum

Nh= Population size for hth stratum

N = Size of the entire population

n = Size of the entire sample

Sampling Procedure

Each stratum ensures that it is unique and represents a distinct segment of the population. The population was then divided into homogeneous, distinct groups called strata, and final members were randomly selected from these strata. The two main strata were the Science department and the Social Sciences department of the chosen universities. Among these, social sciences departments were more suitable for the study and were selected to ensure homogeneity. Table 1 presents the samples drawn from the University of the Punjab, comprising 173 out of 1006, 75 out of 434 from Government College University, and 83 out of 479 from Lahore College for Women University. The sample sizes were determined by visiting various departments such as Sociology, Psychology, Gender Studies, Law, Political Science, Criminology, Geography, Economics, Public Administration, Management Sciences, International Relations, History, Mass Communication, Anthropology, Archaeology, Philosophy, Business Administration, Administrative Sciences, and Linguistics at the University of Punjab, Government College University, and Lahore College for Women University.

Half the data was collected via Google Forms during the pandemic lockdown, then from targeted universities after reopening, with 396 questionnaires distributed. Most teachers filled them out on the spot. Respondents rated their agreement on a five-point Likert scale from 'strongly agree' to 'strongly disagree,' including a neutral midpoint. Sixty-five questionnaires were discarded for being incomplete, resulting in an 83.5% response rate.

Data were collected using validated scales: A scale developed by Diefendorff et al. (2005) was used to measure emotional labor. The scale consisted of 14 items covering each dimension of emotional labor, rated from 1 (strongly disagree) to 5 (strongly agree). The first seven items relate to surface acting, the next four to deep acting, and the last three to genuine emotions. Self-efficacy was assessed using the General Self-Efficacy Scale developed by Sherer et al. (1983). Chen et al. (2001) revised this scale to improve construct validity. It consists of eight items with responses ranging from 1 (strongly disagree) to 5 (strongly agree). To measure job performance, Task Performance items were adapted from Williams and Anderson (1991). There are seven items, rated from 1 (strongly disagree) to 5 (strongly agree), with the last two items being reverse statements. The 8-item Contextual Performance scale was adopted from Koopmans et al. (2014), with responses from 1 (never) to 5 (always). To assess Counterproductive Work Behavior, a 10-item scale from Horan (2016) was used, with a response range of 1 (never) to 5 (very often).

The university's academic committees approved the study. Considering the ethical considerations, the questionnaires were given after the respondents' consent, along with a cover letter that clearly stated and communicated the aim and focus of the study; the researcher exerted no pressure to avoid biases.

Results

This section begins with descriptive and demographic analysis before moving to inferential analysis. Following Hair et al. (2014), the data were analyzed using SPSS 21 and two AMOS models: measurement and structural. CFA was used to examine factor loadings before evaluating the measurement model's reliability, validity, and fit. Hypotheses were tested through the structural model.

The total number of respondents was 331, with 56.8% females and 43.2% males. The largest demographic group consisted of individuals aged 36 to 40 years (n = 87), accounting for 26.3% of the respondents. Only 2.4% (n = 8) were aged 51 years or older. Additionally, just 3.3% (n = 11) of participants were aged 21 to 25 years, while 24.5% fell within the 31 to 35-year-old age group. Among all respondents, 76.7% were married, and 23.3% were single.

According to stratifications, 173 (52.3%) respondents were from the University of Punjab, 75 (22.7%) from Government College University, and 83 (25.1%) from Lahore College for Women University. The number of permanent faculty members (258/77.9%) exceeded that of visiting faculty (73/22.1%). In terms of qualifications, 197 teachers held MS/M.Phil's degrees, accounting for 59.5%. Among 130 teachers, 39.5% held Ph.D. degrees, while only four (1.2%) had completed post-doctorate studies. Regarding experience levels, 104 respondents had 0-5 years of experience, 89 had 6-10 years, and only six had 26 years or more.

The measurement model uses Confirmatory Factor Analysis (CFA) to validate latent constructs, focusing on unidimensionality, reliability, and validity through factor loadings. CFA assesses how sound indicators represent a construct (Hair et al., 2006) and provides insights into the quality of the measure. Zainudin et al. (2019) stated that newly developed items should have a factor loading above 0.5, while established ones should meet or exceed 0.6; items below 0.6 may be removed to enhance the model fit. Therefore, only the factors with standardized factor loadings (SFL) greater than 0.60 were retained. The factors removed were SA1 (0.367), SA5 (0.574), TP5 (0.485), TP6 (0.367), CP1 (0.555), CP2 (0.588), CP8 (0.451), CWB1 (0.501), CWB2 (0.212), CWB3 (0.430), CWB5 (0.410), and CWB8 (0.364), which had lower SFL.

Table 2 compares the first measurement model with the final measurement model (second). It demonstrates that the values align with the threshold values, and the necessary level is achieved after removing low factor loadings in the final measurement model.

Comparison of Model Fit Indices

After completing fitness model indices, the Relative Chi-Square value should be the first consideration, “a ratio of approximately five or less as reasonable" (Wheaton et al., 1977, p. 99). Carmines (1981) proposed that a chi-square (x²) to degrees of freedom ratio within 2:1 to 3:1 indicates an acceptable fit (p. 80). The second model's probability/relative chi-square is 2.81, which is less than 3, indicating that it fits well. GFI, or Goodness of Fit Index, assesses sample conformity, ranging from 0 to 1, with values above .90 considered good (Hoelter, 1983). In the final model, the GFI is 0.0911, indicating a perfect fit, as it is close to 1.

Furthermore, RMSEA assesses the covariance between the model and the selected parameters to address issues related to sample sizes (Browne & Cudeck, 1993). One significant advantage of RMSEA is its ability to construct confidence intervals around its range. For instance, if the RMSEA range is between 0.03 and 0.08, it indicates a 95% confidence level. Lower RMSEA values indicate a better fit (Hair et al., 2006). Therefore, the acceptable threshold for the Root Mean Square Error of Approximation (RMSEA) is below 0.08. In this study, the RMSEA value of 0.069 suggests an excellent model fit. Additionally, the Normed Fit Index (NFI) is calculated as 'the ratio of the difference between the chi-square (x²) value of the fitted model and the null model, divided by the chi-square value of the null model' (Hair et al., 2006, p. 643). Hair et al. (2006) also discussed that RMSEA ranges from 0 to 1, and a "perfect fit would produce an NFI of 1". The NFI (Normed Fit Index) value in this study is 0.913, indicating an excellent model fit. Moreover, if the TLI (Tucker Lewis Index) is below zero or above 1, higher values indicate a better fit. The TLI for this study is 0.910, exceeding the standard threshold of 0.90. CFI (comparative fit index) values above .090 are considered indicative of a perfect model fit. According to Table 5, the final CFI value is 0.948, which is close to 1, indicating an excellent fit.

Once the model's fitness requirements are met, the construct reliability and validity should be verified before proceeding with the structural model. Composite Reliability (CR) must surpass 0.70 for all latent constructs. In this study, the CR for Surface Acting is 0.882, for Deep Acting (DA) is 0.837, and for Genuine Emotions (GE) is 0.937. The Self-efficacy construct has a CR of 0.872. For Job Performance, CR for Task Performance is 0.899, Contextual Performance is 0.871, and Counterproductive Work Behavior is 0.836. According to Hair et al. (2011), CR values between 0.60 and 0.70 are considered good in exploratory research, while those between 0.7 and 0.9 are considered good in advanced research. These CR values are above 0.7 and within the range of 0.8 to 0.9, indicating good scores (see Table 3.0).

According to Hair et al. (2006), convergent validity is established by meeting three criteria. First, the Construct Reliability (CR) should be at least 0.70, a condition satisfied in our study, as evidenced by the values exceeding this threshold (Table 3). Second, standardized factor loadings should be a minimum of 0.60, a requirement also met, as shown in Table 3. Third, the Average Variance Extracted (AVE) should be at least 0.50. In this study, the AVE values are as follows: Surface Acting (0.558), Deep Acting (0.565), Genuine Emotions (0.832), Self-Efficacy (0.534), and Task Performance (0.642), all of which meet or exceed the recommended threshold. For Contextual Performance, the value is 0.579, and the AVE for Counterproductive Work Behavior is 0.509; the values of AVE for each construct exceed the threshold value.

Composite Reliability and Construct Validity

Discriminant validity represents a construct's uniqueness from other constructs in the Structural Equation Model; the vital requirement for discriminant validity is that the correlations must be less than the square root of AVE (Hair et al., 2006). For example, Table 4 represents those correlations among latent constructs (SA, DA, GE, SE, TP, CP, and CWB) that are less than the square roots of AVE.

Discriminant Validity

Structural Equation Modeling (SEM) was conducted using AMOS 24 to test the hypothesized mediation model. Path analysis is being used via CB-SEM to estimate path coefficients of latent constructs, as it "seeks to determine the strength of the parts shown in the path diagrams" (Hair et al., 2006, p. 609). Table 5 presents a significant negative relationship between Surface Acting (SA) and Job Performance (JP), with a correlation coefficient of -0.39, confirming the hypothesis. This indicates that as surface acting increases, teachers' job performance decreases. The results in Table 5 also confirm a negative relationship between Surface Acting (SA) and Self-Efficacy (SE), with a value of -0.22, supporting Hypothesis 2, "there is a negative relationship between SA and SE".

Direct Effects

Note. ** Significant at 1%

Indirect Effects

Note. *Significant at 5%

According to Table 6, the indirect result of SA→SE→JP is -0.11, confirming Hypothesis 8: "Self-efficacy mediates the negative relationship between Surface Acting and Job Performance." The overall change after mediation is 21.3%, supporting the role of Self-Efficacy as a mediator between SA and JP. However, for the indirect result of DA→SE→JP, the path coefficient is 0.043, indicating that self-efficacy does not serve as a mediator in this relationship, thus rejecting Hypothesis 9. Lastly, the indirect result of GE→SE→JP presents a value of 0.148, confirming Hypothesis 10 that "self-efficacy mediates the relationship between genuine emotions and job performance." The overall mediating effect of self-efficacy in the relationship between GE and JP is 36%, which falls within the medium range, emphasizing the significance of SE as a crucial mediator.

Discussion

This study examines the influence of emotional labor strategies on teachers' job performance in public universities, with a particular focus on self-efficacy as a mediating factor. It investigates direct relationships between surface acting, deep acting, genuine emotions, self-efficacy, and performance. Additionally, it examines the mediating role of self-efficacy, highlighting its importance in understanding the effects of emotional labor.

The hypotheses suggest that emotional labor strategies influence teachers' job performance, either positively or negatively, depending on the specific strategy employed. Surface acting harms performance (Alsakarneh et al., 2023; Hori & Chao, 2022). The literature shows that surface acting is the most problematic strategy because it drains resources, such as reducing self-efficacy, and hinders job performance. Continuously feigning emotions can alienate employees, lead to depersonalization, and require increased monitoring. This study finds that employees often view this negatively, feeling deceived, which can lead to burnout, depression, and lower performance, ultimately weakening long-term self-efficacy beliefs. It emphasizes the important mediating role of self-efficacy in the relationship between emotional labor and job performance.

The structural model revealed no statistically significant relationship between deep acting and self-efficacy, nor between deep acting and job performance, suggesting that educators may struggle to internalize the necessary emotions. Deep acting is often described as a labor-intensive method of emotional regulation that requires significant effort to feel the intended emotion genuinely. Employees using deep acting employ various techniques, such as eliciting specific feelings or using trained imagination, which involves recalling personal events linked to the desired emotional state. As a result, both approaches are time-consuming processes for internalizing and experiencing suitable emotions. In teaching, where educators interact with students daily, adopting these strategies can be pretty challenging. Liu et al. (2008) also found that deep acting is more cognitively demanding than surface acting, noting that it involves substantial mental and psychological effort, which may negatively affect job performance. This research suggests that emotional labor can offer benefits, including improved job performance and a healthier work environment, despite some drawbacks (Hao, 2024). Authentic emotions positively influence teachers' performance by enabling effective emotional labor. Employees' internal states, such as resourcefulness and energy, enhance emotional labor and mitigate its potential harm compared to surface or deep acting. Genuine emotions foster sincerity and harmony, require less effort to display, and thereby enhance job performance. 

Among all dimensions of emotional labor, genuine display or naturally felt emotions have a stronger connection to job performance. Teachers use genuine emotions for emotional labor, where individuals effortlessly and unconsciously immerse themselves in experiencing the appropriate emotion for the situation. The results also show a positive relationship between genuine emotions and self-efficacy, as individuals express their true feelings, thereby increasing their efficacy beliefs. According to Humphrey et al. (2015), genuine emotions do not cause emotional dissonance and lead to resource gain rather than resource loss, which further boosts teachers' self-efficacy as they perform their duties.

The current study demonstrates a strong positive relationship between self-efficacy and teachers' job performance. This finding is well-supported by existing research, as many scholars have consistently observed a positive and significant link between self-efficacy and job performance (Iliev, 2024; Na & Isa, 2024). Self-efficacy positively influences job performance, and employees with high self-efficacy are essential for achieving better performance since organizational success ultimately depends on each employee's performance. Therefore, the level of self-efficacy is critical for job performance. Our study shows that self-efficacy is closely and positively associated with job performance, and employees with high self-efficacy tend to perform better. These results have important practical implications for educational institutions, highlighting the significance of self-efficacy in improving teachers' job performance.

The second research objective was to examine "the mediating role of self-efficacy in the relationship between emotional labor strategies and job performance." The findings indicate that self-efficacy mediates the relationship between surface acting and teachers' job performance, suggesting that self-efficacy plays a crucial role in mitigating the negative impact of surface acting on teachers' performance in public sector universities. Consistent with Heuven et al. (2006) results, which found that flight attendants with lower scores on the self-efficacy scale were more likely to experience adverse outcomes, such as discrepancies between their actual and displayed emotions (surface acting) during interactions with passengers, whereas highly self-efficacious flight attendants were less dissonant and could handle highly emotional interactions effectively. The results show no mediation of self-efficacy for deep acting and job performance, indicating that the mental energy and psychological effort required to internalize emotions are genuinely much higher. Teachers often find it challenging to conceal their emotions in this manner.

As Jeung et al. (2017) identified, self-efficacy is a valuable personality trait relevant to understanding the role of emotional labor within the service sector; it acts as a mediator in the relationship between authentic emotional expression and job performance. This indicates that educators with high levels of self-efficacy see the use of genuine emotions as beneficial for improving their professional performance. Educators with higher self-efficacy scores tend to maintain healthier emotional lives, which then positively affect their performance. Additionally, the role of self-efficacy as a mediator does not just show its ability to influence the negative effects linked to emotional labor. Teachers with high self-efficacy demonstrate a stronger sense of mastery, greater self-competence, and a perceived sense of control over their emotional responses, enabling them to perform in ways that boost their effectiveness.

Pakistan, as a developing nation, relies heavily on its educational institutions, which in turn depend upon the competence of teachers and their surrounding environment. Our educational framework discourages the expression of genuine emotions, suppressing feelings that might potentially offend authorities, thereby undermining teachers' self-efficacy and autonomy. Research has shown that expressing sincere emotions can enhance self-efficacy and overall performance. Promoting openness regarding emotional expression within higher education institutions fosters confident educators and leads to improved classroom outcomes. Accordingly, providing training that focuses on the constructive management of emotional labor is essential for cultivating a healthy, growth-oriented academic environment.

Universities can develop programs to train teachers in expressing emotions and handling difficult situations, focusing on skills like attention management, encouraging specific feelings, and using trained imagination. Teachers should also practice these skills independently to enhance their emotional labor performance. This study examines how self-efficacy influences the display of desired emotions at work, improving or regulating performance, and strengthening this connection. It highlights individual factors related to core concepts, helping practical and future teachers improve performance through efficacy strategies and assisting policymakers in understanding emotional job requirements. Theoretically, this confirms that emotional labor strategies affect public sector faculty performance, with self-efficacy serving as a key mediator, thereby adding to the understanding of emotional labor research. 

Although this study highlights the impact of emotional strategies on teachers' job performance by examining their self-efficacy through a quantitative approach, it has some limitations. First, the research was conducted in Lahore, Punjab Province, and its findings are limited to public sector universities in Lahore. The results, focus on variables, and the unit of analysis mainly pertain to individuals (i.e., teachers). The study used a quantitative method with a cross-sectional design. However, using a qualitative approach, such as interviews, could provide deeper insights into the variables. The study could also include private universities, in addition to public sector institutions. Future research should consider conducting a comparative analysis.

Conflict of Interest

The author of the manuscript has no financial or non-financial conflict of interest in the subject matter or materials discussed in this manuscript.

Data Availability Statement

The data associated with this study will be provided by the corresponding author upon request.

Funding Details

No funding has been received for this research.