Grandey (2000) semantically conceptualized emotional labor as the process of emotion regulation and expression by employees for the realization of organizational objectives. According to this definition, emotional labor is the management of emotions by the employees in the workplace, displaying and adjusting feelings and expressions that are needed to fulfill the requirements of the job (Brotheridge and Lee, 2003). In addition to this value-laden feature, emotional labor is also characterized by the responsiveness from the customers, the normative requirements from the employer as well as its omnipresence among employees at different levels (Lucas and Deery, 2004).

As is suggested by Liu et al. (2004), emotional labor is a concept that is of close relevance but yet not limited to the service sector, with the tourism and hospitality industry no exception here. Actually, one of the earliest examination on emotional labor was under the context of the airline sector, indicating the salience of this issue in the tourism and hospitality industry (Hochschild, 1983). It can be said that the job performance of a tourism and hospitality employee is entirely emotion-oriented (Pizam, 2004). On one hand, the facial expressions and body languages are already integrated into the quality of service provided by the employee to the customer, which can be best illustrated by the recognition of the factors of empathy and assurance as the key dimensions of service quality. On the other hand, emotional labor is dynamic rather than static in the service process, open to instant adjustment and optimization in accordance with the feedback from the customer, so that the service product can be successfully consumed. Since customer satisfaction, attitude and loyalty are all best expressed in emotional forms, the resonance functions of the emotional labor of the employee are more and more cherished and anticipated (Yan et al., 2012). Thus, great stakes are placed on the presentation and management of emotional labor for the enterprise, whether in soliciting new customers or the retention of established ones (Brotheridge and Lee, 2003). Hochschild (1983) provided a classical categorization of emotional labor into three streams of surface acting, deep acting and genuine acting. In surface acting, employees fake their motions through observable features in gestures, expressions and voice tones to align with the requirements of their job; deep acting is concerned with more active involvement of the inner states of the employee in addition to surface acting, through the utilization of mental and psychological maneuvers such as thoughts, memories and images. In other words, the employee would ‘labor out’ the expected emotions in deep acting. The third category of genuine emotion is where the felt emotions are congruent with the expressed ones, as the employee express what he or she experiences in a spontaneous and natural manner. Genuine acting is consistent with the term of ‘naturally felt emotions’ offered by Diefendorff et al. (2005). The existence and effects of different forms of emotional labor indicate the dilemma and even conflicts of ‘feeling rules’ and ‘display rules’ in the work environment of many industries, which may easily lead to emotional dissonance and exhaustion and thereby severe mental and psychological challenges to the employee. For instance, in the service industry, job burnout is usually preceded by emotional exhaustion, which is much more noted at middle and senior managerial levels (Bolton, 2005). Conversely, positive effects of emotional labor can arise from the synergy of felt and display emotions such as job satisfaction, self-efficacy and self-esteem (Chu and Murrmann, 2006).

With regard to factors influencing the generation and outcomes of emotional labor, two sources have been identified, namely, personal and organizational (Lucas and Deery, 2004; Yang and Chang, 2008). From a personal perspective, emotional labor has been discovered to be related to the work experiences, personality as well cultural values of the employee. On this account, it has been postulated by Xu (2007) that some employees may be inherently unfit for the emotion labor required by certain professions. This, nevertheless, does not rule out the emotional efforts that can be wielded by the employee to achieve certain expected emotional outcomes without rendering negative emotional consequences. In addition, with the growing level of diversity in the workforce of the tourism and hospitality industry, the cultural underlying of emotional labor are more and more evident (Yan et al., 2012). From the organizational perspective, the consensus has been reached that emotional labor can be effectively regulated and managed so that the positive effects of emotional labor can be induced. Particularly, effective emotional management can introduce energetic and amusing dynamics into the otherwise monotonous tasks and compensate for the ‘dull aspects’ of the job. Meanwhile, emotional labor can be appropriately addressed through the facilitation of job autonomy, which has been empirically examined by Kim et al. (2007) to make positive contributions to the prevention of emotional dissonance and exhaustion.

Perceived organizational support can be defined as the perceptions by the employee of the evaluation of work and extent of support from the organization (Allen, 1995). Perceived organizational support reflects the interactions between the organization and employee, with the aim of realization of an interrelationship that is based on reciprocity. Perceived organizational support is derived from social exchange elaborations, and can be embodied in material, psychological as well as emotional terms. In the professional arena, perception by the employee of perceived organizational support should encompass not only expectations for remunerations that match job performance, but also demand for in-time care and necessary help from the employer. Hutchison (1997) stressed that particularly in scenarios of emergencies or tremendous challenge, the values of perceived organizational support in psychological and emotional forms would overweigh those in monetary forms. In this sense, the psychological contracts between the employee and the employer would be consolidated to such an extent that greater trust, loyalty and sense of belonging would be secured for the organization. Generally speaking, perceived organizational support is recorded to be more salient when such support is perceived as spontaneous, interactive and responsive (Settoon et al., 1996).

Eisenberger et al. (1997) identified three major dimensions of perceived organizational support, namely organizational rewards & job condition, supervisory support and fairness. Organizational rewards & job conditions consist of tangible material rewards as well as those intangible benefits like working experiences beneficial for future career development, autonomy in job, sense of security etc. Such organizational endowments are direct indications of organizational commitments to the employee, and help enhance the in-role and out-role behaviors of the employee toward the achievements of corporate objectives (Hutchison, 1997). Supervisory support is usually presented in the forms of attention to the employee, recognition of the employee’s contribution and communications of corporate values. It was demonstrated by Yoon and Lim (1999) that supervisory support is perceived most strongly from the immediate superiors of the employees, which is an apt reflection of the peculiarities of the hotel sector which are characterized by close and dynamic supervisor-employee interactions. The third dimension of procedural justice, which underpins the equity principle in social exchange elaborations, is related to the perception by the employee support as against those of other employees. From rewards & penalties policies to allocation of organizational resources, procedural justice is actively pursued and appreciated by the employee, and is a key determinant of the balance and harmony of interrelationships among the employees.

Vigorous organizational support has been found out to incease job satisfaction of employees (Hutchison, 1997), a favorable outcome that is strongly associated with sound handling of emotional labor (Asumah et al., 2019; McGinley et al., 2019). In view of the above review of emotional labor and perceived organizational support, the first research hypothesis is proposed as follows:

Eby et al. (2003) defined career competencies as the portfolio of work experiences, skills and qualifications that the employee progressively accumulated during the career. As against those attributes relevant to the required job performance of an employee, career competencies address the employee capacities on at a longitudinal, strategic and sustainable level (Arthur et al., 1995). In this regard, career competencies introduce intelligence aspects into the career management of the employee, and aim for the integration of individual and corporate successes (Eby et al., 2003). Especially, career competencies are significant determinants of positive career changes along the career path of the employee. In addition, with growing emphasis placed on systematic and comprehensive career management, career competencies are purported as significant criteria for career progress to managerial and leadership positions (Sturges et al., 2003).

Three branches of career competencies were articulated by Arthur et al. (1995), which are ‘Knowing why’, ‘knowing who’ and ‘knowing how’. Knowing why concerns the motivational aspirations of the employee in self-identification, self-understanding and possibilities exploration. Knowing why also relates to the ability of adaptation to shifting work contexts. Knowing why can be further accounted for by the three variables of career insight, proactive personality, and openness to experience. Knowing who is synonymous with the networks and contacts from within or beyond the organization, the values of which can be exchanged for the employee’s career progress. Knowing who within the organization encompasses normative relationships as well as those with mentoring natures. Knowing who outside the organization, on the other hand, can extend to contacts such as suppliers, consultants and professional acquaintances. The more conventional concepts of career-relevant knowledge and skills fall into the category of knowing how, which is featured by accumulation and foundation (Sturges et al., 2003). Knowing how consists of two inter-related sub-categories of career-related skills and career identity, with the latter serving as the motivator and director of skills accumulation (Eby et al., 2003).

Career competencies in the tourism and hospitality industry have been opined by Tesone and Ricci (2006) to be one of the most comprehensive and demanding in the service sector, given the diversity of customer needs, combination of physical and psychological inputs and potential of cross-cultural conflicts. The requirements of flexibility and personalization that characterize this industry pose further challenges to the cultivation of career competencies in this industry (Christou, 2002). Major competencies that have been identified are positive attitude, communication skills, effective listening as well as willingness to work hard (Chung, 2000; Brownell, 2008), thereby leading to the formulation of the second research hypothesis of this study as follows:

Career commitment is purported by Blau (1999) to be an attitudinal measurement of one’s motivation to work in a chosen vocation. In this regard, career commitment can be analogized as a psychological contract between an individual and his/her career, indicating the level of attachment and involvement to the career by the individual, the degree of alignment of individual objectives with organizational ones, as well as the extent to which the career takes the central stage of the individual’s daily life. Meanwhile, another paradigm emphasizes the behavioral aspects of career commitment, focusing on the temporal and monetary costs and benefits that are incurred to the individual arising from the commitment behavior. An employee would conduct a subjective person-organization fit evaluation, and the differences between the exact career peculiarities and the expectations of the employee would determine the outcomes of career commitment (Carson and Bedeian, 1994). Career commitment has merits in reliably predicting such important factors as work pressure, job efficacy and satisfaction. Career commitment was firstly applied in the education sector and has been widely adopted by tertiary sectors to examine and tackle employee absenteeism and quitting (Cohen, 1999).

Career commitment is a dynamic concept that is based upon the constant and active interactions between the employee and the career milieu, and may have taken form even before one enters a certain career (Chang, 1999). Cohen (1999) listed out three major factors underlying the emergence of career commitment, namely the predilection for the career, balance of the costs of shifting career and concern over violation of social norms in career change. Respectively, career commitment can be divided into emotional, cost and normative categories. It can be learned from this classification that career commitment can be affected by both intrinsic and extrinsic attributes of the employee, such as demographic features, personality, salary, work environment and work challenges, etc. In the tourism and hospitality contexts, career commitment has been found to be positively related to educational backgrounds, titles, foreign language skills and professional certifications, thereby illustrating the vast spaces that can be explored to enhance the career commitment of tourism and hospitality employees (Leslie and Russell, 2006; Walsh and Taylor, 2007). Therefore, the third to fifth research hypotheses are generated, which are coalesced with the first two hypotheses and presented in the research framework of this study as displayed in Figure 1:

In accordance with relevant literature, emotional labor was measured on a three-dimension scale which included the attributes of surface acting, deep acting and genuine acting. The measurement proposed by Diefendorff et al. (2005) was the primary reference, with attention also paid to the scale developed by Kim and Han (2009) which was of more hospitality features. Altogether, seven, five and two items were proposed for surface acting, deep acting and genuine acting, respectively. Sample items included “In my job, controlling my emotions is very important,” “I more often have an artificial smile than natural smile at work,” “There is a great big difference between my real emotions and expressed emotions” and “I frequently hide my emotions on the job.”

At the same time, the measurement of organizational support was based on the one proposed by Friedman and Greenhaus (2000) with six items, such as “The organization values my contribution to its well-being.,” “The organization really cares about my well-being,” “The organization cares about my general satisfaction at work.,” “The organization takes pride in my accomplishments at work.” Meanwhile, Kong et al. (2012)‘s refined scale of career competences was adapted and adopted, with six items for examination. Sample items were “I seek out opportunities for continuous learning” and “I have a diversified set of job-related skills.” At last, the scale measuring career commitment was refined from that used by Blau (1999), consisting of six items like “If I had all the money I need without working, I would probably still continue in this profession,” “I would recommend a career in the hospitality industry to others” and “I am disappointed that I ever entered this profession.” All the items were measured on a seven-point Likert scale with 1 being “strongly disagree” and 7 being “strongly agree.”

A self-administered questionnaire was developed and distributed to 450 frontline employees in three five star international-brand hotels in Shanghai, China in mid-2022. The benchmark criteria for the interviewee were set at full time employee with more than 6 months of working experiences. In light of a quite internationalized workforce at the investigated hotels, the questionnaire was designed in both English and Chinese, although the nationality and native language were not solicited given the research purposes of this study. While the survey was being filled, trained interviewers would assist with clarification issues when necessary. Out of the 450 copies of surveys distributed, 406 were returned and 387 were determined to be valid for further analysis after initial examination, recording a response rate of 86%, which reached a satisfactory level for further data analysis.

The data analysis process consisted of two steps. Firstly, a confirmatory factor analysis (CFA) was conducted to investigate the internal validity and reliability of the proposed scale. This stage included the examination of dimensionality and convergent and discriminant validity (Anderson and Gerbing, 1988); secondly, the structural equation model (SEM) was applied to test the fitness of the proposed model and the hypotheses developed early. The computer software of SPSS 15.0 and AMOS 17.0 were employed in data categorization and model evaluation.

The valid interview participants served in front office, housekeeping, food & beverage, recreation and marketing departments of the hotel, with 32.4% indicating intra-or cross-departmental experiences. There were 42.3% males and 57.7% females. The age range of 20–30 years recorded the greatest proportion at 68.0%, followed by 31–40 years (18.2%) and 41–50 years (12.4%). The majority of respondents reported holding a bachelor’s degree (46.7%), followed by secondary school (36.2%), graduate (13.2%) and below secondary school (3.9%). In terms of working experiences, the distribution turned out to be quite even among the categories of 1–5 years, 5–10 years and over 10 years. Particularly, employees who had working experiences at more than two hotels accounted for 52.3% of the total, while 37.5% of the respondents reported to have had worked at other service industries before taking the job at the hotel.

The CFA was conducted to investigate the reliability and then validity of the respective proposed measurements with the other subsamples, with the results manifested in Table 1. Overall, as the composite measures of reliability of the examined variables all stayed above 0.70, satisfying the criteria suggested by Hair et al. (2014), it was found that the items concerned were reliable in indicating their corresponding dimensions and that a satisfactory level of internal consistency within each proposed dimension was achieved. As is stipulated by Byrne (2010), standardized factor loadings with a value of less than 0.5 were deemed insignificant and prone to the possibility of cross-loading. Therefore, 5 items were discarded from the proposed scale of emotional labor due to falling below the level of 0.5, including 2 from the surface acting and deep acting dimensions, respectively, and one from the genuine acting dimension; meanwhile, two items were dropped from the original scale for organizational support, while one item each was voided from the proposed measurements of career competencies and career commitment.

All of the measurement models reported goodness-of-fit indices that could establish an acceptable fit between the model and the sample data: (1) emotional labor measurement model (x² = 436.32, df = 27, CFI = 0.91, GFI = 0.94, RMSEA = 0.07); (2) organizational support measurement model (x² = 373.24, df = 22, CFI = 0.93, GFI = 0.91, RMSEA = 0.07); (3) career competences measurement model (x² = 321.61, df = 25, CFI = 0.93, GFI = 0.92, RMSEA = 0.07); (4) career commitment measurement model (x² = 406.97, df = 19, CFI = 0.91, GFI = 0.93, RMSEA = 0.07). According to the criteria offered by Hu and Bentler (1998), CFI and GFI values over 0.9 were deemed as acceptable for model fit. Accordingly, the proposed scales were all uniquely related to their respective dimensions and a satisfactory level of convergent validity was obtained. Then, discriminant validity of the scale was assessed to further validate the proposed scale of switching costs. Here, average variance extracted (AVE) analysis (Henseler et al., 2015), was adopted. It was shown by the results as in Table 2 that the square root of the average variance for each of the proposed dimensions was greater than any of the inter-correlations of the dimensions. This analysis therefore gave additional support for the discriminant validity of the proposed scale.

The overall measurement model, which recorded the goodness-of-fit indices of x² = 1034.77, df = 271, CFI = 0.90, GFI = 0.93, RMSEA = 0.07, demonstrated a considerable degree of good fit between the model and the sample data. As the composite measures of reliability ranged from 0.78 to 0.91, a sound fit had been established between the overall measurement model and the data.

Lastly, the structural model was tested to evaluate the hypothesized relationships among the variables. The model fit indices were reported as follows: x² = 1103.26, df = 282, CFI = 0.93, GFI = 0.91, RMSEA = 0.07, thus representing a fit to the data. It was demonstrated by the results that all structural path estimates reported statistical significance (as indicated by C.R. >1.91), meeting the standards specified by Hu and Bentler (1998), and therefore could be validated. The path coefficient value and significant level, which is delineated in Figure 2, illustrate positive and significant structural paths. Thus, with all the direct positive relationships confirmed, all of the proposed hypotheses were proved to be statistically significant and thereby supported. Specifically, the mediating function of emotional labor was consolidated through the Sober Test, with positive indirect effect coefficients (0.36, 0.39) and t-values (6.73, 7.09). Accordingly, it could be concluded that emotional labor had mediated the effects of organizational support on career competences and career commitment.

Besides, multiple regression analyzes were conducted to scrutinize the specific effects of the respective dimensions of emotional labor and cross-examined with the criterion as suggested by Hair et al. (2014). It was revealed that organizational support exerted a positive effect on the construct of genuine acting (β = 0.63, p < 0.001), a negative effect on the construct of surface acting (β = −0.20, p < 0.001), and little effects on the dimension of deep acting. On the other hand, career competences were discovered to be positively influenced by deep acting and genuine acting (β = 0.38 and 0.40 respectively), while little influenced by surface acting. Finally, the impacts of emotional labor dimensions on career commitment were similar to those on career competences, with deep acting and genuine acting indicating positive influences (β = 0.27 and 0.31 respectively) and surface acting reporting negative effects (β = −0.24).