The Effects of Service Employees’ Emotional Labor on Customer Loyalty in Wealth Management Service Organization

The wealth management firm meets greater competition in generating customer loyalty after China’s 11 opening-up policy. According to emotional contagion and SOR theory, the effect of employees’ competence on customer loyalty was firstly investigated. Secondly, the mediation effect of customers’ trust was tested. Finally, the moderation effect of employees’ emotional labor perceived by customers was investigated. After tested by Process Macro for SPSS, several results were obtained. Employees’ competence positively pre-dicts customer loyalty, and customers’ trust plays a mediation effect. The deep acting (surface acting) strategy of employees plays a positive (negative) moderation effect in the relationship between employees’ competence and customer trust. Theoretical and managerial implications related to emotional labor and trust were generated.


Introduction
The office of the Financial Stability and Development Commission of the State Council of China issued 11 measures to open up the financial industry to the outside world in July, 2019 (China Economic Net, 2019). The "catfish effect" induced by foreign capital forces the regulation and transformation of domestic wealth management enterprises. Wealth management enterprises are typical service organization (Lu et al., 2006) for its intensive interaction with customers. Service organizations are vulnerable to customer switching because of the high level of competition (Lam et al., 2004). A loyal customer is like an asset to an influence customer loyalty instead of directly. Mayer (2008) and Guiso et al. (2004Guiso et al. ( , 2008 show that Financial transactions are very trust-intensive and that trust in financial markets is an important factor for financial development. As the main business is to manage customers' money, it is important to achieve customers' trust for wealth management organization. Without customers' trust, only high competence is not enough to achieve customer loyalty. Creating loyalty for corporate brands has become more challenging due to minimal differentiation among their competitive offerings and similar corporate and organizational values among competitors (Anisimova, 2007;Dawes, Meyer-Waarden, & Driesener, 2015). In this sense, building and maintaining effective relationships with consumers is crucial for gaining customer loyalty (Gbadamosi, 2015). This relationship refers to the emotional component between consumer and firm. Employees' emotion regulation during service interactions affects customer loyalty, all of which are critical for service success (Seger-Guttmann & Medler-Liraz, 2016). Employees are required to regulate emotion expression strategically when interacting with customers, that is, to show positive emotion and suppress negative emotion through facial expression or cognition regulation. This indicates employees have exposed to emotional labor. Ac-Journal of Service Science and Management cording to emotional contagion theory, employees' positive emotion will be experienced by customers. At the same time, customers will have more trust with employees if they show more professional competence. Otherwise, customers will doubt employees' competence if they experience their faked emotion display. Previous study has investigated emotional labor's moderation effect on the relationship between customer participation and buying (Seger-Guttmann & Medler-Liraz, 2016). However, there is a very limited understanding about the moderation effect of emotional labor in the wealth management context. The current study aims to investigate the mechanism of customer loyalty in the wealth management organization from the perspective of emotional labor.

Customer Loyalty
Customer loyalty is an important factor in enhancing the sustainability of a company through maintaining existing customers and strengthening relationships (Hallowell, 1996). Customer loyalty includes two dimensions: attitudinal and behavioral (Chandrashekaran et al., 2007). From the attitudinal perspective, customer loyalty can be interpreted as a degree of positive attitude or preference towards a product or service (Jacoby & Chesnut, 1978). It can be described as the intention to recommend and repurchase (Kumar et al., 2013). Attitudinal loyalty expresses a consumer's desire to establish a relationship with an enterprise. The problem is to verify if this intention will transfer to an action (Nguyen, 2016). From the behavioral perspective, customer loyalty is described as the repeat purchasing or use of a given service or product over time (Leenheer et al., 2007). Behavioral loyalty includes retention, repurchase, duration of the relationship (Kumar et al., 2013). Whether perceived as an attitude or type of behavior, loyalty is one of many elements that influence consumers' buying decisions (Nguyen, 2016).

Competence
The competence of service employees is a body of knowledge gained from training and experience. It constitutes their capacity to provide transactions requested by the customer and to perform in a way to meet his or her expectations (Barclay & Smith, 1997). This competence usually relies on the ability of service employees to acquire and apply this knowledge in accomplishing their job. It is to ensure the success of transactions during the service encounter and help to increase customer's trust towards the firm (Xie & Peng, 2009) and customer retention (Delcourt et al., 2013).
Competence has two components: technical expertise and problem-solving skills (Sirdeshmukh et al., 2002). Technical expertise is considered as a cognitive component of competence. It is the degree of knowledge directly related to the firm's activity. This expertise is requested as a necessary qualification in the recruitment process. It is possessed individually through the employee's professional training. Technical expertise can be accumulated on the job and enhanced by continuing education. Due to its significant impact on the outcome of transactions, technical expertise can help service employees to meet the consumer's expectations and to gain his or her trust towards the firm. The problem-solving skill is the behavioral component of competence. It represents the ability of service employees to tackle conflictual issues, i.e., service failure. In service recovery, employees have to take into consideration the motivations of both parties while attempting to satisfy the customer's needs and protecting the firm's interests. The problem-solving skill of service employees is the result of their individual personality traits and their perceptions of the social interaction with customers during the service encounter (Hartline et al., 2003;Solomon et al., 1985). The two components of competence are the same important components in attracting customers.
Previous research has examined the influences of employee competence on various aspects of the firm, including corporate image (Nguyen & Leclerc, 2011), service quality, customer satisfaction (Arora & Narula, 2018) and customer loyalty (Nguyen, 2016). In the context of wealth management service, the competence of employees is specifically geared towards the success of the transaction that meets the customer's requirements and in turn the loyalty to the firm. Hence the first hypothesis can be stated as follows: H1: Wealth management service employees' competence is positively related to customer loyalty.

Trust
Trust is the expectation held by a consumer that the service provider is dependable and can be relied high on to deliver on its promises and to put the client's interest first. It is an essential ingredient for stable relationships (Kimes, 2010). Trust removes the vulnerability to exploitation (Hansen, 2012). Consumers' perceived trust in a relationship partner has been found to influence overall satisfaction and generate behaviors that increase cooperation, loyalty (Tran & Strutton, 2020) and performance (Kinnel, 2010). The development of trust between the consumer and the financial service provider may promote relational exchanges that prove to add value for both partners (Winchester & Huston, 2017). As trust decisions involve both thinking (cognitive) and feeling (affective) process (Srivastava et al., 2015), the current study examines the concept of trust through its cognitive and affective components. Cognitive trust is driven by knowledge and a rational thought process (Wang et al., 2016), whereas affective trust is driven by feelings and emotional exchanges (Albert & Merunka, 2013;Dowell et al., 2015).
Cognitive trust is an instrumental inference made from information about another's behavior under specific circumstances (Zhang, 2015). The level of cognitive trust may reflect integrity factors including honesty and fairness of the referent (Dirks & Ferrin, 2002). Affective trust relates to the extent to which one feels secure and comfortable about the trustee (Komiak & Benbasat, 2006;Zhang, 2015).

Stimulus-Organism-Response Theory
Stimulus-Organism-Response (SOR) theory was developed by environmental psychologists. The theory posits the environment functions as a stimulus. Organisms' external environments influence their internal reactions. These reactions, in turn, promote external behavioral responses among organisms (Mehrabian & Russell, 1974). These internal organism responses include affective (emotional) and cognitive elements. These emotional and cognitive beliefs mediate approach or avoidance responses toward the environment (Tran & Strutton, 2020). According to SOR theory, employees' competence might function as an environmental stimulus; trust (i.e., customers' internal reaction) as an organism; and customer loyalty functioned as behavioral responses. Hence, we hypothesize as follows: H2: Trust mediates the relationship between employees' competence and customer loyalty.

Emotional Labor
In the workplace, employees should not only pay mental and physical labor, but also emotional labor. In 1983, American sociologist Hochschild proposed the concept of emotional labor on the basis of absorbing Goffman's Parody theory through the analysis of Delta Air flight attendants (Hochschild, 1983). He de- This paper focuses on the customer service situation. The core of emotional labor is an effort to regulate emotions. This regulation process includes two strategies: deep acting and surface acting. Deep acting is a kind of antecedent-focused emotion regulation strategy (Gross, 1998). It needs to reappraise the situation to regulate the inner feelings, and then express the positive emotions required by the organization (through facial expression, tone, gesture and body posture).
Surface acting is reaction-focused, which only needs to regulate the expression of emotions, that is, to suppress negative emotions and show positive ones, although the real feelings of the heart remain the same (Grandey, 2000). Emotions have internal effects, that is, employees engaged in emotional labor will have an impact on themselves, such as loss of self (Hochschild, 1983), emotional exhaustion (Grandey & Melloy, 2017), burnout, health problems (Zapf, 2002), and low job satisfaction. Emotions also have interpersonal effects, that is, employees engaged in emotional labor will have an impact on customers' emotion, attitude, cognition and behavior (Van Kleef, 2016 needs to be combined with certain social and cultural situations (Grandey & Melloy, 2017). The situation of wealth management has its unique characteristics. It makes financial planning for customers by taking customers as the center.
According to the theory of emotional contagion, customers will perceive employees' emotions and be infected. Therefore, the positive emotional display of employees is conducive to the good emotional experience of customers. On the contrary, customers will doubt the service ability of employees if they perceive the fake emotional display of employees. This study supposes that emotional labor moderates the positive effect of employee competence on customer trust.
That is to say, the stronger the employees' professional knowledge and problem-solving ability, the more trust the customers will have, and the sincere service (deep acting) of the employees will strengthen this relationship; on the contrary, the false display (surface acting) will weaken this relationship.
Customer perceived and employee perceived emotional labor are two different concepts (Gong et al., 2020). In the process of establishing the boundary of host and guest emotional communication, customers play a leading role (Christou et al., 2019). Therefore, the current study chooses the employees' emotional labor perceived by customers to study its moderating effect on the positive relationship between employee competence and customer trust. As customers perceive emotional labor in two ways at the same time, deep acting and surface acting are regarded as two independent constructs to explain emotional labor. Accordingly, we hypothesize as follows: H3a: Employees' deep acting strategy positively moderates the relationship between employees' competence and customer trust; H3b: Employees' surface acting strategy negatively moderates the relationship between employees' competence and customer trust.
To sum up, the conceptual model is shown in Figure 1.

Data Collection
Data was gathered using an online survey. To test the tool, a pilot test was first carried out from January 14 to 23, 2020 through WeChat. This initial study sought to improve the questions and remove unclear and/or ambiguous items in order to refine the survey content and structure. Preliminary evidence presented reliable and valid scales. Following the pre-test, the formal research data was collected on Mar. 2020 through social media platforms. After removing 21 incomplete and invalid (income below 4000) questionnaires, we finally get 183 (>150) (Rigdon, 2005) valid responses. Table 1 lists the basic information of the sample. More than 50% of respondents were men; more than 60% of them were older than 36 years of age, and more than 90% of the income were above 5000.
We used Harman's single-factor test to assess common method bias. The first factor explains 35.3% of the covariance amongst all constructs. This is less than 50%, which means that common method bias does not affect our data (Podsakoff et al., 2003(Podsakoff et al., , 2012. This study entails all the common shortcomings involved in using an open access sampling methodology (e.g., self-selection bias, lack of information about non-respondents, and unknown response rate) (Kuss et al., 2014).

Scale Design
The questionnaire items were based on those reported in the literature and adapted for this context. Five items were used for the measurement of competence according to Nguyen's study (Nguyen, 2016 Our control variables included gender and income. Mittal et al. (2019) investigated that female are more satisfied and loyal to their bank compared to male. Including income as a control variable enabled us to control the effect of investing power on consumer loyalty.

Indicator Reliability and Validity
The linkage between constructs was tested and analyzed using SPSS 21.0 and AMOS software. Structural equation modelling (SEM), namely partial least squares (PLS) path modelling was employed. All of the constructs were measured by reflective constructs. As shown in Table 2, the Cronbach's alphas for competence, we analyzed factor loading and average variance extracted (AVE). All of the factor loadings were greater than 0.5. Fornell & Larcker (1981) suggested that the AVE should present a value not lower than 0.5, which reflects adequate convergent validity (Götz et al., 2010). Consequently, and according to the results shown in Table 2, all constructs indicate AVE greater than 0.5.
To investigate the discriminant validity of the variables, we conducted confirmatory factor analysis (CFA) using AMOS and examined the fit indices of the hypothesized 3-factor model (i.e., competence, trust and customer loyalty). The CFA results (Table 3) show that the hypothesized 3-factor model fit the data well (χ 2 /df = 1.10, NFI = 0.92, RFI = 0.94, TLI = 1.00, IFI = 1.00, RMSEA = 0.003). As shown in Table 4, most values of the correlation coefficients between the independent variables were below the cut-off value of 0.7 (Field, 2009;Pallant, 2007). The composite reliability (CR) of each construct ranged from 0.81 to 0.98; the square root of AVE of each construct was greater than its correlations with other constructs (Table 4) which indicates good discriminant validity of the scales.

Structural Model and Hypotheses Testing
Through Pearson's correlation analysis, we evaluated the correlation of the constructs, as shown in Table 4. We can conclude from Table 4 that employee competence was positively related with costomer trust and loyalty (β = 0.424, 0.503, p < 0.01); surface acting was negatively related with customer trust (β = -0.338, p < 0.05); deep acting was positively related with customer trust (β = 0.609, p < 0.01). The amounts of the sample were greater than 100, we used maximum likelihood estimation method (MLE) to evaluate the structure model (Hou et al., 2004). The results of the path analysis is shown in Figure 2 (χ 2 =

Mediation and the Moderated Mediation Effect
To test the mediation and moderated mediation hypotheses, we used the PROCESS macro for SPSS (version 3.0) (Hayes, 2018). In Table 5 and Table 6, we provided estimates of the mediation and moderated mediation effects, along with 95% bias-corrected bootstrapped confidence intervals of our path estimates.

Discussion
The frontline employees play a vital role in wealth management organization.
According to the emotional contagion theory and SOR theory, the conceptual framework of the relationship among employees' competence, customer trust and loyalty was supposed and tested. Several results have reached.
Firstly, employees' competence (including professional knowledge and problem-solving ability) positively influences customer loyalty. This indicates that wealth management service organization should primarily strengthen employees' competence through training and other education. Specific hiring strategy should be employed to select suitable employees.
Secondly, trust plays mediation role in the positive relationship between employees' competence and customer loyalty. Besides competence, wealth management service organization should seek for other practical trust building techniques.
Thirdly, employees' deep acting display strategy positively moderate the positive relationship between employees' competence and customer loyalty, while surface acting plays a negative role. In experience economy era, customers' experiences could be affected by the nature of service employees' emotional labor, which can act as a potentially damaging outside source (Groth et al., 2009).
Wealth management service organization should encourage employees to display genuine emotions (deep acting). Unique HRM policy should be employed, such as recruitment, selection, performance evaluation, etc. organizations especially financial corporates. These corporates pay to train employees necessary emotion expression. For the cost reduction purpose, we advise universities to add emotional labor course in the wealth management-related majors. Through this, the overall cost our society will decrease. This will increase university's popularity in society and the employment rate will be raised. Practitioners sought to raise customer loyalty because it is less costly and more beneficial. This study suggests employees' competence is the primary element to influence customer loyalty. In practice, we should ensure employees' professional knowledge and problem-solving ability firstly. This could be realized by entrance training and timely evaluation. The mediation effect of turst indicates that obtaining customers' trust is also important in fostering loyal customers. Wealth management service organization should put more efforts in earning trust. Sharpe et al. (2007) investigated communication strategies which lead to trust. He found that being comfortable (or calm) with client expressions of strong emotions; actively seeking to understand clients' personalities, family history, and values; and involving the consumer in a systematic process for identifying their values and goals are all practical techniques for building consumer trust.

Limitations and Future Research
This research has some limitations unavoidably. First, the study is subject to all of the shortcomings that exist regarding open access sampling (e.g., self-selection bias, lack of information about non-respondents, unknown response rate) (Kuss et al., 2014). However, the full range of scores on all variables was represented in the data, which normally strengthens the validity of estimated relationships between constructs. Second, the sample size is 183, its representativeness may be not enough. Future research can broaden the sample size to enhance the representativeness. Thirdly, we only tested one antecedent (competence), one mediation factor (trust) and one moderation factor (emotional labor) of customer loyalty, other antecedents, mediation and moderation factors should be included in order to obtain a comprehensive understanding of customer loyalty in future.

Funding
Supported by Wealth Management Project of Shandong Technology and Business University (No: 2019ZBKY042).