Customer incivility, employee emotional exhaustion, and job embeddedness relationship in the Indonesian hospitality sector: The socio-economic status perspective

Abstract This study explores the impact of customer incivility on frontline employees’ job embeddedness in the Indonesian hospitality sector. We also investigate whether emotional exhaustion prevalence mediated, and socio-economic status (SES) moderated the link between customer incivility, emotional exhaustion, and job embeddedness. Data were collected in two stages using an online survey of 192 employees from the hospitality sector in Indonesia. We tested a moderated mediation model using Hayes’ Macro PROCESS version 4.0. Results confirmed that frontline service employees with high levels of uncivil consumer behavior experience have high emotional exhaustion tendencies and low job embeddedness. The findings also reveal that emotional exhaustion partially mediates the relationship between customer incivility and job embeddedness. Moreover, SES mitigates the impact of customer incivility on emotional exhaustion and job embeddedness. A lower level of SES triggers a more substantial effect of customer incivility on exhaustion, whereas a higher SES can reduce the effect of customer incivility on job embeddedness.


Introduction
The denial of acting politely to others is a phenomenon that occurs in almost all societies and has contributed to the deteriorating harmonious interpersonal relationships between individuals and groups in the last two decades (Porath & Pearson, 2012). Since first popularized by Andersson and Pearson (1999), workplace incivility has become a topic of interest for researchers and academics worldwide. A distinctive feature of workplace incivility is that it is subtle, ambiguous, and passive, and sometimes there is no clear intention to harm the victim (Andersson & Pearson, 1999). The

PUBLIC INTEREST STATEMENT
Since workplace incivility (co-workers and superiors) has been widely studied can reduce job embeddedness, this study focuses explicitly on incivility from customers, which is relatively rarely studied. Moreover, we reveal socio-economic status as a complementary explanation for the relationship between customer incivility, emotional exhaustion, and job embeddedness. This study also offers an alternative perspective to explore how the combinations of different SES may predispose employees to attitudes and behaviors in Asian cultures when they become victims of uncivilized behavior from consumers.
term "ambiguous" refers to the perpetrator's unclear purpose, which leads to such behaviors receiving little attention from company leaders; this is considered only a difference in perception and misunderstanding (Cahyadi, Hendryadi, Mappadang et al., 2021). However, over the last ten years, researchers and practitioners have increasingly begun to believe that incivility harms both victims and organizations (Jelavić et al., 2021). Initially, incivility from coworkers/supervisors attracted more attention, but researchers have recently begun studying the effects of uncivil customer behavior, especially in the hospitality sector (Alola et al., 2019;Baker & Kim, 2020;Boukis et al., 2020;Cheng et al., 2020;Sliter et al., 2012;Walker et al., 2014). It is essential to learn from the hospitality industry because employees in this sector are frequently trained to prioritize "customer satisfaction" or that "customer is the king" so that they can continue to deal politely with rude customers (Han et al., 2016;Sliter et al., 2012).
The current study has studied the effect of customer incivility on service employees' attitudes and behaviors; however, there are still limited to a few issues that need to be clarified. For example, customer incivility can drain psychological resources, including burnout, emotional exhaustion, and psychological well-being (Alola et al., 2019;Baker & Kim, 2020;Kim & Qu, 2019;. Moreover, it also affects employee turnover intentions and service performance (Alola et al., 2019;Cheng et al., 2020). On the other hand, job embeddedness reflects a person's strength to maintain their current job compared to quitting (Crossley et al., 2007;Mitchell et al., 2001). The role of job embeddedness in the hospitality industry is also increasingly vital in line with the company's efforts to reduce high employee turnover in this sector. Although both customer incivility and job embeddedness are burgeoning in hospitality research, the linkage between these concepts has never been explored in the existing literature. Hence, our study bridges the gap in customer incivility-job embeddedness relationships in the hospitality sector.
Second, through the moderating and mediating mechanisms, we open space to understand the relationship model between uncivil customer behavior, employees' emotional exhaustion, and job embeddedness. For example, frontline employees' exposure to uncivil behavior is positively related to emotional fatigue and exhaustion (Alola et al., 2019;Baker & Kim, 2020;Kim & Qu, 2019;. Meanwhile, the relationship between emotional exhaustion and job embeddedness has also been explored (Dogantekin et al., 2022;Karatepe et al., 2019). Drawing conservation of resource (COR) theory (Hobfoll, 2001) and affective events theory (Weiss & Cropanzano, 1996), we replicate and expand on previous studies by examining the intermediate role of emotional exhaustion in the customer incivility and job embeddedness relationship.
Third, studies on workplace incivility have highlighted the critical role of power differentiation (Cortina & Magley, 2009;Loh et al., 2021;Peltokorpi & Ramaswami, 2021;Porath & Pearson, 2012). For example, Cortina and Magley (2009) found that victims of incivility experienced more severe when the perpetrator had higher status. Similarly, Porath and Pearson (2012) specifically explored the status differences between perpetrators and victims that could influence employees to respond to experiences of incivility with anger, fear, and sadness. In short, status played a crucial role in victims' responses to anger, fear, and sadness. Since the differentiation of power plays a significant role in incivility studies, we propose that social-economic status may also explain the victim of workplace incivility response in Asia.
The present study responds to Ghosh's (2017) call to explore workplace incivility from a social identity perspective. Ghosh argues that social identity plays a vital role in social interaction in the Asian community. It can be a framework to explain the differences in the responses to workplace incivility, including how this situation can foster psychological distress and employee turnover (Ghosh, 2017). Instead, we used a different approach with the proposed social-economic status (SES) to explore employee responses to uncivil behavior from outside the organization (customers). SES is an individual's view of themselves at the socio-economic level (Easterbrook et al., 2020;Sainz et al., 2021). According to Adler et al. (2000), SES refers to income, occupation, and education level to classify their position between the top and bottom of the ladder. Previous researchers have extensively used SES as a framework that can differentiate individual responses to stressor situations. For instance, several researchers have linked levels of SES with well-being (Sainz et al., 2021) and health problems, stress, and depression (Effati-Daryani et al., 2020;Götz et al., 2018;Vinkers et al., 2020). Thus, the present study provides new insights into differences in SES and how it may affect employee responses to customer incivility behavior on emotional exhaustion and job embeddedness, particularly in Asia.

Theoretical framework and hypotheses
In this study, we use the conservation of resources theory (Hobfoll, 2001), affective events theory (Weiss & Cropanzano, 1996), and Hofstede's national culture theory to provide a theoretical explanation of the proposed model (see, Figure 1). According to the COR theory, individuals maintain, protect, and build resources (Hobfoll, 2001). These resources can be objects, personal characteristics, conditions, or energy. In other words, when a person experiences stress at work, they tend to avoid further stress to maintain their resources. Because a person's available resources, including emotional ones, are limited, exhausting one's resources to the point where they are no longer sufficient to cope with workplace stressors (i.e., workplace incivility), triggers emotional exhaustion and psychological distress (Hobfoll, 2001). The COR theory is commonly used as a theoretical basis to explain how workplace incivility affects job satisfaction, stress, and emotional exhaustion (Anjum et al., 2020;Cortina et al., 2001;Hur et al., 2016;Loh et al., 2021).
Following Walker et al. (2014), we devised AET to add to the COR theory to explain the effect of customer incivility on employee attitudes and behaviors. Briefly, the AET provides a logical framework to explain the formation of negative emotions caused by perceived incivility events in the workplace. For example, customer behavior that violates a sense of respect can trigger affective and emotional reactions during the incident (Weiss & Beal, 2005). However, the affective and emotional responses that arise from a positive or negative event depend on personal interpretation. In other words, other employees may react differently to consumer behavior that can be considered rude by an employee. Similarly, Pearson et al. (2001) stated that perceived workplace incivility could have a negative psychological impact on employees in the form of them being "depressed" and "emotionally hurt."

Customer incivility, emotional exhaustion, and job embeddedness
In the context of a work environment, "uncivil behavior" is a violation of workplace norms encouraging mutual respect and cooperation (Pearson et al., 2001). These norms can originate in society, workplace traditions, and various cultures (in the form of formal and informal organizational policies, rules, and applicable procedures). As a universal norm, recognized standards such as "treating others the way you would like to be treated by others" play a role in perceptions of civility. Workplace incivility violates such norms (Pearson et al., 2001). In this study, we use the concept of incivility, which is sourced from customers in the hospitality sector in Indonesia. Workplace incivility includes workplace mistreatment (Andersson & Pearson, 1999), which may be inflicted by coworkers, superiors, or customers (Sliter et al., 2012). The researchers were initially more interested in the issue of incivility originating from coworkers or superiors (Cahyadi  Welbourne et al., 2020). However, in more recent studies, attention has begun to focus on incivility related to customers. Although customer incivility in their various behaviors when interacting with front-line employees does not seem to be a problem (Walker et al., 2014), the consequences on employee attitudes and behavior have recently been getting attention.
In their theory, Maslach and Jackson (1984) divided burnout into three dimensions: emotional exhaustion, depersonalization, and personal accomplishment. Emotional exhaustion refers to the fatigue experienced by individuals in an organization because of social interactions, which causes emotional exhaustion due to the inability of individuals to manage such demands. The other two dimensions are depersonalization, which refers to a loss of feeling at work, and the third involves attempts to diminish personal achievements, because of this an individual loses the motivation to attain competencies and make accomplishments. Compared to the other two dimensions, emotional exhaustion has been studied more because it is considered a core feature of burnout (Schaufeli et al., 2009), and the experience of emotional exhaustion is often perceived as "burnout" (Schaufeli et al., 2009). Therefore, when exploring psychological reactions to workplace incivility, in this study, we consider only emotional exhaustion. Previous researchers have portrayed perceived incivility from consumers can drain the psychological resources of front-line employees (Boukis et al., 2020) and service performance (Cheng et al., 2020). Moreover, perceived customer incivility is also related to burnout, emotional exhaustion, psychological well-being, and employee turnover intentions (Alola et al., 2019;Baker & Kim, 2020;Kim & Qu, 2019;Shin, Lee et al., 2022). Thus, the first proposed hypothesis is as follows.

H1: Customer incivility positively influences employees' levels of emotional exhaustion.
Job embeddedness was introduced by Mitchell et al. (2001) as a "higher-order aggregate of forces for retention." Job embeddedness has received attention from academics and practitioners because of its uniqueness in explaining one's strength to stay in the organization (as opposed to the concept of turnover intention). Mitchell et al. (2001) outlined three dimensions of embeddedness using a positive psychological perspective: links, fit, and sacrifice. "Links" refers to the social bonds developed and maintained by individuals, groups, and teams, both within and outside the work environment (Mitchell et al., 2001). "Fit" refers to employees' association, affinity, or compatibility with the organization or community where they work (Mitchell et al., 2001). Lastly, "sacrifice" refers to the material or emotional losses encountered by employees at the organizational and community levels when they decide to resign from their jobs (Mitchell et al., 2001).
Although existing studies have not explicitly explored the relationship between customer incivility and job embeddedness, we argue that this relationship may exist based on the assumptions of COR theory. Using the COR theory, we postulate that incivility will reduce job embeddedness because employees will reduce links (take distance from perpetrators); they will facilitate direct contact with perpetrators and lose interest in engaging with social relationships at work, as well as reduce social cohesion (Reisig & Cancino, 2004) and interaction avoidance (Peltokorpi, 2019). In the same vein, individuals who receive inappropriate/disrespectful treatment in their work environment experience disturbances in social relations (Hobfoll, 2001). In other words, individuals who experience a victim of uncivil behavior will question whether they are "fit" with the current organizational environment. As a final result, the combination of fit and high sacrifice as the overall concept of job embeddedness (Mitchell et al., 2001) will weaken along with high customer incivility. Thus, we suspect that workplace incivility is directly and negatively related to job embeddedness. Therefore, the next proposed hypothesis is as follows: H2: Customer incivility negatively influences employees' levels of job embeddedness.
The COR theory and AET can be used as theoretical foundations to explain the relationship between emotional exhaustion and job embeddedness. The emotional exhaustion suffered by someone being treated disrespectfully at work triggers other behaviors, such as avoidance (Cortina et al., 2001;Sliter et al., 2012) and reduced engagement (Tricahyadinata et al., 2020). In other words, to protect their emotional resources, individuals reduce their fit and links and reconsider their organizational membership. In addition, efforts to overcome stressors from disrespectful treatment reduce motivation (Hur et al., 2016). Similarly, employees who are emotionally exhausted by workplace incivility may experience reduced engagement and intrinsic motivation to work, and they may increase avoidance-related behaviors to protect themselves against further depletion of their emotional resources. Correspondingly, previous researchers have documented the fact that emotional exhaustion is related to job embeddedness (Karatepe et al., 2019). Because of the relationship between workplace incivility and emotional exhaustion to job embeddedness, it is also necessary to explore the mediating role of emotional exhaustion in the effect of customer incivility on job embeddedness. Past research has also confirmed the existence of a mediating effect of emotional exhaustion on the link between workplace incivility and job satisfaction. Emotional exhaustion has also been proven to mediate the relationship between abusive supervision and job performance . Based on this argument, the subsequent two proposed hypotheses are as follows.
H3: Emotional exhaustion negatively influences employees' levels of job embeddedness.
H4: Emotional exhaustion mediates the relationship between customer incivility and job embeddedness.

The different responses to customer incivility based on social economics status
In Asian societies, which are generally collectivist and have high power distance and masculine characteristics (i.e., Japan, China, India, and Indonesia), social interaction patterns are strongly influenced by identity-based interactions (Ghosh, 2017;Hofstede et al., 2005). Consequently, social status significantly affects how victims of workplace incivility respond to these unpleasant experiences. For instance, Ghosh (2017) highlights that social-cultural status, including religion, caste, and class, causes the powerlessness of uncivil victims when they are treated inappropriately by individuals from the upper class. In contrast, researchers in Western countries focus more on the status of the job hierarchy (Cortina & Magley, 2009;Mullen et al., 2018), where the power relation between the perpetrator and the victim causes a higher level of severity in the victim of incivility (Jungert & Holm, 2022). In other words, victims may perpetuate their victimizations by tolerating uncivil behaviors from groups with higher social status. Apart from differences in power relations sources originating from the power hierarchy within the organization or society, groups at the bottom line experience a higher level of severity because they do not have the power to fight back, so silence may be the best choice (Lam & Xu, 2019).
SES is a self-assessment of individuals about their position in the context of the economic level. Existing studies have consistently reported that a high level of SES is positively related to wellbeing (Navarro-Carrillo et al., 2020;Sainz et al., 2021;Tan et al., 2020). Similarly, low SES has been reported to be associated with higher levels of psychological distress and depression when faced with stressful situations (Effati-Daryani et al., 2020;Götz et al., 2018;Sainz et al., 2021;Vinkers et al., 2020). Although no studies specifically investigate workplace incivility from the SES perspective, it might be related to how it influences a person's emotional exhaustion and job embeddedness. Based on the idea that the level of SES can affect how individuals respond to various situations (Sainz et al., 2021), we expected that a person's SES level would make them different in responding to uncivil behavior from customers towards them. In line with the COR theory, individuals with low resources do not have sufficient strength to cope with workplace stressors (i.e., customer incivility), triggering higher emotional exhaustion and psychological distress (Hobfoll, 2001;Sainz et al., 2021).
The COR assumption mentioned above is in line with the service context where the customer is considered the "king" so that they have higher power and control. Thus, employees will tend to hold back and avoid confrontation (Cortina & Magley, 2009;Ghosh, 2017). Consequently, this powerlessness will lead to higher exhaustion, especially in the low SES group. Furthermore, we also propose that SES also plays a role in influencing the relationship between customer incivility and job embeddedness. In this relationship between customer incivility and job embeddedness, we argue that a high level of SES can neutralize the effect of unpleasant behavior on job embeddedness. According to the COR assumption, a high level of SES needs to be maintained and preserved by employees. In other words, employees at the top level of SES will still be embedded with the job/organization even though they experience uncivil behavior from consumers. This argument is in line with Perreault et al. (2017) were found that full-time employees were more able to cope with stressors than their part-time counterparts. Moreover, SES was also found to be associated with coping strategies in dealing with deviant behavior (Markova & Nikitskaya, 2017). Recently, researchers (Effati-Daryani et al., 2020;Khademian et al., 2021;Rehman et al., 2021) also found different effects of stressors induced by COVID-19, in which individuals who report high levels of economic status tend to have low levels of stress and depression. Based on these arguments, we propose two hypotheses: H5: SES moderates the effects of workplace incivility on emotional exhaustion H6: SES moderates the effects of workplace incivility on job embeddedness

Participants and procedure
This study conducted a field study on frontline service employees in several hotels in Jakarta, Indonesia. Data collection consisted of two phases, each of which was completed at different times. The research team provided a unique code and separate notes on the questionnaire to identify the respondents to facilitate the data collection process for phases 1 and 2. The first phase of this study was conducted in December 2021; respondents were asked to provide biographical information, customer incivility, and SES. Phase 2 was conducted in April 2022; respondents answered questions related to emotional exhaustion and job embeddedness. After obtaining permits from management, 214 respondents at six hotels in Jakarta filled out a paper-based questionnaire. After discarding 22 incomplete questionnaires, 192 were used in the present study. Table 1 presents the respondents' demographic characteristics.
The respondents included 139 women (42.40 percent) and 53 men (27.60 percent). The majority of respondents (83.85 percent) were 20-29 years old, and the rest were 30-39 years old (6.25 percent) or more than 40 years old (9.90 percent). The educational background of the respondents was as follows: 151 had a basic diploma (78.65 percent), 18 had undergraduate or bachelor's degrees (9.89 percent), and 23 had completed senior high school (11.98 percent). A total of 81.25 percent (156 respondents) had an employment tenure of fewer than five years, and the remaining 18.75 percent had worked for more than five years. Finally, 73.96 percent of the respondents were single, and 26.56 percent were married.

Measurement
This study uses a scale used by previous research to ensure its validity and reliability. All items were rated by participants based on a five-point Likert scale based on context (1 = never/ strongly disagree; 5 = always/strongly agree). Customer incivility was measured using a four-item scale developed by Walker et al. (2014). For example: "customer spoke aggressively toward you" and "customer used a tone when speaking with you." The emotional exhaustion scale was adapted from the Maslach Burnout Inventory (Maslach & Jackson, 1984). Some typical examples included "I feel frustrated by my job" and "I am emotionally drained from my work." Job embeddedness was measured using a seven-item scale from the global job embeddedness scale developed by Crossley et al. (2007). An example of a typical item was "I feel embedded in this organization." All scales have met internal consistency with Cronbach alpha with 0.77, 0.87, and 0.93 for customer incivility, emotional exhaustion, and job embeddedness, respectively (see, Table 2).
The SES was measured using the objective SES approach (Adler et al., 2000), which considered income, occupation, and education level to classify the top and bottom of the ladder in SES. However, because all the respondents were at the same level (frontline service), we changed positions with employee status coded: 1 = contract; 2 = permanent. The education level is coded: 1 = senior high school degree, 2 = diploma degree, and 3 = bachelor's degree. For salary, we classify employee income into two groups based on the Indonesian government income tax regulations: 1 = employees who earn less than 4 million rupiahs per month and 2 = employees who earn more than 4 million rupiahs per month. Following Adler et al. (2000), a composite score from 3 categories (status, education, and salary) was created to obtain an average score by standardizing each indicator.  Table 2 shows that all loading indicators are> 0.70, indicating that all items have good reliability. The AVEs value > 0.50 for all constructs also shows that all measurement models used have met convergent validity. Fornell and Larcker's (1981) criterion is used to assess discriminant validity in this study by comparing the construct's AVEs in all constructs with the squared interconstruct correlation (see , Table 3). Based on these criteria, there is no AVE's < square of the correlation between constructs, indicating that discriminant validity has been satisfied (Hair et al., 2019). Finally, the variance inflation factor (VIF) value is <3, indicating no severe problem with multicollinearity.

Common method bias and descriptive statistics
This study uses single and self-reported data (employees) to have a common method bias (Podsakoff et al., 2012). Considering these conditions, we carried out several procedures. First, we conducted control procedures by taking data at two different times (time-lag) and varying the answer choices on the questionnaire items (1 = never/strongly disagree; 5 = always/strongly agree). Second, statistical control is carried out with the Harman single-factor model, where the average variance extracted (AVE) value is used as a guide to determine whether the data is exposed to common bias. Kock (2017) explains that AVE < 0.40 indicates that the common method variance (CMV) in the data is not a serious violation. The results in Table 3 show that the AVE 0.16 is below the cut-off value of 0.40; thus, the currently used data have no severe CMV violations (see , Table 3). Table 3 shows the research variables' means, standard deviations, and correlations. The results of the descriptive analysis showed that the mean scores for emotional exhaustion, SES, and job embeddedness were above the midpoint on each scale. Meanwhile, perceived customer incivility had an average score slightly below the mean of 2.49. Following the results of the correlation analysis, customer incivility was positively correlated with exhaustion and SES and negatively correlated with job embeddedness. Meanwhile, emotional exhaustion was negatively correlated with job embeddedness.

Hypothesis testing
In this study, we employed hierarchical regression analysis using macro PROCESS version 4.0 (Model 8) developed by Hayes (2017) to test the hypotheses. Table 3 shows that none of the control variables, such as age, sex, education, job tenure, or marital status, significantly affected emotional exhaustion and job embeddedness. The first hypothesis examines the relationship between customer incivility and emotional exhaustion, and the results show a positive and significant relationship (b = .29, p < .01), thus supporting H1. The second hypothesis shows that customer incivility was negatively and significantly related to job embeddedness (b = −.21, p < .01), supporting H2. Similarly, emotional exhaustion in the present study was found to be negatively and significantly related to job embeddedness (b = −.24, p < .01), supporting H3. H4 reflects the indirect effect of customer incivility on job embeddedness via emotional exhaustion. As Table 5 shows, the mediation is significant only at lower level of SES. At high-SES levels, the indirect relationship between customer incivility and job embeddedness was not significant. Therefore, H4 is not fully supported.
The results of testing the moderating role of SES are shown in Tables 4 and 6. First, as shown in Table 4, the value of interaction (customer incivility x SES) has a negative effect on exhaustion (b = −.12, p < .05). Similarly, the interaction effect between customer incivility and SES was also significant (b = .16, p < .05). Therefore, this result supports H5 and H6. Second, the test results of the previous interaction hypothesis demonstrate that SES plays a moderating role in customer incivility-emotional exhaustion and customer incivility-job embeddedness relationships, as detailed in Table 6. The effect of customer incivility on emotional exhaustion is more significant at a low SES (b = .39, p < .01); and its effect is reduced at a high-SES (b = .19, p < .01). The findings suggest that the customer incivility-emotional exhaustion relationship is significant at both lowand high-SES and that the relationship becomes more substantial at a low-SES. Furthermore, with a 95 percent bias correction, the relationship between emotional exhaustion and job embeddedness is insignificant (b = −.08, p > .05) at a high level of SES. Meanwhile, the effect of emotional exhaustion on job embeddedness becomes significant when the SES is low  Index of moderated mediation Index = .03; SE = .02; LLCI = .00; ULCI = .08 Note: N = 192; SES = socio-economic status; CIC = customer incivility; EX = emotional exhaustion; JE = job embeddedness; CL, confidence level; LL = lower level; UL = upper level.
(b = −.34, p < .01). These results indicate that the negative effect of customer incivility on job embeddedness is significant at low-SES levels.

Discussion
The main goal of our study is to explore how frontline service employees respond to customers' uncivil behavior in the hospitality sector. We replicated previous studies to investigate the detrimental effect of workplace incivility on work-related attitudes (emotional exhaustion and job embeddedness). In addition, the present study was intended to explore to what extent the perceived socio-economic status can distinguish employee responses to uncivil behavior from consumers. The mediation moderation analysis was applied to 192 frontline service employees. The results indicated that high levels of perceived customer incivility foster high emotional exhaustion and decrease job embeddedness. The mediation analysis indicates that emotional exhaustion also acts as a mediator between customer incivility and job embeddedness. Furthermore, the results of the study also reveal the vital role of SES in the proposed model. In general, SES originating from social disparities reinforces customer incivility's positive and negative effects on emotional exhaustion and job embeddedness.
The first contribution of this study is that it reveals the relationship between customer incivility and employees' job embeddedness which had yet been unexplored by previous researchers. The results of this study reveal that a high level of perceived customer incivility is associated with a decrease in employees' embeddedness in their job/organizations. In the same vein, our study complements and supports existing studies of perceived customer incivility, which can reduce service performance (Cheng et al., 2020) and increase turnover intention (Alola et al., 2019) in the hospitality sector. The findings of this study provide compelling empirical evidence that perceived uncivil behavior can cause employees to consider whether they are "fit" for their current job; they also support the assumptions of COR theory and AET. In other words, the accumulation of uncivil events received by employees over time will reduce their embeddedness in their current job, particularly in the hospitality sector.
The second contribution to the existing literature is to highlight "how" the link between customer incivility and job embeddedness is mediated by emotional exhaustion. The results of our study confirm that perceived customer incivility directly affects job embeddedness or indirectly through Note: SES = socio-economic status; CL = confidence level; LL = lower level; UL = upper level. Gustiawan et al., Cogent Social Sciences (2023) emotional exhaustion. In other words, employees who have high-frequency uncivil experiences from customers will tend to experience emotional exhaustion, which triggers a decrease in job embeddedness. Hence, we support COR theory and AET as a framework for explaining how customer incivility causes emotional exhaustion in employees, which eventually reduces employees' level of job embeddedness. The present study extant existing research demonstrating the relationship between customer incivility, emotional exhaustion, and other work outcomes (Alola et al., 2019;Baker & Kim, 2020;Kim & Qu, 2019;, while our research shows that customer incivility is directly and indirectly proven to affect job embeddedness significantly. Our findings are consistent with previous studies on the relationship between incivility and emotional exhaustion (Alola et al., 2019;Baker & Kim, 2020;Kim & Qu, 2019; and the link between emotional exhaustion and job embeddedness (Dogantekin et al., 2022;Karatepe et al., 2019). Thus, the results of our study established an alternative model of the antecedents of job embeddedness based on the prevalence of emotional exhaustion (a mediator) as a mechanism for mediating the relationship between customer incivility and job embeddedness.
Third, we introduce SES as a complementary factor affecting how workplace incivility affects emotional exhaustion and job embeddedness. In the context of victims and perpetrators of rudeness, differences in status and hierarchy have been discussed (Cortina & Magley, 2009;Hershcovis et al., 2017;Porath & Pearson, 2012). In this study, we found different roles of SES where the interaction effect was positive for customer incivility and emotional exhaustion and negative for customer incivility and job embeddedness. For emotional exhaustion, participants in the low SES condition suffered more and reported higher emotional exhaustion compared to the middle and high SES groups when receiving impolite behavior from customers. This pattern of findings shows that employees who had a low SES level are significantly more exhausted than those who report SES at a medium or high SES level when they are victims of customer incivility. This result is reasonable considering that a low level of SES has a lower bargaining position (less power) in the organization than the group above it, so they perceived more suffering from the adverse outcomes of customer incivility. In the same vein, in line with the COR assumption, the low SES group has fewer resources and is no longer sufficient to cope with workplace stressors (i.e., customer incivility), triggering higher emotional exhaustion and psychological distress (Hobfoll, 2001;Sainz et al., 2021).
Socioeconomic status also plays an essential role in maintaining employees' job embeddedness when they experience high levels of customer incivility. In the second situation, the negative effect of customer incivility on job embeddedness is significant in the group of employees reporting low and medium SES and becomes insignificant in a high-SES group. This finding indicates that employees at high levels of SES have different responses to customer incivility than the other two groups, particularly regarding job embeddedness. In other words, although employees experience impolite behavior from customers, their attitude toward remaining in the organization is relatively stable. Instead, using the job embeddedness perspective, employees who report low and medium levels of SES are more affected by customer incivility and are more likely to experience "sacrifice" and reconsider their "fit" with the organization. Moreover, our findings support Ghosh's (2017) argument that employees at low social levels tend to experience higher psychological distress and turnover intention when they become victims of incivility at work.
In short, SES is a personal resource that can influence how individuals define themselves In short, SES is a personal resource that can influence how individuals define themselves (Easterbrook et al., 2020;Sainz et al., 2021); these differences in responses based on SES have been extensively studied in the context of organizational behavior. For instance, individuals with low levels of SES tend to report lower well-being (Sainz et al., 2021) and are more prone to health problems, stress, and depression (Effati-Daryani et al., 2020;Götz et al., 2018;Vinkers et al., 2020). SES differences might be considered further for the observed differences in responses to uncivil behavior on employees' emotional exhaustion and job embeddedness. The SES could be an exciting topic for future studies that can offer new insights; as an alternative perspective from hierarchical status and socio-cultural levels (Ghosh, 2017;Lam & Xu, 2019) in studying workplace incivility and its impact on employee work-related behavior, particularly in Asia.

Practical implications
The present study's findings are important for managers and organizations in the hospitality sector. The results show how workplace incivility can increase emotional exhaustion and reduce job embeddedness; thus, policymakers must take anticipatory steps to minimize the harmful effects of workplace incivility. First, the management must identify consumer behavior that can be marked as uncivil behavior. We recommend conducting an internal survey to identify specific types of behavior that employees consider disrespectful. Accordingly, management can devise appropriate anticipatory strategies to reduce potential consumer-employee conflicts. This exploration is essential to determine the response that employees need to make in the face of impolite behavior from consumers. After the management has succeeded in classifying various forms of uncivil (for example, physical or verbal, rude or overtly aggressive words, and other forms), an action plan can be carried out. For example, a consumer who misbehaves because of the influence of alcohol may lead to a widespread and severe escalation that requires security measures to be taken.
Second, the management must equip all frontline service employees with cross-cultural ethics of politeness. Most importantly, before being assigned to the frontline, all service employees could be trained to recognize and differentiate mild to more uncivil severe behavior (e.g., physical aggression), which can be responded to differently. Third, management needs to be aware that uncivil behavior, whatever its form, can cause emotional exhaustion and decrease job embeddedness. Hence, the management needs to open channels for employees to submit their complaints in dealing with consumers. In addition, particularly for inexperienced frontline employees, management must implement policies to provide assistance and supervision from senior supervisors to train them in handling situations in the field more effectively.
Finally, the management needs to provide factual support to raise the level of their employees to a higher level of economic. For example, management can provide scholarship support to young employees to continue their education at the university level, provide measurable career paths for all employees, and provide financial rewards that are commensurate with the workload and results. Increasing self-capacity through motivational training also needs to be considered to increase employee self-esteem.

Limitations and suggestions for future research
Although this study offers several important theoretical and practical implications, such as those for scientific research, we must mention its limitations as a caution for those undertaking future studies. First, in this study, we used a cross-sectional approach twice and in other groups; however, using self-reported data from questionnaires may cause a common bias (Podsakoff et al., 2012). Although the study did not exhibit a common bias according to Harman's single-factor test, the conclusion regarding causality should still be interpreted with caution. Therefore, in future studies, researchers should adopt a longitudinal design to overcome this limitation.
Second, the workplace incivility we investigated came from a single source (customer) to frontline employees in Indonesia's hospitality sector. Our study does not consider disrespect from the coworker/ supervisor perspective (Kim & Qu, 2019;Lam & Xu, 2019). Future researchers could integrate incivility from internal sources (e.g., coworkers) and compare its effects on emotional exhaustion and job embeddedness with those of internal incivility. Third, the majority of respondents in this study are men, who may react differently than women in situations involving customer incivility. We recommend that future studies use a more gender-balanced sample or separate data analyses (e.g., multigroup analysis). Finally, because employee perceptions of workplace incivility differ across cultures (Eka & Chambers, 2019;Ghosh, 2017), this study may be biased based on a specific society's social values and norms. To address this issue, we invite future researchers to compare the results of this study with those obtained in different cultural contexts.