Customer service at altitude: e ﬀ ects of empowering leadership

Purpose – This paper aims to examine how team-level empowering leadership related to service performance through thriving at work and how shared organizational social exchange and customer orientationmoderatedthe latter relationships. Design/methodology/approach – The authors collected the data from 283 ﬂ ight attendants and their supervisors working at a major Korean airline. Multi-level analyses were used to testthe effect of empowering leadership onemployee outcomes. Findings – Both team-level empowering leadership and customer orientation were signi ﬁ cantly and indirectly associated with service performance via thriving at work. Additionally, customer orientation signi ﬁ cantly moderated the relationship between team-level empowering leadership and thriving at work such that the relationship was stronger when customer orientation was low rather than high. In addition, shared organizational social exchange augmentedthein ﬂ uenceofteam-levelempoweringleadershiponserviceperformancebutnotonthrivingatwork. Practical implications – The ﬁ ndings suggest that team-level empowering leadership is more effective in enhancing thriving at work of employees when their customer orientation is low rather than high. In addition, a shared high-quality organizational social exchange augments the effect of empowering leadership onemployees ’ service performance. Originality/value – This paper provides initial evidence of the interaction of team-level empowering leadership and individual ´ -level customer orientation on thriving at work and service performance. Additionally, it documents the differential augmenting effect of shared organizational social exchange on the relationship between empowering leadership and these outcomes. Collectively, the ﬁ ndings explain why and when team-level empoweringleadership relatesto serviceperformance.

The increasingly diverse needs and expectations of customers underscore the importance of discretion in frontline employees' (hereafter employees) service performance and ultimately, customer and organizational outcomes (Ahearne et al., 2005;Jiang et al., 2015). Perhaps because of its implications for fostering employees' experience of discretion (Hartline and Ferrell, 1996), much research has examined as well as documented the role of empowering leadership in enhancing employees' motivational states and the resulting service performance (Ahearne et al., 2005;Lee et al., 2018). However, there is widespread recognition that individual and organizational factors play a significant role in how individuals respond to leadership behaviors (Cheong et al., 2019;Lorinkova and Perry, 2017;Zhang and Bartol, 2010). Given the growing adoption of a customer-oriented strategy among service organizations as well as the documented influence of customer orientation on service performance (Brown et al., 2002;Huo et al., 2019), customer orientation can constitute an important individual difference variable that shapes reactions to empowering leadership. Adopting a unit rather than an individual (Li et al., 2016) level perspective of leadership style, the first objective of this study is to examine the independent and joint influence of team-level empowering leadership and customer orientation on employees' thriving at work and ultimately, their service performance.
In addition, although employees' reactions to leadership can be affected by the quality of the relationship they have with their organizations, prior research has paid little attention to examining social exchange-based team-level boundary conditions for the linkages between empowering leadership and employee work outcomes (Li et al., 2017;Lorinkova and Perry, 2017; for exceptions). Lorinkova and Perry (2017) examined the moderating influence of leadermember exchange on the relationship between empowering leadership and psychological empowerment while Li et al. (2017) examined how organizational support climate moderated the linkages between empowering leadership and citizenship behaviors. Although leaders are perceived as representatives of an organization, employees can also develop an independent or separate relationship with their organizations (Rupp and Cropanzano, 2002). Thus, a second objective of this study is to examine how shared organizational social exchange moderates the relationship between empowering leadership and service performance and its indirect influence on service performance through thriving at work.
Our study provides several important contributions to the service management literature. First, while empowering leadership has been shown to influence employees' attitudinal and behavioral outcomes (Chen et al., 2007;Cheong et al., 2019;Sharma and Kirkman, 2015;Zhang and Bartol, 2010), it is still not clear how it can motivate individuals with different levels of customer orientation. The authors address this gap by examining how team-level empowering leadership and individual-level customer orientation interactively affect thriving at work. Second, by focusing on team-level rather than perceived empowering leadership, it extends Li et al.'s (2016) study by examining leadership as a multilevel phenomenon. Third, by focusing on customer orientation this study adds to prior research that has shown the relational resources of psychological capital (Paterson et al., 2014) and regulation foci (Wallace et al., 2016) as antecedents to thriving at work. Also, although customer orientation has been shown to engender motivational states such as work engagement (Zablah et al., 2012), this paper focuses on a hitherto unexamined motivational pathway, thriving at work, through which it relates to service performance. Figure 1 depicts a schematic representation of the relationships we proposed and examined.
Theory and hypothesis development Team-level empowering leadership and thriving at work. Perhaps as a harbinger of the positive psychology movement (Seligman, 2005), self-determination theory (Deci and Ryan, Customer service at altitude 2000; Gagne and Deci, 2005) seeks to understand the environmental conditions that satisfy psychological needs and ultimately, foster optimal human functioning. Deci and Ryan (2000) posit different regulatory processes ranging from autonomous motivation that foster selfdetermined behavior to controlled motivation that foster prescribed or reward/punishment contingent behavior. The authors focus in this study on autonomous motivation because of the discretion that is required for employees to flexibly respond to the diverse needs and expectations of customers. As a contextual factor, team-level empowering leadership can engender the self-determined behaviors that foster thriving at work. Empowering leadership describes a leadership perspective that provides guidelines for leaders in terms of distributing and exercising their power (Vecchio et al., 2010). Specifically, it denotes a process whereby leaders delegate decision-making power to employees, express confidence in their capabilities in handling challenging responsibilities, hold them accountable, enhance meaningfulness in work, and provide them with the resources including support to perform their responsibilities (Ahearne et al., 2005;Arnold et al., 2000). The empowering leadership dimension of participation in decision-making and the related accountability enhance employees' sense of meaning as well as provide opportunities to learn, leading to thriving at work (cf. Kleine et al., 2019). Expressing confidence in employees' abilities to handle challenging work also provides learning opportunities and a sense of vitality thereby promoting psychological growth or thriving at work (Carver, 1998). The meaning and positive energy that empowering leadership fosters enable employees to be task-focused as well as engage in self-determined behaviors that engender vitality and learning, subsequently leading to thriving at work. While much research has reported that empowering leadership at both team-and individual-level relates to psychological empowerment (Seibert et al., 2011;Zhang and Bartol, 2010), we are aware of only one study that reported individual level empowering leadership to relate to thriving at work (Li et al., 2016). As leaders' interpersonal style has been noted to constitute an ambient social context in which employees operate (Seibert et al., 2011), the authors expect team-level empowering leadership to affect employees' vitality and learning, thereby fostering thriving at work.
H1a. Team-level empowering leadership significantly relates to thriving at work.
Customer orientation and thriving at work. As an individual level manifestation of the marketing concept, customer orientation constitutes a foundational construct in a service organization's efforts to create and sustain competitive advantage (Donavan et al., 2004;   Huo et al., 2019). This study adopted a psychological rather than a behavioral dimension which describes customer orientation as a "tendency or predisposition to meet customer needs in an on-the-job context" (Brown et al., 2002, p. 111) or "the extent to which service providers are willing to put forth time and effort to satisfy their customers" (Susskind et al., 2003, p. 181). The authors expect customer orientation to positively relate to thriving at work because it underpins the agentic behaviors (task focus, exploration and heedful relating) that Spreitzer et al. (2005) identified in their socially embedded model of thriving. High customer orientation employees will put forth time and effort into understanding the needs and expectations of customers. The resulting search or exploratory activities will provide them opportunities to acquire new skills and knowledge as well as feel energized. In support of this, Spreitzer et al. (2005) noted that being task focused encourages learning because such individuals tend to develop and refine routines to do their work better. In addition, inherent in customer orientation is a desire to genuinely understand and satisfy the needs of customers which encourages heedful relating. As Spreitzer et al. (2005) observed, heedful relating enables individuals to provide social support which not only engenders vitality but also, learning, as feel responsible for the outcomes of the larger system.
H1b. Customer orientation positively relates to thriving at work.
Mediating role of thriving at work. Drawing on self-determination theory, the autonomous regulation inherent in thriving at work enables employees to engage in agentic behaviors by taking responsibility for acquiring information/knowledge which they then leverage to effectively address the needs and expectations of customers. Additionally, thriving at work fosters a positive mood that enables employees to meaningfully engage with customers, thereby enhancing their (customers') experience of the service delivery process. As Spreitzer et al. (2005) argued, employees with high thriving at work can produce knowledge resources, positive meaning, positive affective resources and relational resources which they can invest in the service delivery process to enhance the quality of their service performance. In support of these arguments, thriving at work has been reported to have a direct influence on job performance (Paterson et al., 2014) and to mediate the influence of empowering leadership on change-oriented citizenship behaviors (Li et al., 2016) as well as the influence of promotion focus on innovative behavior (Wallace et al., 2016). Taken as a whole, the preceding discussions and prior research imply an indirect effect model where both customer orientation and team-level empowering leadership relate to thriving at work which, in turn, affect service performance.
H2a. Thriving at work mediates the relationship between team-level empowering leadership and service performance.
H2b. Thriving at work mediates the relationship between customer orientation and service performance.
The moderating role of customer orientation. Although both team-and individual-directed empowering leadership have been shown to relate to individual level motivational states (Cheong et al., 2019;Seibert et al., 2011), the question of who benefits from empowering leadership has not been fully addressed. Specifically, while much of the extant research has examined myriad boundary conditions of the empowering leadership-motivational state (psychological empowerment) relationship (Sharma and Kirkman, 2015), there is a dearth of research on when empowering leadership relates to thriving at work. Indeed, the only research known to us examined autonomy orientation as an individual-level moderator for this relationship (Li et al., 2016). This study contributes to this stream of research by Customer service at altitude examining customer orientation as an individual-level moderator of the team-level empowering leadership-thriving at work relationship. Despite the documented importance of customer orientation on employees' motivational states and ultimately performance (Huo et al., 2019;Zablah et al., 2012), not all employees can be said to be high in customer orientation. Consequently, understanding when low customer orientation employees experience the motivational states that engender effective service performance is important. As previously argued, discretion or empowerment is considered a prerequisite in service performance because employees "need the flexibility to make on the spot decisions to completely satisfy customers" (Hartline et al., 2000, p. 56). However, given their predisposition to be focused on the service delivery process and to satisfy customer needs, high customer orientation employees will be motivated to engage in self-determined behaviors such as meaningfully interacting with customers to learn about their needs and possess the emotional resources to satisfy these needs. Additionally, engaging in these self-determined behaviors will contribute to high customer orientation employees' self-verification as individuals who place a high priority on satisfying customer needs and expectations. Consequently, the effects of empowering leadership on contributing to enhanced levels of thriving at work would be marginal for high customer orientation employees.
In contrast, low customer orientation employees do not assume responsibility for learning about the needs and expectations of customers. For this reason, they will only superficially engage with customers and will therefore, not have the knowledge and information to satisfy the needs and expectations of customers. Additionally, they will be less likely to engage in selfdetermined behaviors as they are more passive, less energized, and perhaps more helpless when it comes to responding to the diverse needs and expectations of customers. Relative to high customer orientation employees who desire to satisfy customers, low customer orientation employees are said "to experience an internal resource deficit, face difficulty in coping with job demands [. . .. . .]." (Babakus et al., 2009, p. 484). Consequently, low customer orientation employees will be much more dependent on the resources provided by team-level empowering leadership to engage in self-determined behaviors that energize and enhance learning, leading to thriving at work. Taken together, the preceding arguments suggest a substitutive effect such that team-level empowering leadership plays a less important role in enhancing thriving at work as customer orientation increases.
H3a. Customer orientation moderates the effect of team-level empowering leadership on thriving at work such that this relationship becomes stronger for low rather than high customer orientation employees.
Taken together, the preceding discussions suggest a moderated mediation effect. Specifically, the indirect effect of empowering leadership on service performance through thriving at work is conditional upon customer orientation such that the indirect effect becomes stronger for employees with low (rather than high) customer orientation. Thus, we hypothesize that: H3b. Customer orientation moderates the indirect effect that team-level empowering leadership has on service performance via thriving at work such that the indirect effect becomes stronger for employees with low (rather than high) customer orientation.
Moderating role of shared organizational social exchange. Although team-level empowering leadership shapes the attitudes and behaviors of team members, the exercise of leadership operates in an organizational context. That is, team members' experience of work is shaped IJCHM 31,9 both by their experience of the leader's behaviors as well as those of the organization which, the authors conceptualized as shared organizational social exchange. A social exchange relationship describes a long-term, mutually invested relationship characterized by the exchange of both economic and socio-emotional resources (Rupp and Cropanzano, 2002;Shore et al., 2006). Accordingly, the authors expect team-level empowering leadership to interact with a team's shared organizational social exchange to facilitate employees' receipt of resources such as support, discretion, respect and trust that will engender the agentic behaviors (Spreitzer et al., 2005), which can make employees more energetic. On the other hand, in a context of a low quality shared organizational social exchange, a high team-level empowering leadership may signal conflicting messages to team members about the extent to which they can obtain the resources to promote the agentic behaviors leading to thriving at work. In such context, a high team-level empowering leadership as practiced by a specific leader will not be perceived as an embodiment of the nature of the organization and may therefore, not be sustainable (Eisenberger et al., 2010). As a result, it may not boost team members' sense of self-determination and the resulting agentic behaviors that enhance the motivational state of thriving at work. Taken together, the authors propose that: H4a. Shared organizational social exchange moderates the relationship between teamlevel empowering leadership and thriving at work such that this relationship becomes stronger as shared organizational social exchange increases.
Implicit in the preceding discussion is the view that the indirect effect of team-level empowering leadership on service performance via thriving at work is conditional upon the quality of shared organizational social exchange. Specifically, shared organizational social exchange enhances the indirect effect.
H4b. Shared organizational social exchange moderates the indirect effect that team-level empowering leadership has on service performance via thriving at work such that the indirect effect becomes stronger as shared organizational social exchange increases.
In addition to examining its moderating role in the cross-level motivational implications (thriving at work) of team-level empowering leadership, the authors examined shared organizational social exchange as a boundary condition of the performance implications (service performance) of team-level empowering leadership. As previously discussed, attributes of empowering leadership such as delegating power to make decisions will engender employees' discretion to provide adequate service that can meet customers' unique needs and expectations. However, as the service delivery process takes place in an organizational context, we expect the psychological resources available to teams that have a shared high-quality organizational social exchange to augment those provided by team-level empowering leadership. Collectively, these resources will engender a facilitative context that provides employees discretion to engage in self-determined behaviors to meet customers' expectations. In contrast, a shared low quality organizational social exchange context will signal uncertainty about the receipt of resources or nutriments to sustain self-determined behaviors. Consequently, employees will not be able to meaningfully engage with customers to adapt the service delivery to their expectations. Thus, the authors hypothesize that:

Customer service at altitude
H4c. Shared organizational social exchange moderates the effect of team-level empowering leadership on service performance such that this relationship becomes stronger as shared organizational social exchange increases.

Method
Participants and procedure. Respondents were cabin crew and their inflight supervisors from a large Asian airline based in South Korea. A cabin crew comprises flight attendants that form the customer service unit on a flight, work together in a closed space for the duration of the flight, and report directly to the same inflight supervisor. On consenting to participate in the survey, a human resource (HR) officer prepared a list of flight attendants and their inflight supervisors. To be considered for participation in the study, members of a cabin crew must have worked together on several flights over a specified period under the same inflight supervisor. To facilitate the matching of subordinate-supervisor responses, codes were written at the top right corner of each of the questionnaires. The questionnaires were administered to all cabin crew listed by the HR officer during their non-flight working hours by one of the coauthors with the help of a research assistant. Completed questionnaires were returned sealed in envelopes to the HR officer who coordinated the survey. A week later, inflight supervisors were invited to rate the service performance of each of the members of their cabin crew who had participated in the survey.
Of the 450 questionnaires distributed, a total of 283 matched respondent-supervisor questionnaires were returned and used in the analyses. Employees were 90 per cent female, reported an average age of 29 years (SD = 5.12), average organizational tenure of 6.00 years (SD = 4.87), average formal education of 15.71 years (SD = 1.60), and worked an average of 23.20 hours (SD = 8.08) per week. Supervisors were 80 per cent female, reported average age of 43.72 years (SD = 3.23), and average organizational tenure of 20.81 years (SD = 2.53).
Measures. The questionnaires were originally constructed in English and administered in Korean using Brislin's (1980) recommended back-translation procedure. The authors used a five-point Likert type scale, ranging from (1) "Strongly disagree" to (5) "Strongly agree", except for service performance with a seven-point Likert type scale, ranging from (1) "Highly unsatisfactory" to (7) "Highly satisfactory." A complete list of the items for the key variables is presented in the Appendix.
Team-level empowering leadership. Team-level empowering leadership was assessed with Ahearne et al.' (2005) 12-item scale. An example item is "My immediate supervisor makes many decisions together with me".
Shared organizational social exchange. To measure shared organizational social exchange, Shore et al. (2006) eight-item scale was used. A sample item is "My relationship with my organization is based on mutual trust." Customer orientation. Customer orientation was assessed using Donavan et al.'s (2004) 13-item scale. An example item is "I thrive on giving individual attention to each customer".
Thriving at work. To measure thriving at work, Porath et al.'s (2012) ten-item scale was used. An example item is "I feel alert and awake at work".
Service performance. Service performance was assessed with a seven-item scale reported by Chuang and Liao (2010). Supervisors were asked to assess the extent of their satisfaction with their employees' performance. A sample item is "Helping customers (passengers) when needed".
Control variables. This study controlled for the demographics of tenure and education (both measured in years) and average hours worked per week. Tenure and education were controlled because they have implications for expertise and knowledge of customers which should impact both thriving at work and service performance (Babakus et al., 2009). Given IJCHM 31,9 its implications for fatigue (Chan and Wan, 2012), and therefore thriving at work and service performance, average hours worked per week was controlled. This study also controlled for intrinsic motivation that can influence employee performance (Cerasoli et al., 2014). Intrinsic motivation was measured with a four-item scale reported by Grant (2008).
Analytical strategy. Given the nested nature of the data (i.e. employees were nested within teams), multi-level analyses were conducted with Mplus 7.4 (Muthén and Muthén, 2012). Specifically, we tested team-level empowering leadership and shared organizational social exchange as group level (L2) variables, and customer orientation, thriving at work and service performance as individual level (L1) variables. In addition, to test the hypothesized interactive effects of customer orientation and team-level empowering leadership, group-mean centering was used for customer orientation to eliminate the potential confounding effects of the between group interaction between team-level empowering leadership and customer orientation (Hofmann and Gavin, 1998). Finally, to test the moderating effect of customer orientation and shared organizational social exchange on the hypothesized indirect influence of empowering leadership on service performance via thriving at work, we calculated the confidence intervals (CIs) using the Monte Carlo simulation (Preacher et al., 2010).

Results
To test the distinctiveness of key variables in the study, confirmatory factory analyses (CFAs) were run for the six-factor model (empowering leadership, organizational social exchange, customer orientation, intrinsic motivation, thriving at work and service performance). Following Nasser and Wisenbaker (2003), item-parceling procedure was adopted to achieve an adequate ratio. This study used standardized root mean square residual (SRMR), the comparative fit index (CFI), and the root mean square error of approximation (RMSEA) to assess model fit. The results showed that the six-factor model adequately fits the data (x 2 (174) = 357.79, df = 174, p < 0.01. SRMR = 0.05, CFI = 0.96, RMSEA = 0.06). In addition, given the single-source and nested nature of employee-rated data, we conducted a CFA for the employee-assessed variables using multilevel CFA (Dyer et al., 2005). The five-factor model had a good fit to the data (x 2 (125) = 218.89, p < 0.01, SRMR = 0.05, CFI =0.97, RMSEA = 0.06). These results underscore the distinctiveness of the key variables thereby justifying our decision to test the hypothesized relationships.
Furthermore, we examined whether the data justified aggregation of the team-level constructs (James et al., 1984). Results of a uniform null distribution revealed mean values of rwg(j) across teams of 0.95 for empowering leadership and 0.94 for organizational social exchange indicating adequate within team agreement. In addition, the values of ICC(1) and ICC(2) were 0.22 and 0.59 for empowering leadership, and 0.12 and 0.43 for organizational social exchange, providing sufficient basis to support aggregating empowering leadership and organizational social exchange to the team level.
The descriptive statistics, reliabilities, and correlations for all the study variables were reported in Table I. The reliability estimates for all measures were acceptable (i.e. a > 0.70).
In addition, Table I shows that empowering leadership and organizational social exchange at the individual level were positively and significantly correlated to service performance (r = 0.26, p < 0.01; r = 0.31, p < 0.01, respectively).
The results for H1a, H1b, H2a and H2b are summarized in Table II. Consistent with H1a and H1b, both team-level empowering leadership and customer orientation were positively and significantly associated with thriving at work (b = 0.38, p < 0.001; b = 0.52, p < 0.001, respectively). In addition, thriving at work was positively and significantly associated with service performance (b = 0.45, p < 0.01). In addition, team-level empowering leadership  . These results support H2a and H2b. As shown in Table II, although both team-level empowering leadership and customer orientation were positively and significantly associated with intrinsic motivation (b = 0.46, p < 0.001; b = 0.78, p < 0.001, respectively), intrinsic motivation was not significantly associated with service performance after taking into account thriving at work (b = 0.16, n. s.). These findings indicate that intrinsic motivation did not provide an alternative explanation for the respective influence of empowering leadership and customer orientation on service performance. H3a suggests a cross-level interaction between team-level empowering leadership and individual-level customer orientation on thriving at work. Model 1 in Table III shows that the interaction term was significant (b = À0.31, p < 0.01). Specifically, the relationship between team-level empowering leadership and thriving at work was positive and significant when customer orientation was low (simple slope = 0.33, p < 0.01), but it became non-significant when customer orientation was high (simple slope = 0.01, n.s.). These results are depicted in Figure 2. Thus, H3a was supported.
In addition, we tested whether customer orientation would significantly moderate the indirect effect of empowering leadership on service performance through thriving at work. The Monte Carlo simulation results showed that the moderated mediation effect was significant (moderated mediation = À0.24, 95 per cent CI = [À0.44, À0.06]). Specifically, the indirect effect was not significant when customer orientation was high (indirect effect = À0.04, 95 per cent CI = [À0.18, 0.09]), but significant when customer orientation was low (indirect effect = 0.21, 95 per cent CI = [0.06, 0.38]). These results supported H3b.
To test H4a which posit shared organizational social exchange to moderate the relationship between empowering leadership and thriving at work, a random coefficients Notes: N (employees) = 283; N (team) = 54; * < 0.05; ** < 0.01; *** < 0.001; CI = Confidence Interval Customer service at altitude model was examined. Specifically, thriving at work was regressed on customer orientation at the individual level, team-level empowering leadership and shared organizational social exchange, and their interaction term at the group level (Model 2 in Table III). As shown, the interaction term was not significant (b = À0.04, n.s.), rejecting H4a. In addition, contrary to H4b, the Monte Carlo simulation results showed that shared organizational social exchange did not significantly moderate the indirect effect of team-level empowering leadership on service performance via thriving at work (moderated mediation effect = À0.19, 95 per cent CI = [À0.48, 0.03]).
H4c proposed that shared organizational social exchange would moderate the relationship between empowering leadership and service performance such that this  relationship becomes stronger when shared organizational social exchange is high. In support of H4c and as shown in Model 3, Table III, the relationship between empowering leadership and service performance was significantly moderated by shared organizational social exchange (b = 0.62, p < 0.05). Specifically, the effect of empowering leadership on service performance was positive and significant when shared organizational social exchange was high (simple slope = 0.63, p < 0.01), but not significant when shared organizational social exchange was low (simple slope = À0.11, n.s.). These results are shown in Figure 3.

Discussion and conclusions
Conclusions Given the role of empowering leadership in promoting the discretion that employees need to flexibly respond to customers' unique needs and expectations (Chebat and Kollias, 2000;Hartline and Ferrell, 1996), it is important to understand why and when it influences service performance. Our findings revealed that team-level empowering leadership and customer orientation indirectly related to service performance via thriving at work. Furthermore, customer orientation significantly substituted for the effect of team-level empowering leadership on thriving at work such that this relationship was stronger for employees with low (rather than high) customer orientation. In addition, shared organizational social exchange augmented the influence of team-level empowering leadership on service performance but not its influence on thriving at work.

Theoretical implications
Our finding that team-level empowering leadership and customer orientation indirectly associated with service performance via thriving at work provides support for Spreitzer et al.'s (2005) "socially embedded model" and it's self-determination theory underpinnings.
Although much research has shown customer orientation to influence employees' work outcomes indirectly through motivational pathways such as work engagement (Zablah  et al., 2012), this study provides initial evidence of thriving at work as a pathway through which customer orientation relates to service performance. This is particularly important because although both thriving at work and work engagement are rooted in vitality, it's learning component makes thriving at work a critical attribute in a dynamic workplace (Spreitzer et al., 2010). Additionally, the effect of empowering leadership on service performance through thriving at work adds to the literature that has focused predominantly on psychological empowerment in non-service contexts (Cheong et al., 2019;Li et al., 2017).
Our findings pertaining to the mediating role of thriving at work are reinforced by the fact that intrinsic motivation, which is more proximal to employee outcomes than psychological empowerment (Zhang and Bartol, 2010), did not constitute an alternation explanation of the relationships we examined.
Although much research has documented the performance implications of empowering leadership (Chen et al., 2007), research in a service context has yet to examine when it impacts on employees' motivational state and ultimately, performance. Given the documented positive effect of customer orientation on employee and customer-related outcomes (Zablah et al., 2012), service organizations have understandably invested in the training of employees to become more customer-oriented. Despite these efforts, not all employees are highly customer-oriented; thus, the question is how to motivate low customeroriented employees to engage in self-determined behaviors that lead to enhanced levels of service performance. This study proposed and found that empowering leadership played a more important role in enhancing thriving at work among employees with low but not high customer orientation. Empowering leadership provides them with the confidence to overcome their internal resource deficit and meaningfully, interact with customers, leading to thriving at work. Our finding therefore, extends the current literature by demonstrating the substitution effect of customer orientation in that it makes empowering leadership less important for enhancing high customer orientation employees' thriving at work. It would be interesting to examine whether this substitution effect of customer orientation is stronger in individualistic societies. This is because individuals in such societies relative to those in collectivistic societies tend to view the self as an independent entity that comprises a unique configuration of internal attributes such as preferences, motives and values and to behave primarily based on these internal attributes (Markus and Kitayama, 1994).
The differential effectiveness of shared organizational social exchange as a boundary condition of the motivational and performance implications of team-level empowering leadership contributes to an enhanced understanding of employee reactions to this form of leadership. Consistent with prior research, this study's results revealed that as a relational resource, shared organizational social exchange augments the effects of team-level empowering leadership on employees' service performance. Although they did not examine service performance, Li et al. (2017) reported organizational supportive climate to moderate the effects of team-level empowering leadership on helping or affiliative behaviors. Given that service performance entails helping-oriented behaviors to satisfy the needs of customers (Chuang and Liao, 2010), this study's findings reinforce those reported by Li et al. (2017). These relational resources convey consistent messages about operating in a facilitative context that fosters the self-determined behaviors underpinning the delivery of quality service performance.
However, contrary to our expectation, this team-level interaction did not influence employees' motivational state of thriving at work. This finding is at variance with Lorinkova and Perry (2017) who reported leader-member exchange to moderate the relationship between empowering leadership and the motivational state of psychological empowerment. This may underscore differences between the resources afforded by leader-IJCHM 31,9 member exchange and shared organizational social exchange and the way they complement each other in fostering motivational states and ultimately, performance. Also, given the pervasive influence of leaders on employees' experience of work, employees' motivational state may be more strongly influenced by leadership behaviors relative to the shared quality of the employment relationship. In effect, the resources inherent in a high-quality employment relationship do not augment nor substitute for the effects of those inherent in team-level empowering leadership leading to thriving at work. The differential moderating influence of shared organizational social exchange suggests that performance is driven more by reciprocity concerns whereby organizational relational resources augment those provided by leaders in promoting enhanced service performance but not the motivational state of thriving at work. Despite its intuitive plausibility, future research should subject the preceding explanation to empirical verification.

Practical implications
Given the team-based nature of an airline cabin crews' delivery of customer service, the findings regarding the cross-level effects of team-level empowering leadership provide practical implications that service organizations may leverage to enhance the motivational states and service performance of individual flight attendants. First, the finding that the effect of empowering leadership on thriving at work is contingent upon employees' customer orientation reinforces the importance of organizations considering the attributes of employees when implementing leadership practices. While customer orientation should continue to be an important selection criteria in a service context (Huo et al., 2019;Zablah et al., 2012), this study's findings suggest that team-level empowering leadership is more effective in enhancing the motivation (thriving at work) of employees who are low rather than high in customer orientation.
In addition, this study's results suggest that a shared high quality organizational social exchange can enhance the relationship between team-level empowering leadership and service performance. Consequently, organizations must invest in the development of a high quality of relationship with employees while correspondingly investing in the training of leaders in empowerment practices. Human resource practices that foster employees' perception of a social rather than an economic exchange with their organization provides a facilitative context in which to harness the performance effects of empowering leadership to enhance the quality of employees' service performance. Thus, organizations can reinforce the performance implications of empowering leadership by showing care and concern for their employees if their employees are to reciprocate through enhanced service performance.

Limitations and future research
Like other studies, our study has several limitations. First, the cross-sectional design precludes an inference about the causal status of the relationships we examined. Additionally, with the exception of data on service performance which we obtained from inflight supervisors (a week later), data on the other study variables were obtained from employees. To mitigate the effects of method bias and more confidently establish the causal status of these relationships, future research should use a multi-wave research design.
Second, although this study's findings underscore the utility of self-determination theory as a theoretical grounding for Spreitzer et al.' (2005) socially embedded model, the authors did not obtain data on the agentic behaviors that promote thriving at work. Future research should not only obtain data on these behaviors but also, on psychological need satisfaction to more completely test the self-determination theory underpinnings of the socially embedded model.

Customer service at altitude
Third, this paper focused on service performance which describes in-role performance, a form of affiliative behavior in a service context. Given the documented influence of proactive behaviors such as creative performance (Martinaityte et al., 2019), proactive service performance (Raub and Liao, 2012) and customer-focused voice (Lam and Mayer, 2014) on customer satisfaction, future research should examine whether this study's cross-level model differentially predicts these performance outcomes.
Lastly, our data were obtained within the airline industry from a single country (i.e. South Korea). Although the research hypotheses were largely supported, it is unclear the extent to which the cultural and industry-specific context of this study could have influenced the findings. For example, research has shown South Korea to be a high power distance country (i.e. the extent to which a society accepts the fact that power in institutions and organizations is distributed unequally, Hofstede, 2001;Kim and Leung, 2007). However, this limitation is countered by prior research that has shown empowering leadership to impact both motivational states and performance in high power distance contexts (Li et al., 2016).