Servant leadership and customer service performance: testing social learning and social exchange-informed motivational pathways

ABSTRACT

Drawing on social learning and social exchange theories, we examined a moderated mediation model of the processes through which servant leadership relates to customer service performance. Across two studies, we examined whether frontline employees’ customer-centric attitude (social learning theory) or gratitude towards the organization (social exchange theory) mediates the servant leadership-customer service performance relationship. Additionally, in Study 2, we examined whether initiative climate constitutes a boundary condition of the relationships between the two motivational drivers and customer service performance. Data were obtained from a sample of South Korean (Study 1) and Taiwanese (Study 2) frontline employees. In Study 1, customer-centric attitude and gratitude towards the organization mediate the relationship between servant leadership and service performance but only gratitude towards the organization mediates the servant leadership-customer-directed extra-role performance. The findings of Study 2 reveal that gratitude towards the organization, but not customer-centric attitude mediates the relationship between servant leadership and service performance. Contrary to our prediction, initiative climate did not moderate the relationships between the two motivational drivers and service performance. In sum, our findings provide initial evidence of the importance of social exchange relative to social learning theory in accounting for the customer service performance implications of servant leadership.

KEYWORDS:

• Servant leadership
• gratitude
• customer-centric attitudes
• customer service performance

Introduction

The dynamic and uncertain nature of the service sector coupled with the growing emphasis on customer-centric attitudes in service interactions has engendered much interest in servant leadership because of its serving-oriented attributes (Hunter et al., 2013; Jaramillo et al., 2009; Liden et al., 2014). The steady research interest in servant leadership has uncovered a wealth of insights into the underlying mechanisms through which it relates to employee work-related behaviours (Eva et al., 2019; Kauppila et al., 2022; Lee et al., 2019; Walumbwa et al., 2010; Wu et al., 2021), prompting a call for organizations to implement servant leadership training as an intervention strategy (Van Dierendonck, 2011).

At the core of servant leadership is a focus on a leader’s recognition of “his or her moral responsibility not only to the success of the organization, but also to his or her subordinates, the organization’s customers, and other organizational stakeholders” (Ehrhart, 2004, p. 68). Considering its focus on a concern for the organization’s multiple stakeholders (Ehrhart, 2004; Lemoine et al., 2019), research that explicates the performance implications of servant leadership has, not surprisingly, drawn on either its concern for subordinates by employing social exchange theory or for customers by using social learning theory. Despite the near paradigmatic status of these theoretical perspectives (Madison & Eva, 2019), there is a dearth of research that has simultaneously examined their relative efficacy in accounting for the much-documented relationship between servant leadership and employee performance (Hunter et al., 2013; Madison & Eva, 2019). Consequently, we do not yet know whether employee performance in service contexts is explainable in terms of the transformation of the followers of servant leaders into servant employees (Elche et al., 2020; Ruiz-Palomino & Zoghbi-Manrique de Lara, 2020) or followers’ gratitude to the organization, or both. Addressing this question is of much critical importance because as Lemoine et al. (2019, p. 159) observed, the motivational thrust of servant leadership stems from its caring and other-focused approach which entails a concern for not only employees but also customers. Accordingly, the first objective of this study is to provide a competitive test of gratitude towards the organization (social exchange perspective) and customer-centric attitude (social learning perspective) as theory-informed mediators of the relationship between servant leadership and frontline employees’ customer service performance.

In addition to accounting for the influence of servant leadership on the performance of followers, and in response to calls to investigate the boundary conditions of servant leadership (Donia et al., 2016; Lee et al., 2019), research has examined the contingent nature of the relationship between servant leadership and its motivational drivers (Eva et al., 2019; Neubert et al., 2016; Panaccio et al., 2015; Wu et al., 2021). For example, research has examined leader-follower relationship variables such as organizational embodiment (Wang et al., 2018) as well as individual difference variables such as self-interest (Wu et al., 2021) and proactive personality (Panaccio et al., 2015). There is a growing recognition of the role of context in shaping behaviour in organizational contexts (Johns, 2018). In a service context where the human element is a key competitive advantage, frontline employees should have discretion to meet the diverse needs of customers (Bowen & Lawler, 1995). Although servant leaders empower subordinates, with a few exceptions (Chen et al., 2015; Walumbwa et al., 2010), research has yet to examine the contingent nature of the relationship between the motivational drivers engendered by servant leadership and their performance implications. Accordingly, the second objective of this study is to examine initiative climate as a contextually relevant boundary condition of the relationships between the motivational drivers (customer-centric attitude and gratitude towards the organization) and customer service performance.

In pursuing these objectives, this study makes three contributions to the literature. First, simultaneously examining social learning and social exchange perspectives provides an opportunity to ascertain their relative importance in accounting for the documented relationship between servant leadership and customer service performance. As Madison and Eva (2019, p. 152) noted, such an undertaking helps to understand “ … which path may be stronger in explaining how servant leadership is translating to follower behaviour.” Second, our theory-informed mediators of customer-centric attitude and gratitude towards the organization extend research on prior indicators of the preceding theories. Pertaining to social learning theory, customer-centric attitude provides a test of a fundamental tenet of servant leadership’s motivational thrust in terms of transforming frontline employees into becoming other-oriented and therefore, meeting the needs and expectations of customers (Ruiz-Palomino & Zoghbi-Manrique de Lara, 2020). Pertaining to social exchange theory, we extend Sun et al. (2019) work by conceptualizing gratitude not as a general emotional state of that day but rather, as a positive emotion towards the organization. Lastly, servant leaders have been posited to “give their teams the freedom to envision dynamic new solutions to emerging customer problems” (Lemoine et al., 2021, p. 407). Accordingly, examining initiative climate as a boundary condition provides an opportunity to ascertain whether it strengthens (complements) or weakens (substitutes) for the discretion inherent in servant leadership. This should provide an insight into the ecosystem needed to leverage the motivational implications of servant leadership to promote enhanced customer service performance. Additionally, as the self-direction inherent in initiative climate underpins both social learning and social exchange theories, it provides a parsimonious understanding of the contextual infrastructure needed to leverage the servant leadership-engendered motivational drivers to promote customer service performance.

Theory and hypotheses

At the core of leadership is the ability to influence subordinates to contribute their efforts to the attainment of specified goals (Bass, 1985). However, servant leadership differs from other genres of leadership in two key respects. First, it adopts a stakeholder perspective by which servant leaders exhibit a communal orientation whereby they prioritize the needs and expectations of the organization’s stakeholders including employees, customers, and the community at large. Pertaining to employees, servant leaders treat employees with dignity, provide them with developmental challenges, and encourage them to develop their skills and abilities (Eva et al., 2019; Lemoine et al., 2019). As these behaviours do not only build trust but also enhance the employability of employees, the documented work outcomes of servant leadership have been explained in terms of social exchange by drawing on such constructs as trust, leader-member exchange differentiation, and reciprocity beliefs (Chiniara & Bentein, 2018; Jaramillo et al., 2015; Walumbwa et al., 2010; Zou et al., 2015). Extending this line of research and consistent with recent research that proposes the inclusion of positive emotions in social exchange theorizing (Lawler, 2001), we posit that the competence-building servant leadership behaviours engender gratitude towards the organization which, in turn, transmits the influence of servant leadership onto frontline employees’ customer service performance.

A second tenet of servant leadership is the tendency of servant leaders to transform followers into servants such that they willingly subordinate their self-interest to those of the organization and prioritize serving other organizational stakeholders such as customers. In outlining the defining characteristics of servant leaders, Greenleaf (1977, pp. 13–14) poses the question of whether, while being served, followers become servants themselves. To this end, servant leadership behaviours remind followers of the partnership status of customers and therefore, the importance the organization attaches to meeting their diverse needs and expectations. At the core of the transformation of followers into servants is learning or their imitation of the behaviours of servant leaders (Elche et al., 2020; Ruiz-Palomino & Zoghbi-Manrique de Lara, 2020). Drawing on social learning theory, we posit customer-centric attitude to explicate the influence of servant leadership on frontline employees’ customer service performance (Eva et al., 2019; Liden et al., 2014; Walumbwa et al., 2010).

To provide a more complete understanding of the processes that underpin the documented benefits of servant leadership, research has examined boundary conditions of these relationships (Chen et al., 2015; Eva et al., 2019; Lee et al., 2019; Neubert et al., 2016; Van Dierendonck et al., 2014). This is consistent with Greenleaf’s (1998) observation that certain mindsets and contextual factors might undermine the servant leaders’ efforts to transform followers into servants or become other-oriented. To facilitate prioritizing and meeting the needs of organizational stakeholders such as customers, servant leaders are said to take two key actions. First, they frame how followers think of customers and second, give them autonomy to engage in self-directed behaviour to meet the needs and expectations of customers (Lemoine et al., 2021, p. 407). Accordingly, we examined whether initiative climate (and the inherent opportunities for self-directed behaviour) strengthens (complements) or weakens (substitutes) for the autonomy inherent in servant leadership as a second stage boundary condition of the relationships between our two theory-informed motivational drivers and customer service performance.

In sum, we examined the relative efficacy of social exchange (gratitude towards the organization) and social learning (customer-centric attitude) theories in accounting for the documented performance effects of servant leadership. Additionally, and drawing on the autonomy servant leaders provide followers to understand and satisfy the needs of customers, we examined initiative climate as a second stage boundary condition of the relationship between these two theory-informed motivational drivers and customer service performance. Consistent with the distinction between prescribed and discretionary work role performance (Bettencourt et al., 2001), we defined customer service performance in Study 1 in terms of service performance and customer-directed extra-role. In contrast, we focused in Study 2, on prescribed work role performance or service performance.

Servant leadership, customer-centric attitude, and customer service-related performance

Servant leadership is a genre of leadership which takes the view that one wants to serve, to serve first (Greenleaf, 1977, p. 4) and therefore, constitutes an other-centred leadership style. As Sendjaya and Sarros (2002, p. 61) note, “servant leaders regard their followers as people who have been entrusted to them to be elevated to their better selves and to be what they are capable of becoming.” Following from social learning theory, we expect servant leadership to relate to customer stewardship, a customer-centric attitude which, Schepers et al. (2012, p. 2) define as, “frontline employees’ felt ownership of and moral responsibility for customers’ overall welfare.” Social learning theory posits that individuals learn by modelling the attitudes, values, and behaviours of those who have high status, competence, and power but importantly, are also considered credible (Bandura, 1977). As positive role models, servant leaders inspire frontline employees to emulate their servant-oriented values and behaviours evidencing the contagion effect of leadership. Servant leadership therefore engenders a transformation of the self-concept of followers (frontline employees) from self to other-oriented or customer-centric attitude reflected in a concern for the welfare of customers. Although research has not yet examined the influence of servant leadership on customer stewardship, Jaramillo et al. (2009) reported servant leadership to positively relate to the conceptually synonymous construct of customer orientation. Furthermore, and consistent with our arguments, research at the unit level has shown servant leadership to relate to service climate (Walumbwa et al., 2010) and serving culture (Liden et al., 2014), constructs which are analogous to customer stewardship (orientation) at the individual level. We therefore hypothesize that:

Hypothesis 1a:

Servant leadership positively relates to customer-centric attitude.

The influence of servant leadership on customer service performance has precipitated a steady stream of research to uncover the underlying mechanisms of these relationships (Eva et al., 2019; Lee et al., 2019; Madison & Eva, 2019). As previously noted, and following social learning theory, we focused on customer-centric attitude as one such but hitherto unexamined underlying mechanism. This is because the learning inherent in transforming followers’ self-concept enables followers of servant leaders to become servants themselves whereby, they seek the welfare of customers as external stakeholders (Lemoine et al., 2021). The uncertain nature of the service delivery process suggests that service delivery cannot be completely scripted or formalized. Consequently, when the customer’s needs deviate from the service script, a customer-centric attitude motivates frontline employees to adopt a problem-solving approach to the provision of customer service (Subramony & Pugh, 2015). This entails going the extra mile to engage in a search process to understand these needs in order to customize the service delivery. As uncertainty is inherent in both forms of customer service performance, a customer-centric attitude provides the reason or intrinsic motivation to meet the needs of the customer. Consistent with prior research (Jaramillo et al., 2009; Schepers et al., 2012), we hypothesize that:

Hypothesis 1b:

Customer-centric attitude relates to service performance and mediates the relationship between servant leadership and service performance.

Hypothesis 1c:

Customer-centric attitude relates to customer-directed extra-role performance and mediates the relationship between servant leadership and customer-directed extra-role performance.

Servant leadership, gratitude, and customer service-related performance

Consistent with its core tenet of elevating followers to their better selves, servant leadership shapes perceptions of support through treating them fairly and with dignity as well as through task-related behaviours such as developing employee skills thereby helping them to grow and succeed (Eva et al., 2019; Lee et al., 2019; Liden et al., 2014; Van Dierendonck, 2011). Social exchange theory’s foundational constructs of self-interest and interdependence (Cropanzano et al., 2017; Lawler, 2001), suggest that acts of benevolence as reflected in the preceding tangible and intangible resources, will engender feelings of gratitude on the part of the receiver. We focus on state rather than dispositional gratitude because the former is episodic in nature and can therefore, be shaped by experiences in a social relationship (McCullough et al., 2002; Wood et al., 2010) or organizational experiences (Fehr et al., 2017; Ford et al., 2018; Ng, 2016; Spence et al., 2014). Emmons (2004, p. 9) definition of state gratitude as “the recognition and appreciation of an altruistic gift” chimes with Tsang’s (2006, p. 139) as “a positive emotional reaction to the receipt of a benefit that is perceived to have resulted from the good intentions of another.” McCullough et al. (2001) observe that acts of benevolence engender behaviours that not only directly benefit the benefactor but also a third party. Although their study draws on affective events theory, Ford et al. (2018) report perceived organizational support to positively relate to chronic gratitude. Thus, consistent with Lawler’s (2001) view of injecting positive emotion into social exchange theory and as an agent of the organization, we expect a servant leader’s acts of benevolence to engender gratitude towards the organization.

Hypothesis 2a:

Servant leadership positively relates to gratitude towards the organization.

As a positive emotion felt when one is a beneficiary of an interacting partner’s benevolence, gratitude evokes a desire on the part of the recipient to act in ways that benefit the benefactor or a third party which, in this context, could either be the organization and/or its customers. McCullough et al. (2001) view that gratitude serves a moral motive and engenders the construct’s motivational implications for the beneficiary’s behaviour is consistent with the principle of reciprocity in social exchange theory (Blau, 1964; Gouldner, 1960). As beneficiaries of the tangible and intangible resources received from the servant leader, feelings of gratitude will motivate customer service performance which benefits the organization as well as its customers. Research has shown gratitude to motivate prosocial behaviour (Algoe et al., 2008; Ma et al., 2017) as well as organizational citizenship behaviour (Fehr et al., 2017; Ford et al., 2018; Spence et al., 2014; Sun et al., 2019). Thus, we hypothesize that:

Hypothesis 2b:

Gratitude towards the organization positively relates to service performance and mediates the relationship between servant leadership and service performance.

Hypothesis 2c:

Gratitude towards the organization positively relates to customer-directed extra-role performance and mediates the relationship between servant leadership and customer-directed extra-role performance.

Moderating influence of initiative climate

Servant leaders transform employees not only by enhancing their skills, knowledge and abilities but also cultivating other-oriented attitudes (Lemoine et al., 2019; Ruiz-Palomino & Zoghbi-Manrique de Lara, 2020). From a social learning perspective, an other-oriented or customer-centric attitude developed by emulating leaders’ servant-oriented values and behaviour can be leveraged to enhance performance or model serving others. Similarly, a social exchange-based relationship is anchored in the supportive behaviours of a servant leader such as investment in the follower through developmental challenges that promote the goals of the follower engendering gratitude to the organization. However, the extent to which this motivational state can engender behaviours that satisfy organizational goals may be contingent upon the opportunity to engage in self-directed behaviour. This is consistent with Hofmann et al. (2003, p. 171) observation that “the type of behaviour valued in the work environment will provide the direction for subordinate reciprocation.”

Drawing on the preceding discussion and Ehrhart’s (2004) observation that group-level process variables constitute cross-level moderators of the relationship between motivational state and work behaviour, we posit initiative climate as a second stage moderator of the relationships between the two theory-informed motivational drivers and service performance. Initiative climate constitutes a facet-specific climate which describes a shared perception of the “extent to which self-starting, change-oriented, long-term-oriented, and persistent behaviour is encouraged and rewarded by management” (Raub & Liao, 2012, p. 653). Unlike service climate which focuses on customer-centric attitudes in the service delivery process (Bowen & Schneider, 2014), initiative climate is concerned with proactive execution of work behaviour in general (Raub & Liao, 2012). Social information processing theory suggests that limits to an individual’s information processing capacity enjoins them to rely on social cues to form an understanding of the reality of their work environment (Salancik & Pfeffer, 1978). Collectively, these social cues shape an employee’s socially constructed work environment regarding the attitudinal and behavioural expectations in the service delivery process.

The increasingly diverse needs of customers have injected much uncertainty into the service delivery process. This uncertainty has meant that the service delivery process cannot be meaningfully standardized as frontline employees must engage in a search process to understand the requisite information needed to customize the service delivery (Gilson et al., 2005). In such a context, a frontline employee’s customer-centric attitude and gratitude towards the organization will constitute necessary but not sufficient conditions for meeting the diverse needs of customers. Consequently, social information processing theory would suggest that when initiative climate is high, it signals to frontline employees the appropriateness of exercising discretion and persistence in their search process to uncover the requisite knowledge to customize the service delivery. A high initiative climate therefore constitutes a situational enhancer (Howell et al., 1986) and therefore, strengthens the effect of both customer-centric attitude and gratitude towards the organization on customer service performance. In effect, while the motivational drivers provide a reason to engage in customer service performance, initiative climate energizes frontline employees to engage in this behaviour. In contrast, social information processing theory would suggest that social cues inherent in a low initiative climate discourages the exercise of self-directed actions and persistence in customizing the service delivery to meet the needs of customers. Consequently, a low initiative climate weakens the motivation to engage in the search processes necessary for uncovering the information to customize the service delivery leading to reduced levels of customer service performance.

Hypothesis 3a:

Initiative climate moderates the relationships between (i) customer-centric attitude and (ii) gratitude towards the organization, and service performance such that these relationships are stronger when initiative climate is high but weaker when low.

As we previously posit servant leadership to relate to customer service performance indirectly through the dual motivational mechanisms of customer-centric attitude and gratitude towards the organization, the moderating effect of initiative climate suggests a conditional indirect effect. Specifically, we expect the mediating influence of customer-centric attitude and gratitude towards the organization in the relationships between servant leadership and customer service performance to be conditional upon initiative climate (a second stage moderator).

Hypothesis 3b:

The indirect influence of servant leadership on service performance through (i) customer-centric attitude and (ii) gratitude towards the organization is stronger when initiative climate is high but weaker when low.

Methods

To address our research objectives, we conducted two independent field studies using a time-lagged, multi-source research design. Study 1 draws on a sample of South Korean frontline employees to examine customer-centric attitude and gratitude towards the organization as motivational pathways that transmit the influence of servant leadership on the outcomes of service performance and customer-directed extra-role performance (Hypotheses 1a to 2c). Study 2 draws on a Taiwanese sample of frontline employees to replicate and extend Study 1 (Hypotheses 1a, 1b, 2a, 2b, 3a, and 3b).

Study 1

Sample and procedures

Respondents were frontline employees and their supervisors drawn from franchised retail outlets representing supermarkets, drug stores, and bookshops, located in Seoul, South Korea, and surrounding areas. Data were collected in two waves over a 3-week period. At Time 1, respondents rated the servant leadership style of their franchised retail outlet owner-manager, provided demographic data, and rated other variables not relevant to this study. Of the 580 questionnaire packages sent to respondents across the 112 participating franchised retail outlets, completed questionnaires were returned from 460 representing 92 outlets. At Time 2, questionnaires were distributed to the 460 respondents who returned the Time 1 questionnaires and were requested to rate their customer stewardship, gratitude towards the organization, and other constructs not germane to our objectives in this study. We also distributed questionnaires to 92 supervisors and asked them to rate the service performance and customer-directed extra-role performance of each of their subordinates requested to complete the Time 2 questionnaires. Of the 92 retail outlets, completed and matched questionnaires were returned from 420 respondents representing 84 retail outlets.

Of the final sample of 84 franchised retail outlets, 20 were bookshops, 28 drugstores, and 36 supermarkets.¹ In terms of demographic composition of the sample, 58.8% of the respondents were female, reported a mean age of 33.9 years (SD = 6.40), a mean organizational tenure of 5.1 years (SD = 3.63), and a mean of 47.6 hours worked per week (SD = 4.14). Pertaining to educational attainment, majority of respondents had received an undergraduate degree (61.7%) while 26.7% had received a two-year college degree. The 84 supervisors were predominantly male (95.2%), reported a mean age of 48.2 years (SD = 4.78), a mean organizational tenure of 10.9 years (SD = 4.42), and a mean of 47.3 hours worked per week (SD = 3.26). In terms of educational attainment, 91.7% had received a bachelor’s degree, while 8.3% had obtained a master’s degree.

Measures

The survey instrument was administered in Korean but was originally constructed in English following Brislin’s (1986) recommended back-translation procedure. Unless otherwise indicated, response options ranged from (1) “strongly disagree” to (5) “strongly agree.”

Servant leadership

We treated servant leadership as a perceptual variable and requested respondents to rate the owner-manager’s servant leadership with the aid of Liden et al. (2014) 7-item scale. A sample item is “I would seek help from my manager if I had a personal problem.” The scale’s alpha reliability in this study was .78.

Customer stewardship

We used Schepers et al. (2012) 5-item scale to measure customer stewardship. A sample item is “I feel a sense of responsibility for the customer.” The scale’s alpha reliability in this study was .74.

Gratitude towards the organization

We used Ng’s (2016) 6-item adaptation of McCullough et al. (2002) generalized gratitude scale to measure gratitude towards the organization. A sample item is “I am grateful to this organization for many reasons.” The scale’s alpha reliability in this study was .81

Customer service performance

Supervisors rated the two indicators of customer service performance at Time 2. For service performance, we used Liao and Chuang’s (2004) 7-item scale with response options ranging from (1) “highly unsatisfactory” to (7) “highly satisfactory.” A sample item is “This employee is able to help customers when needed.” The scale’s alpha reliability in this study was .88. To measure customer-directed extra-role performance, we used Netemeyer et al. (2005) 4-item scale. Response options ranged from (1) “never” to (7) “always.” A sample item is “How often does this employee go above the ‘call of duty’ when serving customers?” The scale’s alpha reliability in this study was .85.

Controls

We controlled for the demographics of gender, education, and supervisor-subordinate tenure as they may influence both our mediators and performance indicators (Lemoine & Blum, 2021).

Data analysis

Our participants were nested in 84 retail outlets and every supervisor assessed the customer service-related performance indicators of 5 respondents. Accordingly, we first calculated ICC(1) values for the two service-related outcome variables to evaluate the appropriateness of multilevel analysis. The ICC(1) was .34 for service performance and .31 for customer-directed extra-role performance. These ICC(1) values revealed obvious between-unit variances (Bliese, 2000) so we adopted multilevel structural equation modelling (MSEM) to test the hypothesized relationships and address non-independence of the data. We used Mplus 8.2 software (Muthén & Muthén, 1998–2018) and estimated the path coefficients using the “intercepts-as-outcomes model,” and selected the “Type = twolevel” analytical approach with the maximum likelihood estimation method. Following Preacher and Selig’s (2012) recommendation, we employed the Monte Carlo simulation method to obtain the 95% confidence intervals (CIs) of the mediating effects proposed in the current study with 20,000 repetitions.

Results

Measurement model analysis

To ascertain the construct validities of our measures, we conducted confirmatory factor analyses (CFAs) involving the main research variables and compared the 5-factor model with a number of alternative models. The CFA results show that the proposed 5-factor model had a reasonable fit to the data (χ²(367) = 763.24, p < .001, CFI =.92, TLI =.91, RMSEA =.05, SRMR =.04) and all the scale items significantly loaded on their specified factor (all p < .001). With the exception of two servant leadership items (.35 and .48, respectively), standardized factor loadings for all items exceeded the .50 criterion. Furthermore, the 5-factor CFA model had a significantly better fit than the alternative models (all Δχ² achieved the significance level of p < .001). These results support the discriminant validities of our measures.

Hypothesis testing

Table 1 presents the descriptive statistics, correlations, reliabilities, and the square roots of average variance extracted (AVE) for the variables in Study 1. Table 2 and Figure 1 present the results of the MSEM analysis. As predicted, servant leadership significantly and positively related to the customer-centric attitude of customer stewardship (γ = .34, p < .001), supporting Hypothesis 1a. Furthermore, customer stewardship positively and significantly related to service performance (γ = .27, p < .001), but did not significantly relate to customer-directed extra-role performance (γ = .17, p > .05).

Figure 1. The MSEM results of the theoretical model in study 1.

Note. n= 420 at level 1; n = 84 at level 2. Unstandardized estimates and standard errors (in parentheses) were reported. *p< .05, **p <.01, ***p<.001.

Table 1. Descriptive statistics, correlations, and cronbach’s alpha values of study variables (Study 1).

Variable	Mean	SD	Alpha	1	2	3	4	5	6	7	8
1. Gender	0.59	0.49	—	—
2. Education	2.65	0.69	—	−.16**	—
3. Supervisor-subordinate tenure	3.73	2.38	—	−.06	.14**	—
4. Servant leadership	3.93	0.67	.78	−.01	.13**	.25**	(.58)
5. Customer stewardship	4.10	0.56	.74	.03	.07	.07	.45**	(.59)
6. Gratitude towards the organization	3.93	0.66	.81	.01	.15**	.13**	.56**	.55**	(.66)
7. Service performance	5.39	0.79	.88	.00	.10*	.14**	.46**	.41**	.52**	(.72)
8. Customer-directed extra-role performance	5.00	0.88	.85	−.07	.19**	.06	.41**	.33**	.48**	.70**	(.76)

Note. n = 420 at level 1. The square roots of average variance extracted (AVE) are in parentheses on the diagonal. Gender: 0 = male, 1 = female. Education: 1 = high school or below, 2 = two-year college graduated, 3 = undergraduate degree, 4 = postgraduate degree, 5 = PhD degree.

*p < .05; ** p < .01.

Table 2. Multilevel estimates for the theoretical model (Study 1).

	Dependent variables
	Customer stewardship			Gratitude towards the organization			Service performance			CDERP
	Estimate	SE	p	Estimate	SE	p	Estimate	SE	p	Estimate	SE	p
Intercept	4.10***	.03	<.001	3.93***	.03	<.001	5.39***	.04	<.001	5.00***	.04	<.001
Control variables
Gender	.05	.05	.327	.04	.05	.460	.00	.06	.981	−.11	.08	.130
Education	.03	.04	.374	.07	.04	.113	.02	.05	.662	.13*	.06	.019
Supervisor-subordinate tenure	−.01	.01	.563	−.00	.01	.780	.02	.01	.204	−.00	.02	.829
Independent variables
Servant leadership	.34***	.04	<.001	.54***	.04	<.001
Customer stewardship							.27***	.08	<.001	.17	.09	.060
Gratitude towards the organization							.49***	.06	<.001	.56***	.07	<.001
Level-2 variance		.04			.02			.04			.03
Level-1 residual variance		.21			.28			.38			.53
R^2		.20			.33			.28			.25

Note. n = 420 at level 1; n = 84 at level 2. Unstandardised estimates were reported. SE = standard error. CDERP denotes customer-directed extra-role performance. Servant leadership and control variables were grand-mean centred. As there were no group-level predictors in this model, only individual-level R² were reported. * p < .05; ** p < .01; *** p < .001.

Table 3 shows the results of Monte Carlo simulation for the mediating hypotheses. Servant leadership had a significant indirect effect on service performance through the mediation of customer stewardship (estimate = .091, 95% CI [.039, .151], excluding zero) but did not have a significant indirect effect on customer-directed extra-role performance through the mediation of customer stewardship (estimate = .057, 95% CI [−.002, .123], including zero). Therefore, Hypothesis 1b was supported, but not Hypothesis 1c.

Table 3. The estimated mediating effects (Study 1).

		95% Confidence interval
Paths	Estimate	Lower	Upper
(1) H1b: Servant leadership → customer stewardship → service performance	.091*	.039	.151
(2) H1c: Servant leadership → customer stewardship → CDERP	.057	−.002	.123
(3) H2b: Servant leadership → gratitude towards the organization → service performance	.264*	.191	.343
(4) H2c: Servant leadership → gratitude towards the organization → CDERP	.304*	.218	.396
Path (3) vs. Path (1)	.173*	.061	.285
Path (4) vs. Path (2)	.247*	.117	.376

Note. n = 420 at level 1; n = 84 at level 2. Unstandardised estimates were reported with 3 digits after the decimal point because some estimates were close to zero. CDERP denotes customer-directed extra-role performance. * p < .05.

Table 2 and Figure 1 show that servant leadership positively related to gratitude towards the organization (γ = .54, p < .001), providing support for Hypothesis 2a. Furthermore, the MSEM results show that gratitude towards the organization positively and significantly related to both service performance (γ = .49, p < .001) and customer-directed extra-role performance (γ = .56, p < .001). These findings were reinforced by the Monte Carlo results which show that servant leadership had a significant indirect effect on service performance (estimate = .264, 95% CI [.191, .343], excluding zero) and customer-directed extra-role performance (estimate = .304, 95% CI [.218, .396], excluding zero) through the mediation of gratitude towards the organization. Taken together, Hypothesis 2b and Hypothesis 2c were both supported.²

As our theoretical model had two mediators, we used Monte Carlo simulations to examine the relative strength of these mediating processes. Table 3 shows that servant leadership had a stronger indirect effect on service performance through the mediation of gratitude towards the organization than through the mediation of customer stewardship (estimate = .173, 95% CI [.061, .285], excluding zero). Similarly, servant leadership had a stronger indirect effect on customer-directed extra-role performance through the mediation of gratitude towards the organization than through the mediation of customer stewardship (estimate = .247, 95% CI [.117, .376], excluding zero).

To enhance the rigorousness of our social exchange predictions, we tested a model in which we added direct paths from servant leadership to service performance and customer-directed extra-role performance. Results show that servant leadership had significant direct effects on service performance and customer-directed extra-role performance (γ = .25, p < .001; γ = .27, p < .001). In this model, gratitude towards the organization still had significant mediating effects on the relationships between servant leadership and the two customer service-related performance indicators. In sum, Hypotheses 1a, 1b, and Hypotheses 2a-2c were supported, but Hypothesis 1c was not supported.

Taken together, these findings provide initial support for the social exchange relative to the social learning theory-informed explanation of the influence of servant leadership on the customer service performance indicators. However, we do not yet understand the boundary conditions of the relationships between the motivational drivers and customer service performance. Accordingly, Study 2 replicated and extended the findings of Study 1. Specifically, we examined the relative utility of the dual motivational pathways of the customer-centric attitude and gratitude towards the organization in transmitting the influence of servant leadership on service performance (Hypotheses 1a, 1b, 2a, and 2b). Additionally, we examined initiative climate as a contextual boundary condition of the relationship between the motivational pathways and service performance (Hypotheses 3a and 3b).

Study 2

Sample and procedure

Data were obtained from both frontline employees and unit managers of a large Taiwanese brokerage firm. At Time 1, respondents assessed the servant leadership of the unit manager, the customer-centric attitude of customer orientation, initiative climate, and their demographic background. Of the 661 questionnaires distributed to respondents across 62 units, completed questionnaires were obtained from 644 respondents representing a 97.4% response rate. At Time 2 (a month later), respondents rated gratitude towards the organization, and other variables not germane to this study’s objectives. Completed responses were obtained from 485 respondents representing a 73.4% response rate. At Time 3 (two weeks later), we asked unit managers to assess the service performance of the 485 respondents who completed both the Time 1 and Time 2 questionnaires.

The final sample comprised 485 respondents and 62 managers from 62 units. Of the 485 respondents, 78.8% were female, reported an average age of 45.0 years (SD = 8.90), an average tenure of 15.2 years (SD = 7.26), and the overwhelming majority (79.4%) had received a college/undergraduate degree. Of the 62 managers, 75.8% were female, and reported an average age of 47.5 years (SD = 7.16). Pertaining to educational attainment, 59.7% of the managers had a college degree while 38.7% had a master’s degree.

Measures

Questionnaires were administered in Chinese using Brislin’s (1986) back translation procedure to translate the original English language version. Unless otherwise indicated, response options ranged from (1) “strongly disagree” to (5) “strongly agree.”

Servant leadership

We used the same 7-item scale (Liden et al., 2014) as in Study 1 to measure perceived servant leadership. Response options ranged from (1) “strongly disagree” to (7) “strongly agree.” The scale’s alpha reliability in this study was .97.³

Customer orientation

We used a 5-item scale reported by Schepers et al. (2012) to measure customer orientation. A sample item is “I take a problem-solving approach to selling products or service to customers.” The scale’s alpha reliability was .93.

Gratitude towards the organization

We used the same 6-item scale (Ng, 2016) as in Study 1 to measure gratitude towards the organization. The scale’s alpha reliability in this study was .90.

Initiative climate

We used a 16-item scale developed by Raub and Liao (2012) to measure the four subdimensions of initiative climate. These dimensions which include self-starting behaviour, change orientation, long-term focus, and persistence reflect the construct domain of Frese et al. (1996) conceptualization of personal initiative. A sample item is “In this work unit, employees are recognized or rewarded for taking the initiative to solve work-related problems.” To examine its factor structure, we followed Raub and Liao (2012) to conduct a second-order CFA for this measure (χ² (100) = 730.56, p < .001, CFI =.90, TLI =.88, RMSEA =.11, SRMR =.05). Although some of the fit indexes from the CFA results were not ideal, the CFI and SRMR values were acceptable. All items loaded significantly onto their specified dimension (from .45 to .87, all p < .001), and the factor loading of the four dimensions onto the second-order initiative climate construct were high (from .91 to .99, all p < .001). Thus, we followed previous research to calculate the initiative climate score by averaging across the four dimensions (Hong et al., 2016; Raub & Liao, 2012). The scale’s alpha reliability was .95. Results of aggregation statistics revealed a mean r_wg(j) of .97, a median r_wg(j) of .98 (calculated using the slightly skewed distribution), ICC(1) of .15, and ICC(2) of .56, which justified treating initiative climate as a unit level construct.

Service performance

Following Raub and Liao (2012), we used Griffin et al. (2007) 3-item scale to measure service performance. Unit managers were asked to rate the service performance of each of their subordinates who completed surveys over the two time points. A sample item is “This employee carries out the core parts of his/her job well.” The scale’s alpha reliability was .89 in this study.

Controls

As in Study 1, we controlled for the demographics of gender, education, and supervisor-subordinate tenure (Lemoine & Blum, 2021).

Data analysis

We first examined the ICC(1) value for the outcome variable – service performance. The ICC(1) value was .35, revealing substantial between-unit variance. As in Study 1, we adopted MSEM using Mplus 8.2 software (Muthén & Muthén, 1998–2018) to test the hypothesized relationships. Our theoretical model had cross-level interaction effects, so we employed the “intercepts-and-slopes-as-outcomes model” and selected the “Type = twolevel random” analytical approach with the maximum likelihood estimation method. A Monte Carlo simulation with 20,000 repetitions was conducted to obtain the 95% CIs for the mediating as well as the moderated mediating effects.

Results

Measurement model analysis

We conducted a CFA model for the 5 main constructs and compared this model with several alternative models. Before conducting the CFA, we formed item parcels as indicators for initiative climate following the default theoretical dimensions (Raub & Liao, 2012). The CFA results show that the proposed 5-factor model had a good fit to the data (χ²(265) = 716.23, p < .001, CFI =.96, TLI =.95, RMSEA =.06, SRMR =.04). All items and indicators had a standardized factor loading that exceeded the .50 criterion (all p < .001). The results of model comparison show that the 5-factor CFA model had a significantly better fit than the alternative models (all Δχ² achieved the significance level of p < .001). These results support the discriminant validities of our measures.

Hypothesis testing

Table 4 presents the descriptive statistics, correlations, reliabilities, and the square roots of average variance extracted (AVE) for the variables in Study 2. Table 5 and Figure 2 present the results of the MSEM analysis. Results show that servant leadership significantly and positively related to customer orientation (γ = .14, p < .001) and gratitude towards the organization (γ = .12, p < .001). These findings replicate the findings of Hypothesis 1a and Hypothesis 2b in Study 1. The MSEM results further replicated Hypothesis 2b as gratitude towards the organization had a positive and significant relationship with service performance (γ = .14, p < .05). However, customer orientation did not have a positive and significant relationship with service performance (γ = −.07, p > .05), which is inconsistent with the finding of Hypothesis 1b in Study 1.

Figure 2. The MSEM results of the theoretical model in study 2.

Note. n= 485 at level 1; n = 62 at level 2. Unstandardized estimates and standard errors (in parentheses) were reported. *p < .05, **p < .01, ***p < .001.

Table 4. Descriptive statistics, correlations, and cronbach’s alpha values of study variables (Study 2).

Variable	Mean	SD	Alpha	1	2	3	4	5	6	7	8
1. Gender	0.79	0.41	—	—
2. Education	1.90	0.44	—	−.05	—
3. Supervisor-subordinate tenure	2.75	3.81	—	.04	.03	—
4. Servant leadership	4.27	1.04	.97	−.01	−.05	.08	(.91)
5. Customer orientation	4.23	0.50	.93	.05	.09*	−.04	.27**	(.84)
6. Gratitude towards the organization	3.72	0.63	.90	.02	.18**	.06	.22**	.28**	(.78)
7. Initiative climate	3.91	0.51	.95	.03	.11*	−.08	.39**	.53**	.30**	(.79)
8. Service performance	3.95	0.65	.89	.16**	.14**	.04	.07	.03	.21**	.08	(.85)

Note. N = 485 at level 1. The square roots of average variance extracted (AVE) are in parentheses on the diagonal.

Gender: 0 = male, 1 = female. Education: 1 = high school or below, 2 = undergraduate degree, 3 = postgraduate degree.

* p < .05; ** p < .01.

Table 5. Multilevel estimates for the theoretical model (Study 2).

	Dependent variables
	Customer orientation			Gratitude towards the organization			Service performance
	Estimate	SE	p	Estimate	SE	p	Estimate	SE	p
Intercept	.00	.02	1.000	.00	.02	1.000	3.97***	.06	<.001
Control variables
Gender	.07	.05	.131	.07	.06	.199	.21**	.06	.001
Education	.06	.05	.169	.15**	.05	.004	.12*	.06	.038
Supervisor-subordinate tenure	−.01	.01	.069	.00	.01	.512	.00	.01	.649
Level-1 independent variables
Servant leadership	.14***	.02	<.001	.12***	.03	<.001
Customer orientation							−.07	.06	.270
Gratitude towards the organization							.14*	.06	.017
Level-2 independent variable
Initiative climate							.25	.19	.195
Cross-level interactions
Customer orientation × initiative climate							.02	.23	.920
Gratitude towards the organization × initiative climate							−.14	.21	.503
Level-2 variance		—			—			.14
Level-1 residual variance		.19			.24			.26
Level-2 R^2		—			—			.58
Level-1 R^2		.24			.36			.41

Note. n = 485 at level 1; n = 62 at level 2. Unstandardised estimates were reported. SE = standard error. Initiative climate and control variables were grand-mean centred. Servant leadership, customer orientation, and gratitude towards the organization were group-mean centred. R² values were calculated using Snijders and Bosker’s (2012) formulas.

*p < .05; ** p < .01; *** p < .001.

To examine the mediating relationships, we conducted a Monte Carlo simulation. The results, shown in Table 6, reveal a significant indirect effect of servant leadership on service performance through the mediation of gratitude towards the organization (estimate = .017, 95% CI [.003, .036], excluding zero) replicating the results of Hypothesis 2b in Study 1. However, servant leadership did not have a significant indirect effect on service performance through the mediation of customer orientation (estimate = −.009, 95% CI [−.028, .007], including zero), a finding which is inconsistent with the finding of Hypothesis 1b in Study 1. Table 6 also shows that servant leadership had a stronger indirect effect on service performance through the mediation of gratitude towards the organization than through the mediation of customer orientation (estimate = .027, 95% CI [.001, .055], excluding zero). Taken together, these findings provided support for Hypothesis 2b as well as replicated the findings in Study 1.

Table 6. The estimated mediating effects (Study 2).

		95% Confidence interval
Paths	Estimate	Lower	Upper
(5) Servant leadership → customer orientation → service performance	−.009	−.028	.007
(6) Servant leadership → gratitude towards the organization → service performance	.017*	.003	.036
Path (6) vs. Path (5)	.027*	.001	.055
H3b (i): Servant leadership → customer orientation → service performance
High-level initiative climate (+1 SD)	−.008	−.034	.016
Low-level initiative climate (−1 SD)	−.010	−.037	.015
Index of moderated mediation	.003	−.060	.068
H3b (ii): Servant leadership → gratitude towards the organization → service performance
High-level initiative climate (+1 SD)	.012	−.008	.035
Low-level initiative climate (−1 SD)	.022*	.001	.049
Index of moderated mediation	−.017	−.073	.034

Note. n = 485 at level 1; n = 62 at level 2. Unstandardised estimates were reported with 3 digits after the decimal point because some estimates were close to zero. * p < .05.

Hypothesis 3a proposed a moderating effect of initiative climate on the relationships between our focal theory-informed mediators (i.e., customer orientation and gratitude towards the organization) and service performance. Results in Table 5 and Figure 2 show that initiative climate neither moderated the relationship between customer orientation and service performance (γ = .02, p > .05), nor between gratitude towards the organization and service performance (γ = −.14, p > .05). Therefore, Hypothesis 3a was not supported.

Hypothesis 3b proposed that initiative climate moderates the indirect effects of servant leadership on service performance through customer orientation and gratitude towards the organization. As there were no significant cross-level interactions between the mediators and initiative climate, we did not find significant moderated mediation effects. The indexes of moderated mediation (IMM) for Hypothesis 3b were not significant (for the mediator of customer orientation, IMM =.003, 95% CI [−.060, .068], including zero; for the mediator of gratitude towards the organization, IMM = −.017, 95% CI [−.073, .034], including zero). Therefore, Hypothesis 3b was not supported.

To enhance the rigorousness of our findings, we tested a model in which servant leadership had a direct path to service performance. The results show this direct path to be nonsignificant (γ = −.03, p > .05) while gratitude towards the organization still mediated the relationship between servant leadership and service performance. Furthermore, the results show that unlike gratitude towards the organization, customer orientation did not mediate the servant leadership-service performance relationship. Taken together, the findings replicated support for the social exchange rather than social learning-informed explanation of the influence of servant leadership on customer service performance.

Discussion

Across two studies, our findings revealed that servant leadership related to gratitude towards the organization and the customer-centric attitudes of customer stewardship (Study 1) as well as customer orientation (Study 2). Second, both customer stewardship and gratitude towards the organization mediated the influence of servant leadership on service performance but only gratitude towards the organization mediated the servant leadership-customer-directed extra-role performance relationship (Study 1). The mediating role of gratitude towards the organization in the servant leadership-service performance relationship observed in Study 1 was replicated in Study 2 but not that of customer-centric attitude measured by customer orientation. Lastly, and contrary to our prediction, initiative climate did not moderate the relationship between the motivational drivers (customer orientation and gratitude towards the organization) and service performance (Study 2). We discuss the implications of our findings in the succeeding sections.

Theoretical implications

First, by testing the relative effectiveness of social learning and social exchange theories in accounting for the influence of servant leadership on customer service performance, we respond to Madison and Eva’s (2019) call to examine the relative importance of the different pathways that explicate the relationship between servant leadership and its documented performance outcomes in a service context. Consistent with the predictions of social learning theory, frontline employees emulate the serving nature of servant leaders thereby instigating a cycle of service as manifested in the customer-centric attitudes of customer stewardship and customer orientation. Indeed, this interpretation dovetails with Liden et al. (2014) finding that servant leadership promotes a serving culture at the unit level as well as Lemoine and Blum’s (2021) finding that servant leaders engender a prosocial motivation, which ultimately motivates followers to become servant leaders themselves. However, across both studies, our findings reveal social exchange theory to offer a stronger explanation of the influence of servant leadership on the customer service performance indicators relative to social learning theory. This suggests that it is the servant leader’s promotion of followers’ need to grow, develop, and prosper, and the ensuing social exchange-based relationship that accounts for the documented performance effects of servant leadership. Considering the documented relevance of both theories, future research should investigate additional mechanisms through which social learning and social exchange theories transmit the outcomes of servant leadership.

Second, our findings directly address the question of whether servant leaders promote customer service performance through inculcating customer-centric attitudes or meeting the needs of followers which evoke gratitude towards the organizations. Although both customer stewardship and gratitude towards organization transmit the influence of servant leadership on service performance, only the latter transmits the influence of servant leadership on customer-directed extra-role behaviour (Study 1). In contrast, while the findings of Study 2 replicate the mediating influence of gratitude towards the organization, it did not replicate the mediating influence of the customer-centric attitude of customer orientation relative to customer stewardship. While this may highlight conceptual differences between the two forms of customer-centric attitudes as well as the context of the studies (Study 1 in nonfinancial service outlets while Study 2 in a financial service (brokerage firm)), gratitude towards the organization appears to offer a stronger explanation of the performance implications of servant leadership. Although this finding reinforces the primacy of social exchange theory in this research domain (Eva et al., 2019; Lee et al., 2019; Panaccio et al., 2015; Sun et al., 2019), prior research has not examined gratitude as a driver of customer service performance. However, as research has shown gratitude to relate to organizational citizenship behaviour (Spence et al., 2014; Sun et al., 2019) and helping behaviour (Sawyer et al., 2022), the gratitude-driven customer service performance (service performance and customer-directed extra-role performance) we uncovered helps to deepen social exchange-based employment relationships. These findings provide initial support for Fehr et al. (2017) call for research to examine the role of gratitude in domains such as servant leadership (Sun et al., 2019) and customer service where it could play a potentially central role.

Lastly, and contrary to our expectation, initiative climate did not moderate the relationships between the motivational drivers and service performance (Study 2). Prior research has reported mixed findings regarding the ability of facet-specific climate to moderate the relationships between motivational drivers engendered by servant leadership and its performance outcomes (Chen et al., 2015; Eva et al., 2019; Walumbwa et al., 2010). The lack of support for the hypothesized moderating role of initiative climate suggests that the prosocial motivation that gratitude engenders negates the influence of initiative climate in strengthening the gratitude towards the organization-service performance relationship. Alternatively, rather than complementing or strengthening the autonomy-inducing effect of servant leadership in leveraging its motivational implications, initiative climate substitutes for (or weakens) the motivational implications of servant leadership in promoting enhanced service performance. Considering the suggestion that organizations adopt training in servant leadership as an intervention strategy (Van Dierendonck, 2011), our findings suggest a need for research to investigate the ecosystem of servant leadership to better understand contextual factors that strengthen (complement) or weaken (substitute) the motivational implications of servant leadership for enhanced service performance.

Practical implications

Three key practical implications for the management of frontline employees follow from the findings across the two studies. Our finding regarding the importance of social exchange relative to social learning theory suggests that service organizations should ascertain and satisfy the needs of followers (Chiniara & Bentein, 2016; Mayer et al., 2008) in order to enhance the motivational underpinning of servant leadership in promoting customer service performance. Satisfying the needs of followers may engender the moral emotion of gratitude which has been noted to be effective in such domains as servant leadership and customer service (Fehr et al., 2017). Second, the increasingly competitive service sector has underscored the importance of leadership styles that are not only suited to this context but also, serve multiple stakeholders. Consistent with prior research, our findings reinforce calls for training in servant leadership as an intervention strategy for developing the internal capability needed to satisfy not only the goals of the organization but also of its internal and external stakeholders. As research has shown personality traits such as core self-evaluations and pro-socially motivated conscientious people to relate to servant leadership, service organizations may need to consider these traits in the selection and development of servant leaders (Eva et al., 2019; Panaccio et al., 2015). Lastly, the finding pertaining to the role of initiative climate as a boundary condition suggests a need for service organizations to understand the ecosystem of servant leadership to ensure the investment in servant leadership training can be effectively leveraged to enhance not only its motivational but also performance implications. This is because the non-significant moderating role of initiative climate suggests that the simultaneous use of servant leadership and initiative climate does not strengthen the extent to which the motivational implications of servant leadership can be leveraged to promote enhanced performance. This has managerial practice implication because it shapes the decision to invest in either training in servant leadership as an intervention strategy or practices that promote initiative climate.

Limitations and directions for future research

Despite the methodological strengths of our research including the multi-study, multi-source, and multi-wave design, we acknowledge several limitations. First, although we draw on the multiple stakeholder tenet of servant leadership to examine a concern for customers (customer orientation) and for employees (gratitude), future research in service contexts should not only replicate our findings but also examine other mediators informed by social learning and social exchange theories. Following Madison and Eva’s (2019, p. 152) and in addition to replicating our findings, research that examines the social learning path may examine prosocial motivation while the social exchange path may examine follower reciprocation in the form of duty orientation. Second, we adopted a dyad perspective of leadership which precludes examining our focal relationships at the unit level. Given the documented influence of servant leadership and associated motivational processes on unit-level performance (Liden et al., 2014), future research should focus on unit level servant leadership and how the ensuing ambient environment engenders the motivational processes at both unit and individual levels. Third, future research should control for other leadership styles such as transformational leadership to more confidently ascertain its incremental contribution relative to other leadership styles (Chen et al., 2015). Lastly, although the findings of servant leadership research have been shown to be largely invariant across cultures (Mittal & Dorfman, 2012), data for the two studies were obtained from South Korea and Taiwan, respectively. As these two countries share a collectivistic culture where individuals attach low priority to self-interest, future research should ascertain the generalizability of our findings to an individualistic culture where individuals tend to be high in self-interest. This is because research has shown those high in self-interest to be most changed by a servant leader than those low in self-interest (Wu et al., 2021).

Conclusion

We examined the efficacy of a social exchange relative to a social learning explanation of the documented effects of servant leadership on customer service performance. Across two studies, our findings provide initial evidence that gratitude towards the organization (social exchange) and not customer-centric attitudes (social learning) to be more efficacious in accounting for the performance implications of servant leadership. Additionally, Study 2 revealed that initiative climate failed to moderate the influence of the preceding mediators on service performance. Taken together, our findings suggest that service organizations should not only invest in servant leadership as an intervention strategy but also underscore the importance of understanding the ecosystem of this genre of leadership if service organizations are to effectively leverage its motivational implications for enhanced frontline employee service performance.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The data that support the findings of this study are available from the corresponding author, upon reasonable request.

Notes

1. We compared the ratings of respondents across the three different retail sectors using one-way analysis of variance (ANOVA). The results did not reveal significant differences in their ratings of the study variables.

2. In Study 1 and Study 2, we tested the models with controls and without controls. Removing the controls only slightly changed the strength of estimates but the results of hypothesis testing remained the same.

3. Two of our measures (servant leadership and gratitude towards the organization) in Study 1 and Study 2 evidenced differences in alpha reliability, but all reliabilities as well as their 95% confidence intervals were satisfactory and higher than the .70 criterion (cf., Iacobucci & Duhachek, 2003). Considering that respondents for the two studies shared the same Confucian heritage, these differences (in alpha reliability) may be attributed to their occupation as well as the length of the survey (Rolstad et al., 2011; Taber, 2018).