Servant Leadership and Innovative Behaviour: An Empirical Analysis of Ghana’s Manufacturing Sector

In the competitive global market, innovation is vital to a firm’s longevity. To this end, organisations seek new and alternative ways to motivate employee innovation. This study examines the role of servant leadership as an antecedent to innovation. Drawing on the social identity model, this study examines the effect of servant leadership, team-member exchange (TMX) and perceived insider status on employee innovative behaviour. Primary data were collected from six manufacturing companies in Ghana. Using a sample of 213 employees and their immediate supervisors, a confirmatory factor analysis was conducted to test the discriminant validity of our measurement model. Hierarchical multiple regression was then used to determine direct and interaction effects, followed by bootstrapping tests to identify mediation and moderated mediation effects. The results showed that servant leadership and TMX are significantly related to perceived insider status. The bootstrapping indirect test and Sobel test demonstrated that perceived insider status mediates the relationship between servant leadership and innovative work behaviour. Moreover, the mediated relationship is only significant when TMX is low. This study empirically validated servant leadership as an antecedent to employee innovative behaviour. The findings demonstrated that perceived insider status is a mediating mechanism in this relationship, with TMX as its boundary condition.


Introduction
Since the 1990s, there has been a steady increase in the number of publications relating to innovation [1]. The growth reflects an increasingly dynamic global market where the ability to develop and implement new products, services or work processes gives one organisation a competitive edge over another [2]. Due to this competitive business climate, organisations are increasingly concerned with ensuring a lasting business [3]. To maintain a sustainable business, organisations seek multiple means to achieve a sustainable advantage. Innovation has been identified as an alternative that fosters sustainability because it allows for the development of strategic tools to support management in organising operations [4]. This has motivated researchers and businesses to look for creative ways to stimulate and motivate innovative behaviour among followers and employees [5].
A review of the extant literature points to leadership style as a key antecedent to innovation [6][7][8]. Leaders play multiple roles through their guidance, support, initiating structure, motivating, and championing behaviour to cultivate innovation [9]. However, Shalley, Zhou and Oldham [10] argue that the extant empirical literature is inconclusive, and highlight the need for further research into the

Servant Leadership and Innovative Work Behaviour
Eva et al. [13] define servant leadership as an 'other-oriented' leadership concept wherein the leader prioritises the individual needs and interests of followers, and is oriented towards concern for others, including society at large. This definition reflects key dimensions of the servant leader as having a deliberate focus on the interest and well-being of others, personalised interactions with followers, and a broader concern for organisational actors and larger community. In a review, van Dierendonck [15] presents six characteristics of servant leadership: (1) empowering and developing followers by cultivating a self-confident attitude and giving followers a sense of personal power; (2) humility to seek the help of others, and retreat to the background when necessary; (3) authenticity to one's self, and adherence to a moral code; (4) interpersonal acceptance of the perspectives and feelings of others; (5) providing direction to ensure that followers know what their roles and responsibilities are; and (6) stewardship to act as role models and caretakers of followers.
Innovation is a multistage process involving the generation, promotion, and realisation of new ideas, with each stage requiring different activities and individual behaviours [25]. During the initial stage, the individual generates ideas or solutions to identified problems. These ideas and solutions can be novel or adopted from elsewhere. The next stage involves trying to legitimise the idea or solution by seeking the support of potential allies. In the final stage, the idea is brought to life by the creation of a prototype that can be implemented within the work group or organisation [26]. This study focuses on innovative work behaviour-that is, the behaviours directed towards the intentional introduction of new and useful ideas within a work role, group or organisation [27].
There is some empirical evidence linking servant leadership to employee innovative behaviour. For instance, Panaccio et al. [22] contend that because servant leaders focus on employees' needs rather than their own selfish interests, they enhance their followers' psychological contract fulfilment, and motivate innovation. Such leaders build relationships with employees that go beyond the economic exchange associated with contractual employment relationships. As a result of the genuine intentions of servant leaders, their followers are keen to put in the effort required to realise new ideas [28].
This study examines the relationship between servant leadership and innovative work behaviour within the framework of social identity theory. According to social identity theory, an individual's self-concept is derived from their perception of their membership to a relevant social group [29]. The theory predicts how perceived social identities influence intergroup interactions based on perceived legitimacy and group status differences. Thus, while the servant leader's focus is not on influencing organisational outcomes per se, especially when that outcome does not serve the follower's interest, we argue that the strong sense of belonging and acceptance that arises from the dyadic relationship with the servant leader will result in employees working to advance organisational goals. More specifically, we argue that because servant leaders build personalised relationships with their followers, followers are likely to hold perceptions of belonging to the inner circle, thus providing strong intrinsic motivation to engage in innovative behaviour. In this regard, innovative behaviour could serve a functional role insofar as such behaviour serves a dual purpose: to enhance the image of the group to which the individual belongs, and to maintain the individual's status within the said group.

Servant Leadership and Perceived Insider Status
Perceived insider status can be defined as an employee's perception of being a part (insider) of the organisation. As such, perceived insider status reflects an employee's perception of being part of the 'in-group' and having earned acceptance within a work organisation, which translates into a sense of personal relatedness with other organisation members [19]. Employees who feel like insiders believe that they have earned a personal space within the organisation. Such employees freely share their relevant needs and feelings with other insiders, resulting in the creation of feelings of intimacy [30].
Social identity theory suggests that servant leaders treating their followers with integrity and respect will result in followers engendering a more favourable self-concept through social categorisation and comparison processes. Tajfel and Turner [29] propose three cognitive processes-namely, social categorisation, social identification, and social comparison-underlying the evaluation of others as being part of the 'in-group' or 'out-group'. The social categorisation process involves putting labels on and categorising others (including ourselves) to help understand the social environment. This process constitutes the basis from which we ascribe behaviours that we believe are expected of groups to which we and others belong. As people tie their self-worth to group membership, social identification results in individuals behaving in ways that serve the interest of the group(s) to which they belong. After categorising and identifying with a group, people engage in intergroup comparisons, attaching prestige to some groups but not others.
We seek to extend the social identity theory by anticipating that servant leadership could influence employees' sense of belonging to the group. Social identity theory argues that leaders can influence employees' self-identity insofar as employees consider themselves as ingroup as opposed to outgroup members [31]. The extant literature points to a leader's interpersonal treatment as cues for followers' perceptions of status within the work group [32]. For servant leaders, their follower-centric and authentic nature helps create strong bonds with their employees, making them feel like partners in the organisation [13]. Accordingly, Liden et al. [33] suggest that because servant leaders have a unique relational style that requires one-on-one attention to their followers, servant leaders build strong relationships with followers that develop into identities interlinked with the work group represented by the leader.
Group identification reflects a trickle-down process whereby the leader's qualities may engender group identification [34]. By extension, we anticipate that servant leadership will promote perceptions of insider status because servant leaders prioritise individual needs and have a genuine interest in nurturing and investing in the long-term goals of their followers. Servant leaders are humble and authentic; they empower and guide their followers, accept them for who they are, and generally seek the betterment of society-qualities that help create mutual trust [16]. Thus, we propose the following hypothesis: Hypothesis 1. Servant leadership is positively related to perceived insider status.

The Mediating Role of Perceived Insider Status
Stamper and Masterson [19] conceptualised perceived insider status based on the notion of citizenship [35], arguing that being an insider comes with certain rights and responsibilities. Being an insider implies an acceptance of the responsibilities that come with this status. This inherent sense of responsibility may explain why employees undertake extra-role activities, including innovative behaviour. This feeling of obligation to the group to which an individual belongs is consistent with social identity theory.
A key premise of social identity theory is that the desire for the positive distinction of one's own group explains why group members undertake activities they believe to be beneficial to the group image-a desire hinging on subjective beliefs about the legitimacy and stability of status within groups [29]. We make two assertions regarding innovative behaviour. First, employees are likely to view group success as linked to their own success due to the emotional connection to the group. To identify with a group is to personally experience the losses or successes of the group [34]. Consequently, employees will strive to be innovative in order to improve the work group's image and ensure its continued success. Innovation will also provide an opportunity for employees to reaffirm their social status as insiders. In this regard, Chen and Aryee [36] identify perceived insider status as a mechanism explaining innovation. More specifically, they argue that employees belonging to the 'in-group' explore ways to advance their individual and organisational goals, thereby promoting innovative behaviour. [19] explain that the various relationships between employees and the organisation that give rise to perceived insider status are produced by differences in inducements such as promotions, benefits, and training. Access to opportunities enhances employees' skills and abilities [37]. This suggests that the servant leader's tendency to create a supportive environment by providing the resources-both tangible and intangible-required by employees will send signals to employees that they are valued members of the organisation, motivating them to be innovative.

Stamper and Masterson
Similarly, Tokay and Eyupoglu [38] posit that employee clarity regarding work-related procedures and decisions will encourage extra-role behaviour as the trust between employees and the organisation is strengthened. We also expect that, because servant leaders abide by a moral code, followers will have faith in the organisational systems, trusting that they will be rewarded fairly and recognised for the work they do. We thus predict that perceived insider status is a psychological mechanism explaining how servant leaders influence innovative behaviour. Therefore, we propose the following hypothesis: Hypothesis 2. Perceived insider status mediates the relationship between servant leadership and innovative work behaviour.

TMX and Perceived Insider Status
Tajfel and Turner [29] suggest that social identities are more influential when individuals have an emotional connection to a group and when they consider membership to this group as central to their self-identity. Employees interact with their peers, often working on related tasks with shared organisational resources to achieve communal work goals. Thus, co-workers could serve as social models for employees in that these relationships are valuable to the execution of work roles. In fact, Ashforth and Meal [34] suggest that factors such as interpersonal interaction, close proximity and shared goals (factors typically associated with group formation) are likely to influence how people identify with a group. Thus, we contend that the exchange relationship with peers-that is, TMX-may be central to the formation of social identities within the work group.
TMX reflects the quality of the relationship with members in a group. TMX is defined as the reciprocal exchange of ideas, resources, feedback, and recognition. In high-quality TMX relationships, the levels of reciprocity go beyond what is expected for the completion of a task [39]. Employees who enjoy high-quality TMX believe themselves to be more respected and hold a higher status compared to others [40]. They have also been shown to exhibit organisational commitment [41]. Such individuals see their interactions with team members as indicative of their team identity and ignore the differences that exist between unique dyadic exchanges with individual team members [42]. An experiment conducted by Farmer, Van Dyne and Kamdar [43] also found that individuals who believed they had positive social comparison reported higher-quality relative TMX.
Employees who enjoy good-quality relationships with their peers become psychologically attached to their work group [44], paving the way for social identification and comparison processes that allow for the linking of employee self-esteem and perceptions of insider status. Moreover, employee perceptions of inclusion in an organisation are formed by the treatment they receive from both leaders and co-workers [41]. Accordingly, we posit the following hypothesis: Hypothesis 3. TMX is positively related to perceived insider status.

The Moderating Role of TMX on the Relationship between Servant Leadership and Perceived Insider Status
An employee's social network at work consists of parallel relationships with the leader and their peers. There is empirical evidence to support the notion that relationships with the leader and co-workers complement and play an equally important role in influencing multiple attitudinal and task performance indicators in the work environment [45,46]. Farh, Lanaj and Ilies [45] found that TMX can enhance performance even when the quality of resources provided by the leader is low. They argue that employees who enjoy high TMX may be protected from overdependence on low-quality resources as a result of their obligation to use their colleagues' resources.
The extant research asserts that individuals with high TMX receive more resources and job-specific tasks than peers with low TMX [39]. These employees perform their roles more effectively because they are more adept at utilising and anticipating the actions of co-workers [42]. They also earn the trust of their peers, helping them to gain credibility within the team [47]. These findings highlight how employee perceptions of belonging could be satisfied by high-quality exchange relationships with peers, leading us to propose that high TMX could substitute servant leadership in facilitating perceptions of insider status within an organisation.
More specifically, we argue that servant leadership will be more important in facilitating in-group perceptions for members who experience a lower-quality exchange with their peers. Another study supporting this rationale suggests that individual perceptions of relatively high TMX in work teams satisfy the need for differentiation and identification [43], as indicated by the social identity theory. Hence, we propose the following hypothesis: Hypothesis 4. TMX moderates the relationship between servant leadership and perceived insider status, such that the relationship is stronger for employees who experience low rather than high TMX.

The Moderating Mediation Role of TMX
While we acknowledge the significance of servant leaders in motivating employees to be innovative, we contend that support from team members is just as important at each stage of the innovation process. According to Reagans and McEvily [48], employees in high-quality relationships engage in frequent knowledge exchange with one another, thereby creating knowledge pools from which employees can draw in developing new ideas. As employees have more contact with their peers than their leader, they could engage in more work-specific interactions with colleagues. Tiwana and Mclean [49] suggest that individuals exhibit innovative behaviour when they work in an environment where people often share and bounce ideas off one another. As mentioned, getting others on board with a novel idea can be a daunting task for the innovator [50]. However, for employees in high-quality TMX, the emotional support received from peers serves as a support system, thus enabling their pursuit of innovation [51].
Nonetheless, engaging in innovative behaviour involves an inherent risk as ideas may go against the status quo and challenge existing structures and routines. Thus, the support of peers becomes vital in dealing with the stress of the job [52]. If interpreted as perceptions of in-group status, the support could provide the psychological safety net needed to forge ahead despite the challenges. According to Jokisaari [53], co-workers are more likely to empathise, and subsequently provide the support needed for innovation. Workers tend to seek social support from the people with whom they spend time in close proximity [54]. Thus, we contend that the support of those whose work processes might be affected will be key to the success of innovation.
Essentially, the extant literature suggests that the impact of the servant leader on innovative behaviour becomes less salient for employees enjoying good relationships with their colleagues. In contrast, employees with low TMX-who may lack emotional support, trust, frequent interaction, and knowledge exchange with peers-are likely to become more reliant on the servant leader for a sense of belonging and the resources required to motivate the generation, promotion and implementation of new ideas. Tajfel [55] suggests that individuals in high-status groups have a lower need for positive affirmation. On the contrary, people in low-status groups feel implicitly threatened and go to great lengths to differentiate themselves from high-status groups. We extend this rationale of the social identity theory and argue that employees who experience low TMX may feel a greater need to prove themselves as worthy ingroup members, thus strengthening the saliency of servant leadership as well as the perceived insider status to affect innovative behaviour. In order to explain why and when servant leadership facilitates innovative work behaviour, we propose a moderated mediation model depicted in Figure 1. Thus, we hypothesise that: Hypothesis 5. TMX moderates the indirect effect of servant leadership on innovative work behaviour via perceived insider status, such that the mediated relationship is stronger under low rather than high TMX.

Sample and Procedure
The sample for this study consisted of employees from Ghana's manufacturing sector. While much of the day-to-day business for manufacturing workers is routine, employers encourage innovative behaviour from "ordinary" employees to managers [56]. By using manufacturing workers, we sought to reduce effects that are due to an orientation towards innovative-related work roles. Additionally, because supervisors maintain close relationships with employees to ensure efficiency, and employees in turn work closely with each other to accomplish their tasks, we selected manufacturing sector employees to be our participants. Prior to commencing our research, the first author sent letters to various firms explaining the research project and assuring interested organisations of confidentiality of the respondents and their responses. The data were collected from full-time employees in six manufacturing firms in a major city in Ghana. All firms were privately owned companies. The data were collected during regular working hours with the help of a research assistant and the human resource personnel at the respective firms.
In order to reduce the common method bias, we employed a multisource and multitime approach to data collection. Separate questionnaires were administered to employees and their immediate supervisors. In the first wave, the focal employee measured their perception of their supervisor's servant leadership, their own competence, perceived insider status, and TMX. In the second wave, the supervisor measured employees' innovative work behaviour. Following previous research on innovative behaviour [46], each wave was separated by two weeks. Of the 250 surveys administered, 219 surveys were returned from employees and 26 immediate supervisors. After deleting missing data, 213 matched dyads remained with a response rate of 85.2%.
In this sample, 48.8% of respondents were male, while 51.2% were female. The average age of respondents was 32 years (SD = 5.96), the mean organisational tenure was 3.14 years (SD = 2.25) and the average length of education was 14.62 years (SD = 2.09).

Servant Leadership
Servant leadership was measured using the seven-item scale employed by Liden et al. [57]. A sample item is "My leader makes my career development a priority." The response scale was measured using a five-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree). The Cronbach's alpha was 0.88 (see Appendix for complete items).

Sample and Procedure
The sample for this study consisted of employees from Ghana's manufacturing sector. While much of the day-to-day business for manufacturing workers is routine, employers encourage innovative behaviour from "ordinary" employees to managers [56]. By using manufacturing workers, we sought to reduce effects that are due to an orientation towards innovative-related work roles. Additionally, because supervisors maintain close relationships with employees to ensure efficiency, and employees in turn work closely with each other to accomplish their tasks, we selected manufacturing sector employees to be our participants. Prior to commencing our research, the first author sent letters to various firms explaining the research project and assuring interested organisations of confidentiality of the respondents and their responses. The data were collected from full-time employees in six manufacturing firms in a major city in Ghana. All firms were privately owned companies. The data were collected during regular working hours with the help of a research assistant and the human resource personnel at the respective firms.
In order to reduce the common method bias, we employed a multisource and multitime approach to data collection. Separate questionnaires were administered to employees and their immediate supervisors. In the first wave, the focal employee measured their perception of their supervisor's servant leadership, their own competence, perceived insider status, and TMX. In the second wave, the supervisor measured employees' innovative work behaviour. Following previous research on innovative behaviour [46], each wave was separated by two weeks. Of the 250 surveys administered, 219 surveys were returned from employees and 26 immediate supervisors. After deleting missing data, 213 matched dyads remained with a response rate of 85.2%.
In this sample, 48.8% of respondents were male, while 51.2% were female. The average age of respondents was 32 years (SD = 5.96), the mean organisational tenure was 3.14 years (SD = 2.25) and the average length of education was 14.62 years (SD = 2.09).

Servant Leadership
Servant leadership was measured using the seven-item scale employed by Liden et al. [57]. A sample item is "My leader makes my career development a priority." The response scale was measured using a five-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree). The Cronbach's alpha was 0.88 (see Appendix A for complete items).

Perceived Insider Status
We employed the six-item scale by Stamper and Masterson [19] to measure employees' perceived insider status. A sample item is "My work organisation makes me believe that I am included in it." Responses were measured on a five-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree). The Cronbach's alpha was 0.86 (see Appendix A for complete items).

Team-Member Exchange
TMX was measured using the three-item scale developed by Chae, Seo, and Lee [58]. A sample item is "Other members on this team recognise my potential." Responses were measured on a five-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree). The Cronbach's alpha was 0.74 (see Appendix A for complete items).

Innovative Work Behaviour
Supervisors assessed subordinates' innovative work behaviour using the three-item scale adapted from Schuh et al. [59]. A sample item is "This employee searches out new working methods, techniques, or instruments." The responses ranged on a Likert scale of 1 (strongly disagree) to 5 (strongly agree) with a Cronbach's alpha of 0.84 (see Appendix A for complete items). We examined the validity of the scale by conducting principal axis factor analysis with varimax rotation. The results show that all three items have high loadings (average loading 0.78) on the single factor. The factor also explained 86% of the total variance in the items.

Control Variables
We included employee age, gender, organisational tenure and education as controls in order to account for demographic differences in perceived insider status, leadership perceptions, TMX and innovative behaviour [24,46]. Age was measured in years. Gender was measured as a dichotomous variable, coded as 0 for male and 1 for female. Education was measured as years of education. Additionally, we controlled for competence, which has been shown to be highly correlated with innovative behaviour [60]. Employees were asked to measure their competence on three items using a five-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree). This scale represents the competence dimension of the psychological empowerment construct developed by Spreitzer [61]. A sample item is "I am confident about my ability to do my job." The Cronbach's alpha was 0.72 (see Appendix A for complete items).

Common Method Bias
In the research design stage, we employed a multisource and multitime approach to data collection to reduce the common method bias (CMB). However, because all the variables apart from innovative work behaviour were measured by the same participants, CMB may result in false internal consistency and lead to potentially misleading results [62]. To assess the effects of common method variance, we followed the recommendation by Podsakoff, MacKenzie, Lee and Podsakoff [63] and conducted Harman's single-factor test by loading all the items of the study constructs into an exploratory factor analysis. The results indicate that no single factor explained more than 34% of the covariance among the variables. Again, a confirmatory factor analysis (CFA) with all the variables loaded on one factor revealed that the single-factor model exhibits significantly worse fit indices (χ 2 = 909.02.26; df = 209; CFI = 0.64; TLI = 0.60; RMSEA = 0.13) than the hypothesised five-factor model. Thus, we concluded that CMB did not significantly alter the validity of the study results.

Analytical Strategy
We first conducted a confirmatory factor analysis to determine the distinctiveness of the study variables and then carried out a chi-square model comparison test using STATA 14.1. All the variables in the study were captured as individual-level variables, and thus the analyses were conducted at the individual level of analysis. We proceeded to use hierarchical multiple regression analyses to examine the direct and interaction effects. With regard to the interaction term, we mean-centred the values for the independent variable and moderator before creating the product term in order to reduce the collinearity between the composite variables of the product term. We used the bootstrapped indirect effect proposed by Preacher and Hayes [64] to test the mediating effect predicted by Hypothesis 2. The moderated mediation of Hypothesis 5 was analysed following the three interlinking procedures proposed by Muller, Judd and Yzerbyt [65]. Table 1 presents the mean, standard deviation, bivariate and internal reliability of the study variables. All measurement scales had good reliability (0.72 to 0.88). Consistent with our predictions, servant leadership has a significant correlation with perceived insider status, and innovative work behaviour (i.e., r = 0.42, p < 0.01; r = 0.47, p < 0.01, respectively). TMX is also significantly correlated with perceived insider status, and innovative work behaviour (i.e., r = 0.37, p < 0.01; r = 0.30, p < 0.01, respectively). The variance inflation factor for all independent variables in the regression analyses were below 10, suggesting that multicollinearity is not an issue [66].

Confirmatory Factor Analysis and Chi-Square Difference Test
The stringent cutoffs for the model fit indices are comparative fit index (CFI) and Tucker-Lewis index (TLI) ≥ 0.95, as well as root-mean square error of approximation (RMSEA) ≤ 0.05 [67]. The hypothesised research model (a five-factor model) showed sufficient fit indices (χ 2 = 270.26; df = 199; CFI = 0.96; TLI = 0.96; RMSEA = 0.04) and differed significantly in comparison with other alternative models, as shown in Table 2.

Hypotheses Testing
Hypotheses 1 and 3 predicted a positive relationship between servant leadership (H1) and perceived insider status, as well as between TMX (H3) and perceived insider status. The hierarchical regression results presented in Table 3 show a significant, positive relationship between servant leadership and perceived insider status (β = 0.30, p < 0.01; Model 2) as well as between TMX and perceived insider status (β = 0.21, p < 0.01; Model 2) after controlling for gender, age, education, tenure and competence. In order to test the mediating effect predicted by Hypothesis 2, we utilised the bootstrapping indirect effect [64]. We used 10,000 replications and found support for the indirect effect of servant leadership on innovative work behaviour through perceived insider status. The coefficient effect was 0.16, and the 95% bias-corrected bootstrap confidence interval did not contain zero [0.08, 0.26]. In addition, the Sobel [68] test showed significant effect (z = 3.28, p < 0.01). These results support full mediation.
Hypothesis 4 proposes that TMX moderates the relationship between servant leadership and perceived insider status, such that the relationship is stronger under low rather than high TMX. In the regression table, the interaction between servant leadership and TMX was found to be significant. The interaction co-efficient was negative. (β = −0.17, p < 0.05; Model 3). The post hoc simple slope analysis reveals that servant leadership is related to perceived insider status under low TMX (simple slope = 0.34, SE = 0.07, p < 0.01), but not under high TMX (simple slope = 0.12, SE = 0.10, ns). Figure 2 depicts this interaction. We followed the three-step procedures developed by Muller et al. [65] to test the conditional indirect effect predicted in Hypothesis 5. In the first step, the cross-product term between the independent variable and moderator should be significantly related to the mediator. For the second step, the mediator should be significantly related to dependent variable, while controlling for all independent variables and the interaction term. In the final step, the confidence interval from the bootstrapping analysis for the conditional indirect effect must not include zero.
The results of Hypothesis 4 provide support for step one (β = −0.17, p < 0.05; Model 3). The second step is supported (β = 0.32, p < 0.01; see Model 6). Perceived insider status was positively related to innovative work behaviour after controlling for demographics, competence, servant leadership and the interaction term. In the final step, we ran the conditional indirect effect with 10,000 resamples and a 95% bias-corrected confidence interval. Results show that when TMX is high, the mediated relationship between servant leadership and innovative work behaviour via perceived insider status is not supported (indirect effect = 0.05, SE = 0.06, BCa CI = [−0.05 to 0.17]). In contrast, the mediating role of perceived insider status is supported in the case of low TMX (indirect effect = 0.13, SE = 0.05, BCa CI = [0.05 to 0.23]). Together, these results support Hypothesis 5.

Discussion
Employing social identity theory, this study sought a better understanding of why and how servant leadership facilitates innovative work behaviour. We examined the role of perceived insider status and TMX to explain this relationship. The results showed that, after controlling for age, gender, education, tenure and competence, servant leadership was significantly related to perceived insider status. The results supported the notion that perceived insider status plays a mediating role. As predicted, the relationship between servant leadership and perceived insider status was stronger for employees perceiving low TMX. Furthermore, the conditional indirect effect was significant for low TMX but not high TMX. The theoretical and practical implications are discussed below.

Theoretical Contributions
This study sought to contribute to the growing empirical research on servant leadership and its outcomes and its mediating factors. First, to the best of our knowledge, no study has empirically examined perceived insider status as a mediating mechanism explaining the relationship between servant leadership and innovative work behaviour. Recent studies have examined the role of psychological contract fulfilment, employee engagement and empowerment as mediators in the relationship [13]. We contribute to the current literature by providing new insights into We followed the three-step procedures developed by Muller et al. [65] to test the conditional indirect effect predicted in Hypothesis 5. In the first step, the cross-product term between the independent variable and moderator should be significantly related to the mediator. For the second step, the mediator should be significantly related to dependent variable, while controlling for all independent variables and the interaction term. In the final step, the confidence interval from the bootstrapping analysis for the conditional indirect effect must not include zero.
The results of Hypothesis 4 provide support for step one (β = −0.17, p < 0.05; Model 3). The second step is supported (β = 0.32, p < 0.01; see Model 6). Perceived insider status was positively related to innovative work behaviour after controlling for demographics, competence, servant leadership and the interaction term. In the final step, we ran the conditional indirect effect with 10,000 resamples and a 95% bias-corrected confidence interval. Results show that when TMX is high, the mediated relationship between servant leadership and innovative work behaviour via perceived insider status is not supported (indirect effect = 0.05, SE = 0.06, BCa CI = [−0.05 to 0.17]). In contrast, the mediating role of perceived insider status is supported in the case of low TMX (indirect effect = 0.13, SE = 0.05, BCa CI = [0.05 to 0.23]). Together, these results support Hypothesis 5.

Discussion
Employing social identity theory, this study sought a better understanding of why and how servant leadership facilitates innovative work behaviour. We examined the role of perceived insider status and TMX to explain this relationship. The results showed that, after controlling for age, gender, education, tenure and competence, servant leadership was significantly related to perceived insider status. The results supported the notion that perceived insider status plays a mediating role. As predicted, the relationship between servant leadership and perceived insider status was stronger for employees perceiving low TMX. Furthermore, the conditional indirect effect was significant for low TMX but not high TMX. The theoretical and practical implications are discussed below.

Theoretical Contributions
This study sought to contribute to the growing empirical research on servant leadership and its outcomes and its mediating factors. First, to the best of our knowledge, no study has empirically examined perceived insider status as a mediating mechanism explaining the relationship between servant leadership and innovative work behaviour. Recent studies have examined the role of psychological contract fulfilment, employee engagement and empowerment as mediators in the relationship [13]. We contribute to the current literature by providing new insights into understanding how servant leadership unfolds (i.e., by invoking feelings of identification and belonging) to influence innovative work behaviour. More specifically, servant leaders could enhance perceptions of insider status among followers, thereby promoting innovative work behaviour as employees work to maintain their status as organisational insiders by engaging in behaviours that benefit the work group. In this regard, Zelditch [69] has argued that individuals gain legitimacy by engaging in socially accepted behaviours that match the categorisation and expectations of the groups with which they are identified.
Secondly, while we acknowledge that servant leadership is important for organisational outcomes, researchers have not paid much attention to contexts or conditions that may neutralise the effects of servant leadership. To address this issue, this study examines the moderating role of TMX in the servant leadership literature. The results on the simple and moderated mediating role of TMX may be the most important contribution of our study. First, we hypothesised that TMX will moderate the relationship between servant leadership and perceived insider status, such that the relationship is strengthened when employees do not experience high-quality exchange relationships with co-workers. Second, we predicted that the indirect relationship between servant leadership and innovative behaviour via perceived insider status will be conditional to TMX, such that the indirect relationship will be significant under low TMX but not high TMX. Both predictions are supported by the results of this study.
These findings may be attributable to the fact that employees interact more with their co-workers than with their leader. As such, the need to belong and gain in-group status could be met if they receive the necessary support from peers. However, this suggests that for members experiencing low TMX, the role of the leader becomes critical in facilitating the perceptions of insider status that will motivate engagement in extra-role activities beneficial to the work group, including innovative work behaviour. Our study suggests that there may be other moderating variables that alter the effects of servant leadership.
Thirdly, the results of our analysis support the empirical evidence linking servant leadership to followers' innovative behaviour in a new context. Although some researchers have connected servant leadership to innovation at the individual and team level, the majority of these studies have been conducted in Western countries, thus limiting the generalisability of the results. Ghana is a country with high collectivism [70], meaning that members belonging to a group or association are willing to go the extra mile to fulfil their obligations to the in-group [71]. Although the culture has high power distance [72], which traditionally implies that decision-making is expected to be more decentralised, it also ranks highly on uncertainty avoidance [70]; employees prefer to relegate decision-making responsibilities to the leader so long as their interests are not taken for granted. Thus, the findings of our study suggest that, even in a culture with such a unique blend, the relevance of servant leadership still holds. Our study also helps to dispel long-held notions that servant leadership may not be appropriate in cultures with high power distance [15]. Our data were collected from the manufacturing sector in order to contribute to the generalisability of the servant leadership and innovative work behaviour literature in a new social-economic context, namely West Africa.
Finally, although Yoshida et al. [28] used the social identity theory to examine the image of the leader as a group prototype embodying the team's core values, beliefs and norms, which increases the team's collective efficacy and motivation to be innovative, our study expands the social identity theory and suggests that how employees identify with the servant leader affects their perception of acceptance within the organisation, which in turn stimulates individual-level innovation.

Practical Implications
The results of our study have several implications for managers and their organisations. First, it is important for managers to invest in followers' growth and to engage with employees at a personal level to understand their abilities and capabilities and thus promote innovative behaviours. In the current business climate of globalisation and competitive advantage, it is even more important for organisations to look for multiple ways to inspire employees' innovative behaviour. By reaffirming the importance of servant leaders in the workplace, this study points to the need for organisations to recruit and develop leaders who exhibit servant leadership tendencies like humility, authenticity and stewardship, and who are not motivated by a lust for power. Moreover, the results of our study show that perceived insider status is a mediating mechanism, thus providing a framework through which leaders can motivate innovative behaviour.
Second, this study points to the need for managers to pay greater attention to promoting high-quality exchange relationships among team members. One way of achieving this would be to recruit employees with personality traits such as openness to experience, extraversion and agreeableness. As noted, organisations need to nurture a team-oriented environment by scheduling events and activities that promote both job-and non-job-specific interactions to build trust and respect among co-workers. Such activities need not be limited to orientation sessions or the early stage of team development, but can be made a part of the team's work routine. Managers also need to pay more attention to employees who, as a result of either their personality or job role, do not have significant interactions with their colleagues. For such workers, the relationship with the leader provides cues informing their perceptions of in-group status within the team [32].
In summary, our study highlights the potential direct and indirect effects of servant leadership behaviour. We propose that servant leadership is similar to other value-based leadership theories linked to both in-role and extra-role behaviour in the workplace. We provide evidence indicating that organisations must implement complementary structures that address both the vertical dyadic relationship with the leader as well as the horizontal relationships between team members.

Limitations and Future Research
This study has several limitations that can be addressed in future research. First, although we attempted to use a multitime approach by separating the ratings of the employees and the supervisor by two weeks, the data on servant leadership, TMX and perceived insider status representing the independent variable, moderator and mediator, respectively, were collected at the same time. Consequently, the data are still susceptible to CMB. While we conducted Harman's one-factor test to ascertain the effect of CMB on our results, this was by no means adequate and limits our ability to establish causality.
Second, we conducted our research at the individual level because we were interested in how employee perceptions of leadership influence behaviour in the workplace. However, leadership was originally theorised as a group-level construct grounded in the psychology and social relations of its members [73]. Future research should utilise multilevel or team-level analysis.
Although there is empirical evidence to support the assertion that servant leadership has incremental validity over other leadership theories [74,75], which guided our decision not to control for other leadership styles, this serves as a third potential limitation of our study. Further investigations should use other leadership approaches as covariates to confirm that servant leadership explains more variance beyond these leadership approaches. Particular attention should be given to theories that share similarities with servant leadership in terms of their conceptualisation, including authentic leadership, ethical leadership, empowering leadership and spiritual leadership.
Moreover, we treated innovative work behaviour as a stable variable by measuring it at a single time. However, innovation is a complex process that requires different activities and individual behaviours at each stage [25]. In this regard, Paulus [76] argues that innovation is a nonlinear, cyclical, repetitive process of idea generation and implementation. As such, it is reasonable to assume that the leadership impact on each stage of the innovative process (i.e., idea generation, idea promotion and idea implementation) will differ. Future research needs to address this gap by conducting longitudinal studies, as well as by distinctly analysing the leadership impact at each stage of the innovation process.
As previously noted, due to the growth of multinational corporations, there is a need for universally acceptable leadership approaches to guide practitioners. Consequently, it is necessary to empirically test leadership theories beyond the Western context. While our study has attempted to address this gap by using data from Ghana's manufacturing sector, the generalisability remains limited as the manufacturing sector may not be representative of other work contexts. Therefore, there is still a need to expand the literature on servant leadership to other sectors and non-Western cultures.
Finally, while scholars have paid significant attention to the outcomes of servant leadership, research on its antecedents is lacking and generally limited to leaders' personality traits and gender. Future research must be directed towards the antecedents of servant leadership, particularly for practitioners, in order to help nurture more servant leaders in organisations.

Conclusions
This study sought to contribute to the growing literature on innovation [77,78] by examining the role of servant leadership in facilitating innovative work behaviour. We predicted and empirically proved that perceived insider status is a mediating factor explaining the relationship between servant leadership and innovative work behaviour. We also introduced TMX as a boundary condition in the model. We found that, in addition to the role of servant leadership, the quality of TMX is important in facilitating innovative behaviour through perceptions of in-group status, such that the mediating role of perceived insider status is significant under low TMX but insignificant when TMX is high. Theoretically grounded in social identity theory, this study highlights the functional role of innovation. The findings of this study underscore the need for organisations to implement policies that encourage better interaction among employees. Despite the potential limitations of this study, these findings contribute to the growing research on servant leadership and innovative work behaviour.