The effect of servant leadership on employees’ self-reported performance: Does public service motivation play a mediating explanatory role?

Article history: Received: October 9, 2019 Received in revised format: November 25 2019 Accepted: December 31, 2019 Available online: January 2, 2020 Previous studies have established that leadership has a decisive role in shaping organizational outcomes in general and performance in particular. However, few efforts have gone into explaining the effect demonstrated by servant leadership on how employees do perceive their performance and whether such a relationship is mediated by public service motivation. This paper addresses servant leadership, public service motivation and employee performance among staff in the context of Jordanian civil servants. Data was gathered from a random sample drawn among 3000 employees of the Jordanian customs department. Structure equation modelling techniques were used to statistically analyze the dataset. This draws scholars’ attention into how servant leadership plays a role affecting civil servants perceived selfreported performance while motivating employees to serve the cause of governmental institutions which in turn enhances an increase on the reported performance levels. This paper produces further evidence to support servant leadership and public service motivation adoption to affect performance in previously neglected area of Jordanian public service within the Jordanian customs department. © 2020 by the authors; licensee Growing Science, Canada


Introduction
Evidence accumulated from researchers for the past decades on the effects of servant leadership is indebted to Greenleaf (1970) seminal work for paving the way to study the servant form of leadership (Laub, 1999). Ever since, different avenues for understanding the effects of servant leadership on organizations have been explored. Servant leadership is grounded in the principle that leaders who care less about achieving personal goals and more on fulfilling followers wants are more inclined to motivate people around them (Greenleaf, 1970). Leaders who display concern regarding others more than self-related concerns do show signs of humbleness. This situation stimulates establishing strong links with followers which makes it easier for followers to embrace their work-related goals and targets (Owens & Hekman, 2012). Servant leadership represents a positive view within organizational behavior as it underlines the behavior of leaders which pushes followers on realizing their professional potential (Cameron & Spreitzer, 2012). Such an approach goes a long way to explain how implementing positively oriented approach to human resources is adequately measured, developed, and controlled to attain performance advantages in the competing world of organizational settings today (Luthans, 2002). Despite the fact that different views to leadership do also incorporate supporting followers, servant leadership highlight this strong "followers supporting role" through leading by offering service to followers, as displayed in its name, is what sets this approach to leadership apart from other schools of thought in leadership. Servant leadership theory lends support to the notion that placing a prioritized weight on giving tangible and intangible support to followers.
Although other approaches to leadership include supporting followers, the strong emphasis on leading by serving followers, captured in the name, servant leadership, is unique among leadership approaches. It is thought that when leaders place a priority on providing tangible that guides followers to reach their goals will encourage followers to accept leaders as role models which stimulates appropriate behaviors out of feeling the need to do so, as opposed feeling coerced to behave in a certain way (Greenleaf, 1970). Against the background of today's organizational settings, public organizations in particular are set to benefit from having servant-minded leaders who might influence their organizational behaviors. This research examines the developing role of servant leadership among civil servant in the context of the Jordanian customs department and its effects on employees' performance. Also, our theory predicts that such a form of leadership does support public service motivation that acts a mediating force in explaining part of the effect through an indirect relationship through public service motivation. This paper is organized as follows: first the study model is theorized to produce three hypotheses. Second data gathering and methodological choices are explained. Third data analysis and interpretation are offered to test the theorized hypotheses. Fourth discussion and concluding remarks are revealed.

Servant leadership
Researchers examining servant leadership have drawn similar results in different countries and cultures upholding the positive association with induvial and organizational consequences (Hanse et al., 2016;Liden et al., 2015). In particular, scholarly evidence points to the impact of servant leadership in organizations that are characterized with high bureaucratic structures whereby empowering and mentoring staff is at premium despite varying degrees of centralization and formalization (Eva et al., 2013).
Researchers have grounded their examinations of servant leadership in different theories (Eva et al., 2019). Social learning theory seems to get primal place in scholars' usage to explain how servant leadership affects organizational repercussions (Hunter et al., 2013;Liden et al., 2014). In explaining the effects of servant leadership on follower outcomes, prior research has typically drawn on social learning theory (Hunter et al., 2013;Liden et al., 2014). Social learning theory contemplates that subordinates are inclined to watch and emulate attitudes and behavior as they start to share the values of perceived role models within a certain environment (Bandura,1977). Within a professional organizational setting, leader signifies an important role model that can be attained with power and status over subordinates (Hunter et al., 2013). For instance, leaders disseminate to followers the criteria to judge accepted behavior and attitude through formal commands and rewards systems, or align their own behavior that they expect the other employees to embody (Newman et al., 2017).
When taken in contrast to other forms of leadership, servant leadership is well positioned to promote leaders as role models within organizations. This relies on the evidence pointing to followers' beliefs that leaders have a tendency to act in employees' best interest while having motivation to serve employees well-being without expecting anything as in a form reciprocal interaction with others (Sendjaya et al., 2008). This builds on prior work where social learning theory applications provided evidence that acting in harmony with leaders is a key feature of servant leadership (Graham, 1991). This is supported by researchers' reporting of positive impacts of servant leadership in numerous ways that enhances organizational effectiveness. These include manifestations like facilitating work performance, nurturing creative thinking and innovation, and boosting organizational citizenship behaviors of employees (Eva et al, 2019).
Such an understanding, does point that servant leadership would encourage employees to display higher degrees of public serve motivation in the context of public institutions while bringing performance levels up a notch. Based on the literature review conducted in this research, we propose the two following hypotheses: H1: Servant leadership has a significant effect on the levels of employee performance.
H2: Servant leaders are significantly correlated to public service motivation as exemplified by employees.

Public service motivation
Economists have supported theorized management scholars about the relationship between public service motivation and performance gains. The height of this economists' thinking went as far as calling for public service motivation to be accounted for as a raison d'êtres for providing governmental to support national services (François, 2000).
The premise behind this logic is that motivation to serve in public organizations does supplement their daily jobs. As public service motivation assures that their motivated attitudes do translate into a servicing behavior that lends caring for the collective best interests of governmental institutes and the general public more likely to be met (Grant et al., 2009). Within this perspective employees performance refers to what employees do in their typical time at work which breaks down into actions that are aimed at achieving organizational targets (Campbell & Wiernik, 2015). As such the interest of organizations is theorized to be maintained in the case of highly motivated employees who would invest more time and effort into achieving institutional targets. Based on this understanding, it is reasonable to suggest that public service motivation enhances employee service levels in public organizations. Also, public service motivation can act as a bridge that facilitates the role of servant leaders in enhancing the performance of subordinates.
The arguments made here are used to formulate the following two hypotheses: H3: Public service motivation is associated with employees' self-reported performance in the Jordanian customs department.
H4: Public service motivation does mediate the relationship between servant leadership and employees' self-reported performance.

Fig. 1. study framework
The relationships among study variables are demonstrated in Fig. 1, where direct and indirect relationships are depicted.

data collection
To ensure a relevant level of PSM practices, we targeted general customs department in Jordan to collect data for this study. All employee's customs department currently operating the Jordanian territories constitute the population of this study. This paper employed a simple random sampling technique. Primarily due to its no chance of human bias during data collection, and its ability to free classification error, simple and easy to interpret (Bryman, 2012). The questionnaire was first developed in English, and because the majority of Jordanians citizen mother tongue was Arabic, the research followed (Perrewe et al., 2002) recommendations for back-translation. The study population consisted of 3000 employees currently operating at the Jordan Customs Department (JCD) (Al-A'wasa, 2018). The number of the study's population was identified through using the following equation (i.e. the Steven Thompson equation): where N is the population size, and N=3000, the number of sample size is calculated as n=340. Hence, the general director of the Jordanian customs department was contacted for permission to participate in the study. Based on the sampling calculator suggestion above 340 sample is adequate, however, the researchers distributed five hundred questionnaires (500) to increase reliability and also this was conducted face to face (hand-delivered) to the employees of each customs department. 286 surveys were returned back, and 16 surveys were abandoned because of incomplete questionnaires and/or unengaged responses. The remaining 270 surveys with a response rate of 54.0% were used in order to analyze the data collected from the customs department in Jordan.

research measurement
The questionnaire developed for this research was adapted from the existing leading literature in the field. A 5-Likert scale was used as followed by previous studies, and answers ranged from strongly disagree to strongly agree. A survey instrument was developed in several stages using a short version of Perry's (1996) PSM scale was measured using 13 items adapted from Perry et al., (2008). Servant leadership style was measured using 7 items adapted from Liden et al. (2015). Employee selfreported performance was measured using 4 items based on Vandenabeele (2009).

Model measurement
Using IBM AMOS V24 dataset has gone through confirmatory factor analysis to confirm the underlying factor structure of the data. Additionally, this analysis provided evidence to reliability and validity issues surrounding this research as shown in Table 1   Values of Cronbach alpha reported above show that the validity and reliability are two criteria used to test the measurement model (Hair et al., 2010). Next, we checked for item components reliability using Cronbach's α estimation. Our results revealed that the Cronbach's α value were ranged from 0.72 to 0.89 for all study constructs. The acceptable cutoff point of Cronbach's α value is 0.70 (Hair et al., 2009). These results were adequate to support composite reliability. Next, the AVE was calculated in order to check the convergent validity (Fornell & Larcker 1981). As the scholars noted that the value of AVE must be greater than 0.5 in order to be acceptable (Malhotra & Dash, 2011;Bagozzi & Yi, 1988), however the above Table 2 shows that the AVE values were ranged from 0.5 to 0.7, the AVE of PSM civic duty construct is slightly below the acceptable threshold of 0.50, (0.470). We considered this deviation as minor limitation; this research meets all the criteria for reliability and validity. As a final step in checking the construct validity, the CR was calculated in order to have a significant internal reliability all factors must be greater than 0.70 (Fornell & Larcker, 1981;Hair et al., 2010). However, the above Table  1 shows that all factors acceding the acceptable threshold and were ranged from 0.72 to 0.90, these results seems to be satisfactory and the construct validity results were existed among various constructs during the CFA model measurement. The below Table 2 represents the correlation matrix among study constructs.

Path analysis and hypotheses testing
As mentioned before, in order to test the proposed hypotheses, we must apply the final step in SEM which is conducting the structural model in order to achieve the main goals of this dissertation. Due to that, structural equation modeling was conducting because of the nature of the thesis model, the interaction and mediation effects. According to Hair et al., (2010), SEM was applied due to its nature in allowing the researchers in testing a more sophisticated research model mediation-moderation effect when comparing it with the traditional hierarchical regression analysis. In addition, bias-corrected bootstrapping with 2000 samples method was applied in order to test the indirect effect between the independent (public service motivation) and the dependent (employee performance) constructs. As Preacher and Hayes, (2004), note that, bootstrapping technique is a strong method in testing the mediation effect, due to its ability to resampling the dataset in order to create a confidence interval (CI) for a mediating effect. The first step in conducting a structural model is looking for a good model fit measure. The results of a model fit measures are as following: (χ2/df) = 1.62, CFI = 0.92, IFI = 0.92, TLI = 0.91, RMSEA = 0.048, SRMR = 0.065, PClose = 0.694, (see Table 4.5 in the CFA section 4.5 for more information about the acceptable cutoff point and illustration of the model fit measures). Our research model met all the criteria for a good model fit (Bagozzi & Yi, 1988;Hair et al., 2010). Fig. 2 below shows that the research direct effect hypothesis results were statistically significant at P<0.05. Table 3 along with the above Fig. 2 show that, as was expected, hypothesis one which was proposed that public service motivation has a significant positive relationship with the employee self-reported performance. The results reveled that, there is a statistically significant relationship between PSM and employee performance (β = -0.148, p < .008). Therefore, H1 was accepted negatively. Regarding our second hypothesis which was hypothesized that servant leadership is significantly related to public service motivation (β = 0.276, p < .000), H2 was also accepted positively. Pertaining hypothesis three which was stated out that public service motivation was significantly related to employee performance, the results obtained shown that (β = 0. .499, p < .000) H3 was accepted positively. The results are depicted in Table 3 below.

mediation analysis
Next, we analyze our mediation hypotheses. We used Baron and Kenny's (1986) mediation technique. To test for hypothesis six, we obtain the indirect effect of PSM on employee performance based on the bias-corrected bootstrapping (sampling 2000) and Bias-corrected confidence level of 0.90 were used in this dissertation for testing the indirect effect of Servant leadership to employee performance through Public service motivation. We observed that PSM is significantly and positively associated with employee performance (β = 0.259, p < 0.001, CI 0.176:0.357). It follows that H4 was accepted. As public service motivation does mediate the relationship between servant leadership and employee performance.

Conclusion
The results of this research have shown that there is a significant relationship between servant leadership and employee performance. However, it was a weak negative relationship which contradicts the positive relationships as reported in the literature. The result requires future researchers to conduct further investigations in the public sector in Jordan in general and/or in the customs department in particular to confirm the existence of this relationship direction. On the other hand, the results depicted that there is a statistically positive and significant relationship between servant leadership and public service motivation. In addition, the results have shown that, there is a statistically significant relationship between public service motivation and employee performance was indicating a positive sign, which was supported positively, this is because of the nature of public organizations in general and customs department in particular. Finally, public service motivation does mediate the relationship between. These results enhanced and improve the body of literature in public organization and human resource research by providing evidence from customs administration in Jordan.