Servant leadership,leader effectiveness,and the role of political skill: A study of interscholastic sport administrators and coaches

Abstract

Servant leadership is a model of leadership based on ethics and benevolent service to others and has been associated with numerous positive outcomes for employees and organizations. Due to a limited number of studies examining servant leadership (SL) within sport, the purpose of this study was to investigate the relation of servant leadership and leader effectiveness outcomes in sport administration and to examine if political skill (i.e., how people influence others), was a moderator of servant leader effectiveness. A multilevel model was used to examine these questions within a sample of interscholastic athletic directors and head coaches. Findings revealed servant leadership was directly related to leader effectiveness (LE), affective organizational commitment (AOC), and job satisfaction (JS) of head coaches. Moreover, there was a significant negative interaction between servant leadership and political skill. The negative interaction may indicate that extreme values of servant leadership have opposite relations to leader effectiveness and other outcomes when political skill is present. Though a minimal effect, political skill detracts from LE in those who are perceived as exhibiting strong servant leader behaviors. Findings and limitations are discussed; questions for future research are suggested.

Keywords

Affective organizational commitment,athletic director,job satisfaction

Introduction

Motivated by organizations’ search for pro-social forms of leadership, the focus of leadership studies and practice has recently shifted from traditional authoritarian leadership styles to more positive and less-egoistic forms of leadership aimed at developing subordinates by meeting their needs and promoting the growth of others.^1,2 This shift has increased attention being given to servant leadership, a style of leadership structured on foundational principles of other-centeredness and ethical behavior that suggests truly effective and legitimate leaders prioritize service to others ahead of personal power and control.³ Though research from a variety of settings provides support for servant leadership’s important associations with follower attitudes and behaviors, as well as unit and organizational outcomes^1,2,4 only recently has servant leadership received consideration as an effective leadership model in sport management and the coaching sciences.^5–9

Most definitions of leadership tend to be based on the assumption that leadership involves a process of intentional influence being exerted over others to provide structure, facilitate activities and relationships, and guide the coordinated efforts of followers.¹⁰ Contingency and person-situation theories of leadership assert a leader’s effectiveness in influencing others depends on the interaction of a leader’s behavior with characteristics of the context. However, social influence theory specifically posits individuals use influence, be it intentional or unintentional, to achieve desired social outcomes.¹¹ The social influence literature also encompasses a broad spectrum of social effectiveness constructs (e.g. social intelligence, emotional intelligence, self-efficacy) that describe an individual’s ability to self-monitor, read social situations, and adapt to specific requirements of the environment to influence one’s own behaviors or the behavior of others. Political skill, first popularized in early 2000, is one such construct.

The concept of political skill is defined, “as the ability to effectively understand others at work, and to use such knowledge to influence others to act in ways that enhance one’s personal and/or organizational objective” (Ahearn et al.,¹² p. 311). In addition to positive associations with leader effectiveness,¹³ having high levels of political skill has been associated with numerous positive outcomes for leaders and employees, such as higher job performance ratings from supervisors, increase in team performance, less experienced stress at work, and career success.^13–16 Collectively, political skill appears to be an important concept for leaders to understand and utilize, not only to be an effective leader, but to nurture the development of a successful career.^17,18 Still, key questions remain. How do servant leaders exude influence? How does the use of political skill impact measures of servant leader effectiveness and follower attitudes? Thus, to improve how servant leadership functions in the context of sport, examining the role of political skill as a potential moderator of the relationship between servant leadership and key organizational outcomes (e.g., leadership effectiveness, commitment and satisfaction) represents a worthwhile contribution to the study and practice of leadership in sport.¹⁹

Accordingly, the purpose of this study is to improve understanding of servant leadership in the context of sport. This purpose is accomplished by building on and addressing research recommendations found in the extant sport management and servant leadership literature^1,7,19 in two specific ways. First, athletic director-head coach relationships are explored in the context of interscholastic sport administration to extend the external validity and contextual boundary conditions of servant leadership research. Second, moderating factors that may allow servant leadership to have its greatest possible influence are investigated.

Servant leadership research model

The process of leadership influence happens through multilevel interactions between a leader and followers.²⁰ Thus, a multilevel examination is essential for advancing servant leadership theory¹ and leadership theory in the realm of sport.²¹ The sum of intended explorations in the proposed multilevel model of interscholastic athletic directors and head coaches is illustrated in Figure 1. In the model, the individual outcomes of affective organizational commitment (AOC) and job satisfaction (JobSat) are measured along with perceptions of servant leader behavior (SL) on an individual-level. To measure servant leader effectiveness, scores are aggregated to the school-level for servant leadership (agg.SL) and leader effectiveness (agg.LE) in order to provide results based on group association rather than individual perceptions. Finally, athletic director political skill (PSI) is evaluated at the school-level and analyzed as a moderator between servant leadership and the focal outcomes.

Figure 1.

Hypothesized model. Note. SL: servant leadership; AOC: affective organizational commitment; JobSat: job satisfaction; PSI: political skill inventory; agg.SL: aggregated servant leadership; agg.LE: aggregated leader effectiveness.

Research hypotheses development

The influence of servant leadership

Although several different theories may offer explanations for why servant leaders influence follower attitudes, behaviors, and performance, the most common theories utilized are social exchange theory²² and self-determination theory.²³ Social exchange theory hinges upon a universal principle of reciprocity that engenders positive responses to positive treatment. According to this principle, a follower who receives positive treatment from a leader in the form of trust, guidance, respect, and support will be likely to reciprocate by increasing their commitment to and positive contributions toward the organization. Additionally, when servant leaders satisfy follower’s basic needs for competence, autonomy, and relatedness, servant leaders can engender positive attitudes such as job satisfaction and enhance perceptions of leadership effectiveness among their followers or direct reports.²⁴

Servant leadership research consistently reveals positive associations with team performance,^25,26 employee job satisfaction,²⁴ and positive perceptions of how a group is treated.²⁷ Further, servant leadership engenders a mindset in employees that contributes to enhanced creativity and helping behavior.²⁸ In the context of sport, initial empirical studies investigating servant leadership and its associations with positive outcomes have focused on coach-player relationships. Such research conveys that servant leader coaches tend to produce athletes that are more satisfied, have higher intrinsic motivation, are more task oriented, demonstrate stronger athletic coping skills, and possess more self-confidence than athletes with non-servant leader coaches.^29,30 Servant leadership behaviors of coaches also positively affect athlete satisfaction,³¹ and significant correlations between servant leadership and cohesion, resilience, and coaching success have been reported.⁶

Because the servant leadership model is strongly linked to ethics, virtues, and morality,² it has been recommended as a model of leadership that may provide an avenue to alleviating the ethical imbalance prevalent in intercollegiate athletics.⁵ Indeed, a recent study of college athletic directors supports the notion that servant leadership is effective at influencing the development of an ethical climate within a collegiate sport organization.⁷ When subordinates perceive an organization’s leadership and climate to be ethical, characterized by honesty, concern for others, and interpersonal fairness, attachment to the organization and attitudes toward individual roles and responsibilities are positively influenced.³² Thus, it is expected that perceived servant leadership behavior of athletic directors will be related to perceived leader effectiveness and follower attitudes, specifically affective organizational commitment and job satisfaction.

Hypothesis 1: Athletic directors who exhibit high levels of servant leader behaviors will be positively associated with head coaches’ perceptions of athletic directors’ effective leadership.

Hypothesis 2: Athletic directors who exhibit high levels of servant leader behaviors will be positively associated with head coaches’ affective commitment to their schools.

Hypothesis 3: Athletic directors who exhibit high levels of servant leader behaviors will be positively associated with head coaches’ satisfaction with their jobs.

Political skill and servant leadership

Political skill is an important component of effective leadership because of the many social interactions and influential relationships facing leaders.³³ Politically skilled individuals understand the dynamics of social situations and can combine social astuteness with the ability to make behavioral adjustments to different situational demands in the workplace. This behavior is adapted to situations in a way that appears to be sincere, inspires trust, and results in effective influence over others, allowing objectives to be accomplished without leaders appearing overtly manipulative or controlling.¹³

A 2015 review of political skill research purports positive outcomes associated with political skill fit in one of four categories; career success, individual performance, stress management, or leader effectiveness,¹⁷ and a recent meta-analysis validates political skill as positively related to an individual’s self-efficacy, job satisfaction, organizational commitment, work productivity, and personal reputation and negatively related to physiological strain.³⁴ Other positive outcomes resulting from effective use of political skill are higher job performance ratings from supervisors, improved team performance, less experienced stress at work, and career success.^13–15 Follower perceptions of leader political skill have significantly predicted ratings of leader effectiveness.¹³ As well, leader self-rated political skill has been shown to positively influence follower attitudes, job satisfaction, and commitment when mediated by leader effectiveness.¹⁹

Political skill can influence leader effectiveness through a positive effect on charisma,³⁵ which may play a part in transformational and transactional leadership behaviors serving as mediators through which political skill impacts perceptions of leader effectiveness.³⁶ However, servant leadership does not include a charismatic component, and unlike other types of leaders, a servant leader does not initially aspire to lead but seemingly answers an altruistic calling to be “servant first”. [Greenleaf³, p. 27] Nevertheless, political skill is expected to enhance the effectiveness of servant leaders because of its relation to leader and follower effectiveness that happens through quality leader-follower relationships.³⁷ In fact, many of the skill sets associated with servant leadership and political skill complement one another, such as the need to appear genuine and sincere. Thus, relying again on social exchange theory and given that servant leaders already influence the attitudes, motivation, and satisfaction of subordinates through the processes of social exchange and need satisfaction,^1,2,4 this influence is expected to be enhanced via the social influence of political skill. With that in mind, the following hypotheses are proposed.

Hypothesis 4: Political skill will moderate the relation between servant leadership and leader effectiveness, such that more political skill strengthens the positive relation between servant leadership and leader effectiveness.

Hypothesis 5: Political skill will moderate the relationship between servant leadership and affective organizational commitment of subordinates, such that more political skill strengthens the positive relation between servant leadership and affective organizational commitment.

Hypothesis 6: Political skill will moderate the relation between servant leadership and job satisfaction of subordinates, such that more political skill strengthens the positive relationship between servant leadership and job satisfaction.

Method

Sampling

Data collection consisted of two phases. Phase one utilized a purposive sample of interscholastic athletic directors (ADs) invited via e-mail to complete an on-line questionnaire through Qualtrics. Purposive sampling is a nonprobability sampling technique that researchers use when they desire to recruit study participants from specific, predefined groups (e.g. interscholastic ADs).³⁸ Initial invitations were distributed to a list of ADs in the southwestern United States whose e-mail addresses were gathered from school websites. The initial distribution of invitations was enhanced with a second e-mail invitation sent by the National Interscholastic Athletic Administrators’ Association to its active members. In phase 2 of data collection, e-mail questionnaires were sent to all head coaches on staffs of participating ADs.

Athletic director variables

The athletic director instrument included two control variables. Age was utilized as a control variable given that age might play a role in athletic directors being perceived as servant leaders by head coaches, as well as their self-perception of political skill. Additionally, school enrollment was utilized with the idea that larger programs likely offer a greater number of sport options and employ a greater number of head coaches. In this situation, head coaches might have less individual social exchange with their athletic director, thus possibly impacting their commitment and perception of athletic director servant leadership.

Athletic directors completed the Political Skill Inventory (PSI),³⁹ a questionnaire that contains items designed to assess social astuteness, interpersonal influence, apparent sincerity, and networking. The 18-item instrument utilizes a 7-point Likert scale to measure participants’ perceived levels of political skill where 1 = ‘‘strongly agree’’ and 7 = ‘‘strongly disagree’’. The internal consistency reliability for this scale was α = .90.

Head coach variables

The head coach instrument included two control variables: age and school enrollment. Age was utilized with thoughts that it might impact the bottom-line mentality of head coaches, and enrollment was used with the idea that larger programs might not allow for as much social exchange between athletic directors and head coaches. Other than the control variables, all of the items in the head coach instrument utilized a 7-point Likert scale where 1 = ‘‘strongly agree’’ and 7 = ‘‘strongly disagree”. Head coaches evaluated the servant leadership behaviors of their school's AD by completing the 14-item Servant Leadership Scale.²⁷ The scale reliability was α = .97.

Head coaches evaluated the leader effectiveness of their respective athletic directors by responding to six items adapted from Douglas and Ammeter’s¹³ leader effectiveness scale. The reliability for this scale was α = .90. Affective organizational commitment and job satisfaction were also evaluated by head coaches. Affective organization commitment measures the strength of one’s emotional attachment and identification with an organization. The scale used to measure commitment consisted of five items.⁴⁰ The reliability for this scale was α = .79. Job satisfaction represents a feeling of enjoyment that is derived from one’s job. A four-item scale developed by Eisenberger, Cummings, Armeli, and Lynch⁴¹ was used to assess employee’s job satisfaction. The reliability for this scale was α = .90.

Data analysis

All analyses were completed using R Studio statistical analysis program (RStudio Team, 2015).⁴² Prior to analysis, data was checked for missing entries. Less than 1% missing data on any individual response was not considered a threat and was listwise deleted during analysis as a missing rate of 5% or less is considered inconsequential.⁴³ A confirmatory factor analysis (CFA) of head coach reported scales was executed as well as a test for the influence of common method variance.^44,45 A CFA is used to verify the factor structure of a set of observed variables and allows a researcher to test the hypothesis that a relationship exists between observed variables and their underlying latent construct. The model demonstrated acceptable fit (X²(399, N = 807) = 2774.461, p<.01; CFI = .88, TLI = .88, RMSEA = .08, SRMR = .07). A test of the influence of common method variance (CMV) was performed by running the same CFA but allowing all items to load onto an uncorrelated latent variable representing method. The average variance explained by the method factor was 12%, which is less than the 25% average in the literature (Williams et al., 1989),⁴⁴ suggesting CMV was not a major threat in this sample.

The investigation utilized a multilevel model (Figure 1) and analysis strategy appropriate to the levels involved in each specific hypothesis. First examined was the association of the individual-level predictor variable servant leadership (SL) and outcomes, affective organizational commitment (AOC) and job satisfaction (JS), utilizing linear regression to test Hypotheses 2 and 3_. Measuring outcomes at the individual-level reveals more information about individual perceptions of a leader’s effectiveness in relation to outcomes. Servant leadership was group-mean-centered by subtracting the unit mean (i.e., school mean) from each individual score to create a pure individual-level variable. Group-mean-centering provides similar results to grand-mean-centering but allows for a test of separate level effects.⁴⁶ Thus, group-mean-centering holds value for testing cross-level interactions,⁴⁷ and provides the purest estimate of the within-group slope.⁴⁶

Next, Hypotheses 1 and 4 were tested; this process involved examining the association of school-level predictors, servant leadership and political skill, and the leadership effectiveness outcome. Individual-level servant leadership and individual-level leader effectiveness scores were aggregated to the school-level to provide an evaluation of predictors based on group association rather than individual differences. Regression analysis was used to test the relationship between servant leadership and leader effectiveness, as well as any interaction between servant leadership and political skill that might be present. Already a school-level variable, political skill was grand mean centered prior to analysis to aid in interpreting the interaction.⁴⁸

Finally, testing Hypotheses 5 and 6 involved multilevel modeling to examine the association of school-level predictors, servant leadership and political skill, and cross-level interactions that might influence individual-level outcomes, organizational commitment and job satisfaction. Multilevel modeling can be thought of as an ordinary regression model that has additional variance terms for handling non-independence due to group membership and explains how nesting HCs in subsets may produce additional sources of variance.⁴⁹ Variance partitions coefficient (VPC) and Akaike information criteria (AIC) scores were evaluated to determine the model that explained the most variance. VPC is a measure of the amount of variance explained by the model, calculated by dividing the between-school variance by the total variance. Use of AIC index comparison approach is quite common within the sociological literature. This approach was utilized to quantify the degree to which the given model represents improvement over comparison models.

Results

Data were collected from 250 interscholastic athletic directors (ADs) and corresponding subsets of head coaches (HCs). The sample of respondents consisted of secondary schools covering 47 states in the United States of America. Descriptive statistics, level 1 and level 2 intercorrelations, and scale reliabilities are presented in Tables 1 and 2.

Table 1.

Descriptive statistics.

Level 1 - Head Coaches (n = 809)
	Age	Enrollment	Years with AD	Gender
Mean	43.14	1063	4.8	229 Female
Median	43	950	3	577 Male
Range	21–77	85–4822	<1–25	3 na

Level 2 - Athletic Directors (n = 250)
	Age	Enrollment	Years as AD	Gender	Number of HCs	School Classification
Mean	47.9	1065.2	11	41 Female	19.63	211 Public
Median	48	832.5	10	209 Male	17.5	39 Private
Range	26–66	85–8000	1–38		2–126

Table 2.

Intercorrelations.

Level 1 Correlations
Scale	α =	1.00	2.00	3.00	4.00	5.00	6.00
1. Enrollment
2. Age		−0.01
3. Servant Leadership	0.97	−0.04	0.00
4. Leader Effectiveness	0.90	−0.06	−0.05	0.73
5. Affective Org. Commitment	0.79	−0.02	−0.04	0.45	0.54
6. Job Satisfaction	0.90	0.00	−0.04	0.42	0.55	0.54
Mean		1064.00	43.14	2.03	2.12	2.26	1.84
Standard Deviation		739.86	11.41	1.03	0.94	0.93	0.85
Range		85−	21−	1.00−	1.00−	1.00−	1.00−
		4822.00	77.00	7.00	6.33	6.00	7.00
Level 2 Correlations
Scale		1.00	2.00	3.00	4.00	5.00
1. Enrollment
2. Age		0.06
3. Political Skill	0.90	−0.03	−0.03
4. Servant Leadership (agg)		−0.01	0.08	0.09
5. Leader Effectiveness (agg)		−0.04	0.01	0.13	0.80
Mean		1065.20	47.90	1.91	2.04	2.16
Standard Deviation		967.92	9.09	0.47	0.78	0.80
Range		85−	26−	1.00−	1.00−	1.00−
		8000.00	66.00	3.50	5.50	5.67

Prior to testing the hypotheses, Kernel density plots deemed the assumption of normality tenable and no outliers were identified. Diagnostics tools confirmed assumptions of linearity in the regression models by plotting the residuals of the model to the predicted values and the independent variables. Graphs showed the linearity assumption was met. The normality assumption was checked using QQ plots and histograms, which showed the normality assumption was tenable.

Aggregation statistics were utilized to ensure appropriate aggregation of the multilevel variables (servant leadership and leader effectiveness), and to determine if multilevel modeling was appropriate for individual-level outcomes. First, r_wg(j) was calculated. This is useful for determining within group agreement and reliability.⁵⁰ The mean of group r_wg(j) scores for servant leadership (r_wg(j) = .76) and leader effectiveness (r_wg(j) = .75) exceeded the strong interrater agreement threshold of .70, and affective organizational commitment (r_wg(j) = .52) and job satisfaction (r_wg(j) = .61) are considered moderate agreement (LeBreton & Senter, 2008).⁵² This evidence supports aggregating servant leadership and leader effectiveness to the school-level.

Next, intraclass correlation coefficients (ICC) were calculated as an empirical test of whether sufficient variance existed at the school-level to employ multilevel analyses.⁵² The ICC indicates the percentage of variance in individual measures that can be attributed to group membership. Results indicated the ICC for servant leadership (.19), leader effectiveness (.20), organizational commitment (.08), and job satisfaction (.05) were sufficient for multilevel modeling to be deemed appropriate considering ICCs as low as .05 have been shown to provide evidence of group level effects.⁵¹

Hypothesis tests

Step one utilized linear regression for testing Hypotheses 2 and 3 concerning the direct effects of individual-level predictor servant leadership on individual-level outcomes of affective organizational commitment and job satisfaction. An AIC model was calculated using group mean centered servant leadership and control variables (age and enrollment) to find the model of best fit for each outcome, identifying SL as a lone predictor variable to be the best fitting model. Therefore, control variables were dropped from the final regression analysis. Results indicated the predictors explained 9% of the variance in both organizational commitment (R² = .094, F(1,807) = 84.11, p < .00) and job satisfaction (R² = .091, F(1,807) = 80.71, p < .00). Servant leadership was significantly associated with both organizational commitment (β = .37, p < .00, 95% CI [.30, .46]) and job satisfaction (β = .34, p < .00, 95% CI [.26, .41]) in the predicted direction. Servant leadership’s significant association with organizational commitment and job satisfaction was confirmed, supporting Hypotheses 2 and 3.

Step two involved testing Hypotheses 1 and 4 at the school-level. Because these hypotheses include only school-level predictors, servant leadership and political skill, and the aggregated outcome leader effectiveness, associations were tested using multiple linear regression. An AIC model was calculated to determine the model of best fit using school-level predictor variables servant leadership and political skill and control variables age and enrollment. The model containing an interaction term between servant leadership and political skill provided the best fit. However, the difference in the top three models was negligible. Results of AIC analysis are presented in Table 3. The model containing servant leadership as a stand-alone predictor scored only .16 higher on AIC analysis, suggesting that political skill may not play much of a role in servant leader effectiveness.

Table 3.

Model selection based on AIC.

Model	K	AICc	Delta_AICc	AICcWt	CumWt	LL
SL.PSI	4	346.95	0.00	0.29	0.29	−169.39
SL.Age	4	347.08	0.13	0.27	0.57	−169.46
SL	3	347.11	0.15	0.27	0.84	−170.50
SL.Enr	4	348.11	1.16	0.16	1.00	−169.97
PSI	3	597.48	250.53	0.00	1.00	−295.69
PSI.Enr	4	599.14	252.19	0.00	1.00	−295.49
PSI.Age	4	599.51	252.56	0.00	1.00	−295.68
SL.PSI.Enr.Age	3	600.67	253.72	0.00	1.00	−297.29
Sl.PSI.Age	3	601.19	254.24	0.00	1.00	−297.54
Enr	3	601.30	254.35	0.00	1.00	−297.60
SL.PSI.Enr	3	601.48	254.53	0.00	1.00	−297.69
Age	3	601.76	254.80	0.00	1.00	−297.83

Results of school-level regression are presented in Table 4. School-level servant leadership was positively associated with leader effectiveness, thus supporting Hypothesis 1. Though not significant, political skill was positively associated with leader effectiveness. Results also reveal a significant antagonistic interaction between servant leadership and political skill (β = −.15, p < .10). An antagonistic interaction indicates that extreme values of the predictor variable have opposite relations to the outcome than it would have if there was no moderating effect. However, Hypothesis 4 predicted that political skill would moderate the relationship between servant leadership and leader effectiveness such that higher levels of political skill enhances the positive relationship between servant leadership and leader effectiveness. Thus, Hypothesis 4 was not supported.

Table 4.

Level 2 regression results (school-level).

Leader Effectiveness
	B	SE B	p		95% CI
Intercept	2.169	0.030	0
School-level SL	0.813	0.039	<0.000	***	0.736	−0.889
Political Skill (PSI)	0.106	0.064	0.101		−0.021	−0.234
School-level SL x PSI	−0.150	0.080	0.061	*	−0.307	−0.007

*p < .10, ***p < .001.

Residual standard error: 0.4769 on 246 degrees of freedom.

Multiple R-squared: 0.6472, Adjusted R-squared: 0.643.

F-statistic: 150.4 on 3 and 246 DF, p-value: < 2.2e-16.

The third step of analysis involved testing H₅ and H₆ concerning the cross-level interaction of school-level servant leadership and political skill on individual-level outcomes, organizational commitment and job satisfaction. Results of the multilevel regression are displayed in Tables 5 and 6. First, the simplest multilevel model which allows for school effects on outcomes was constructed. The variance partition coefficient (VPC), calculated by dividing the between-school variance by the total variance, was .103 for commitment and .078 for job satisfaction, indicating that 10.3% (AOC) and 7.8% (JS) of variance in outcome variables can be attributed to school effects.

Table 5.

Multilevel modeling results for affective organizational commitment (AOC).

Level and variable	Null(Step 1)	95% C.I.	Random InterceptFixed Slope (Step 2)	95% C.I.	Random InterceptRandom Slope (Step 3)	95% C.I.	Random InterceptRandom SlopeCross-level Interaction (Step 4)	95% C.I.
Level 1
Intercept	2.26	2.18–2.33	2.26	2.19–2.33	2.26	2.18–2.33	2.26	2.17–2.33
Individual-level SL			0.38	0.30–0.45	0.44	0.35–0.53	0.45	0.03–0.87
Level 2
Variance components
Within-unit (L1) variance	0.78	−0.94–2.5	0.67	−0.94–2.27	0.63	−0.93–2.19	0.63	−0.92–2.18
Intercept (L2) variance			0.12	−0.55–0.79	0.13	−0.57–0.83	0.13	−0.57–0.83
Unit-level SL					0.05	−0.37–0.57	0.04	−0.34–0.42
Unit-level SL x PSI							0.05	−0.37–0.47
Variance Partition Coefficient	0.103		0.150		0.170		0.137
n (indiviual-level)	803.00		802.00		802.00		802.00
n (unit-level)	249.00		249.00		249.00		249.00
Fit Indices
AIC	2155.00		2065.00		2061.10		2066.30
BIC	2169.00		2083.70		2089.20		2108.50
logLik	−1074.50		−1028.50		−1024.60		−1024.10
Deviance	2149.00		2057.00		2049.10		2048.30

Table 6.

Multilevel modeling results for job satisfaction (JobSat).

Level and variable	Null(Step 1)	95% C.I.	Random InterceptFixed Slope(Step 2)	95% C.I.	Random InterceptRandom Slope (Step 3)	95% C.I.	Random InterceptRandom Slope Cross-level Interaction (Step 4)	95% C.I.
Level 1
Intercept	1.84	1.77–1.91	1.84	1.77–1.91	1.84	1.77–1.91	1.84	1.77–1.91
Individual-level SL			0.34	0.27–0.41	0.34	0.25–0.43	0.34	0.25–0.43
Level 2
Variance components
Within-unit (L1) variance	0.66	−0.63–1.95	0.58	−0.91–2.07	0.51	−0.89–1.90	0.51	−0.89–1.91
Intercept (L2) variance			0.08	−0.46–0.62	0.10	−0.53–0.73	0.10	−0.52–0.72
Unit-level SL					0.08	−0.48–0.64	0.08	−0.47–0.63
Unit-level SL x PSI							0.00	−0.93–0.07
Variance Partition Coefficient	0.078		0.116		0.145		0.145
n (indiviual-level)	805.00		804.00		804.00		804.00
n (unit-level)	249.00		249.00		249.00		249.00
Fit Indices
AIC	2014.30		1930.70		1897.70		1903.60
BIC	2028.40		1949.40		1925.90		1945.80
logLik	−1004.10		−961.30		−942.90		−942.80
Deviance	2008.30		1922.70		1885.70		1885.60

Next, to build the multilevel model, the intercept was allowed to vary across schools by adding the individual-level variable servant leadership. Then, the intercept and slope were allowed to vary by adding the school-level variable servant leadership. Finally, to test the cross-level interaction, the school-level interaction term political skill, was added to the model. In the case of both organizational commitment and job satisfaction, an evaluation of VPC and AIC scores indicated the model containing the cross-level interaction explained no more variance than the model allowing for random intercept and random slope. Results also reveal a very small improvement in fit when allowing for a random slope versus a fixed slope. Thus, Hypotheses 5-6 were not supported because political skill did not positively enhance servant leadership’s association with commitment and job satisfaction.

Discussion

The current study incorporates a multilevel model approach which answers a call to extend research examining servant leadership¹ as well as leadership in sport management²¹ beyond a single level. The absence of multilevel studies has been noted as a primary limitation to studying leadership.^1,20 Further, to develop leadership theory focused specifically on sport and understand the contextual boundaries best suited for each theory, it is critical that multilevel investigations be implemented in sport leadership research.²¹

The current study answers a second call, that of investigating the effects of servant leadership across multiple sport contexts.²¹ Complimenting earlier findings that the servant leadership of administrators nourishes an ethical climate in sport organizations,⁷ the results of this study support the notion that servant leadership behaviors exhibited by interscholastic athletic directors do lead to increased head coach perceptions of leader effectiveness, organizational commitment, and job satisfaction. Thus, servant leadership is revealed as a positive and effective style for leading interscholastic sport organizations.

Finally, this study extends understanding of the influence mechanisms involved in leadership by examining the interaction of servant leadership and the social effectiveness construct, political skill. Several hypotheses proposed that political skill would moderate the effectiveness of servant leadership, such that more political skill would enhance the positive relationship between servant leadership and outcomes. These hypotheses were not supported.

Previously, political skill research has shown a positive association with leader effectiveness¹³ and follower attitudes³³ when not considered in conjunction with servant leadership. In the current study, it appears that the most effective servant leaders, as perceived by head coaches, evaluated themselves as possessing less political skill, and servant leaders who rated themselves as having more political skill, were perceived by head coaches as less effective servant leaders. Though this result initially may seem counterintuitive (i.e., that more effective servant leaders are less socially effective), it is reasonable to assume that the humility present in servant leaders^4,9,29 may be a factor in servant leader athletic directors evaluating themselves as having less political skill. Meaning, though these individuals are socially effective servant leaders, they appear to be reticent to rate themselves as highly socially effective. Thus, the presence of humility in servant leadership may be a key factor in increased leader effectiveness scores from subordinates.

Alternatively, this outcome regarding political skill as a moderator of servant leader effectiveness may be a result of influence from the Dunning-Kruger effect.⁵³ This effect is a cognitive bias based on illusory superiority resulting in a miscalibration of one’s ability. In this case, athletic directors who rated themselves as high in political skill were perceived as less effective leaders by head coaches. Perhaps those athletic directors who are not as socially effective in the workplace may not realize their lack of political skill. In turn, athletic directors perceived as effective leaders, who rated themselves lower on political skill, may have done so because they possess a greater self-awareness and understanding of social effectiveness in the workplace. Whether the explanation for our findings is related to humility or false conceptions of self, our contrary findings deserve further attention in future research.

Practical implications

The findings of this study hold important ramifications for athletic directors in that servant leader athletic directors may have the capacity to influence head coaches and student-athletes through role-modeling. The servant leadership of athletic directors can influence goal orientation development, which may ultimately make student-athletes the greatest beneficiaries of having a servant leader athletic director and head coaches who mirror servant leader behaviors (e.g., focusing on personal development, academic success, and constructive social dynamics). Further, because of servant leadership’s positive association with development of an ethical climate,⁷ this form of leadership may be an important and effective style for athletic directors to consider when presented with the opportunity to build or re-build an athletic program based on ethics while also enhancing follower commitment and satisfaction levels.

The movement toward servant leadership practice in amateur sport begins in sport management classrooms by preparing future sport management practitioners. Robinson et al.⁹ for example, present a useful tool for sport management professors to implement the practice of servant leadership in sport management courses. The authors present a three-sphere model of servant leadership based on a core altruistic calling and central qualities of genuine caring, empathy, and humility. Additionally, they offer suggestions for teaching and activities for developing servant leadership in undergraduate and graduate sport management students.

The relationship between servant leadership and political skill also warrants consideration. Rather than not utilizing political skill, which at its core is focused on social effectiveness, perhaps servant leaders should consider how engaging in forms of political skill may benefit others. Being socially effective should not be discouraged; a lot of positive outcomes can stem from leader social effectiveness. Thus, aspiring servant leaders need to consciously balance the possible positive impacts of political skill on career advancement with the potential negative impact of political skill on servant leader effectiveness.

Limitations and directions for future research

A potential limitation to this study is that athletic directors made the decision whether to participate or not. A large sample size was provided by testing this model in a national population, but athletic directors may have chosen not to participate if they felt the relationships with their head coaches were weak and lacking in trust. Additionally, responses from every head coach on each staff were not available as some head coaches chose not to participate and some athletic directors who chose to participate had no head coaches respond to the survey. In future studies, screening questions for athletic directors and coaches could be included to help explain a potential non-response bias. Additional control variables should be considered as well, such as tenure for athletic directors and head coaches as a case can be made that a new head coach might not be able to accurately rate his or her athletic director.

This study is a step toward answering the call to place focused attention on the study of servant leadership¹ and its effects on outcomes of interest in multiple sport contexts.²¹ Another sport-based population in which servant leadership is ripe for investigation is athletic training. Positive benefits from servant leadership may exist for athletic trainers as seen in other service-oriented industries such as nursing.²⁶

Researchers should also continue the investigation of servant leadership in sport utilizing a multi-level design and extend the investigation to 3-levels, inclusive of student-athletes. For example, in the current study, a 3-level design would allow the opportunity to measure the trickle down benefits from the servant leader behaviors of athletic directors to student-athletes. Neubert et al.²⁶ conducted this sort of study, only instead of athletic directors and student-athletes, they explored servant leadership with nurse managers and patient outcomes.

There is also a need to continue exploring the contextual variance of servant leadership effectiveness by considering the influence of follower variables. Indeed, one area to be addressed is the extent to which servant leaders are effective when their subordinates are not of a similar, servant leader mindset. A bottom-line mentality⁵⁴ existing as a one-dimensional win-at-all-costs attitude in head coaches, for example, may deflect the positive benefits of servant leadership espoused by sport administrators.

In the current study, political skill did not emerge as a mechanism that improves servant leader effectiveness, but that does not automatically mean political skills is unimportant to sport leaders. Social effectiveness is a key component of effective leadership,^15,18 and political skill skill is also a notable contributor to career success.^14,16,17 Because of the heightened levels of interpersonal and social aspect of the sport industry¹⁸ as well as the increased visibility of sport administrators and coaches, a study of the career success implications of political skill for sport professionals is needed.

Conclusion

This study extends research on servant leadership by demonstrating the association of servant leadership behaviors with perceptions of leader effectiveness as well as the affective organizational commitment and job satisfaction of employees in a sport administration setting. Political skill was evaluated as an influence mechanism that may enhance servant leader effectiveness, but no positive moderation effect was found. While numerous questions remain for future researchers regarding the influence and practice of servant leadership, this study mirrors accumulating evidence that benefits abound for leaders who practice servant leadership by promoting the development and interests of others.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Marshall J Magnusen

References

Eva

Robin

Sendjaya

, et al. Servant leadership: a systematic review and call for future research. Leader Q 2019; 30: 111–132.

Parris

Peachey

JW.

A systematic literature review of servant leadership theory in organizational contexts. J Bus Eth 2013; 113: 377–393.

Greenleaf

RK.

Servant leadership: a journey into the nature of legitimate power and greatness. New York: Paulist, 1977.

van Dierendonck

Servant leadership: a review and synthesis. J Manage 2011; 37: 1228–1261.

Burton

Welty Peachey

JW.

The call for servant leadership in intercollegiate athletics. Quest 2013; 65: 354–371.

Gillham

, and Hansen

Examining athlete, team, and coach variables across levels of coaching success. Int Sport Coach J 2015; 2: 233–247.

Burton

Welty Peachey

, and Wells

The role of servant leadership in developing an ethical climate in sport organizations. J Sport Manage 2017; 31: 229–240.

Kim

Wells

. Being a servant-leader in sport: servant leadership as the key to the coach-athlete relationship. Int J Sport Manage 2017; 18: 19–43.

Robinson

Neubert

Miller

Servant leadership in sport: a review, synthesis, and applications for sport management classrooms. Sport Manage Ed J 2018; 12: 39–56.

10.

House

Hanges

Ruiz-Quintanilla

, et al. Cultural influences on leadership and organizations: project GLOBE. In: Mobley

Gessner

Arnold

(eds) Advances in global leadership. Stamford: JAI, 1999, pp.171–233.

11.

Levy

Collins

Nail

PR.

A new model of interpersonal influence characteristics. J Soc Behav Personality 1998; 13: 715–733.

12.

Ahearn

Ferris

Hochwarter

, et al. Leader political skill and team performance. J Manage 2004; 30: 309–327.

13.

Douglas

Ammeter

AP.

An examination of leader political skill and its effect on rating of leader effectiveness. Leader Q 2004; 15: 537–550.

14.

Magnusen

Kim

Perrewé

PL.

Gaining a competitive edge when recruiting student-athletes: the role of political skill. Int J Sport Sci Coach 2014; 9: 1291–1310.

15.

Magnusen

Kim

Thriving in the political sport arena. J App Sport Manage 2016; 8: 15–42.

16.

Todd

Harris

, et al. Career success implications of political skill. J Soc Psych 2009; 149: 179–204.

17.

Kimura

A review of political skill: current research trend and directions for future research. Int J Manage Rev 2015; 17: 312–332.

18.

Magnusen

Perrewé

PL.

The role of social effectiveness in leadership: a critical review and lessons for sport management. Sport Manage Ed J 2016; 10: 64–77.

19.

Robinson

Magnusen

Kim

JW.

The socially effective leader: exploring the relationship between athletic director political skill and coach commitment and job satisfaction. Int J Sport Sci Coach 2019; 14: 197–205.

20.

Yammarino

Leadership past, present, and future. J Leader Org Studies 2013; 20: 149–155.

21.

Welty Peachey

Damon

Zhou

, et al. Forty years of leadership research in sport management: a review, synthesis, and conceptual framework. J Sport Manage 2015; 29: 570–587.

22.

Blau

PM.

Exchange and power in social life. New York: John Wiley, 1964.

23.

Ryan

Deci

EL.

Self-determination theory and the facilitation of intrinsic motivation, social development, and well-being.

Am Psych 2000; 55: 68–78.

24.

Mayer

Bardes

Piccolo

RF.

Do servant-leaders help satisfy follower needs? An organizational justice perspective. Eur J Work Org Psych 2008; 17: 180–197.

25.

Liden

RC.

Antecedents of team potency and team effectiveness: an examination of goal and process clarity and servant leadership.

J App Psych 2011; 96: 851–862.

26.

Neubert

Hunter

Tolentino

RC.

A servant leader and their stakeholders: when does organizational structure enhance a leader's influence?

Leader Q 2016; 27: 896–910.

27.

Ehrhart

MG.

Leadership and procedural justice climate as antecedents of unit-level organizational citizenship behavior. Pers Psych 2004; 57: 61–94.

28.

Neubert

Kacmar

Carlson

, et al. Regulatory focus as a mediator of the influence of initiating structure and servant leadership on employee behavior. J Appl Psychol 2008; 93: 1220–1233.

29.

Hammermeister

Burton

Pickering

, et al. Servant leadership in sport: a concept whose time has arrived. Int J Servant Leadership 2008; 4: 185–215.

30.

Kim

, and Wells

(2017). Being a servant-leader in sport: servant leadership as the key to the coach-athlete relationship. Int J Sport Manag 2017; 18: 19–43.

31.

Azadfada

Besmi

Doroudian

AA.

The relationship between servant leadership and athlete satisfaction. Int J Basic Sci App Res 2014; 3: 528–537.

32.

Neubert

Kacmar

Carlson

, et al. The virtuous influence of ethical leadership behavior: evidence from the field. J Bus Eth 2009; 90: 157–170.

33.

Treadway

Bentley

Williams

, et al. The skill to lead: the role of political skill in leadership dynamics. In Day

(ed) The Oxford handbook of leadership and organizations. New York: Oxford, 2014, pp.505–523.

34.

Munyon

Summers

Thompson

, et al. Political skill and work outcomes: A theoretical extension, meta‐analytic investigation, and agenda for the future. Pers Psych 2015; 68: 143–184.

35.

Blickle

Meurs

Wihler

, et al. Leader inquisitiveness, political skill, and follower attributions of leader charisma and effectiveness: test of a moderated mediation model. Int J Selection & Assess 2014; 22: 272–285.

36.

Ewen

Wihler

Blickle

, et al. Further specification of the leader political skill–leadership effectiveness relationships: Transformational and transactional leader behavior as mediators. Leader Quart 2013; 24: 516–533.

37.

Brouer

Douglas

Treadway

, et al. Leader political skill, relationship quality, and leader effectiveness: a two-study model test and constructive replication. J Leader & Org Studies 2013; 20: 185–198.

38.

Trochim

Donnelly

JP.

Research methods knowledge base. Mason, OH: Thomson Custom Publishers, 2007.

39.

Ferris

Treadway

Kolodinsky

, et al. Development and validation of the political skill inventory. J Manage 2005; 31: 126–152.

40.

Meyer

Allen

NJ.

Commitment in the workplace: theory, research and application. Thousand Oaks, CA: Sage, 1997.

41.

Eisenberger

Cummings

Armeli

, et al. Perceived organizational support, discretionary treatment, and job satisfaction. J App Psych 1997; 82: 812–820.

42.

RStudio Team. RStudio: integrated development for R. Boston, MA: RStudio Inc, 2015.

43.

Schafer

JL.

Multiple imputation: a primer.

Stat Meth Med Res 1999; 8: 3–15.

44.

Williams

Cote

Buckley

, et al. Lack of method variance in self-reported affect and perceptions at work: Reality or artifact. J App Psych 1989; 74: 462–468.

45.

Podsakoff

MacKenzie

Podsakoff

NP.

Sources of method bias in social science research and recommendations on how to control it. Ann Rev Psych 2012, 65, 539–569.

46.

Hofmann

Gavin

MB.

Centering decisions in hierarchical linear models: implications for research in organizations. J Manage 1998; 24: 623–641.

47.

Bryk

Raudenbush

SW.

Hierarchical linear models: applications and data analysis methods. Newbury Park, CA: Sage, 1992.

48.

Bliese

PD.

Multilevel modeling in R (2.6): a brief introduction to R, the multilevel package and the nlme package. Online Referencing, https://cran.r-project.org/doc/contrib/Bliese_Multilevel.pdf (2016, 3 August 2020).

49.

Jackson

DL.

Reporting results of latent growth modeling and multilevel modeling analyses: some recommendations for rehabilitation psychology.

Rehab Psych 2010; 55: 272–285.

50.

James

Demaree

Wolf

Estimating within-group interrater reliability with and without response bias. J App Psych 1984; 69: 85–98.

51.

LeBreton

Senter

JL.

Answers to 20 questions about interrater reliability and interrater agreement. Org Res Meth 2008; 11: 815–852.

52.

Klein

Kozlowski

SW.

Multilevel theory, research, and methods in organizations: foundations, extensions, and new directions. San Francisco: Jossey Bass, 2000.

53.

Kruger

Dunning

Unskilled and unaware of it: how difficulties in recognizing one’s own incompetence lead to inflated self-assessments. J Personal Soc Psychol 1999; 77: 1121–1134.

54.

Greenbaum

Mawritz

Eissa

Bottomline mentality as an antecedent of social undermining and the moderating roles of core self-evaluations and conscientiousness. J App Psych 2012; 97: 343–360.