Literature shows that the concept of servant leadership is explained differently by different authors and organisations. Servant leadership is a leadership philosophy focused on follower development and represents a plausible leadership paradigm for the 21st-century organisation (DeHaven 2007:114). The very concept of servant leadership is based on the values of humility and respect for others (Oforchukwu 2013:53). The emerging approach to leadership and service is called ‘servant leadership’ (Greenleaf 1998:2). Servant leaders are those who make a deliberate choice to serve others and to put other people’s needs first (French et al. 2011:451). The concept of servant leadership emphasises increased service to others, a holistic approach to work, a sense of community and shared decision making (Bodenhamer & Barrows 1994:1202). Servant leadership is a question of inner motivation, of a deeply felt mission; and everyone can become a servant leader regardless of whether he or she already is a leader because the combination of opposites lies within everyone’s reach (Trompenaars & Voerman 2009:13). Servant leadership includes deliberate actions that strive to enhance the best of others. Servant leadership is a humane effort that is people-focused instead of profit-driven. The central theme of servant leadership is working to improve others (Baron 2010:1–87).

A servant leader lives, loves and leads by conscience – the inward moral sense of what is right and what is wrong (Sipe & Frik 2009:17). The servant leader leads people in a manner that helps them grow and increase their capacity to contribute (useful products and services) and in the process gain the satisfaction of making a greater contribution to the success of the organisation (Neuschel 2005:12). Servant leaders feel that their role is to help people achieve their goals; they constantly try to find out what their people need to perform well and live according to the vision (Blanchard 2010:262). The servant leader ensures a constant connection between experience and learning and tests these according to the next planned steps (Trompenaars & Voerman 2009:54). The natural instinct of a servant leader is to provide information freely and readily to workers which helps create a work climate based on inclusivity and partnership, which are key facets of a spiritually rich workplace (Giacalone & Jurkiewicz 2010:133).

Developed upon Greenleaf’s original findings, a list of 10 characteristics of a servant leader were identified as being of critical importance and central to the development of servant leaders (Spears & Lawrence 2002:5). The authors listed and explained the characteristics as follows:

All leaders, in particular servant leaders, have an unwavering commitment to their values (Baron 2010:4). Values are the core elements of servant leadership; the values of servant leaders not only yield observable attributes, but they also affect the leader’s organisation (Oforchukwu 2013:53). Leader values significantly affect followers and ultimately influence organisational performance (Oforchukwu 2013:53). Servant leaders facilitate the growth, goals and development of others to liberate their best qualities in pursuing the organisation’s mission (Daft & Marcic 2011:395). In organisations these leaders’ top priority is to service employees, customers, shareholders and the general public (Daft & Marcic 2011:176). Servant leadership helps create a high-performing organisation (Blanchard 2010:282).

Furthermore, servant leaders want to make a difference in the lives of their people and, in the process, create an impact on the organisation (Blanchard 2010:262). Blanchard (2010:262) further said that ‘Servant leadership can occur in any organisation’. The essential quality of a servant leadership culture will be evident by the way employees embrace the vision, live out the core values and relearn key virtues. The servant-led organisation is completed with the knowledge that the financial bottom line is not the essential bottom line (Baron 2010:134):

Servant leadership succeeds or fails based on the personal values of the people who employ it (Oforchukwu 2013:53). The most persistent barrier to being a servant leader is a heart motivated by self-interest that looks at the world as a ‘give a little, take a lot’ proposition (Blanchard 2010:271).

South African researchers have identified similar important servant leadership characteristics which are critical to the field of servant leadership. These characteristics include authenticity, humility, integrity, listening, compassion, accountability, courage and altruism (Coetzer, Bussin & Geldenhuys 2017:19). Servant leadership has also been identified as an important role player in terms of creating positive employee attitudes and improved organisational performance (Coetzer et al. 2017:19). Furthermore, South African research has indicated that servant leadership is a critical type of leadership required because of our unique political landscape (Kgatle 2018:8). According to Kgatle (2018:8), servant leadership can provide a solution to improve governance, trustworthiness and accountability within the South African environment.

With South Africa’s unique landscape, it would also be critical to consider how different demographical groups view servant leadership. A recent South African study found that Africans view servant leadership differently when compared to their White counterparts. African’s reported higher levels of aggregate servant leadership observations, including different opinions about interpersonal support, building community, altruism and moral integrity when compared to their White colleagues (McCallaghan, Jackson & Heyns 2019:1029). With regards to gender differences, McCallaghan et al. (2019:1029) found that males reported higher observations for the interpersonal support and egalitarianism servant leadership variables. Differences are not only reported for gender and race groups but also include differences amongst age groups and generational cohorts.

Researchers have found evidence that we may experience changes in how individuals view leadership, especially between intergenerational groups (Haeger & Lingham 2013:299). However, previous South African examinations could not find any evidence that servant leadership and employee age would be related (Dannhauser & Boshoff 2006:11).

Taking into consideration the preceding paragraphs, the study considered the objectives as discussed further.

The primary objective of this research is to gain an understanding of servant leadership in a South African power utility.

The specific secondary objectives of this research are:

This study followed a quantitative approach. A cross-sectional survey design was used to reach the stated objective. According to Spector (2019:130), cross-sectional studies are especially valuable to determine a specific observation at a single point in time. The study further made use of a self-administered questionnaire. Permission was obtained from the South African power utility. The agreement stipulated that the identity of the business unit in the power utility may not be disclosed and should be referenced as a power utility in the scope. The questionnaires were distributed to the participating respondents. Two methods were used to distribute the questionnaires. A hard copy questionnaire and an electronic questionnaire were administered. The rationale behind using both a hard copy and electronic questionnaire was to avoid exclusion should an employee not have access to a computer. The type of method chosen by the respondent was based on the convenience of the respondent. All completed questionnaires were then sent to the researcher via e-mail, facsimile or in hard copy format. Ethical considerations were made by attaching a covering letter to the questionnaire. The cover letter explained the objectives of the research and confirmed the anonymous, confidentiality and voluntary nature of the study.

The target population or study sample comprised all of the employees at a specific business unit of the power utility under investigation. A total of 771 respondents from a convenience sample completed questionnaires. Respondents were employed at a specific business unit of a power utility situated in the Free State province. Table 1 represents the demographical characteristics of the respondents who participated in the study.

A closer of inspection of Table 1 reveals the fact that the majority of the participating respondents fall in the age group category of 25–34 years (45.26%), with the second largest group comprising people aged 55+ years (22.82). The majority of the respondents were male (75.4%), followed by the females (24.25%), while the majority of the employees (34.89%) were Sesotho speaking, followed by Afrikaans (19.33%) and English (14.66%). The majority of the workforce (55.6%) has completed their tertiary education with only 25.81% of the respondents having qualifications below grade 12.

The measuring instrument comprised two sections. The first section collected data on the demographical information, and the second section collected data on servant leadership.

A 30-item instrument on servant leadership, which includes eight dimensions, namely Standing Back, Forgiveness, Courage, Empowerment, Accountability, Authenticity, Humility and Stewardship (Van Dierendonck & Nuijten 2011:249–267), was applied to collect opinions on servant leadership from a follower perspective. The specific instrument has been validated by Van Dierendonck and Nuijten (2011:249–267), who explained that the correlations between the different items on the instrument were good. According to Van Dierendonck and Nuijten (2011:249–267), the results demonstrate that the Servant Leadership Survey (SLS) has convergent validity with other leadership measures and also adds unique elements to the leadership field.

The SLS uses a 4-point Likert scale ranging from ‘disagree strongly’ (1), ‘disagree’ (2), ‘agree’ (3) to ‘strongly agree’ (4), regarding the inclusion of the eight dimensions. A typical statement from the SLS reads as follows: ‘My manager holds me and my colleagues responsible for the way we handle a job’.

The data were analysed by using the Statistical Package for Social Sciences (SPSS) version 26 (IBM Corp. 2020). Internal consistency and reliability of the questionnaire was assessed by calculating Cronbach alpha coefficients and inter-item correlations. Internal consistency as measured with Cronbach alpha was considered sufficient at a moderate level (α = 0.60) (Taber 2018:1278). According to Pallant (2007:95), measuring scales with low scale items may present lower Cronbach alpha scores, and therefore inter-item correlation analysis should also be conducted. Inter-item correlations were considered sufficient at r = 0.30 (Field 2013:821). Once reliability was determined, the variables that met the minimum thresholds were subjected to an exploratory factor analysis (EFA). According to Pallant (2007:190), sample adequacy would be acceptable when the Kaiser–Olkin Measure of Sampling Adequacy (KMO) is above 0.60 and the Bartlett’s test of sphericity is significant (p < 0.05). The principal axis factoring was used as the extraction method with the Oblimin Kaiser Normalisation as rotation. Criteria for factor selection was set at 1.00 for eigenvalues (Pallant 2007:190). Factor loadings were considered adequate at 0.30 (Pallant 2007:196). Once the internal structures were confirmed, we conducted an analysis for convergent validity as part of construct validity. Composite reliability (CR) may be examined for this purpose (Hair et al. 2010:619). Critical values for CR should be above 0.70 (Hair et al. 2010:619).

Frequency distributions, mean values and standard deviations were calculated for every leadership variable in order to determine central tendencies with regards to servant leadership, including servant leadership features.

Analysis of variance (ANOVA) was calculated in order to determine statistical differences between group. Post hoc tests were conducted where more than two groups were present. The Welch’s test for significance was also considered as this test is robust and does not assume equal variances. Additionally, Cohen’s theory was used to interpret the effect-size (d-values): small effect or practical non-significant differences (d-value ≈ 0.2), medium effect or practical visible differences (d-value ≈ 0.5) and large effect or practical significant differences (d-value ≈ 0.8) (Ellis & Steyn 2003:51–53).

Ethical clearance was obtained from the NWU Business School.

Table 3 contains the descriptive and CR results from the variables under investigation.

Forgiveness recorded the highest mean score of m = 3.26 and SD = 0.53, which is within the scale’s ‘agree’ range. This result implies that respondents are of the opinion that their managers or leaders demonstrate forgiveness. The variable Standing back recorded the lowest mean value of m = 1.31; SD = 0.54 which is within the scales disagree range. This would indicate that participants disagreed that their manager or leader gives credit to others, is not chasing recognition and appears to enjoy his colleagues’ success more than his own. The scores for Empowerment (m = 1.77; SD = 0.45), Authenticity (m = 1.52; SD = 0.52) and Humility (m = 1.50; SD = 0.51) are also within the measurement scales ‘disagree’ range. In reality this result indicates that respondents are of the opinion their managers or leaders are not demonstrating empowerment, authenticity or humility. Both the Courage (m = 3.07; SD = 0.45) and Stewardship (m = 2.51; SD = 0.55) variables were in the measurements ‘agree’ range. The specific result indicated respondents were of the opinion their managers or leaders are demonstrating courage and stewardship. The CR scores were also above the 0.70 threshold, which indicates adequate composite reliability.

The final section examined the view of servant leadership across demographical groups. The statistical analysis in terms of group comparisons was restricted to only the gender and age groups. Table 4 contains the results of comparisons between the male and female respondents. The Lavene’s test for homogeneity was not significant for any of the variables, and therefore ANOVA was conducted for the gender groups assuming equal variances.

The results from Table 4 indicate that only Stewardship reveals a statistical difference when comparing means between the male and female groups (p < 0.01). The male group recorded a higher observation for stewardship (m = 2.56; SD = 0.51) when compared to the female group (m = 2.38; SD = 0.27). The result was considered as a medium practically visible difference. The remaining gender comparisons did not yield any significant difference when comparing mean scores between the male and female respondents.

The final group comparison was conducted for the age groups. The test for homogeneity was significant for age group ANOVA’s. Subsequently, the post hoc analysis was conducted by making use of the Tamhane multiple comparison test as this method does not assume equal variances.

A closer inspection of the Table 5 indicates that only Stewardship did not yield a significant result (p = 0.09) when comparing servant leadership observations between age groups. The remaining variables, including the aggregate servant leadership variable, revealed a statistical significance when comparing means between age groups (p < 0.01). Consequently, post hoc tests were conducted on Empowerment, Standing back, Forgiveness, Courage, Authenticity, Humility and Servant leadership.

The post hoc test for Empowerment revealed a statistical difference when comparing the mean scores for Age group 1 and Age group 3 (p < 0.01; d = 0.57; Medium effect), Group 2 and 3 (p < 0.01; d = 0.40; Medium effect). In both cases, Age group 3 (35–54) recorded higher observations for Empowerment. In reality, respondents who are between 35 and 54 years old reported higher opinions of Empowerment when compared to employees who are 34 years old and younger. Similarly, a significant result was observed when comparing Age group 1 and Age group 4 (p < 0.01; d = 0.42; Medium effect) with Age group 4 (55+) reporting higher opinions of Empowerment.

For Standing back there was visible difference between Age group 1 and Age group 3 (p < 0.01; d = 0.65; Medium effect) and between Age group 1 and Age group 4 (p < 0.01; d = 0.40; Medium effect). In both cases, Age group 3 (35–54) and Age group 4 (55+) recorded higher observations for Standing back when compared to Age group 1 (18–24). In reality this result meant that older respondents reported higher observations of Standing back when rating their immediate leader or manager. There was also a difference between Age group 2 and Age group 3 for Standing back (p < 0.01; d = 0.40; Medium effect), with Age group 3 (35–54) reporting higher observations of Standing back.

With regards to Forgiveness, there was difference between Age group 1 and Age group 3 (p < 0.01; d = 0.50; Medium effect) and between Age group 1 and Age group 4 (p < 0.01; d = 0.43; Medium effect). In both cases, Age group 1 (18–24) reported higher observations of Forgiveness. Similarly, there were visible differences when comparing Age group 2 and Age group 3 (p < 0.01; d = 0.37; Medium effect), and between Age group 2 and Age group 4 (p = 0.04; d = 0.28; Small effect). Age group 2 (25–34) reported higher observations for Forgiveness in both situations. In reality, employees 34 years old and younger would be inclined to report higher recordings of Forgiveness when rating their supervisor or manager when compared to employees older than 35 years old.

In terms of Courage, there were reported differences between Age group 1 and Age group 3 (p < 0.01; d = 0.63; Medium effect) and between Age group 2 and Age group 3 (p < 0.01; d = 0.46; Medium effect). In both comparisons Age group 3 (35–54), reported lower observations of courage when compared to the other age groups. This result suggests that employees between 35 and 54 years old report lower recordings of Courage for their immediate supervisor or manager when compared to employees 34 years and younger. There was further a difference between Age group 3 and Age group 4 (p < 0.01; d = 0.40; Medium effect), with Age group 4 (55+) recording higher opinions of Courage when rating their immediate supervisor or manager.

For Authenticity, there was differences between Age group 1 and Age group 2 (p = 0.03; d = 0.30; Medium effect) and between Age group 2 and Age group 4 (p < 0.01; d = 0.43; Medium effect). In both comparisons, Age group 2 (24–34) reported higher annotations of Authenticity when rating their respective leader or manager. Furthermore, there was a difference between Age group 3 and Age group 4 (p = 0.04; d = 0.34; Medium effect) with Age group 3 (35–54) indicating higher observations for Authenticity when compared to employees older than 55 years old.

The comparisons for the Humility variable also revealed differences when comparing observations between age groups. There was a difference between Age group 1 and Age group 3 (p < 0.01; d = 0.50; Medium effect) and between Age group 2 and Age group 3 (p < 0.01; d = 0.48; Medium effect) with Age group 3 (35–54) indicating higher opinions of Humility in respect of their immediate leader or manager in both situations. Furthermore, there was a difference for Humility between Age group 2 and Age group 4 (p < 0.01; d = 0.21; Small effect), with Age group 4 (55+) reporting a slightly higher opinion for Humility.

In terms of the aggregate Servant Leadership recording, we found differences between Age group 1 and Age group 3 (p < 0.01; d = 0.48; Medium effect), with Age group 3 (35–54) reporting higher observations of Servant Leadership when compared to Age group 1 (18–24). In reality, this result would mean that employees aged between 35 and 54 years old were inclined to record higher observations of Servant Leadership when rating their immediate supervisor or manager when compared to employees aged between 18 and 24 years old. There were also differences between Age group 2 and Age group 3 (p < 0.01; d = 0.35; Medium effect) with Age group 3 (35–54) also reporting higher observations of Servant Leadership when compared to employees aged between 25 and 34 years old.