Organisational citizenship behaviour has been defined as ‘individual behaviour that is discretionary, not directly or explicitly recognised by the formal reward system, and that in the aggregate promotes the effective functioning of the organisation’ (Organ, Podsakoff & MacKenzie 2006). OCB can be described as discretionary, contextual or extra-role performance (Asgari et al. 2008). According to Organ et al. (2006), OCB consists of five dimensions, namely: (1) altruism (e.g. providing voluntary assistance to other employees with a heavy workload), (2) courtesy (e.g. displaying behaviour that maintains good work relationships such as empathy and understanding, even when provoked), (3) civic virtue (e.g. taking part in volunteer projects that promote the good reputation of the organisation), (4) sportsmanship (e.g. not complaining about little inconveniences in the workplace) and (5) conscientiousness (e.g. working late to finish a project even though there is no direct financial reward).

As relational constructs, trust and OCB play a critical role in the social exchange context (Engelbrecht & Chamberlain 2005; Rubin, Bommer & Bachrach 2010). In a trust relationship, parties to the relationship are prepared to be vulnerable to one another’s actions because of the expectation that when either party acts in a way that influences the other, the rights and interests of the other will not be violated (Hassan & Ahmed 2011; Joseph & Winston 2005).

Trusting the leader to not violate the rights and interests of the follower is important because leaders are in a position where they make decisions and take actions, which directly affect followers. Trust is established through inferences that followers make about the nature of their relationship with the leader, based on the actions and character of the leader (Dirks & Ferrin 2002). Mutual trust develops over time as employees and their leader engage in the exchange of socio-emotional benefits through the display of concern, respect and care for each other (Zhu et al. 2013). This should in turn strengthen the emotional bond between followers and their leader.

When followers perceive their leader to be capable, benevolent and to have integrity, they will be more willing to reciprocate with desirable behaviours like OCB (Dirks & Ferrin 2002; Organ et al. 2006; Rubin et al. 2010; Zhu et al. 2013). A trusting relationship between employees and their leader should also enhance group cohesion, which can lead to a greater willingness of employees to go above and beyond their job role to help others and support their organisation on a voluntary basis (Zhu et al. 2013). In contrast, when the leader is perceived to be untrustworthy, followers will divert energy to protecting themselves, instead of focussing on their work performance (Dirks & Ferrin 2002).

Several researchers have found a positive relationship between trust in leadership and OCB (Asgari et al. 2008; Chen et al. 2008; Dirks & Ferris 2002; Engelbrecht & Chamberlain 2005; Newman et al. 2014; Rubin et al. 2010; Shahzad et al. 2013). Employees are more likely to demonstrate OCB when they trust their leader and when they experience an exchange of socio-emotional benefits in their relationship with the leader. This leads to the formulation of the first research hypothesis:

If the trust relationship between leader and follower is dependent largely on the character and actions of the leader, what character and actions should a leader display to foster the trust relationship? From the above it would seem that behaviours such as trustworthiness, integrity, benevolence, compassion and honesty are important.

These types of behaviours are typical of what the researchers have grouped as value-based leader behaviours, that is, leader behaviours that are based on the ‘golden rule’ of do unto others as you would have them do unto you (Kinnier, Kernes & Dautheribes 2000; Lennick & Kiel 2008). When examining the most prominent leadership theories, those that align most closely to these behaviours are transformational, servant, authentic and ethical leadership. Ample studies have been conducted that provide proof of the positive relationship between each of these theories and trust in the leader.

Transformational leaders were found to build trust through individualised concern for followers, which led to care for and development of the follower and strong emotional bonds between follower and leader (Asgari et al. 2008; Dirks & Ferrin 2002; Hemsworth et al. 2013; Zhu et al. 2013).

Relational transparency and self-awareness were found to be key ingredients to building trust between authentic leaders and their followers. Relational transparency leads to openness and the expression of true thoughts and feelings about issues and a forthrightness of communication, while self-awareness leads the leader to be honest about their strengths and weaknesses and how these affect others (Hassan & Ahmed 2011; Joseph & Winston 2005; Walumbwa et al. 2008).

Servant leaders build trust through their insight, empathy, concern for the welfare of their followers, as well as the ethical example they set for followers (Joseph & Winston 2005; Miao et al. 2014; Sendjaya & Pekerti 2010).

Ethical leaders seem to enhance mutual trust by behaviour that is typically marked by the care, equity, fairness and respect they show towards their followers (Engelbrecht et al. 2017; Hassan & Ahmed 2011; Hemsworth et al. 2013; Keck et al. 2018; Miao et al. 2014; Newman et al. 2014; Sendjaya & Pekerti 2010; Zhu et al. 2013).

In analysing the behaviours of value-based leaders, the researchers found that there was a great deal of overlap in the behaviours inherent to the four value-based leadership theories mentioned above. While each placed an emphasis on specific moral values, they all essentially measure moral leadership (Xu, Caldwell & Anderson 2016). To measure the underlying behaviours of transformational, servant, authentic and ethical leadership, Hendrikz and Engelbrecht (2019) developed principled leadership as a holistic, value-based leadership construct, defining principled leadership as follows:

Trust is an essential component of the social exchange theory (Sendjaya & Pekerti 2010). Social exchange is the voluntary actions of employees who are motivated by an expectation of support and benefits from their leader (Blau 1964). Principled leaders may thus influence follower trust in leaders through social exchange. Furthermore, principled leaders may gain follower trust through role modelling, a phenomenon that can be understood using social learning theory (Brown & Treviño 2006). According to this theory, individuals learn by observing the behaviour of role models chosen, based on their attractiveness and credibility. Principled leaders are particularly likely to become role models, because their behaviours, specifically integrity, concern for others and strong ethics, may enhance their attractiveness and trustworthiness, resulting in greater credibility (Brown & Treviño 2006; Liden et al. 2014; Walumbwa et al. 2008). Such leadership behaviours provide a psychologically secure environment in which employees can develop high levels of trust in the leader (Lu 2014). Consequently, the following hypothesis can be postulated:

The discussion to this point has highlighted several behaviours that would be expected from principled leaders. However, not all leaders behave in a principled manner. The Enron debacle of 2001 and more recently the scandals involving Volkswagen, FIFA, Steinhoff, Bosasa and the alleged state capture of the South African government indicate that immoral leader behaviour is alive and well. A question that begs answering in this regard is why some people are more inclined to act morally than others?

To explore this question, the researchers investigated the concept of moral intelligence, which has been defined as follows: ‘Moral intelligence is the mental capacity to determine how universal human principles should be applied to our values, goals and actions’ (Lennick & Kiel 2008:7). The understanding of moral intelligence requires a brief discussion of research relating to the origins of morality.

In Plato’s opinion, the development of the ability to reason brought about morality, which in turn controlled the wayward emotions of humankind (Narvaez 2010). Plato’s pupil, Aristotle, alleged that humans possess within themselves the emotional and rational abilities necessary to achieve moral excellence, which can be developed through proper guidance by society and family (Dow 1998).

The theologian Augustine argued that, essentially, man desires happiness and peace. However, when his will is directed towards selfish desires instead of towards God, he will act immorally (Wogaman 2011). To counteract this inherent selfishness, morality must be developed through training of good habits (Narvaez 2010).

Similarly, Confucius in China, and Buddha in India, held the view that people within themselves want to be just, self-controlled, and temperate as this is critical for their well-being and happiness (Hass 1998).

In addition to the philosophical viewpoints stated above, empirical science has begun to prove that an innate propensity to be moral is part of the physical make-up of all human beings (Lennick & Kiel 2008; Narvaez 2010). As suggested by the ancient philosophers and modern psychology, nurturing, training and development of this inborn morality is, however, necessary for humans to function effectively in terms of their morality (Lennick & Kiel 2008).

The most defining benefit of moral intelligence is that it is viewed as the ‘central intelligence’, which provides an anchor for the cognitive, emotional and social intelligences, ensuring that the right thing is done, rather than acting in possible destructive ways (Beheshtifar et al. 2011; Lennick & Kiel 2008).

Liden et al. (2014) analysed moral maturity and moral conation, which are related concepts to moral intelligence. Leaders who possess moral maturity are likely to behave in a way that is consistent with their personal ethical norms (i.e., their moral identity). Moral conation is the capacity to believe one is morally responsible and act in a moral way, even in the face of obstacles. Liden et al. (2014) view moral maturity and moral conation as antecedents of the ethical behaviours of servant leaders. Furthermore, leaders’ cognitive moral development has been found to be positively linked to ethical decision-making (Ashkanasy, Windsor & Treviño 2006).

In today’s world of work, leaders will inevitably be faced with ethical decisions and the more adept they are at handling these, the more successful they will be (Beheshtifar et al. 2011; Liden et al. 2014). Morally intelligent leaders thus have an edge on those who have an underdeveloped inner moral compass (Beheshtifar et al. 2011; Engelbrecht et al. 2018). This led to the postulation of the following hypothesis:

Based on the analysis of the above-mentioned literature a conceptual model of the relationships among moral intelligence, principled leadership, trust in the leader and OCB was developed (see Figure 1).

Purposive, non-probability sampling was used to select participants from various companies in South Africa. The participating organisations included software development and consulting services (80%), retail (6%), wine making (5%), construction (5%) and public service (4%). The population can be defined as middle to top management in organisations. The sample of 300 respondents consisted of employees who reported to managers in such positions. Of the respondents, 58% were working in non-managerial positions, 31% in lower to mid-level management and 11% in senior to top management level positions. The average age of participants was 33 years.

Electronic or paper-based questionnaires were sent either to a contact person at the organisation for further distribution, or directly to participants. In the covering letter, the purpose of the research was clearly explained to participants; confidentiality and anonymity were guaranteed and feedback would be provided in aggregate format only. Furthermore, electronic data was directly sent to the statistical centre of the university.

The researchers utilised the following scales to measure the four components of the structural model:

The Research Ethics Committee of Stellenbosch University granted ethical clearance for the study: SU-HSD-000617. Furthermore, informed consent was sought from participating employers and respondents.

Table 1 presents the means, standard deviations, reliability coefficients (α values) and Pearson correlation coefficients for all four measures included in the study.

Reliability analysis was also conducted on each of the dimensions of the scales comprising the structural model. All the Cronbach’s alpha values were above the acceptable cut-off value of 0.70 (Nunnally 1978) except for the subscale Civic Virtue of the OCBS, for which a value of 0.66 was obtained. As per Nunnally (1978) this deems the dimension to have low but acceptable reliability. In addition, the corrected item-total correlations were satisfactory (> 0.30) for all subscales (Pallant 2013) (see Table 2).

The fit of the measurement models was investigated by performing confirmatory factor analysis (CFA) to determine whether the theorised models were supported by the data collected.

All fit indices of the overall measurement model underlying the structural model indicated good fit. In addition, the p-value for close fit supported the close fit hypothesis (see Table 3).

The structural model also showed good fit. This was supported by all the fit indices. Support was also found for the close fit hypothesis (see Table 3).

Hypothesis 1 postulated a positive relationship between trust in the leader (ŋ₂) and OCB (ŋ₃). This was confirmed by this study (β = 0.50; p < 0.05) (see Table 4). This corroborates the results of prior studies in which a positive, significant relationship between trust in leadership and OCB was found (Asgari et al. 2008; Chen et al. 2008; Engelbrecht & Chamberlain 2005; Newman et al. 2014; Rubin et al. 2010; Shahzad et al. 2013). Thus, if employees believe their leaders to be highly trustworthy, they will in turn display greater levels of OCB.

The study supported Hypothesis 2, which proposed that a positive, significant relationship exists between principled leadership (ŋ₁) and trust in the leader (ŋ₂) (β = 0.92; p < 0.05) (see Table 4). This confirms prior findings of a positive relationship between value-based or principled leadership and trust in leaders (Engelbrecht et al. 2017; Hassan & Ahmed 2011; Hemsworth et al. 2013; Keck et al. 2018; Miao et al. 2014; Newman et al. 2014; Sendjaya & Pekerti 2010; Zhu et al. 2013). Principled leaders will therefore create high levels of trust between subordinates and their leaders.

Support was found for Hypothesis 3, which suggested that a positive, significant relationship exists between moral intelligence (ξ₁) and principled leadership (ŋ₁) (β = 0.98; p < 0.05) (see Table 4). This supports the positive relationship established between integrity, as well as cognitive moral development and value-based or principled leadership (Ashkanasy et al. 2006; Beheshtifar et al. 2011; Engelbrecht et al. 2018; Lennick & Kiel 2008; Liden et al. 2014). Consequently, leaders with a high level of moral intelligence will exhibit strong principled leadership.

Based on the results of the direct relationships, two ad hoc analyses were conducted to test mediator effects formally. Firstly, the significance of the mediation effect of trust in the leader in the principled leadership–OCB relationship was tested (as both Hypothesis 1, the effect of trust in the leader on OCB, and Hypothesis 2, the effect of principled leadership on trust in the leader, were supported). The results of the Sobel test (Preacher & Hayes 2004) revealed that trust in the leader significantly mediated the relationship between principled leadership and OCB (z = 2.88; p < 0.01). Secondly, as both Hypothesis 2 and Hypothesis 3 (the effect of moral intelligence on principled leadership) were supported, principled leadership as a mediator in the moral intelligence–trust in the leader relationship was formally tested. The results (z = 5.21; p < 0.01) confirmed this mediation effect.

Analysis of the modification indices did not show any other significant, positive paths that could be considered within the structural model, other than those postulated as part of the study.