Original Research

Is the leadership performance of public service executive managers related to their emotional intelligence?

Shanil J. Haricharan

Received: 16 Sept. 2021; Accepted: 02 Mar. 2022; Published: 19 Apr. 2022

Copyright: © 2022. The Author(s). Licensee: AOSIS.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Abstract

Orientation: Growing evidence indicates a positive relationship between emotional intelligence (EI) and leadership performance. However, in non-Western public service contexts, scholarship on the nature of this bivariate relationship trails behind.

Research purpose: Using the behavioural EI model, this study examined the relationships between EI competencies and leadership performance of executive managers in the South African public service.

Motivation for the study: A significant bivariate relationship using the behavioural EI model implies empirical significance and practical implications for policy and leadership development in the public service.

Research approach/design and method: The multi-rater Emotional and Social Competence Inventory (ESCI) measured EI competencies of 35 executive managers rated by 230 respondents. Multi-source nominations from 371 respondents measured leadership performance. Five study hypotheses were tested using Spearman’s rank correlation coefficients and analysis of variance.

Main findings: The results indicated significant positive correlations between leadership performance and all four EI clusters of competencies: self-awareness, self-management, social awareness and relationship management. Of the 12 EI competencies, adaptability, inspirational leadership, emotional self-awareness and positive outlook displayed the strongest correlations. Also, a significant negative relationship between the managers’ competency gap and (self-other agreement) their performance was observed.

Practical/managerial implications: The results have implications for management and leadership development and recruitment in the public service.

Contribution/value-add: Using the behavioural method, this quantitative study validated the positive relationship between EI and leadership performance in the South African public service

Keywords: emotional intelligence; emotional and social competence inventory; competencies; executive managers; public service; leadership performance; self-other agreement; South Africa.

Introduction

A growing body of literature on emotional intelligence (EI) suggests that a person’s ability to perceive, identify and manage emotion is essential in making good decisions, taking optimal action to solve problems and coping with change (Ashkanasy & Daus, 2005; Caruso & Salovey, 2004; Cherniss, Extein, Goleman, & Weissberg, 2006; Goleman, 1998, 2004; Goleman, Boyatzis, & McKee, 2002). Notably, EI is postulated as an important predictor of success in the workplace and is widely considered a fundamental element of effective leadership (Boyatzis, 2009; Dulewicz & Higgs, 2003; George, 2000; Goleman et al., 2002; Goleman & Boyatzis, 2017; Mohammed, 2018). In addition, several studies show a positive relationship between EI and transformational leadership (Barling, Slater, & Kelloway, 2000; Brown & Moshavi, 2005; Görgens-Ekermans & Roux, 2021; Mills, 2009; Orazi, Turrini, & Valotti, 2013; Sosik & Megerian, 1999), charismatic leadership (Shamir, House, & Arthur, 1993) and authentic leadership (Miao, Humphrey & Qian, 2018).

Emotional intelligence is regarded as an emerging theory at an early stage in development (Cherniss et al., 2006) and the study of emotions is receiving greater consideration in leadership models (Dinh et al., 2014). Over the past two decades, despite the growth in Western and private sector scholarship on EI and leadership, there is a deficit of studies on the bivariate relationship in public sector and non-Western literature, with scholars calling for further research into the EI field (Boyatzis, 2009; Kotzé & Venter, 2011; Newman, Guy, & Mastracci, 2009). In addition, House and Aditya (1997) draw attention to the assumptions and cultural disposition of leadership literature, which they claim is predominantly Western-centric, focussed on developed nations, distinctly American in character, positivistic and individualist rather than collectivist.

Cognisant of cultural relativism and to address the cultural biases and deficits in the scholarship, there is a need to test these claims in non-Western public service contexts. As a result of probable conceptual distinctions across cultures, scholars have proposed further research to test the universality of EI constructs and metrics, and the possibility of contextual differences, in specific emotional and behavioural expressions and regulation (Boyatzis, 2009; Emmerling & Boyatzis, 2012) and relationships with leadership styles and effectiveness (Miao et al., 2018).

Congruently, public sector scholars Denhardt and Denhardt (2007) and Kellis and Ran (2013) call for greater attention to the significant gap in the literature to embrace developing leadership models and research appropriate for the public service. Many scholars call for greater research endeavours that redefine the structures, tools, processes, competencies and functions of leadership in public organisations (Kellis & Ran, 2013; Van Wart, 2011; Wright, 2011).

Research purpose and objectives

The purpose of this quantitative research study was to test claims on the nature of the relationships between EI competencies and executive managers’ leadership performance in the South African public service, framed within the behavioural EI model (Boyatzis, 2009; Boyatzis, Rochford, & Cavanagh, 2017; Goleman et al., 2002) and measured with the Emotional and Social Competence Inventory (ESCI).

The objectives of the study were threefold:

1. Firstly, to determine the nature of the relationship between the four EI clusters and 12 individual EI competencies with leadership performance in the South African public service context.

2. Secondly, to determine those EI competencies that differentiate high performing executive managers.

3. Thirdly, to establish the relationships between the executive managers’ competency gap or self-other agreement (SOA) and their leadership performance.

Literature review

Emotional Intelligence: Concepts, models and metrics

The concept of EI builds on a long history of theory and research on personality, social and positive psychology and particularly on the importance of non-cognitive aspects of intelligence, drawing from antecedents since the early 1900s (e.g. the work of Robert Thorndike, David Wechsler, Howard Gardner, David McClelland and Robert Sternberg).

Generally, EI research is classified into three streams, identified by Ashkanasy and Daus (2005) as the cognitive-based ability models using objective test items (Mayer-Salovey-Caruso Emotional Intelligence Test, MSCEIT), self-report or peer-report measures based on the four-branch model of EI (e.g. Wong-Law Emotional Intelligence Scale, WLEIS; Swinburne University Emotional Intelligence Test, SUEIT) and ‘mixed models’ of emotional competencies (Emotional Intelligence Quotient Inventory, EQ-I; Emotional Competence Inventory, ECI; ESCI). In their ability-based EI model, Mayer, Salovey and Caruso (2000, p. 396) defined EI as the ‘ability to perceive and express emotion, assimilate emotion in thought, understand and reason with emotion and regulate emotion in the self and others’. Conversely, framing EI as behavioural competencies instead of abilities (‘mixed model’, ESCI), Goleman (1998) defines EI as the capacity to understand one’s emotions and to manage them effectively and to understand and effectively manage the emotions of others. Later ‘mixed models’ were further classified into trait EI and behavioural EI (Boyatzis et al., 2017).

However, despite these differences in metrics, there are noteworthy similarities between the three EI conceptual streams. All three include two common domains of EI: firstly, awareness of and managing one’s own emotions and secondly, awareness of and managing others’ emotions. Cherniss (2010) maintains that across the major EI models, three principles endure: firstly, emotions play a vital role in life; secondly, people differ in their ability to perceive, understand, use and manage emotions; and thirdly, these differences affect how an individual adapts in a variety of contexts, including the workplace.

Distinctions between cognitive ability, personality traits and emotional intelligence

The pursuit for characteristics to adequately explain workplace success has been continuous within social psychology and the organisational field (Ramo, Saris, & Boyatzis, 2009). Research indicates that cognitive intelligence, personality and EI influence work-related outcomes (O’Boyle, Humphrey, Pollack, Hawver, & Story, 2011; Walter, Cole, & Humphrey, 2011). Cherniss et al. (2006, p. 240) claimed that EI characterises a set of abilities that are distinct from either intelligence quotient (IQ), a cognitive ability measure or the ‘Big Five’ personality traits in the five-factor model. O’Boyle et al. (2011) confirmed that all three constructs (cognitive ability, personality traits and EI) are distinctly different. Several research findings consistently demonstrate that traditional cognitive intelligence tests such as IQ by itself are neither good nor the only predictor of job performance and leadership effectiveness (Judge, Colbert, & Ilies, 2004; McClelland, 1998; Miao et al., 2018; O’Boyle et al., 2011). Specifically, McClelland (1998, p. 331) argued that traditional tests of cognitive intelligence, such as IQ, were ‘failing to account for successful performance, especially in high-level executive positions’.

Furthermore, several meta-analytical studies reveal that EI predicts performance and displays incremental validity over and above the five-factor model of personality traits and cognitive ability (e.g. Van Rooy & Viswesvaran, 2004; Van Rooy, Viswesvaran, & Pluta, 2005; O’Boyle et al., 2011; Walter et al., 2011). In their meta-analysis, Joseph and Newman (2010) showed that all three streams of EI measures have greater incremental validity compared with the five-factor model of personality traits and cognitive ability.

The behavioural approach to emotional intelligence (mixed emotional intelligence stream)

The drive to optimise an individual’s performance has led to the quest of understanding human talent. A major part of this search has focused on leadership talent, which is considered as a key determinant of organisational success. Several studies and meta-analyses indicate that EI competencies based on the behavioural approach to EI (‘mixed model’) are valid constructs in the study of human talent and performance, as they account for a significant variance in predicting or understanding job performance (Boyatzis, 2009, 2018; Dulewicz & Higgs, 2003; Goleman et al., 2002; O’Boyle et al., 2011; Ryan, Emmerling, & Spencer, 2009; Sharma, 2012) and leadership effectiveness (e.g. Görgens-Ekermans, 2009; Piel, 2008). As such, EI competencies enable the study of a leader’s personal characteristics related to behaviour.

Boyatzis (1982, p. 12, 23) defined competencies as ‘underlying characteristics of an individual, which are causally related to effective and superior performance in a job’. The EI behavioural model views EI as a set of competencies that drive leadership and job performance and comprises four clusters of competencies, namely, self-awareness: recognising and understanding own emotions; social awareness: recognising and understanding the emotions of others; self-management: effectively managing own emotions; and relationship management: applying emotional understanding in dealings with others (Goleman et al., 2002; Boyatzis, 2018). The four clusters comprise a total of 12 EI competencies, such as emotional self-awareness, empathy, emotional self-control, adaptability and conflict management. Citing affective neuroscience research, Goleman and Boyatzis (2008) submitted that each of the EI clusters derives from different neurological mechanisms that are distinct from the cognitive domains of ability (see also Rochford, Jack, Boyatzis, & French, 2017).

In the behavioural EI model, exemplary leaders display strengths in the 12 EI competencies, as these competencies are antecedents to visionary, democratic, affiliative and coaching leadership styles, and therefore these leaders are regarded as resonant or transformational leaders (Boyatzis & McKee, 2005). These effective leaders promote a collaborative, inclusive, relational, empathetic, compassionate, caring and supportive institutional culture (Boyatzis, 2018; Goleman et al., 2002). As anticipated, these leadership approaches were regarded as extremely valuable in responding to the Coronavirus disease-2019 (COVID-19) pandemic, including being ‘less autocratic, arrogant, blustering, boastful, impatient, impulsive, indecisive … swaggering and top-down’ (Huang, 2020, p. 11). Likewise, Garikipati and Kambhampati’s (2020) concluded in their study on head of states during the COVID-19 pandemic that generally women leaders were more effective in their responses to the pandemic as they draw on their EI competence, such as empathy and tended to adopt a more democratic and participative leadership styles and a less autocratic or directive style compared with their male counterparts.

Leadership in the public service context

Upholding democratic principles, public service values and effective implementation of public policy continue to defy many governments, particularly during recurrent crises (Kellis & Ran, 2013), as observed during the current COVID-19 pandemic (Garikipati & Kambhampati, 2020; Huang, 2020). Beinecke and Spencer (2009) declared that there is a crisis in public leadership with both political and administrative public leaders viewed with suspicion and cynicism. Globally, these negative perceptions have increased over the past few decades as evidenced in waves of public discontent and resistance as trust and confidence in the public sphere has declined remarkably across continents (Holzer & Illiash, 2009).

For over a century, strong hierarchical traditions have characterised the public service, dominated by the ‘machine’ image of organisations (Robledo, 2009). This has led to the erosion of personal and social facets of organisational life. Social factors were only introduced into organisational behaviour models after the Hawthorne study in the 1930s. The command-and-control approach to public service management still prevails in the 21st century public service institutions. Consistently, scholarship for most of the 20th century was dominated by cognitive, rational, transactional and task-focused leadership approaches (Benington & Hartley, 2011; Robledo, 2009). This approach was reinforced in the last quarter of the 20th century by the dominance of new public management (NPM) thinking and practice drawing on private sector models, despite the distinctiveness of the public service context as a defining issue in the public governance literature (Chipkin & Meny-Gibert, 2012; Mau, 2009).

The metaphor for the new millennium public service has implications on how public servants are managed and led, as earlier theories focused mainly on controlling and manipulating employees and their work environment with the objective of maximising efficiency and productivity (Denhardt & Denhardt, 2007). In the turbulent 21st century public service context, alternative theoretical approaches and practice in governance, and leadership are necessary, with several calls for new models for governance and types of political, administrative and citizen leadership (Mau, 2009; Morse & Buss, 2008; Van Wart, 2011).

Denhardt and Denhardt (2006, p. 83) stressed that ‘human emotions and qualities such as empathy and intuition’ have a significant role to play in public leadership. Likewise, Vigoda-Gadot and Meisler (2010, p. 72) contest the hegemony of rationality in public administration, suggesting that ‘… emotions in management and the management of emotions play a significant role in the outcomes of public administration personnel …’ For instance, emotionally intelligent public leadership is considered ‘important in moving both leader and follower towards the public good’ (Newman et al., 2009, p. 9). The current COVID-19 pandemic has further emphasised these observation (Garikipati & Kambhampati, 2020; Huang, 2020).

In the South African context, the National Development Plan 2030 (RSA, 2012) asserted that changes in the public service require new kinds of knowledge, abilities, behaviours, attitudes and professionalism to shift the image of the state. However, despite the noble intent of public policy frameworks and initiatives, there is insufficient discourse on how to promote this democratic developmental public service ethos (Edigheji 2010). In this transforming context, effective public service leadership and governance are critical (Chipkin & Lipietz, 2012). Notably, there is still a lack of rigorous scholarship on public sector transformation and leadership, except for scholarly work on political leaders – illustrated by the many books on South African presidents.

Amongst the few South African studies on transformational leadership, Mokgolo, Mokgolo and Modiba’s (2012, p. 7) study on the relationships between transformational leadership behaviours and employee attitudes and performance in the provincial government sphere indicated that departmental leaders ‘do not have the appropriate transformational skills to manage their departments effectively and efficiently’. This accordingly leads to ‘negative perceptions of employees about how leaders manage their departments and about their leadership’ (Mokgolo et al., 2012, p. 7). The authors further claimed that the low level of public service delivery is the result of incompetent management and leadership. Transformational leadership styles place huge emphasis on the emotional and social competence of leaders (Goleman et al., 2002; Görgens-Ekermans & Roux, 2021). Moreover, Kotzé and Venter (2011, p. 412), using the Bar-On EQ-I measure, found a significant difference in overall EI scores between effective and ineffective middle-level South African public managers, concluding that there is ‘a lack of certain EI competencies at both managerial and entry level’.

Research questions and hypotheses

Given the conceptual distinctions and the empirical evidence presented, this study’s theoretical grounding is the competency theory of action and performance. The study assessed competencies as behavioural manifestation of EI that predicts job performance, particularly leadership performance (Boyatzis, 2009, 2018; Goleman, 2004; Goleman et al., 2002).

The primary research question: ‘What is the nature of the relationship between EI and leadership performance amongst South African public executive managers?’ was examined through testing the following five hypotheses:

Hypothesis 1: The four EI clusters are positively related to leadership performance.

Hypothesis 2: Each of the 12 EI competencies is positively related to leadership performance.

Hypothesis 3: A significant difference exists between the EI competency scores of high-performing and low-performing executive managers.

A secondary research question: ‘What is the nature of the relationship between the EI competency gap and leadership performance?’ was tested through two hypotheses:

Hypothesis 4: A significant negative relationship exists between the executive managers’ EI competency gap and leadership performance.

Hypothesis 5: A significant difference exists in the EI competency gap between high-performing and low-performing executive managers.

Research design

This study used a quantitative research design to test the five hypotheses. Theoretically, the behavioural EI model (Goleman et al., 2002) and the attendant measurement tool, the ESCI, was applied to measure EI competencies and to relate these to leadership performance using a nominations survey.

The empirical framework of this study, illustrated in Figure 1, comprises the variables to be tested, that is, the four emotional and social intelligence clusters, comprising 12 EI competencies (independent variable) and leadership performance nominations (dependent variable).

FIGURE 1: The study’s empirical framework: The relationship between emotional and social intelligence (EI) competencies and leadership performance.

Sampling framework and population

Emotional and Social Competence Inventory survey raters sample

The population for this study was 67 executive managers (Chief Director, Deputy Director-General, and Head of Department, i.e., rank levels 14–16), from four departments in a provincial government in South Africa and were selected using convenience sampling. Three departments represented the centre of government departments with province-wide transversal functions, whilst the fourth department represented the largest social service delivery line department. The 67 managers were invited to participate in the survey and provided with guidelines on the online ESCI survey. The sample for the survey was 35 (response rate = 52%).

The demographic profile of the sample comprised 60% males, almost half (46%) between the ages of 46–55 years, predominantly post-graduate (86%) black African (80%) Chief Directors (rank level 14, 77% ) largely from the three transversal departments (84%) and in the employ of the provincial government for the past 10 years (72%). In addition, 43% of the managers had fewer than 11 subordinates.

The final sample of ESCI rater respondents was 230, comprising five rater groups, namely manager or supervisor (n = 28), subordinate (n = 70), peer (n = 65), other (n = 37) and client (n = 30).

Leadership nominations survey sample

Using convenience sampling, middle-level and senior managers (n = 400; 100 per department) in the four departments, with close working contact with the 35 executive managers, were invited to nominate up to a maximum of three managers in an online nominations survey comprising a list of the managers’ names, with the question: ‘Choose the executive manager/s in your department whom you view as outstanding in their leadership performance’. A final total of 191 respondents’ (response rate = 48%) 371 nominations was used, ranging from a high of 32 nominations (outstanding leadership performance) to a low of zero (poor leadership performance). The descriptive statistics for the ESCI survey and performance nomination scores are presented in Table 1.

Measuring instruments

Independent variable measurement

The independent variable was EI competency score that was measured using the quantitative survey instrument, the ESCI (Boyatzis, 2009; Goleman et al., 2002). The ESCI is a self-administered, joint self-report and observer-rated online tool designed to assess EI competencies of individuals and organisations. The ESCI collects and reports on ratings of a leader’s competence in EI behaviours from multiple sources, such as supervisors, subordinates, peers, clients and others, including the leader’s self-ratings. All rater group respondents have a close working relationship with the participants assessed by the ESCI. The ESCI questionnaire comprises a set of 68 pre-determined closed questions using a five-point Likert-type scale (‘consistently’, ‘often’, ‘sometimes’, ‘rarely’ and ‘never’). Each question relates to one of the 12 EI competencies, which are factored independently and organised into four clusters. The researcher is an accredited user of the ESCI and is qualified to administer the tool and interpret the results.

The ESCI is specifically located within organisational life and work outcome contexts and thus was considered a relevant metric for this study; its emphasis on personal and professional development constitutes the cornerstone for the competency development process. Numerous empirical studies, displaying strong evidence of high reliability and validity of the ESCI have employed the ESCI to explore relationships between EI and leadership or management performance, as well as assessment, development and training of EI (Boyatzis, 2009; Görgens-Ekermans & Roux, 2021; O’Boyle et al., 2011; Piel, 2008; Ramo et al., 2009; Sharma, 2012). The average Cronbach’s alpha for the ESCI is 0.87 for ratings by rater respondents (ranging from 0.79 to 0.92), based on a sample of 52 363 (Hay Group, 2011). A principal axis exploratory factor analysis with promax rotation showed the factor analytic properties of the ESCI to be outstanding (Boyatzis, 2009).

Leadership nominations

Leadership performance was measured using multi-source nominations, which have high criterion validity, that is, they predict hard job performance outcomes (Boyatzis, 2009; McClelland, 1998; Ramo et al., 2009). Furthermore, relevant to this study, nominations relate significantly to work-output measures and strongly correlate with each other. Nominations are regarded as an effective and valid measure to predict solid job performance outcomes, as shown in a number of competency studies (Boyatzis, 2009; McClelland, 1998; Ramo et al., 2009).

Statistical analyses

Two quantitative analytical methods, namely Spearman’s rank correlation and analysis of variance (ANOVA), were used to test the five study hypotheses (Table 2). Spearman’s rank correlation coefficient determines the strength and direction of the bivariate linear relationship, using a monotonic function (Leedy & Ormrod, 2010). To examine the relationship between the EI competency gap and leadership performance, the descriptive statistics for the competency gap or SOA were calculated. Self-other agreement is defined as the difference between the score of each self-rater and the total other raters’ score. To test hypotheses 3 and 5, two sub-samples were extracted from the total sample of executive managers, based on their performance nomination scores, namely the top 35% (n = 12) and the bottom 35% (n = 12).

Research procedure and ethical considerations

Permission for the study was granted by the provincial government’s Office of the Premier and support was received from the HR branches in each department. The voluntary participation of all participants was sought through informed consent. The identity of all participants and the data collected is strictly anonymous and confidential. The researcher complied with the ethical procedure guidelines set out by a Western Cape university. The Hay Group granted permission to apply the ESCI for academic research purposes.

Ethical considerations

This article followed all ethical standards of research without direct contact with human or animal subjects.

Results

Emotional and Social Competence Inventory reliability analysis

Cronbach’s alpha coefficients for the ESCI competencies in this study show an overall alpha value of 0.84, with a range from 0.77 to 0.89 for the 12 EI competencies indicating acceptable internal consistency reliability of the scale in this setting (see Table 3).

The bivariate relationships between emotional intelligence and leadership performance

Hypothesis 1: The four emotional intelligence clusters are positively related to leadership performance

An overall strong, positive and significant correlation (r = 0.56; p < 0.001) was found in the relationship between EI and leadership performance of executive managers in this public service setting. As shown in Table 4, strong correlations were displayed in two EI clusters: self-awareness (r = 0.58) and self-management (r = 0.55) whilst the relationship management (r = 0.54) and social awareness clusters showed moderate relationships. Thus, Hypothesis 1 is supported.

Hypothesis 2: Each of the 12 emotional intelligence competencies is positively related to leadership performance

Four EI competencies showed strong, positive and highly significant (p < 0.001) correlations with leadership performance nominations, namely adaptability (r = 0.64), inspirational leadership (r = 0.61), emotional self-awareness (r = 0.58) and positive outlook (r = 0.55) (see Table 5). Furthermore, five EI competencies displayed moderate, positive and highly significant (p < 0.01) correlations with leadership performance, namely (in descending order of coefficient magnitude): influence (r = 0.50), teamwork (r = 0.48), coach and mentor (r = 0.48), achievement orientation (r = 0.48) and conflict management (r = 0.47). In addition, organisational awareness (r = 0.43) and empathy (r = 0.39) displayed weak, positive, significant (p < 0.05) correlations with leadership performance. Based on these results, Hypothesis 2 was confirmed for 11 EI competencies, however, it was not confirmed for the emotional self-control EI competency.

Hypothesis 3: A significant difference exists between the emotional intelligence competency scores of high-performing and low-performing executive managers

A two-way between-groups ANOVA was conducted to explore the impact of the EI competencies on the two levels of leadership performance. The ANOVA analysis yielded statistically significant effects at p < 0.01 for the four EI clusters and 11 EI competencies and at p < 0.05 for emotional self-control. Thus, Hypothesis 3 is supported.

Overall, the high-performing executive managers’ EI mean scores were significantly higher with three competencies yielding the highest mean differences and F-values, namely adapatability (M = 0.61; F (1,22) = 29.04, p < 0.01), emotional self-awareness (M = 0.58; F (1,22) = 28.05, p < 0.01), inspirational leadership (M = 0.67; F (1,22) = 26.43, p < 0.01). It is noteworthy that the competencies of coach and mentor (M = 0.58; F (1,22) = 19.10, p < 0.01), positive outlook (M = 0.57; F (1,22) = 16.48, p < 0.01) and teamwork (M = 0.57; F (1,22) = 12.28, p < 0.01) also displayed high mean differences and the next three highest F-values.

The relationships between the emotional intelligence competency gap ratings and leadership performance

Hypothesis 4: A significant negative relationship exists between the executive managers’ emotional intelligence competency gap and leadership performance

A moderate, negative and significant correlation (r = -0.47, p < 0.01) between the public service executives’ EI competency gap and their leadership performance was recorded. The EI competency gap mean for the executive managers was 0.28 and the range was between 1.68 (the highest) and -0.73 (the lowest). Twenty-five per cent of the gap scores were below -0.04, with a median of 0.24 (50% below median), and 25% were above 0.51. Therefore, hypothesis 4 is accepted.

Hypothesis 5: A significant difference exists in the emotional intelligence competency gap between high-performing and low-performing executive managers

The low-performing executive managers’ overall mean self-EI score (4.35) was higher than that of their high-performing colleagues (4.18). Conversely, the high-performing executive manager-group was rated higher (M = 4.26) by the total other respondent groups than the low-performing executive managers (M = 3.73). The mean competency gap for the high-performing group was -0.08 and for the low-performing group was 0.62.

The ANOVA result for the EI competency gap between these two groups of executive managers revealed a F-value (1.22) = 15.81 at p < 0.01. Therefore, a statistically significant difference exists in the EI competency gap between high- and low-performing executive managers. Accordingly, hypothesis 5 is accepted. In addition to the ANOVA results, executive managers with a higher mean emotional self-awareness score had smaller mean EI competency gaps and higher mean performance nomination scores (refer to Table 6).

Discussion

The bivariate relationships between emotional intelligence and leadership performance

Overall, the quantitative analyses showed strong, positive and significant correlations between public service executive managers’ EI competencies and their leadership performance. All four ESCI clusters displayed either strong (self-awareness and self-management) or moderate (relationship management and social awareness) positive correlations with leadership performance. Thus, Hypothesis 1 was confirmed.

Nine of the 12 EI competencies showed a positive and highly significant (r = 0.47–0.64; p < 0.01 to p < 0.001) correlations with leadership performance, whilst two of the 12 EI competencies showed a positive and relatively moderately significant (r = 0.39–0.43; p < 0.05 to p < 0.01) correlations with leadership performance. Emotional self-control was the only competency that did not show a significant (p > 0.05) association with performance. Thus, Hypothesis 2 was confirmed for 11 competencies.

The ANOVA analysis indicated significant difference between the EI competency scores of high-performing and low-performing executive managers at p < 0.01 for the four EI clusters and 11 EI competencies and at p < 0.05 for emotional self-control. Thus, Hypothesis 3 was supported. In addition, the ESCI indicated high reliability in this setting.

These results are in general agreement with previous research findings in the fields of EI and leadership. Several meta-analyses confirm the significant relationship between EI and performance (Joseph & Newman, 2010; O’Boyle et al., 2011; Van Rooy & Viswesvaran, 2004; Van Rooy et al., 2005). Both meta-analyses by Joseph and Newman (2010) and O’Boyle et al. (2011) confirmed that EI relates to job performance over and above cognitive ability and personality for all three EI streams, and notably, that the ECI/ESCI displayed the largest incremental validity (O’Boyle et al., 2011).

In particular, this study confirms claims that adopt a behavioural approach to EI on which the theoretical framework of this study was based (Boyatzis, 2009, 2018; Cherniss, 2010; Dulewicz & Higgs, 2003; Goleman & Boyatzis, 2008; Havers, 2010; Hopkins, O’Neil, & Williams, 2007; Piel, 2008; Ramo et al., 2009; Rosete & Ciarrochi, 2005; Ryan et al., 2009). For example, a study on Spanish executives in three public sector organisations show that EI competencies are valuable predictors of job performance as measured by performance nominations (Ramo et al., 2009). In addition, Cherniss et al.’s (2006) review of EI studies show superior performers scoring higher in all four ECI clusters.

Several studies show a positive relationship between leader EI and transformational leadership (e.g. Barling et al., 2000; Brown & Moshavi, 2005; Görgens-Ekermans & Roux, 2021; Mills, 2009; Piel, 2008; Sosik & Megerian, 1999) and charismatic leadership (Shamir et al., 1994). Transformational leadership scholars have theorised the significance of a leader’s impact on their followers’ emotional states (Bass & Avolio, 1994). For example, Piel (2008) found that the EI relationship management cluster was a potential predictor of transformational leadership and subsequently demonstrated that all ESCI clusters were positively correlated with the transformational leadership factors but negatively associated with transactional leadership. Recently, in a South African study, Görgens-Ekermans and Roux (2021) observed strong effects between three ESCI clusters (relationship management, self-management and social awareness) with transformational leadership (idealised influence, inspirational motivation and individualised consideration, respectively).

The four EI competencies displayed the strongest positive correlations with leadership performance and the highest significant mean differences and F-values between the EI competencies of high- and low-performing managers, namely emotional self-awareness, adaptability, inspirational leadership and positive outlook are of noteworthy significance. The self-awareness cluster showed the strongest correlation with leadership performance. The findings highlight the significance of self-awareness, which is considered the foundational competency or cornerstone of EI (Goleman et al., 2002; Görgens-Ekermans & Roux, 2021) and the fundamental component of effective leadership (Day, Fleenor, Atwater, Sturm, & McKee, 2014). This study supports Havers’ (2010) claim that participants with high emotional self-awareness display more ESCI competencies at strength. In addition, Piel (2008) using the ESCI found that self-awareness was strongly correlated with the transformational leadership ability of idealised influence (Bass & Avolio, 1994).

The adaptability competency displayed the strongest correlation with leadership performance and the second-largest significant difference in means between high- and low-performing executive managers. Strength in this competency illustrates top-performing executive managers’ adeptness in situations of complex change and in times of crises, such as the current COVID-19 pandemic, through behaviours and actions related to adaptive leadership, such as flexibility, open-mindedness and ambiguity (Heifetz, 1994; Huang, 2020). Adaptability is strongly associated with transformational leadership (Humphrey, 2002; Tucker & Russell, 2004), as this competency is effective during times of constant and rapid change, uncertainty and crises (Humphrey, 2002). Furthermore, the importance of adaptability for public service executive managers in their turbulent environment is essential (Lan & Hung, 2018; Newman et al., 2009), with Van Wart (2011) referring to adaptability as a key component of flexibility and a positive leadership trait, critical in all change functions and essential in the more complex and ambiguous public service environment.

The EI competency of inspirational leadership displayed the second strongest correlation with leadership performance and the largest significant difference in means for high- and low-performing executive managers. Inspirational leadership is strongly related to transformational leadership elements such as modelling the way, instilling pride, focusing on followers’ needs, enunciating a clear and shared vision and trust (Orazi et al., 2013). For example, significant correlations were found between EI competencies (using the ESCI) and transformational leadership scores using the Multifactor Leadership Questionnaire (Piel, 2008).

The relationships between the emotional intelligence competency gap and leadership performance

Several researchers (Atwater & Yammarino, 1992; Fleenor Smither, Atwater, Braddy, & Sturm, 2010; Harris & Schaubroeck, 1988) claimed that self-ratings tend to be inflated compared with other ratings. This study confirms their claim as half the self-rating participants overestimated their scores. In addition, all the EI clusters and competencies mean self-ratings were higher than those of the ratings by others (e.g. managers, subordinates, peers, etc.).

Fleenor et al.’s (2010) review confirmed the general positive relationship between SOA and leadership effectiveness. Atwater and Yammarino’s (1992) research concluded that over-raters (i.e. those with self-ratings that differ greatly from other ratings) are poorer performers compared with in-agreement raters or under-raters. This study confirms this conclusion, as the correlation analysis indicated a moderate negative significant relationship between the executive managers’ competency gap and their performance. In addition, the ANOVA analysis indicated that low-performing executive managers’ self-ratings were on average seven times larger than that of their top-performing counterparts. The results also compare favourably with other studies (e.g. Fleenor et al., 2010).

Atwater and Yammarino’s (1992) research also concluded that individuals with high SOA ratings are considered more self-aware. This study corroborates their claim, as the executive managers with high emotional self-awareness scores displayed strong SOA, that is, their competency gap was on average nine times lower than that of the low-scoring executive managers. The highly self-aware executive managers were significantly superior performers receiving an average of 70% more performance nominations from their colleagues. These findings have significance for studies that only use self-ratings.

Practical implications

The results of this exploratory study have significant theoretical, methodological and practical implications for public leadership. The first major implication is that transformational and emergent leadership models that include EI offer the public service appropriate exemplars in the pursuit of distinctive public leadership models. However, public service leadership models remain extremely managerial and transactional and pay little attention to emotions and feelings (Denhardt & Denhardt, 2006; Newman et al., 2009; Vigoda-Gadot & Meisler, 2010). Day et al.’s (2014) comprehensive review of 25 years of leadership development research and theory found that EI competencies are needed for leadership and managerial development initiatives.

The methodological implication is the importance of using a 360-degree feedback process based on self- and multi-source rating instruments as these have greater reliability than single-source instruments and are appropriate for development purposes. As shown in this study (and several other studies), self-ratings are significantly inflated, and therefore have limitations for development purposes. Day et al. (2014) also found support for 360-degree feedback tools (such as the ESCI) as a process to facilitate leader and leadership development because feedback from multisource rating fosters self-awareness and the development of individual leaders.

Limitations and recommendations

There are a few limitations of this study to be observed. Firstly, the small sample size of executive managers (n = 35) as the initial sample size of 67 participants did not materialise because of attrition of participants during the research and too few raters completing the surveys. The lukewarm interest by public managers in participating in research studies is a cause for concern given the paucity of evidence on public leadership. Secondly, the study was limited to the context of four departments within a provincial government. Thirdly, the dependent variable (leadership performance) was measured through a broad measure, which limits the examination of the multiple facets of this complex phenomenon. Therefore, the generalisability of the results to other settings may be limited.

However, the research results provide a number of opportunities for future research. As this was the first study on EI competencies and leadership performance using the research design and methods in the South African public service, a large-scale study is warranted. This will have the benefit of further testing the research design and the results.

A fertile area for future EI research in the public service is to test measurable elements of transformational and emergent (e.g. authentic, servant, ethical and spiritual) leadership models (considering their growth in recent literature) as variables (Miao et al., 2018). These leadership approaches are appropriate as they are concerned with emotions, values, ethics and long-term relationships, as well as followers’ motives, needs and satisfaction.

Conclusion

The primary objective of this quantitative study was to examine the nature of the relationships between executive managers’ EI competencies and their leadership performance in the South African public service, framed within the behavioural EI model (Boyatzis, 2009; Boyatzis et al., 2017; Goleman et al., 2002) through applying metrics of the ESCI and leadership nominations. Related to this aim, the study determined the EI competencies that differentiate high performing executive managers. The secondary methodological aim was to examine the relationship between EI competency gap/SOA and leadership performance. The study examined five hypotheses that were all supported. The results indicated significant and positive correlations between EI competencies and leadership performance with the competencies of adaptability, inspirational leadership, emotional self-awareness and positive outlook displaying the strongest correlations. A significant negative relationship between the managers’ SOA and their performance was observed.

This study contributes empirical evidence on the nature of the bivariate relationship in the African public service context. The high Cronbach’s alpha coefficients for the EI competencies indicate the suitability of the ESCI in this context, therefore implying the applicability of Western EI concepts and metrics. However, further research is warranted in other African public sector settings to compare the competencies with the strongest correlations with performance. The results have implications for management and leadership development and recruitment in the public sector.

Acknowledgements

The author would like to express his special gratitude to Prof. Erwin Schwella (University of Stellenbosch) for supervising his original thesis; Dr. Annie McKee (University of Pennsylvania) and Prof. Richard Boyatzis (Case Western Reserve University) for their guidance on the research design and access to relevant literature; and Prof. Martin Kidd, Ms. Magriet Treurnicht and Ms Adéle Burger (University of Stellenbosch) for their guidance and support with the statistical analysis of the nominations and the ESCI survey data.

The author declares that he has no financial or personal relationship(s) that may have inappropriately influenced him in writing this article.

S.J.H. was responsible for the final research design, data collection, analyses and article write-up.

This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.

The data that support the findings of this study are available from the corresponding author, upon reasonable request.

The views and opinions expressed in this article are those of the author and do not necessarily reflect the official policy or position of any affiliated agency of the author.

References