Existing research suggests that CEO personality traits, including narcissism, can significantly affect a firm's performance and strategic decision-making. This study evaluates the direct outcome of CEO narcissism on ESG performance in Indonesia.
This research is using non-financial firms listed on the Indonesian Stock Exchange from 2017 to 2022, a total of 132 firm-year observations, and the data are processed by ordinary least squares (OLS) regression fixed effect.
Empirical results exhibit that narcissistic CEO significantly influences a firm's social performance positively, whilst indicating insignificance towards overall ESG performance, environmental and governance performance in Indonesia.
While many empirical works have assessed the relationship of CEO narcissism and ESG performance in other countries, very few studies have been conducted in Indonesia and none is using LinkedIn profiles to measure CEO narcissism levels.
1. Introduction
Sustainability has emerged as a hot topic in the academic and corporate communities during the previous several decades. Institutions and multinational corporations (MNCs) have increasingly changed their responsiveness to corporate social responsibility (CSR) towards environment, social and government (ESG) practices. The concept of ESG has piqued global interest and has evolved into a new concept within the global economic framework (Bildirici et al., 2024; Dabbebi et al., 2022; Effah et al., 2025). Existing literature also supports the fact that ESG activities positively influence a firm's performance (Agarwala et al., 2024; Dakhli, 2022; Fahad and Busru, 2021; Yook and Lee, 2020).
The global ESG concern has also reached Indonesia, as shown by the nation's commitment to supporting the UN's Sustainable Development Goal (SDG) by 2030. In January 2022, President Joko Widodo introduced Indonesia's Green Taxonomy during Financial Services Industry Annual Meeting 2022, acknowledging the significance of a sustainable finance taxonomy (PwC, 2022).
However, only 17.5% firms listed on the Indonesian Stock Exchange (IDX) disclosed the sustainability report in 2016. The Financial Service Authority (OJK) issued OJK Regulation No.51/2017 (OJK, 2017) which mandates that Indonesian firms are required to disclose a sustainability report within two-year transition (Gunawan et al., 2022).
Acknowledging the urgency of ESG performance for a corporation, scholars across the regions have delved into researching the variables that may influence it, such as financial performance (2023; Dakhli, 2022), digital transformation (Xu and Yin, 2025), culture (Nguyen et al., 2024), the board of directors (Yadav and Prashar, 2023), including CEO personalities (Awuah et al., 2024; Bildirici et al., 2024; Bouzouitina et al., 2021; Kind et al., 2023; Nasih et al., 2023; Lassoued and Khanchel, 2023).
Research has consistently demonstrated that the Chief Executive Officer (CEO) is a key member of the management board and plays a pivotal role in the company's most significant strategic choices (Naaman and Sun, 2022). Upper Echelon Theory (1984) posits that the values and attributes of the top management team (TMT) members are reflected in the organization's decisions. The theory further elaborates on how individual characteristics affect the decision-making process and, consequently, the firm's performance. This understanding provokes other authors to delve into how CEOs and the level of their characteristics influence a firm's ESG activities such as masculinity (Nasih et al., 2023), hubris and narcissism (Lin et al., 2022; Martínez-Ferrero et al., 2024), overconfidence (Bouzouitina et al., 2021) and power (Awuah et al., 2024). Narcissism is a more complex psychological concept, and our evaluation reveals specific behaviors linked to exploitiveness and entitlement, as opposed to optimism.
Past literature reveals how CEO narcissism has been established to increase a firm's ESG performance in developed countries such as Europe (Awuah et al., 2024; Riera and Iborra, 2023) and the United States (Bildirici et al., 2024; Lassoued and Khanchel, 2023; Petrenko et al., 2016; Al-Shammari et al., 2019; Dabbebi et al., 2022). However only few studies were conducted in developing countries (Agnihotri and Bhattacharya, 2023; Falah and Mita, 2020; Ledi et al., 2025). Moreover, other studies observed that CEO narcissism negatively affected the ESG performance (Effah et al., 2025; Kind et al., 2023; Al-Shammari et al., 2019; Ahn et al., 2020) and found insignificant results. There is inconclusive result regarding CEO narcissism's effect on ESG performance and these are the research gaps to be filled by the current study.
Research on CEO narcissism has traditionally relied on self-reported measures prone to social desirability bias (Cragun et al., 2020), and to overcome this limitation, scholars have utilized secondary data sources (Chatterjee and Hambrick, 2007), then extended by Aabo and Eriksen (2018), who developed LinkedIn profiles as a tool for measuring CEO narcissism level. Using LinkedIn profiles can give you more information, like how visible and important the CEO is on their social media presence, which can be a sign of narcissism (Junge et al., 2024). Given that few studies employed this method in previous research (Aabo et al., 2022; Riera and Iborra, 2023), the LinkedIn approach constitutes a novel contribution to this field.
This current study has the objective of addressing the direct relationship between a narcissistic CEO and a firm's ESG performance, with a particular focus on the scope of Indonesia, by using publicly listed firms on the Indonesian Stock Exchange from 2017 to 2022.
The proposed framework leads to three contributions: (1) it develops the existing literature about CEO narcissism within the Upper Echelons framework by investigating its association with ESG, (2) by applying a new measurement (the LinkedIn approach), it expands the body of literature on CEO narcissism, (3) the study provides indispensable insight into what a self-centered CEO does to a company's ESG performance in the Indonesian context.
The authors will discuss the literature review and create the hypothesis in the following part. Later, this paper will delve into the sample and research method, including how LinkedIn profiles will be utilized to measure CEO narcissism. It will then be followed by a section to discuss research findings, including the robustness test. Finally, the paper will conclude with recommendations for future research and limitations.
2. Literature review
2.1 CEO narcissism
The Upper Echelons Theory (UET), proposed by Hambrick and Mason (1984), posits that the psychological and observable attributes of top executives significantly affect strategic choices. Top executives' characteristics and strategic actions can be impacted by their external atmosphere or conditional company characteristics, which are antecedents of top executive characteristics and company outcomes that will reinforce strategic decisions.
There are several traits stated in UET, and the personalities determine the cognitive traits, personal values and dispositions of CEOs, which may either promote or deter them from engaging in actions that are considered socially irresponsible practices (Dabbebi et al., 2022) and our emphasis is directed on Narcissism – a key factor that is found to affect CEO's decision making (Dabbebi et al., 2022; Martínez-Ferrero et al., 2024; Yook and Lee, 2020).
The American Psychiatric Association further defines narcissism in the Diagnostic and Statistical Manual of Mental Disorder as a “multifaceted personality trait that combines grandiosity, attention-seeking behaviour, an inflated self-view, a need for continuous self-regulation, and a general lack of consideration for others” (American Psychiatric Association, 2013). Narcissistic individuals are characterized by an excessive sense of self-assurance regarding their intelligence and judgment, as well as a condescending and disagreeable demeanor. Narcissistic CEOs may make decisions that damage the company's finances and other strategic decisions to achieve their own personal goals but narcissism may also be a “double-edged” quality that can be good for an organization because, due to the fact that narcissistic CEOs are frequently perceived as outgoing, charismatic, positive and very worried about their image, reputation and how people see them (Martínez-Ferrero et al., 2024).
2.2 CEO narcissism and ESG performance
ESG, introduced by the United Nations in 2004 (Gillan et al., 2021), signifies how a company considers environmental, social and governance elements during decision-making (Rau and Yu, 2023). This study characterizes ESG performance as a criterion for assessing a company's efforts to integrate environmental, social and governance considerations into its operations and strategy. Previous studies show several positive impacts, stating that firms that achieved higher ESG scores are superior in terms of firm profitability (Treepongkaruna and Suttipun, 2024) and maintain a firm's good reputation or satisfy stakeholders' expectations (Eliwa et al., 2021).
The top management team is in charge of making strategic decisions, and the Upper Echelon Theory supports this by showing how should the CEO be involved in ESG activities or disclosures (Agnihotri and Bhattacharya, 2023; Al-Shammari et al., 2019; Awuah et al., 2024; Bildirici et al., 2024; Dabbebi et al., 2022; Kind et al., 2023; Lassoued and Khanchel, 2023).
The CEOs set the company's strategic direction and are the main decision makers (Chatterjee and Hambrick, 2007; Naaman and Sun, 2022), which shaped by their characteristics and experiences as the UET stated (Hambrick and Mason, 1984). Narcissistic CEOs are likely to spend more time and money on activities that draw more attention than on those that are less likely to get attention. Past research (Cragun et al., 2020; Dabbebi et al., 2022) signified that narcissistic CEOs prefer to be involved in decision-making processes, prioritizing their image and reputation and focusing on initiatives that enhance ESG performance. CEO narcissism can increase ESG performance by motivating firms to disclose social and corporate governance activities that bolster their public image (Yook and Lee, 2020).
Using data from 55 US energy companies from 2003 to 2022, Bildirici et al. (2024) look at the connection between quantile-specific effects, and the results demonstrate that CEOs who admires themselves have a big impact on ESG for all quantiles over time. Ledi et al. (2025) also discovered the same when they looked at how the personalities of top team managers affected ESG performance in Ghana. Awuah et al. (2024) posited that CEOs often determine firms' sustainability initiatives, and found that self-admiring CEOs may allocate additional resources and time to the ESGs to fulfill their personal desire for adoration and attention by capitalizing on the ESG activities.
ESG performance of 81 listed firms on the Johannesburg stock market is studied by Effah et al. (2025) in relation to the CEO narcissism, resulting in a significantly negative relationship. Firms managed by narcissistic CEOs steer the firm toward ESG activities by disclosing social and corporate governance activities (Aabo and Eriksen, 2018; Dabbebi et al., 2022) to gain media attention and enhance their own reputation. As a result, narcissistic CEOs accomplish their objectives by bolstering their public persona and inspiring self-esteem, legitimacy and reverence via these disclosures. Hence the authors propose the hypothesis as follows:
CEO narcissism does significantly influence overall ESG performance.
2.3 Individual environmental, social and governance performance
Eliwa et al. (2021) say that Environmental, Social and Governance (ESG) practices are a complicated, multi-dimensional idea where the effects of one dimension may cancel out those of another and the level of impact on each component may change depending on how different the scenario is (Kind et al., 2023; Martiny et al., 2024). The environmental part of ESG is about the natural world that businesses are attempting to protect. Governance is more about how a company behaves, whereas the social side is more about how individuals get along and how happy they are at work (Rau and Yu, 2023). To understand what impact they have, we need to look at each of the three ESG factors separately (Lassoued and Khanchel, 2023; Kind et al., 2023; Dabbebi et al., 2022).
Dabbebi et al. (2022), who examined 118 US S&P 500 firms, found that a CEO exhibiting elevated levels of self-admiration tends to reveal social and corporate governance initiatives and aligned with Lassoued and Khanchel (2023). Kind et al. (2023) found that both environmental and governance performance are affected by CEO narcissism, but not the case with social performance. CEO narcissism significantly affects a firm's environmental performance (Lin et al., 2022). CEO narcissism also influences social performance significantly, as it allows them to increase their social desirability level (Bildirici et al., 2024). Gao et al. (2023) has demonstrated that narcissistic CEOs are motivated by their need for reputation for public welfare, including the government, consumers and the media.
Lastly, CEO narcissism is also believed to significantly impact a firm's governance performance as it may bring a high sense of charisma and leadership (Yang et al., 2023). Khanchel and Bentaleb (2022) posited that narcissistic CEO will place a higher emphasis on the disclosure of corporate governance to appease market participants. Conversely, governance-related activities are generally quiet and do not draw the public's attention, hence a narcissistic CEO discourages doing them (Kind et al., 2023; Dabbebi et al., 2022). Those activities characterize the firm's dedication to excellent governance and strong internal affiliation, such as audit committee structure, board composition, etc. (Kind et al., 2023; Cragun et al., 2020; Dabbebi et al., 2022). These previous results signify the magnitude of looking into how CEO narcissism affects individual ESG components, which is stated in the hypotheses further in the text.
CEO narcissism influences a firm's environmental performance significantly.
CEO narcissism influences a firm's social performance significantly.
CEO narcissism influences a firm's governance performance significantly.
3. Research methods
3.1 Population and sampling
This research used secondary data from 132 observations of publicly listed firms on the Indonesian Stock Exchange from 2017 to 2022 (Table 1). The starting period of 2017 was selected when sustainability report disclosure became mandatory, as the Indonesian Financial Services Authority (OJK) stipulates regulation number 51/POJK.03/2017 (OJK, 2017). The sample number is based on (1) not including banks and other financial companies, (2) requiring companies to have ESG performance index scores from Bloomberg and (3) requiring the CEOs who were in charge during the time to have active LinkedIn profiles.
Details of sample
| Category | Number of observations |
|---|---|
| Total firms listed in IDX (2017–2022) | 4,071 observations |
| Minus banks and financial institutions (2017–2022) | 264 observations |
| Minus firms with data unavailability (e.g. missing ESG score, missing LinkedIn profile) | 3,675 observations |
| Final sample | 132 observations |
| Category | Number of observations |
|---|---|
| Total firms listed in IDX (2017–2022) | 4,071 observations |
| Minus banks and financial institutions (2017–2022) | 264 observations |
| Minus firms with data unavailability (e.g. missing ESG score, missing LinkedIn profile) | 3,675 observations |
| Final sample | 132 observations |
3.2 Dependent variable
Firms' ESG scores are taken from the Bloomberg database, ranging from 0 to 100, indicating no disclosure to a full one (ESG Advising LLC, 2024). A higher Bloomberg score means that companies are giving out more ESG data, and the higher the ESG score, the better the company's ESG performance. It has been used in other studies (Yadav and Prashar, 2023; Dabbebi et al., 2022; Kind et al., 2023).
3.3 Independent variable
This research is focused on utilizing LinkedIn as a narcissism measurement tool as established by Aabo and Eriksen (2018). It consists of six items as seen in Table 2.
LinkedIn narcissism measurement
| LinkedIn section | Connection with narcissism | Measurement |
|---|---|---|
| Skills | Indicate their level of self-promotion behavior (Buffardi and Campbell, 2008) | Continuous; Log of skills total in LinkedIn profile (0–1) |
| Connections | Reflects the number of connections and popularity (Buffardi and Campbell, 2008) | Continuous; Total connections in Linkedin profile (0–1) |
| Previous positions | Suggest a desire to show their prior achievements (Aabo and Eriksen, 2018) | Continuous; Log of previous position total in LinkedIn profile (0–1) |
| Profile picture | Presence or absence of a profile picture serve as an indicator of narcissistic tendencies (Chatterjee and Hambrick, 2007) | Binary; 0 as none, 1 as available |
| Summary | Focus on their detailed qualifications and capabilities (Aabo and Eriksen, 2018) | Binary; 0 as none, 1 as available |
| Interests | Desire for self-expression and attention-seeking behavior (Aabo and Eriksen, 2018) | Binary; 0 as none, 1 as available |
| LinkedIn section | Connection with narcissism | Measurement |
|---|---|---|
| Skills | Indicate their level of self-promotion behavior ( | Continuous; Log of skills total in LinkedIn profile (0–1) |
| Connections | Reflects the number of connections and popularity ( | Continuous; Total connections in Linkedin profile (0–1) |
| Previous positions | Suggest a desire to show their prior achievements ( | Continuous; Log of previous position total in LinkedIn profile (0–1) |
| Profile picture | Presence or absence of a profile picture serve as an indicator of narcissistic tendencies ( | Binary; 0 as none, 1 as available |
| Summary | Focus on their detailed qualifications and capabilities ( | Binary; 0 as none, 1 as available |
| Interests | Desire for self-expression and attention-seeking behavior ( | Binary; 0 as none, 1 as available |
3.4 Control variables
Adapting the pattern from preceding research (Aabo et al., 2022; Aabo and Eriksen, 2018; Riera and Iborra, 2023), the control variables consist of two CEO attributes and four firm-oriented variables. CEOs who are older are less inclined to undertake difficult assignments (Awuah et al., 2024). CEO gender has also positively influenced ESG performance, through a 19% improvement in the ESG score (Aabo and Giorici, 2023). Size of a firm has been reported to affect its ESG performance significantly (Agarwala et al., 2023). Older firms may perform better ESG performance due to longer experience and exposure to stakeholders' demand (Lassoued and Khanchel, 2023). Larger boards have greater expertise and skills, which gives them more points of view on corporate strategy and affects ESG performance (Tibiletti et al., 2021). Companies that make more profit tend to do better on ESG (Gunawan et al., 2022; Yook and Lee, 2020).
3.5 Models and technique of analysis
Ordinary Least Square (OLS) fixed-effects model is applied since we have panel data and robust standard error is applied to avoid autocorrelation and heteroscedasticity (Aabo and Eriksen, 2018). The Hausman specification test is conducted, and the fixed-effect (p-value:0.000) is selected. The following models are established to test H1-H2:
where ESGPij is firm's ESG performance score for year (i) in each firm (j). Narcissismij is the level of CEO narcissism for year (i) in each firm (j). EnvPij firm's environmental performance score for year (i) in each firm (j); SocPij is firm's social performance score for year (i) in each firm (j), GovPij is firm's governance performance score for year (i) in each firm (j) while the details of control variables and measurement are depicted in Table 3.
Control variables measurement
| Symbol | Variable name | Measurement |
|---|---|---|
| CAge | CEO Age | CEO's age during their tenure as CEO |
| CGender | CEO Gender | Binary; 0 as female and 1 as male |
| FSize | Firm Size | Natural log of firm's Total Assets |
| ROA | Return on Asset | Firm's profitability (Income before Discontinued Operations and Extraordinary Items/Total Assets) |
| FAge | Firm Age | Length of year from firm's IPO until current period |
| BSize | Board Size | Number of people on a firm's board of director |
| Symbol | Variable name | Measurement |
|---|---|---|
| CAge | CEO Age | CEO's age during their tenure as CEO |
| CGender | CEO Gender | Binary; 0 as female and 1 as male |
| FSize | Firm Size | Natural log of firm's Total Assets |
| ROA | Return on Asset | Firm's profitability (Income before Discontinued Operations and Extraordinary Items/Total Assets) |
| FAge | Firm Age | Length of year from firm's IPO until current period |
| BSize | Board Size | Number of people on a firm's board of director |
4. Results and discussion
4.1 Results
Table 4 exhibits the descriptive statistics results of 132 observations from publicly listed Indonesian firms from 2017 to 2022, except financial institutions. Excluding the EnvP score and ROA, all the variables have similar mean and median values, showing that the data are positively skewed. The average of ESG score is 42.096, whilst the median is 41.48, showing very minimal variance between. The difference between the minimum (value of 22.08) and maximum is also comparatively high (value of 77.86).
Descriptive statistics
| Mean | SD | Median | Min | Max | |
|---|---|---|---|---|---|
| ESGP | 42.096 | 11.743 | 41.48 | 22.08 | 73.866 |
| EnvP | 24.595 | 21.238 | 17.895 | 0 | 78.6 |
| SocP | 29.043 | 12.093 | 28.87 | 0 | 58.615 |
| GovP | 72.537 | 11.117 | 75.196 | 26.66 | 98.615 |
| Narcissism | 3.941 | 1.171 | 3.65 | 2.172 | 5.77 |
| CAge | 51.826 | 6.353 | 52 | 37 | 69 |
| CGender | 0.917 | 0.272 | 1 | 0 | 1 |
| FSize | 30.879 | 0.915 | 30.945 | 28.13 | 32.58 |
| ROA | 0.05 | 0.165 | 0.038 | −1.287 | 0.353 |
| FAge | 20.692 | 9.524 | 23 | 0 | 40 |
| BSize | 6.159 | 1.824 | 6 | 3 | 12 |
| Mean | SD | Median | Min | Max | |
|---|---|---|---|---|---|
| ESGP | 42.096 | 11.743 | 41.48 | 22.08 | 73.866 |
| EnvP | 24.595 | 21.238 | 17.895 | 0 | 78.6 |
| SocP | 29.043 | 12.093 | 28.87 | 0 | 58.615 |
| GovP | 72.537 | 11.117 | 75.196 | 26.66 | 98.615 |
| Narcissism | 3.941 | 1.171 | 3.65 | 2.172 | 5.77 |
| CAge | 51.826 | 6.353 | 52 | 37 | 69 |
| CGender | 0.917 | 0.272 | 1 | 0 | 1 |
| FSize | 30.879 | 0.915 | 30.945 | 28.13 | 32.58 |
| ROA | 0.05 | 0.165 | 0.038 | −1.287 | 0.353 |
| FAge | 20.692 | 9.524 | 23 | 0 | 40 |
| BSize | 6.159 | 1.824 | 6 | 3 | 12 |
Note(s): ESGP: ESG Performance; EnvP: Environmental Performance; SocP: Social Performance; GovP: Governance Performance; CAge: CEO Age; CGender: CEO Gender; FSize: Firm Size; ROA: Return on Asset; FAge: Firm Age; BSize: Board Size
A Pearson correlation test was conducted to check for multicollinearity (Table 5). All variables displayed a score lower than 0.8 (Gujarati and Porter, 2020), which indicates that no multicollinearity is found in our data.
Pearson correlations
| Variables | (1) | (2) | (3) | (4) | (5) | (6) | (7) | (8) | (9) | (10) |
|---|---|---|---|---|---|---|---|---|---|---|
| (1) ESGP | 1.000 | |||||||||
| (2) EnvP | 0.907* | |||||||||
| (3) SocP | 0.839* | 0.698* | ||||||||
| (4) GovP | 0.524* | 0.208 | 0.240* | |||||||
| (5) Narcissism | 0.207 | 0.127 | 0.224* | 0.170 | ||||||
| (6) CAge | −0.163 | −0.130 | −0.101 | −0.159 | −0.327* | |||||
| (7) CGender | −0.029 | −0.005 | −0.077 | 0.000 | 0.064 | 0.052 | ||||
| (8) FSize | 0.129 | 0.030 | 0.103 | 0.237* | −0.072 | −0.123 | 0.049 | |||
| (9) ROA | 0.353* | 0.315* | 0.274* | 0.220 | 0.131 | 0.084 | 0.013 | 0.058 | ||
| (10) FAge | 0.216 | 0.250* | 0.108 | 0.091 | 0.123 | 0.051 | 0.016 | 0.126 | 0.385* | |
| (11) BSize | 0.181 | 0.198 | 0.159 | 0.022 | −0.060 | 0.021 | 0.026 | −0.085 | 0.180 | 0.243* |
| Variables | (1) | (2) | (3) | (4) | (5) | (6) | (7) | (8) | (9) | (10) |
|---|---|---|---|---|---|---|---|---|---|---|
| (1) ESGP | 1.000 | |||||||||
| (2) EnvP | 0.907* | |||||||||
| (3) SocP | 0.839* | 0.698* | ||||||||
| (4) GovP | 0.524* | 0.208 | 0.240* | |||||||
| (5) Narcissism | 0.207 | 0.127 | 0.224* | 0.170 | ||||||
| (6) CAge | −0.163 | −0.130 | −0.101 | −0.159 | −0.327* | |||||
| (7) CGender | −0.029 | −0.005 | −0.077 | 0.000 | 0.064 | 0.052 | ||||
| (8) FSize | 0.129 | 0.030 | 0.103 | 0.237* | −0.072 | −0.123 | 0.049 | |||
| (9) ROA | 0.353* | 0.315* | 0.274* | 0.220 | 0.131 | 0.084 | 0.013 | 0.058 | ||
| (10) FAge | 0.216 | 0.250* | 0.108 | 0.091 | 0.123 | 0.051 | 0.016 | 0.126 | 0.385* | |
| (11) BSize | 0.181 | 0.198 | 0.159 | 0.022 | −0.060 | 0.021 | 0.026 | −0.085 | 0.180 | 0.243* |
Note(s): ESGP: ESG Performance; EnvP: Environmental Performance; SocP: Social Performance; GovP: Governance Performance; CAge: CEO Age; CGender: CEO Gender; FSize: Firm Size; ROA: Return on Asset; FAge: Firm Age; BSize: Board Size
CEO narcissism does not show a statistically significant relationship towards overall ESG performance; hence, it fails to accept the H1 (p-value>0.05) as indicated in Table 6, aligning with past literature findings (Al-Shammari et al., 2019) [1]. That means a firm that has a CEO with a narcissistic character exhibits no relationship with ESG performance. The finding fails to validate UET, and this contradicts previous literature conducted (Bildirici et al., 2024; Ledi et al., 2025; Awuah et al., 2024). Chatterjee and Hambrick (2007) contend that the psychology of narcissism significantly influences organizational strategy, but these ramifications are not straightforward.
Regression results
| Dependent variable | ESGP | EnvP | SocP | GovP |
|---|---|---|---|---|
| Narcissism | 1.056 | 0.230 | 1.803* | 1.068 |
| (0.802) | (1.606) | (0.874) | (0.843) | |
| CAge | −16.16* | −27.28* | −7.559 | −13.12 |
| (6.534) | (12.34) | (7.307) | (8.751) | |
| CGender | −3.757 | −3.743 | −6.137* | −1.527 |
| (3.064) | (4.954) | (3.089) | (2.890) | |
| FSize | 1.209 | −0.417 | 1.472 | 2.639** |
| (0.987) | (1.840) | (1.021) | (0.947) | |
| ROA | 52.10*** | 84.95*** | 43.22*** | 27.47*** |
| (8.964) | (15.39) | (9.061) | (7.712) | |
| FAge | −0.0378 | 0.114 | −0.164 | −0.0796 |
| (0.102) | (0.189) | (0.102) | (0.114) | |
| BSize | 0.238 | 0.315 | 0.529 | −0.113 |
| (0.533) | (1.015) | (0.573) | (0.622) | |
| _cons | 59.56 | 129.5 | 6.745 | 38.58 |
| (48.04) | (93.11) | (52.57) | (53.76) | |
| N | 132 | 132 | 132 | 132 |
| Adj. R2 | 0.377 | 0.301 | 0.256 | 0.113 |
| Year FE | Yes | Yes | Yes | Yes |
| Industry FE | Yes | Yes | Yes | Yes |
| Dependent variable | ESGP | EnvP | SocP | GovP |
|---|---|---|---|---|
| Narcissism | 1.056 | 0.230 | 1.803* | 1.068 |
| (0.802) | (1.606) | (0.874) | (0.843) | |
| CAge | −16.16* | −27.28* | −7.559 | −13.12 |
| (6.534) | (12.34) | (7.307) | (8.751) | |
| CGender | −3.757 | −3.743 | −6.137* | −1.527 |
| (3.064) | (4.954) | (3.089) | (2.890) | |
| FSize | 1.209 | −0.417 | 1.472 | 2.639** |
| (0.987) | (1.840) | (1.021) | (0.947) | |
| ROA | 52.10*** | 84.95*** | 43.22*** | 27.47*** |
| (8.964) | (15.39) | (9.061) | (7.712) | |
| FAge | −0.0378 | 0.114 | −0.164 | −0.0796 |
| (0.102) | (0.189) | (0.102) | (0.114) | |
| BSize | 0.238 | 0.315 | 0.529 | −0.113 |
| (0.533) | (1.015) | (0.573) | (0.622) | |
| _cons | 59.56 | 129.5 | 6.745 | 38.58 |
| (48.04) | (93.11) | (52.57) | (53.76) | |
| N | 132 | 132 | 132 | 132 |
| Adj. R2 | 0.377 | 0.301 | 0.256 | 0.113 |
| Year FE | Yes | Yes | Yes | Yes |
| Industry FE | Yes | Yes | Yes | Yes |
Note(s): Standard errors in parentheses; *p < 0.05, **p < 0.01, ***p < 0.001
ESGP: ESG Performance; EnvP: Environmental Performance; SocP: Social Performance; GovP: Governance Performance; CAge: CEO Age; CGender: CEO Gender; FSize: Firm Size; ROA: Return on Asset; FAge: Firm Age; BSize: Board Size
This study conducts a further analysis by separating the dimension of ESG as shown by Kind et al. (2023). Accordingly, Ahn et al. (2020) posit that it is expected that narcissistic CEOs are prioritize certain component of ESG over others. Table 6 shows the statistical result that variable narcissism did not depict a significant effect on EnvP (p-value>0.05). Hence, it means that a firm led by a narcissistic CEO has no effect on its environmental performance and aligned with (Lassoued and Khanchel, 2023; Dabbebi et al., 2022). The results possibly due to environmental performance of Indonesian firms merely show a mean of 24.595 (min. = 0 and max = 78.6 – Table 4). The discrepancy between these values shows that although some Indonesian firms have shown effort in executing environmental activities, many firms have not prioritized them. The absence of effective environmental management in developing countries (e.g., Indonesia) may be attributed to inadequate law enforcement (Rini et al., 2024). Another reason is that 72% of Indonesians lack attentiveness towards environmental issues as reported by the Indonesian Ministry of Environment and Forestry (Dewi, 2022), which made the firm to put less priority on taking care of this aspects (Wahyuningrum et al., 2024) as less contribution was made to the bottom line. Hence, the CEO's narcissistic trait fails in influencing the firm to conduct more environmental activities in Indonesia.
A similar finding was also found in the component of corporate governance, resulting in a failure to accept hypothesis 2c. As seen in Table 6, the variable narcissism shows an insignificant result toward GovP (p-value>0.05), indicating that the performance of corporate governance is not affected by the presence of narcissistic CEOs in firms. This finding is not able to prove the Upper Echelon Theory and consistent with Ahn et al. (2020), who conducted a study in the non-manufacturing industry of the United States. This may be because the non-manufacturing industry comprises 56% of the current study.
Highly self-admiring CEOs are preoccupied with the people's view of their persona, even if it necessitates selective disclosure to conceal their immoral behaviors (Kind et al., 2023). Then, this regulation allows the existence of other board members to ensure that the firm's strategic decisions remain objectively made, aligned with Tibiletti et al. (2021). Hence, whenever a narcissistic CEO desires to execute actions to improve his popularity in society, it may not be approved, hindering the connection of CEO narcissism and corporate governance performance in Indonesia.
Finally, CEO narcissism exhibits a significant positive relationship with a corporation's social performance in Indonesia (β = 1.803 and p-value<0.05), meaning that firms under a narcissistic CEO are proven to boost their firm's social performance. Hence, hypothesis H2c is accepted. This validates UET and supports the previous literature claiming a positive significant relationship between these variables (Riera and Iborra, 2023; Bildirici et al., 2024; Dabbebi et al., 2022).
It is due to narcissistic CEOs focusing on creating a better public image and avoiding misconduct that may damage their reputation (Donker et al., 2023) and this will be done by engaging in ethical and philanthropic firm-wide activities, including those that level up the firm's social performance (Petrenko et al., 2016). To be more precise, the increasing media attention and stakeholder interest in social activities present an ideal opportunity for a self-admiring CEO to receive the acclaim and admiration they think is rightfully theirs, thereby enhancing CEOs' visibility and presenting them as noble saviors and protagonists (Dabbebi et al., 2022; Gao et al., 2023).
4.2 Robustness test
4.2.1 Measurement
Additional analysis was performed to exhibit the robustness of our findings, utilizing different measures for CEO narcissism by assigning the CEO narcissism to a dummy variable (Awuah et al., 2024). As shown in Table 7, CEO narcissism does not affect ESG Performance (p-value>0.05), and it is similar to Table 6. Furthermore, the separate pillar of ESG performance also exhibits the same findings as proven previously in Table 6.
Robustness test results
| Dependent variable | ESGP | EnvP | SocP | GovP |
|---|---|---|---|---|
| Narcissism_ binary | 1.637 | −2.047 | 4.337* | 3.644 |
| (1.832) | (3.611) | (1.919) | (2.031) | |
| CAge | −0.482*** | −0.833** | −0.158 | −0.347* |
| (0.135) | (0.257) | (0.148) | (0.159) | |
| CGender | −3.467 | −2.679 | −7.359* | −1.232 |
| (3.247) | (5.711) | (3.209) | (2.069) | |
| FSize | 1.058 | −0.468 | 1.666 | 2.122* |
| (1.002) | (1.869) | (1.017) | (0.876) | |
| ROA | 29.31** | 44.99* | 41.43*** | 34.94*** |
| (10.81) | (18.18) | (9.633) | (7.970) | |
| FAge | 0.00527 | 0.189 | −0.266* | 0.0168 |
| (0.100) | (0.189) | (0.108) | (0.106) | |
| BSize | 0.682 | 1.010 | 0.632 | −0.276 |
| (0.577) | (1.086) | (0.581) | (0.562) | |
| Constant | 25.57 | 60.85 | −13.65 | 24.63 |
| (35.15) | (65.80) | (36.32) | (30.52) | |
| N | 132 | 132 | 132 | 132 |
| Adj. R2 | 0.317 | 0.258 | 0.279 | 0.184 |
| Year FE | Yes | Yes | Yes | Yes |
| Industry FE | Yes | Yes | Yes | Yes |
| Dependent variable | ESGP | EnvP | SocP | GovP |
|---|---|---|---|---|
| Narcissism_ binary | 1.637 | −2.047 | 4.337* | 3.644 |
| (1.832) | (3.611) | (1.919) | (2.031) | |
| CAge | −0.482*** | −0.833** | −0.158 | −0.347* |
| (0.135) | (0.257) | (0.148) | (0.159) | |
| CGender | −3.467 | −2.679 | −7.359* | −1.232 |
| (3.247) | (5.711) | (3.209) | (2.069) | |
| FSize | 1.058 | −0.468 | 1.666 | 2.122* |
| (1.002) | (1.869) | (1.017) | (0.876) | |
| ROA | 29.31** | 44.99* | 41.43*** | 34.94*** |
| (10.81) | (18.18) | (9.633) | (7.970) | |
| FAge | 0.00527 | 0.189 | −0.266* | 0.0168 |
| (0.100) | (0.189) | (0.108) | (0.106) | |
| BSize | 0.682 | 1.010 | 0.632 | −0.276 |
| (0.577) | (1.086) | (0.581) | (0.562) | |
| Constant | 25.57 | 60.85 | −13.65 | 24.63 |
| (35.15) | (65.80) | (36.32) | (30.52) | |
| N | 132 | 132 | 132 | 132 |
| Adj. R2 | 0.317 | 0.258 | 0.279 | 0.184 |
| Year FE | Yes | Yes | Yes | Yes |
| Industry FE | Yes | Yes | Yes | Yes |
Note(s): Standard errors in parentheses; *p < 0.05, **p < 0.01, ***p < 0.001
ESGP: ESG Performance; EnvP: Environmental Performance; SocP: Social Performance; GovP: Governance Performance; CAge: CEO Age; CGender: CEO Gender; FSize: Firm Size; ROA: Return on Asset; FAge: Firm Age; BSize: Board Size
4.2.2 Method
To validate the robustness of our findings, another analysis was done by using Generalized Estimating Equations (GEE). It is an extension of generalized linear models (GLMs) created by Liang and Zeger (1986), that assumes responses are independent and follow (Dabbebi et al., 2022; Lassoed and Khancel, 2023; Petrenko et al. 2016). As displayed in Table 8, the result is consistent with those in Tables 6 and 7.
Robustness test GEE results
| Dependent variable | ESGP | EnvP | SocP | GovP |
|---|---|---|---|---|
| Narcissism | −0.368 | −1.284 | 6.521* | 0.523 |
| (1.564) | (2.760) | (2.868) | (1.062) | |
| CAge | −13.42 | −7.772 | −11.05 | −31.14** |
| (10.63) | (20.94) | (27.39) | (11.03) | |
| CGender | −6.738 | −2.203 | 37.81** | 2.044 |
| (5.936) | (7.884) | (12.55) | (2.510) | |
| FSize | 1.868 | 2.731 | 5.022 | 0.935 |
| (1.489) | (3.035) | (3.039) | (1.196) | |
| ROA | 16.11* | 19.70 | 27.62* | 15.81* |
| (8.039) | (13.41) | (13.90) | (6.700) | |
| BSize | −0.255 | −1.544* | −1.141 | −0.127 |
| (0.857) | (0.661) | (0.686) | (0.723) | |
| FAge | −0.0167 | 0.102 | −0.372 | 0.102 |
| (0.151) | (0.254) | (0.346) | (0.130) | |
| Constant | −4479.7*** | −7021.2*** | −7082.8*** | −2459.9 |
| (1262.6) | (1979.0) | (1746.0) | (1269.7) | |
| N | 132 | 132 | 132 | 132 |
| Wald χ2(p-value) | 0.0043 | 0.0001 | 0.000 | 0.000 |
| Dependent variable | ESGP | EnvP | SocP | GovP |
|---|---|---|---|---|
| Narcissism | −0.368 | −1.284 | 6.521* | 0.523 |
| (1.564) | (2.760) | (2.868) | (1.062) | |
| CAge | −13.42 | −7.772 | −11.05 | −31.14** |
| (10.63) | (20.94) | (27.39) | (11.03) | |
| CGender | −6.738 | −2.203 | 37.81** | 2.044 |
| (5.936) | (7.884) | (12.55) | (2.510) | |
| FSize | 1.868 | 2.731 | 5.022 | 0.935 |
| (1.489) | (3.035) | (3.039) | (1.196) | |
| ROA | 16.11* | 19.70 | 27.62* | 15.81* |
| (8.039) | (13.41) | (13.90) | (6.700) | |
| BSize | −0.255 | −1.544* | −1.141 | −0.127 |
| (0.857) | (0.661) | (0.686) | (0.723) | |
| FAge | −0.0167 | 0.102 | −0.372 | 0.102 |
| (0.151) | (0.254) | (0.346) | (0.130) | |
| Constant | −4479.7*** | −7021.2*** | −7082.8*** | −2459.9 |
| (1262.6) | (1979.0) | (1746.0) | (1269.7) | |
| N | 132 | 132 | 132 | 132 |
| Wald χ2(p-value) | 0.0043 | 0.0001 | 0.000 | 0.000 |
Note(s): Standard errors in parentheses; *p < 0.05, **p < 0.01, ***p < 0.001
ESGP: ESG Performance; EnvP: Environmental Performance; SocP: Social Performance; GovP: Governance Performance; CAge: CEO Age; CGender: CEO Gender; FSize: Firm Size; ROA: Return on Asset; FAge: Firm Age; BSize: Board Size
5. Conclusion
5.1 Conclusion
Within the current study, the relationship between CEO narcissism and ESG Performance is examined and three elements of ESG Performance, namely environmental, social and governance performance. The narcissism was measured using LinkedIn components for all Indonesian-listed companies from 2017 to 2022. The outcomes suggest that while a self-admiring CEO has a positive effect on social performance specifically, it does not significantly have an unfavorable effect on the overall ESG performance or its environmental and governance components. This nuanced understanding highlights the complexity of CEO narcissism in affecting ESG performance.
Many important things will happen because of the study. First, it says that the attributes of a CEO might help investors figure out the risks that come with ESG regulations. As firms put more and more weight on ESG elements in their plans, it will be crucial to know how these factors work to promote responsible leadership and long-term business practices in Indonesia and other places. Second, Indonesian public companies should earn a better ESG Score so they may do more to benefit the environment, society and good government. Overall, the data reveal that CEO narcissism has a substantial positive effect on social performance, but it does not seem to have an effect on environmental or governance performance. The results reveal that CEOs who enjoy themselves are more likely to work on socially responsible projects and this trend indicates how vital it is to grasp what makes each CEO different when selecting whether or not a company should pursue ESG projects.
5.2 Managerial implications
This study's results make a substantial addition to the current body of knowledge regarding CEO narcissism and its influence on ESG performance. This research plays a role in the understanding of the UET by establishing that CEO personality traits, particularly narcissism, play an essential role in shaping corporate strategies related to ESG initiatives. This study highlights the nuanced relationship between CEO characteristics and ESG performance within the Indonesian context. Additionally, the benefit of LinkedIn profiles as a measurement tool for assessing CEO narcissism offers a more objective approach than traditional self-reported measures.
From a practical point of view, knowing how CEO narcissism affects ESG performance can help investors and other stakeholders figure out what dangers and possibilities their investments might bring. Companies with narcissistic CEOs may be more likely to take part in socially responsible projects and how important it is to look at both environmental and governance issues in order to have a balanced approach to ESG practices, no matter what the CEO is like. For business people, making people aware of how different CEO traits affect strategic decisions can help businesses become more sustainable.
5.3 Limitations and future research
First, the limited number of samples due to using LinkedIn profiles to measure CEO narcissism introduces significant limitations. Out of the 1,004 potential observations, the authors could take only 132 observations. LinkedIn Accounts found to be active are limited, and ESG scores are unavailable on Bloomberg. Second, the framework adopted in this study is simple and establishes a boundary condition in revealing the effects of CEO narcissism on ESG performance. This suggests that future research should explore including either moderating or mediating variables, in explaining about a not so straightforward relationship. Third, the study's empirical framework did not show a strong link between CEO narcissism and overall ESG performance in Indonesia. There may be non-linear correlations that were not taken into account in the model. Future research could build on this study by looking into psychological qualities such as power and overconfidence that might affect ESG performance.
Note
Note that we also separated the data based on the median score of CEO narcissism and median score of ESG Performance, then run the panel regression again. However, the impact of narcissism remained insignificant.

