When Leadership Meets Worldwide Governance: The Role of CEO Characteristics in Environmental, Social, and Governance Performance

Abstract

This study investigates how CEO demographic characteristics, including age, gender, and nationality, and cognitive characteristics, including tenure, education, and multiple directorships, influence firms’ ESG performance, with a focus on the moderating role of Worldwide Governance Indicators (WGIs). Using a regime/smooth transition approach with panel data from STOXX Europe 600 firms spanning the years 1999 and 2023, the results show that demographic characteristics exert a more consistent effect than cognitive effects in the full sample and in non-sensitive industries. In sensitive industries, however, both demographic and cognitive CEO traits significantly affect ESG performance. Older and female CEOs enhance ESG performance under strong worldwide governance indicators (WGIs) in the full sample and sensitive industries, whereas foreign CEOs perform better under weaker worldwide governance conditions. In non-sensitive industries, the patterns for female and foreign CEOs are reversed. Cognitive traits such as tenure and multiple directorships show limited influence, while higher educational qualifications improve ESG outcomes under weak governance but reduce them under strong governance across all samples. Overall, the findings highlight the importance of aligning CEO characteristics with the institutional governance environment to enhance corporate sustainability performance. This study contributes by examining how CEO demographic and cognitive characteristics affect ESG performance under varying country-level governance conditions. It also highlights sectoral differences between sensitive and non-sensitive industries and, by using a nonlinear (PSTR) approach, uncovers regime-dependent effects with implications for governance-aware CEO selection and ESG strategy. This study extends upper echelons and institutional theories by showing that the effect of CEO characteristics on ESG performance depends on country governance quality, offering insights for boards and policymakers seeking to align leadership selection with governance contexts to strengthen sustainability and accountability.

1. Introduction

In an era marked by mounting political, environmental and social challenges, corporate sustainability has evolved from a peripheral concern to a strategic imperative. Stakeholders, including investors, regulators, and civil society, demand greater transparency and accountability from firms, particularly regarding environmental, social, and governance (ESG) performance. As ESG practices become central to long-term value creation [1], understanding the determinants of ESG engagement is more critical than ever.

Over the past few decades, EU regulations have increasingly emphasized the importance of sustainability, significantly influencing CEO behavior and corporate governance practices in response to heightened accountability for environmental, social, and governance (ESG) disclosures. Prior regulations, such as the Non-Financial Reporting Directive (NFRD) of 2014, laid the groundwork for current requirements, emphasizing the need for companies to disclose non-financial information related to their operations [2,3]. More recently, the Corporate Sustainability Reporting Directive (CSRD) requires comprehensive ESG reporting that is standardized, audited, and aimed at enhancing transparency [4]. While they strengthen stakeholder trust, they can also pose implementation challenges [5].

In line with this institutional emphasis, ref. [6] found that CEO social capital contributes more positively to ESG performance in state-owned enterprises, which similarly operate under stronger institutional control and public scrutiny. Likewise, the Sustainable Finance Disclosure Regulation (SFDR) reinforces these objectives by improving the quality of ESG information, supporting sustainable investment [7], and fostering accountability across value chains [8].

Amid this increasing scrutiny of ESG disclosure, the CEO has emerged as a pivotal force in shaping a company’s ESG performance by directing strategic vision, allocating resources, and coordinating reporting activities [9,10]. However, the impact of CEO characteristics on ESG outcomes remains understudied and inconsistent [11]. While certain characteristics such as female leadership and higher education have been linked with improved ESG disclosure, other characteristics like age and tenure show mixed or insignificant effects [9,12,13,14,15]. Additionally, a CEO’s foreign experience may enhance ESG performance in well-governed settings [16,17], though counterevidence also exists [18,19]. Few studies have systematically explored how national governance quality measured through indicators like rule of law, voice and accountability, control of corruption, and regulatory quality interacts with CEO characteristics to influence ESG outcomes [19,20].

Developing economies, especially those characterized by weak governance, remain underrepresented in the literature, and the moderating role of institutional quality has received limited attention [11]. Although this study focuses on European firms, it captures substantial variation in institutional quality, particularly across Eastern Europe, where governance tends to be comparatively weaker. Rather than treating Europe as institutionally homogeneous, this study exploits cross-country variation in Worldwide Governance Indicators (WGIs) within the sample, enabling analysis of both relatively strong and weaker governance environments. This contextual gap limits the generalizability of existing theoretical frameworks. Even within Europe, institutional contexts vary significantly, shaping how CEO demographic and cognitive characteristics translate into ESG actions. As ESG becomes increasingly central to corporate strategy and reporting, understanding this interaction is essential because institutional environments influence both CEOs’ strategic flexibility and the effects of their characteristics on sustainability outcomes. Addressing this gap enables more effective, location-sensitive policies and leadership strategies; for example, by incentivizing ESG compliance in well-governed countries, or strengthening institutions before introducing CEO-focused initiatives [21]. Drawing on upper echelons theory [22], institutional theory [23,24], and stakeholder theory [25], this paper examines how CEO characteristics influence ESG performance across different governance levels, with particular attention given to the moderating role of Worldwide Governance Indicators. This approach captures the country-level institutional conditions under which CEOs operate, including accountability, political stability, government effectiveness, regulatory quality, rule of law, and control of corruption [26]. Strong national governance promotes transparency, accountability, and ethical behavior while creating a more predictable business environment that provides both constraints and incentives for CEOs to pursue ESG objectives. Studies that do not include WGIs as moderating variables implicitly assume that CEOs operate under the same institutional context everywhere. Evidence suggests that external institutional factors such as WGIs alter the relationship between cognitive characteristics and ESG performance [20]. In addition, focusing on the moderating role of WGIs helps explain inconsistencies in findings related to the CEO-ESG relationship. Moreover, the sample is categorized into sensitive and non-sensitive industries. Sensitive industries include basic materials, energy, industrials, utilities, consumer discretionary, consumer staples, financials, and telecommunications. In contrast, non-sensitive industries comprise real estate, technology, and healthcare sectors [27,28].

This study makes four key contributions to the sustainability literature. First, it advances strategic perspectives on sustainability by empirically examining how both cognitive and demographic CEO characteristics influence firms’ ESG performance, using regression analyses that capture the direct effects of executive traits on sustainability outcomes. Second, it adopts a multi-level analytical lens by integrating macro-level WGIs as moderators of micro-level CEO influences, employing a smooth transition regression approach to capture governance-dependent strategic behaviors. Third, by distinguishing between sensitive and non-sensitive industries, this study captures sectoral heterogeneity through sub-sample analyses and formal statistical tests, demonstrating how industry-specific pressures shape sustainability strategies and stakeholder expectations. Finally, drawing on cross-national panel data, this study enhances the generalizability of its findings by validating results across diverse institutional contexts and conducting robustness checks to ensure consistency.

This paper is organized as follows: Section 2 outlines the theoretical framework and hypothesis development. Section 3 details the research method, including sampling procedures, data collection, and variable measurement. Section 4 presents the empirical findings and discussion. Finally, Section 5 offers the conclusion, contributions, and practical implications.

2. Theoretical Background and Hypothesis Development

Firms’ ESG strategies result from a mix of CEO characteristics, organizational processes, and institutional pressures. A multi-theoretical framework drawing on upper echelons, stakeholder, and institutional theories [11] better explains how CEOs shape sustainability performance.

2.1. Theoretical Background

This study adopts a multi-theoretical perspective by integrating upper echelons, stakeholder, and institutional theories to explain how CEO characteristics influence ESG performance across varying governance contexts. While each theory offers a distinct lens, their integration provides a more comprehensive understanding of how executive attributes shape strategic sustainability outcomes under different institutional conditions.

Upper echelons theory [22] posits that organizational outcomes reflect top executives’ cognitive biases, values, and experiences. Both demographic and cognitive significantly shape ESG-related decisions [29,30,31,32,33]. Studies show that gender and education [34,35], as well as age and education [36], influence environmental performance. Ref. [37] finds that CEOs from disadvantaged backgrounds drive stronger ESG results while attributes like multiple directorships and tenure reflect institutional exposure and stakeholder responsiveness, supporting ESG initiatives [31,38,39]. Ref. [40] confirms the link between CEO characteristics and firm strategy. Cognitive and demographic characteristics are thus tied to sustainability engagement [41,42]. The theory underpins this study by emphasizing how CEO characteristics interact with governance quality [43,44] to shape ESG commitment. However, while upper echelons theory predicts that CEO traits influence strategic outcomes, it does not specify a uniform direction for this influence, as the same characteristic may lead to different ESG outcomes depending on contextual conditions.

Stakeholder theory, introduced by [25], stresses the importance of addressing the interests of all stakeholders, not just shareholders, including employees, customers, suppliers, and communities [45]. It is crucial to ESG performance, which inherently reflects how firms manage responsibilities across ESG performance [6,46,47]. Ref. [48] outlines three dimensions of stakeholder theory—normative, instrumental, and descriptive—which, respectively, focus on ethical duties, performance outcomes, and actual behavior [49]. Recent work emphasizes relationship-based engagement to align mutual sustainability goals and resolve trade-offs between stakeholder groups [50,51,52]. In terms of CEO characteristics, stakeholder theory helps explain how individual demographic and cognitive attributes shape executive sensitivity to stakeholder interests and societal expectations. For instance, gender, age, and education may affect how CEOs assess stakeholder needs and broader social concerns [34,53,54]. CEOs in governance systems that prioritize stakeholder responsiveness are more likely to reflect their attributes in ESG decision making [55,56]. Accordingly, the effect of CEO characteristics on ESG performance may vary depending on how these attributes shape responsiveness to competing stakeholder demands, which can differ across governance environments.

Institutional theory posits that firms adopt ESG practices to align with societal norms, regulations, and expectations, gaining legitimacy and minimizing reputational or regulatory risks [23,24,57]. Ref. [58] argues that such behavior intensifies when regulatory frameworks, stakeholder oversight, and norms of corporate responsibility are strong. CEOs play a critical role in interpreting and responding to institutional demands; ref. [20] shows that cognitive characteristics like education influence ESG performance, with their impact moderated by institutional environments such as local fiscal policies and governance strength. Modern perspectives within institutional theory emphasize strategic agency over passive conformity [59,60], viewing CEO attributes—education, tenure, and international experience—as lenses shaping organizational responses to institutional demands [31,38]. For example, the positive ESG effects of female CEOs are accentuated in countries with legal gender equality and diminished in those with societal bias [61]. Overall, institutional theory offers a valuable framework for analyzing how governance conditions (e.g., WGIs) moderate the link between CEO characteristics and ESG performance [27,62,63,64]. Importantly, this perspective implies that the same CEO characteristic may generate different ESG outcomes depending on the strength and configuration of institutional environments, thereby introducing variability in the CEO–ESG relationship across WGI contexts.

2.2. Hypothesis Development

2.2.1. Direct Effects of CEO Demographic Characteristics on ESG Performance

CEO demographic characteristics influence ESG performance, with prior studies exploring characteristics such as marital status, religion, political affiliation, reputation and power [14,37,65,66,67,68,69,70,71,72]. This study focuses on age, gender, and nationality, aligning with upper echelons and stakeholder theories [22,34,73], and addresses gaps in integrated ESG research [11,74]. These characteristics capture core social and cultural dimensions that shape how CEOs perceive and respond to ESG issues—age reflects generational outlooks, gender relates to stakeholder engagement styles, and nationality embodies institutional norms—making them theoretically grounded and practically relevant choices for this study.

Age is a critical factor influencing ESG outcomes. Younger CEOs often support progressive ESG efforts and innovation but may prioritize short-term financial goals, limiting environmental initiatives [41]. Older CEOs, by contrast, tend to enhance environmental performance due to a long-term legacy focus [75]. These studies align with upper echelons theory, which emphasizes the role of age and career stage in shaping strategic orientation [36]. Evidence from European firms confirms older CEOs’ stronger alignment with corporate social responsibility (CSR) goals [76]. However, the impact of age may vary with firm risk or sector sensitivity [11,77], and may not always translate into proactive ESG engagement under different contextual conditions.

Gender plays a key role in ESG outcomes. Female CEOs are generally linked to stronger ESG performance, especially in the social pillar [9]. This effect is amplified by strong governance and a critical mass of female directors [69,78,79,80], in line with stakeholder theory, which connects gender diversity with greater stakeholder responsiveness [34,54]. Institutional theory also emphasizes the importance of gender equality laws in shaping the effectiveness of such characteristics. However, some empirical findings suggest that when institutional support is weak, gender effects may be limited or negative [6,61], reflecting variation in how gender-related leadership attributes translate into ESG outcomes across governance contexts.

Nationality is a key demographic characteristic, though its direct impact on ESG remains limited and inconclusive. Some research suggests that CEO nationality affects ESG priorities and stakeholder relations, especially in firms with strong governance frameworks [81]. This aligns with institutional theory, where national contexts shape how demographic characteristics influence ESG [31,82]. Cross-national comparisons reveal that a CEO’s origin can explain differences in ESG and financial performance [83], though this influence may be muted in globalized industries [84]. Still, the interplay between nationality and other CEO attributes remains underexplored [11] and may depend on the degree of alignment between a CEO’s background and the institutional environment.

The influence of CEO demographic characteristics on ESG performance is shaped by the surrounding institutional environment. Strong governance systems can enable these characteristics to translate more effectively into ESG engagement. This relationship is examined in the following section through the moderating role of WGIs.

2.2.2. Moderating Role of WGIs for the Relationship Between CEO Demographic Characteristics and ESG

While CEO demographic characteristics have a direct influence on ESG outcomes, their impact is not uniform across institutional contexts. Institutional theory highlights that firms adopt ESG practices to align with dominant institutional norms and legitimacy concerns [23,24], suggesting that the effectiveness of CEO characteristics is contingent on the strength of the surrounding governance environment.

Empirical evidence shows that improvements in WGIs enhance countries’ ability to implement sustainability strategies, boosting environmental and social performance where governance is strong [85,86]. In such contexts, CEOs with pro-social demographics such as younger age, female gender, or diverse nationality are better positioned to advance ESG efforts, backed by supportive regulations and accountability frameworks [69,75]. Cross-national research indicates that female CEOs drive stronger CSR performance when legal gender-equality protections exist even in socially biased settings [61]. These results highlight institutional theory’s interactive view [20,59], where governance conditions shape how CEO attributes translate into ESG outcomes. CEO characteristics translate into ESG outcomes only within favorable institutional environments. In contrast, institutional pressures can weaken CEOs’ ESG impact when governance is frail—e.g., duality, concentrated ownership, fiscal/regulatory gaps, or political ties—sometimes reversing positive leadership effects [6,20,67,68,72,87]. Moreover, variation in institutional environments means CEO characteristics like age or nationality influence ESG outcomes only within context, depending on national governance quality, cultural expectations, or legal safeguards. Strong governance (WGI) amplifies their positive ESG effects, while weak governance constrains or distorts them.

Based on the previous discussion, the following hypotheses are derived:

H1. 

Demographic CEO characteristics have different effects on ESG under different levels of Worldwide Governance Indicators as moderators.

H1a. 

CEO age has different effects on ESG under different levels of Worldwide Governance Indicators as moderators.

H1b. 

Female CEOs have different effects on ESG under different levels of Worldwide Governance Indicators as moderators.

H1c. 

Foreign CEOs have different effects on ESG under different levels of Worldwide Governance Indicators as moderators.

2.2.3. Direct Effects of CEO Cognitive Characteristics on ESG Performance

Various studies link CEO cognitive characteristics—such as functional experience and professional background, motivational orientation, dynamic capabilities, reputation, social capital, and career horizon—to ESG outcomes [6,41,66,80,88,89,90,91]. This study focuses specifically on CEO educational qualifications, multiple directorships, and tenure. Drawing on upper echelons theory [22], these characteristics represent deep-level cognitive attributes that shape how CEOs interpret sustainability demands, perceive risk, and exercise moral reasoning—factors especially salient under conditions of high discretion, where individual decision-making plays a critical role. These characteristics shape how CEOs interpret pressures, manage risks, and allocate ESG resources [20,34,80,92]. Despite mixed empirical findings, evidence supports the significance of cognitive characteristics in ESG performance and calls for integrated, multi-theoretical analysis [11]. By examining varied European institutional contexts, our study helps mitigate geographic bias.

CEO educational qualifications are frequently studied as a key cognitive trait affecting ESG performance. Research consistently shows that CEOs with advanced degrees—especially in business, engineering, or science—tend to champion ESG initiatives and CSR investments [39,76,93]. Higher qualifications in education can enhance cognitive capacity, strategic vision, ethical orientation, and awareness of social and environmental issues, leading to stronger CSR performance and environmental disclosures [34,41,83,89,92]. These findings align with upper echelons theory, which links formal training to executive values and priorities [20,34,36], further highlighting education’s indirect environmental benefits. However, ref. [94] reports no significant results, indicating that education alone may not fully drive ESG outcomes.

CEOs holding multiple board directorships can bring valuable cognitive diversity and external exposure that inform ESG decision-making. Empirical evidence suggests that “busy” CEOs with board memberships in other firms enhance sustainability initiatives through knowledge spillover, benchmarking, legitimacy signaling, and stakeholder responsiveness [34,75,77]. These dynamics resonate with stakeholder theory, highlighting how external board affiliations expose CEOs to broader stakeholder norms and ESG expectations [72]. However, excessive directorships may dilute managerial attention and oversight, potentially weakening ESG responsiveness [95], suggesting that the benefits of external exposure may diminish beyond certain thresholds.

CEO tenure is a cognitive attribute linked to ESG outcomes, though evidence remains mixed. Longer-tenured CEOs accumulate firm-specific knowledge and political capital, potentially leading to strategic rigidity or prioritizing financial stability over ESG experimentation [35,95]. However, tenure can also empower long-term sustainability agendas [75,96], increase confidence in reputational ESG risks [39], and foster stakeholder relations that support ESG initiatives [83]. Family firm studies show a positive link between tenure and ESG reporting, reinforcing long-term orientation benefits [73]. Despite the limited research on tenure’s governance interactions [11], evidence increasingly suggests that accumulated knowledge and familiarity enable informed, sustained ESG contributions. Recent studies highlight that directors with prior CEO experience can enhance firms’ environmental innovation by aligning ESG strategies with industry norms and regulatory contexts [97,98].

Conversely, evidence from U.S. firms indicates that co-opted independent directors tend to weaken environmental performance due to reduced board monitoring [99]. This negative effect is amplified by CEO power but mitigated under strong corporate governance, underscoring how governance quality shapes leadership effects on ESG outcomes. Collectively, these findings reinforce the need to consider institutional moderation in sustainability performance. The effect of CEO cognitive attributes on ESG outcomes also depends on institutional context. High-quality governance frameworks can strengthen the influence of characteristics such as education, tenure, and external exposure, as discussed next through the moderating role of WGIs.

2.2.4. The Moderating Role of WGIs for the Relationship Between CEO Cognitive Characteristics and ESG

The institutional environment critically shapes how CEO cognition affects ESG performance. Institutional theory suggests that regulations, governance frameworks, and societal norms influence corporate behavior by defining the boundaries of legitimacy and strategic action [23,24,31], thereby conditioning how cognitive CEO attributes translate into ESG outcomes. WGIs signal to CEOs how to interpret ESG imperatives, amplifying or constraining cognitive characteristics. Yet [11] notes that governance’s moderating influence on CEO cognition–ESG links remains underexplored, underscoring the need for further research.

In strong governance environments, institutional pressures promote transparency, stakeholder engagement, and long-term strategic thinking, motivating CEOs to apply their cognitive strengths toward ESG alignment [20,49]. Robust legal and regulatory frameworks heighten ESG visibility and reduce uncertainty around sustainability investments [75], enabling cognitively capable CEOs—those with advanced education, long tenure, or global experience—to capitalize on these contexts. Enhanced governance also fosters better monitoring and stakeholder collaboration, reinforcing ethical leadership behaviors supportive of ESG goals [39,96]. Research further indicates that CEO power, when coupled with strong governance, strengthens both ESG and financial performance [72], while other studies demonstrate that governance structures and fiscal policies influence how executive education and incentives impact ESG outcomes, underscoring institutional theory’s moderating role [20]. Conversely, where governance is weak or ineffective, opportunistic behavior and short-term pressures can diminish or even reverse the positive effects of CEO cognitive characteristics on ESG [85,95], highlighting the critical role of cross-national institutional variation.

Based on the preceding discussion, the following hypotheses are proposed:

H2. 

Cognitive CEO characteristics have different effects on ESG under different levels of Worldwide Governance Indicators as moderators.

H2a. 

CEOs with multiple directorships have different effects on ESG under different levels of Worldwide Governance Indicators as moderators.

H2b. 

CEOs with higher qualifications have different effects on ESG under different levels of Worldwide Governance Indicators as moderators.

H2c. 

CEO tenure has different effects on ESG under different levels of Worldwide Governance Indicators as moderators.

Based on the preceding discussion, the study’s conceptual framework is illustrated in Figure 1.

3. Research Methodology

3.1. Sampling and Data Collection

We have a panel of yearly data for 1361 firms for the period 1999–2023. After removing firms with missing data on ESG, CEO, or governance variables, the usable sample was reduced to 743 firms. The number of countries is 41. The empirical analysis focuses on firms listed in the STOXX Europe 600 index, which comprises 600 companies, over the period from 1999 to 2023. The initial sample includes 1361 firms, corresponding to the availability of both ESG metrics and WGIs. Regime parameter estimates (μ, σ, θ) and transition diagnostics are provided in Appendix A.

The sample spans multiple European countries exhibiting substantial variation in Worldwide Governance Indicators (WGIs), capturing both relatively strong and weaker institutional environments within the region.

The dataset is structured around three primary components: the dependent variable is ESG performance, sourced from the Refinitiv database; the independent variables consist of various CEO characteristics, extracted from BoardEx; and the moderating variables are the WGIs, obtained from the World Bank’s WGI dataset (as detailed in

Table 1. Variable measurement.

Variables	Abbreviations	Measurement	Source
Panel A: dependent variables:
Sustainability performance	ESG	ESG index	Refinitiv DataStream
Panel B: independent variables:
CEO Demographic Characteristics
Age	Age	Years from birth	Board EX
Gender	Gen	A dummy variable indicating CEO gender (assigned a value of 1 if the CEO is female, and 0 if male).	Board EX
Nationality	Nat	A dummy variable indicating CEO nationality (assigned a value of 1 if the CEO is foreign, and 0 if domestic).	Board EX
CEO Cognitive Characteristics
Tenure	Ten	Number of years since the CEO has been in his/her current position.	Board EX
Multiple directorship	Multip	Number of quoted boards the director has served on over their career.	Board EX
Education	Edu	Number of qualifications CEO has obtained.	Board EX
Panel C: moderating variables (Worldwide Governance Indicators):
Political stability	PolSt	Measures the perceived likelihood of political unrest, violence, or terrorism within a country.	World bank
Rule of law	RuLaw	Measures perceived confidence in and adherence to societal rules, especially regarding contract enforcement and property rights.	World bank
Voice and accountability	VoAc	Measures perceptions of citizens’ ability to choose their government and enjoy freedoms of expression, association, and access to free media.	World bank
Control of corruption	ConCo	Measures perceptions of the extent to which public power is used for private gain, including petty and grand corruption and state capture by elites or private interests.	World bank
Government effectiveness	GoEf	Measures perceptions of public service quality, civil service independence, policy formulation and implementation, and the government’s credibility in committing to policies.	World bank
Regulatory quality	RegQu	Measures perceptions of the government’s ability to design and implement effective policies and regulations that support private sector development.	World bank
Panel D: control variables:
Profitability	Prof	The company’s Return on Assets (ROA) ratio.	Refinitiv DataStream
Firm size	Fsize	The natural logarithm of the company’s total assets.	Refinitiv DataStream
Leverage	Lev	The ratio of total debt to total assets.	Refinitiv DataStream

).

3.2. Variable Measurement

Table 1 presents the variable measurements utilized in this study. Panel A outlines the ESG score. Panel B details the CEO demographic characteristics and cognitive attributes. Panel C introduces the WGIs. Lastly, Panel D lists the control variables (profitability, leverage and firm size) (as detailed in Table 1).

3.3. Research Model

3.3.1. The Model

We use a smooth transition model to capture the effect of the regressors on the dependent variable. That is, we use the following model:

$$y_{\mathrm{i},\mathrm{t}}={\beta }_0+p_{\mathrm{j},\mathrm{t}-1}({\beta }^{\prime }{X}_{\mathrm{i},\mathrm{t}-1}+{\alpha }^{\prime }{C}_{\mathrm{i},\mathrm{t}-1})+(1-p_{\mathrm{j},\mathrm{t}-1})({\gamma }^{\prime }{X}_{\mathrm{i},\mathrm{t}-1}+{\delta }^{\prime }{C}_{\mathrm{i},\mathrm{t}-1})+{\mathrm{ƒ}}_{\mathrm{i}}+{\xi }_{\mathrm{i},\mathrm{t}}$$

(1)

where $\beta$ and $\alpha$ are vectors of the sensitivities of the dependent variable $y_{\mathrm{i},\mathrm{t}}$ (ESG indicator) to the vector of explanatory variables $X_{\mathrm{i},\mathrm{t}}$ and the vector of control $C_{\mathrm{i},\mathrm{t}}$ variables successively in regime one, and similarly, $\gamma$ and $\delta$ are vectors of the sensitivities of the dependent variable to the vector of explanatory variables $X_{\mathrm{i},\mathrm{t}}$ and the vector of control $C_{\mathrm{i},\mathrm{t}}$ variables successively in regime two. The component ${\xi }_{\mathrm{i},\mathrm{t}}$ is the unexplained part of the error term and $p_{\mathrm{j},\mathrm{t}}$ is the probability of country j being in a bad institutional regime (regime one when the WGI is below its mode $\mu$) as defined below. Similarly, (1$-p_{\mathrm{j},\mathrm{t}})$ is the probability of country j being in a bad institutional regime (regime two when the WGI is above its mode $\mu$) as defined below. The term ${{\displaystyle \mathrm{ƒ}}}_{\mathrm{i}}$ captures the unobserved heterogeneity in the model (fixed effect/random effect).

To model the institutional moderator’s distribution $m_{\mathrm{j},\mathrm{t}}$, we use the rescaled split (two piece) normal distribution of [100].

$$\mathrm{f}\left(m_{\mathrm{j},\mathrm{t}}\right)={\displaystyle \frac{1}{\sqrt{2\mathsf{\pi }}\mathsf{\sigma }}}\left\{\begin{array}{c}\begin{array}{cc}\mathrm{e}\mathrm{x}\mathrm{p}\left(-{\displaystyle \frac{1}{2(1-\mathsf{\theta }){\sigma }^2}}{(m_{\mathrm{j},\mathrm{t}}-\mu )}^2\right)& \mathrm{i}\mathrm{f} m_{\mathrm{j},\mathrm{t}}\le \mathsf{\mu }\end{array}\\ \\ \begin{array}{cc}\mathrm{e}\mathrm{x}\mathrm{p}\left(-{\displaystyle \frac{1}{2(1+\mathsf{\theta }){\sigma }^2}}{(m_{\mathrm{j},\mathrm{t}}-\mu )}^2\right)& \mathrm{O}\mathrm{t}\mathrm{h}\mathrm{e}\mathrm{r}\mathrm{w}\mathrm{i}\mathrm{s}\mathrm{e}\end{array}\end{array}\right.$$

(2)

where ${\mathrm{m}}_{\mathrm{t}}$ is the institutional moderator, $\mu$ is its mode (the most frequent value of the institutional factor ${\mathrm{m}}_{\mathrm{t}}$), $\mathsf{\sigma }$ is the dispersion parameter and $\mathsf{\theta }$ is the skewness parameter ($\mathsf{\sigma }>0$ and $-1<\mathsf{\theta }<1$). If θ < 0, then the distribution is asymmetric to the left, and if θ > 0 the distribution is skewed to the right. If θ = 0, the distribution becomes symmetric and the two regimes collapse into one regime. That is, we have

$$\mathrm{f}\left(m_{\mathrm{j},\mathrm{t}}\right)={\displaystyle \frac{1}{\sqrt{2\mathsf{\pi }}\mathsf{\sigma }}}\mathrm{e}\mathrm{x}\mathrm{p}\left(-{\displaystyle \frac{1}{2{\sigma }^2}}{(m_{\mathrm{j},\mathrm{t}}-\mu )}^2\right),$$

(3)

where (3) is just the normal probability distribution function.

The distribution in (2) can be rewritten as a mixture of two half normal distributions (truncated normal distributions):

$$\mathrm{f}\left(m_{\mathrm{t}}\right)=\mathsf{\tau }{\displaystyle \frac{2\mathrm{e}\mathrm{x}\mathrm{p}\left(-\frac{1}{2(1-\mathsf{\theta }){\sigma }^2}{(m_{\mathrm{j},\mathrm{t}}-\mu )}^2\right)}{\sqrt{2\mathsf{\pi }}\sqrt{1-\mathsf{\theta }}\sigma }}{I}_{\mathrm{t}}+(1-\mathsf{\tau }){\displaystyle \frac{2\mathrm{e}\mathrm{x}\mathrm{p}\left(-\frac{1}{2(1+\mathsf{\theta }){\sigma }^2}{(m_{\mathrm{j},\mathrm{t}}-\mu )}^2\right)}{\sqrt{2\mathsf{\pi }}\sqrt{1+\mathsf{\theta }}\sigma }}{(1-I}_{\mathrm{t}})$$

(4)

where $I_{\mathrm{t}}$ is equal to 1 if ${\mathrm{m}}_{\mathrm{t}}\le \mathsf{\mu }$ and zero otherwise and, $\mathsf{\tau }$ is the unconditional probability of being in regime one, given by $\mathsf{\tau }={\displaystyle \frac{1-\mathsf{\theta }}{2}}.$

The distribution in Equation (2) is more flexible. Actually, when σ is very high, the countries are more dispersed in terms of institutional performance; when it is low, the countries are more concentrated in the middle of the distribution, meaning that most of them have the same performance. The parameter θ captures the degree of asymmetry of the distribution. That is, when θ is negative, most countries have a performance below the mode, and when it is positive, most of the countries have a performance higher than the mode. In addition, if θ = 0, then 50% of the countries have a performance below the mode and 50% are higher than the mode.

Figure 2 illustrates these facts. For illustrative purposes and to prevent loss of generality, we fix the mode at 0 and we vary the dispersion parameter σ and the asymmetry parameter θ. Changing the values will not change our analysis since it will only shift the figure to the left (θ < 0) or to the right (θ > 0) without altering the shape of the distribution.

Comparing plots (a) and (b) of the figure, we notice that the distribution is symmetric, with higher dispersion in plot (b) (θ = 0). Assigning moderate negative (positive) values to θ (plots 3 and 4 in the figure) helps capture the moderate concentration of the countries with low (high) institutional performance. On the other hand, when θ is close to −1 (extreme left asymmetry), most of the countries underperform relative to the modal country (plot 5 in the figure). Furthermore, when θ is close to 1 (extreme right asymmetry), most of the countries overperform relative to the modal country (plot 6 in the figure).

We define the conditional probability of a country being in a low-performing regime as follows:

$$p_{\mathrm{j},\mathrm{t}}={\displaystyle \frac{\mathrm{e}\mathrm{x}\mathrm{p}\left(-\frac{1}{2(1-\mathsf{\theta }){\sigma }^2}{(m_{\mathrm{j},\mathrm{t}}-\mu )}^2\right)}{\mathrm{e}\mathrm{x}\mathrm{p}\left(-\frac{1}{2(1-\mathsf{\theta }){\sigma }^2}{(m_{\mathrm{j},\mathrm{t}}-\mu )}^2\right)+\mathrm{e}\mathrm{x}\mathrm{p}\left(-\frac{1}{2(1+\mathsf{\theta }){\sigma }^2}{(m_{\mathrm{j},\mathrm{t}}-\mu )}^2\right)}}$$

(5)

Equation (5) states that when $p_{\mathrm{j},\mathrm{t}}$ is equal to 1 in the limit (not significantly different from 1), then country j is in a low institutional performance regime and when $p_{\mathrm{j},\mathrm{t}}$ is equal to 0 in the limit (not significantly different from 0), then the country is in a high institutional performance regime. Moreover, when ${0<p}_{\mathrm{j},\mathrm{t}}$< 1, then the country is in a transition period between the two regimes, where $p_{\mathrm{j},\mathrm{t}}$> 0.5 means that the country is more likely to transit toward a bad regime and vice versa when $p_{\mathrm{j},\mathrm{t}}$< 0.5.

We also notice from Equation (5) that the probability of being in a bad regime varies with respect to the cross-sectional level (country level) and time dimension, allowing our model to capture bi-dimensional endogenous macroeconomic heterogeneity in addition to unobserved exogenous firm/time-specific heterogeneity characteristic of random/fixed effects.

3.3.2. Estimation Strategy and Hypothesis Testing

We use a two-step estimation strategy in our model. In the first step, we estimate the parameters ($\mu$, $\mathsf{\theta }$ and $\mathsf{\sigma }$) in (4) by using a maximum likelihood estimator; that is, we solve the following problem:

$$\left(\mu ,\mathsf{\theta },\mathsf{\sigma }\right)=argmax\left[\sum _{t=2}^{T}ln\left(\mathsf{\tau }{\displaystyle \frac{2\mathrm{e}\mathrm{x}\mathrm{p}\left(-\frac{1}{2(1-\mathsf{\theta }){\sigma }^2}{(m_{\mathrm{j},\mathrm{t}}-\mu )}^2\right)}{\sqrt{2\mathsf{\pi }}\sqrt{1-\mathsf{\theta }}\sigma }}{I}_{\mathrm{t}}+(1-\mathsf{\tau }){\displaystyle \frac{2\mathrm{e}\mathrm{x}\mathrm{p}\left(-\frac{1}{2(1+\mathsf{\theta }){\sigma }^2}{(m_{\mathrm{j},\mathrm{t}}-\mu )}^2\right)}{\sqrt{2\mathsf{\pi }}\sqrt{1+\mathsf{\theta }}\sigma }}{(1-I}_{\mathrm{t}})\right)\right]$$

Then, we compute the conditional probability in (5) using the estimated values of $\mu ,\mathsf{\theta },\mathrm{a}\mathrm{n}\mathrm{d} \mathsf{\sigma }$. In the second step, we use the estimate of $p_{\mathrm{j},\mathrm{t}}$ and estimate the parameters of the main model in (1). The estimators of our model are consistent under standard conditions, and the standard hypothesis testing procedure is valid. For joint hypothesis testing of block parameters, we used the Wald test for joint significance.

4. Empirical Results and Discussion

4.1. Descriptive Statistics

Table 2. Descriptive analysis.

	Mean	Median	Maximum	Minimum	Std. Dev.
ESG	46.133	46.020	95.100	1.230	18.332
Ten	4.856	3.600	31.900	0.000	4.544
Multip	2.371	2.000	17.000	1.000	1.879
Age	53.008	53.000	76.000	31.000	6.276
Edu	1.846	2.000	8.000	0.000	1.145
Gen	0.042	0.000	1.000	0.000	0.201
Nat	0.462	0.000	1.000	0.000	0.499
PolSt	61.906	61.058	100.000	0.000	13.026
RuLaw	89.371	92.381	100.000	0.000	12.230
VoAc	89.635	92.019	100.000	0.000	11.697
ConCo	89.849	93.333	100.000	0.000	12.722
GoEf	88.632	90.909	100.000	0.000	10.373
RegQu	91.584	94.737	99.524	0.000	11.176
Prof	6.159	5.520	269.110	−138.230	14.501
Fsize	14.686	14.462	22.164	7.428	1.947
Lev	26.083	24.240	167.240	0.000	20.160

This table shows the descriptive analysis for the variables: ESG (sustainability performance), Ten (CEO tenure), Multip (CEO multiple directorship), Age (CEO age), Edu (CEO education), Gen (CEO gender), Nat (CEO nationality), PolSt (political stability), RuLaw (rule of law), VoAc (voice and accountability), ConCo (control of corruption), GoEf (government effectiveness), RegQu (regulatory quality), Prof (profitability), Fsize (firm size), and Lev (leverage).

presents descriptive statistics for the variables used in the analysis. The average ESG score is 46.13. Regarding CEO demographics, 4.2% of the CEOs are female, 46.2% are foreign directors, and the average age is 53 years. Examining cognitive CEO characteristics, the average CEO tenure is 4.8 years; on average, CEOs hold 1.84 qualifications, and there are two CEOs with multiple directorships per firm.

Ultimately, this study utilizes the WGIs as moderator variables. Among these, political stability exhibits the lowest average percentile rank at approximately 61.9%, indicating relative instability compared to other governance dimensions. Conversely, regulatory quality demonstrates the highest average percentile rank at about 91.5%, reflecting strong perceptions of the government’s ability to formulate and implement sound policies and regulations. The remaining WGI dimensions—rule of law, voice and accountability, control of corruption, and government effectiveness—each show an average percentile rank of approximately 89%, suggesting consistent performance across these areas.

4.2. Correlation Analysis

Table 3. Correlation Analysis.

	Ten	Multip	Age	Edu	Gen	Nat	Prof	Fsize	Lev
Ten	1
Multip	−0.002	1
Age	0.218	0.169	1
Edu	−0.011	0.189	−0.023	1
Gen	−0.042	0.032	−0.028	0.063	1
Nat	−0.044	0.165	0.047	0.193	0.008	1
Prof	0.047	−0.027	0.000	−0.071	−0.041	−0.046	1
Fsize	−0.055	0.346	0.195	0.219	0.022	0.413	0.059	1
Lev	−0.037	0.093	0.012	0.079	0.004	0.136	−0.030	0.297	1

This table shows the correlation analysis for the variables: Ten (CEO tenure), Multip (CEO multiple directorship), Age (CEO age), Edu (CEO education), Gen (CEO gender), Nat (CEO nationality), Prof (profitability), Fsize (firm size), and Lev (leverage).

provides the correlation analysis among the independent variables. The highest observed pairwise correlation is 0.41, while the lowest is −0.07. These values suggest that the matrix has full rank, indicating no issues of multicollinearity. Consequently, the estimators are well-behaved, and the regression results are reliable.

4.3. Result, Analysis, and Discussion

Fixed-Effect Model

The results of the Hausman test yield a statistic of 197.61, with a p-value of 0.000. This highly significant result leads us to reject the null hypothesis that the random-effects model is appropriate. Consequently, the fixed-effects model is preferred for this analysis. We also used Benjamini–Hochberg-Adjusted p-values with a False Discovery Rate (FDR) of 10%. We did not find any false positive.

Table 4. ESG performance for full sample.

		Constant	Cognitive CEO Characteristics			Demographic CEO Characteristics			Prof	Fsize	Lev
			Ten	Multip	Edu	Age	Gen	Nat
PolSt	Low	−91.72 ***	−0.72	2.71	28.86 ***	−2.17 **	−171.96 ***	19.69	−0.58	9.622 **	−0.56
	High		0.88	−2.50	−29.42 ***	2.67 **	191.85 ***	−17.61	0.57	7.33 *	0.63
RuLaw	Low	−93.10 ***	−0.57	−0.73	6.05 **	−0.69	−54.46 **	13.53 **	−0.10	10.08 ***	−0.06
	High		0.74	1.02	−6.16 **	1.18 ***	73.89 ***	−11.78 *	0.08	7.01 ***	0.11
VoAc	Low	−93.60 ***	−0.53	−1.45	6.49 **	−0.78 *	−53.67 **	14.74 **	−0.15	10.66 ***	−0.11
	High		0.70	1.79	−6.61 **	1.28 ***	72.97 ***	−13.2 **	0.13	6.47 ***	0.17
ConCo	Low	−94.12 ***	−3.15	12.95	12.03	5.25 **	−186.81 **	42.59	0.36	−10.16	0.97 **
	High		3.32	−12.66	−12.10	−4.79 *	206.63 ***	−41.38	−0.39	27.56 ***	−0.93 *
GoEf	Low	−92.95 ***	−0.96	−0.86	10.39 **	−0.96	−102.34 **	22.09 **	−0.27	9.86 ***	−0.28
	High		1.13	1.13	−10.48 **	1.44 *	121.62 ***	−20.27 *	0.25	7.25 ***	0.33
RegQu	Low	−93.05 ***	−0.33	−0.72	6.42 **	−1.01 **	−53.35 **	14.79 ***	−0.10	11.28 ***	−0.13
	High		0.50	1.04	−6.56 **	1.52 ***	73.02 ***	−13.06 **	0.08	5.78 ***	0.18

This table shows the fixed-effect model for the full sample under high WGIs and low WGIs (Worldwide Governance Indicators): Ten (CEO tenure), Multip (CEO multiple directorship), Age (CEO age), Edu (CEO education), Gen (CEO gender), Nat (CEO nationality), PolSt (political stability), RuLaw (rule of law), VoAc (voice and accountability), ConCo (control of corruption), GoEf (government effectiveness), RegQu (regulatory quality), Prof (profitability), Fsize (firm size), and Lev (leverage). *, **, and *** represent significance at 0.1, 0.05, and 0.01 levels, respectively.

shows the effect of CEO characteristics on the ESG score for the full sample. Overall, the results provide positive effects under strong governance conditions; several relationships exhibit context-dependent variations, indicating that the influence of CEO characteristics on ESG performance is not uniform across institutional settings. Regarding demographic CEO characteristics, when WGIs are low, consistent with [101], CEO age negatively affects ESG performance, except for rule of law and government effectiveness, where the effect is not significant, and except for the “control of corruption” indicator, where it is positive, possibly due to older CEOs’ reliance on traditional practices or hesitance in unstable environments. Conversely, when WGIs are high, CEO age has a positive effect on ESG performance, except for the “control of corruption” indicator, where the effect is negative. As a result, in well-governed environments, older CEOs may enhance ESG outcomes due to their leadership experience and better alignment with institutional frameworks. This affirms our contention that older CEOs often drive stronger environmental performance and focus on long-term legacy [75], especially when supported by favorable regulatory conditions and mechanisms that promote public accountability. These findings align with upper echelons theory, in which the age and career stage shape the strategic orientation and social priorities of leaders [36]. This pattern further supports prior empirical findings that the positive effect of CEO age on sustainability outcomes is more pronounced in structured and stable governance environments, while weaker institutional contexts may limit the translation of experience into ESG engagement [75].

Furthermore, when WGIs are low, the presence of a female CEO has a significant negative effect on ESG performance across all governance indicators. Female CEOs face bias, weak support, and limited influence, hindering ESG implementation. Our findings align with previous studies, showing weak or negative female leadership–ESG links in such contexts [6,61,102]. However, our results indicate that when WGIs are high, female CEOs have a positive significant effect on ESG performance across all indicators. Supportive institutional frameworks—strong rule of law and gender-inclusive cultures—enable female CEOs to fully utilize their strengths (stakeholder engagement, ethical decision-making), enhancing ESG outcomes. Our findings align with [103,104]. The discrepancy in our findings related to the effect of gender on ESG performance is explained by the governance environment within which the CEO operates, highlighting the potency of a supportive institutional framework for the success of female CEOs. Our results align with stakeholder and institutional theories [34,54,61]. This pattern is also consistent with prior empirical evidence that the impact of female CEOs on ESG outcomes is contingent on the institutional environment, becoming more positive under stronger governance conditions [9,69].

Interestingly, foreign CEOs have a significantly positive effect on ESG performance when WGIs are low, except in the areas of political stability and control of corruption, where their impact is not significant. This result is contrary to our expectation and may be due to the fact that the foreign CEOs’ international experience fosters higher ESG performance—introducing global best practices, ethical standards, and governance expertise—helping offset weak local systems and significantly improve sustainability. However, when WGIs are high, foreign CEOs have a significant negative effect on ESG performance, with no significant effect observed on the political stability and control of corruption indicators. This can be explained by the fact that in well-governed environments, local CEOs better align with national ESG norms, while foreign CEOs may struggle adapting—especially amid political instability or corruption—due to weaker local networks and institutional understanding. These results reflect institutional theory’s insight that contextual pressures can not only weaken but, in some cases, reverse the expected influence of leadership characteristics on ESG outcomes. This pattern is also consistent with prior research suggesting that the impact of CEO characteristics on ESG outcomes varies across institutional contexts, although the direction and magnitude of this effect may differ depending on the nature of the leadership attribute [14,87].

Regarding cognitive CEO characteristics, under low WGIs, the number of qualifications held by the CEO has a positive effect on ESG scores; this positive result is consistent with [105], except for the control of corruption and government effectiveness indicators, where the effect is not significant. Highly qualified CEOs contribute to improved ESG performance by applying strategic insight, ethical awareness, and environmental responsibility, particularly in weak institutional settings. The result is consistent with [20,34,36,83,89,92]. Therefore, when WGIs are high, the number of CEO qualifications has a negative effect on ESG scores, except for control of corruption, where the effect remains insignificant. Despite strong governance, highly qualified CEOs often favor profitability, sidelining ESG and focusing on financial and technical priorities. This is in line with [94] who report that education alone may not significantly influence CEO impact on ESG outcomes. Taken together, these findings suggest that although prior research generally supports a positive association between CEO education and ESG engagement, this relationship may vary across governance contexts and, in some cases, become less favorable in more structured institutional environments.

Therefore, neither CEO tenure nor multiple directorships influenced ESG scores across all WGIs. Despite stakeholder theory predicting broader norms from external affiliations [72], this effect did not materialize. Additionally, longer-serving CEOs showed no positive impact on ESG performance, leaving tenure’s influence ambiguous. This finding is consistent with prior empirical evidence reporting mixed or insignificant effects of CEO tenure and external affiliations on ESG outcomes, suggesting that these characteristics may not systematically translate into sustainability engagement across different institutional contexts [94].

Table 5. ESG performance for sensitive industries.

		Constant	Cognitive CEO Characteristics			Demographic CEO Characteristics			Prof	Fsize	Lev
			Ten	Multip	Edu	Age	Gen	Nat
PolSt	Low	−79.88 ***	−1.51	3.33	35.66 ***	−1.73	−177.83 ***	35.41 **	−0.41	6.52	−0.97 **
	High		2.04	−3.33	−35.99 ***	2.10 *	196.04 ***	−35.65 **	0.38	9.07 **	1.06 **
RuLaw	Low	−80.60 ***	−0.80	−1.05	7.96 ***	−0.43	−62.48 **	21.31 ***	−0.01	7.96 ***	−0.08
	High		1.35 *	1.13	−7.78 **	0.79	80.31 ***	−21.93 ***	−0.02	7.69 ***	0.17
VoAc	Low	−81.26 ***	−0.73	−1.72	8.38 ***	−0.56	−61.33 **	21.94 ***	−0.06	8.66 ***	−0.13
	High		1.26 *	1.87	−8.20 ***	0.93 *	78.97 ***	−22.84 ***	0.03	7.03 ***	0.22
ConCo	Low	−81.70 ***	−5.83 *	0.46	−14.66	10.82 ***	−164.06 **	61.71	0.51	−17.49 *	0.76
	High		6.39 *	−0.32	14.93	−10.55 ***	181.95 **	−62.59	−0.56	33.52 ***	−0.69
GoEf	Low	−80.37 ***	−1.45	−1.11	14.27 ***	−0.55	−114.78 **	36.37 ***	−0.15	6.86 **	−0.34
	High		1.99	1.17	−14.06 **	0.90	132.37 ***	−36.88 ***	0.13	8.79 ***	0.42
RegQu	Low	−80.49 ***	−0.48	−1.01	8.30 ***	−0.91 *	−59.26 **	21.75 ***	−0.02	9.65 ***	−0.14
	High		1.02	1.12	−8.16 ***	1.28 ***	77.31 ***	−22.375 ***	−0.01	5.93 ***	0.22

This table shows the fixed-effect model for the sensitive industries under high WGIs and low WGIs (Worldwide Governance Indicators): Ten (CEO tenure), Multip (CEO multiple directorship), Age (CEO age), Edu (CEO education), Gen (CEO gender), Nat (CEO nationality), PolSt (political stability), RuLaw (rule of law), VoAc (voice and accountability), ConCo (control of corruption), GoEf (government effectiveness), RegQu (regulatory quality), Prof (profitability), Fsize (firm size), and Lev (leverage). *, **, and *** represent significance at 0.1, 0.05, and 0.01 levels, respectively.

presents the effects of CEO characteristics on ESG performance in sensitive industries, and, regarding demographic CEO characteristics, CEO age has a positive significant impact on ESG performance under high WGIs, particularly in political stability, voice and accountability, and regulatory quality. Older CEOs navigate complex regulations and stakeholder demands through deep insight. However, no significant relationship is found for rule of law and government effectiveness, while under control of corruption, CEO age has a negative and significant effect. Under low WGIs, CEO age generally shows no significant impact under all the WGIs, except for a positive effect for control of corruption and a negative effect for regulatory quality. This suggests that in weak anti-corruption environments, older CEOs may have a less favorable influence, possibly due to entrenched managerial behaviors.

Moreover, female CEOs under low WGIs have a negative significant impact on ESG performance across all WGIs, possibly due to challenges in weak institutional environments. Under high WGIs, female CEOs show a positive significant effect, likely because robust governance structures provide necessary support and accountability.

Furthermore, foreign CEOs have a positive significant effect on ESG performance under low WGIs across most indicators, except for control of corruption where there is no significant effect. The positive effect of foreign CEOs is possibly due to their diverse perspectives and international experience. Under high WGIs, their effect becomes negative and significant, suggesting challenges in aligning with local expectations or navigating complex regulations.

Regarding cognitive CEO characteristics, tenure CEO has a positive significant effect on ESG performance under high WGIs, especially for rule of law, voice and accountability, and control of corruption, likely due to accumulated experience. However, under low WGIs, tenure CEO shows a negative effect, particularly for control of corruption, suggesting that without strong frameworks, long-tenured CEOs may become complacent or resistant to change, and have no significant effect for all other WGIs. On the other hand, CEOs with multiple directorships do not significantly impact ESG performance under any WGIs.

Moreover, CEOs with a higher number of qualifications are positively associated with ESG performance under low WGIs (except for control of corruption) but negatively associated under high WGIs (except for control of corruption). These findings are consistent with prior research indicating that the effects of CEO characteristics on ESG performance vary across institutional settings, particularly in industries exposed to higher regulatory and stakeholder pressures [9,69].

Table 6. ESG performance for non-sensitive industries.

		Constant	Cognitive CEO Characteristics			Demographic CEO Characteristics			Prof	Fsize	Lev
			Ten	Multip	Edu	Age	Gen	Nat
PolSt	Low	−126.54 ***	6.52	−14.75	25.82	−0.14	193.27	−158.61 ***	−2.37 *	3.79	2.72 *
	High		−7.34	17.11	−26.63	0.80	−165.52	172.21 ***	2.40 *	18.21 *	−2.82 *
RuLaw	Low	−124.34 ***	1.90	−2.14	35.43 **	−1.51	1061.13 *	−84.12 ***	−0.92	5.21	3.00 **
	High		−2.65	4.00	−36.09 **	2.16 *	−1032.90 *	95.19 ***	0.89	16.55 ***	−3.09 **
VoAc	Low	−124.24 ***	1.81	−4.87	23.16 *	−1.04	1338.42 **	−70.67 ***	−1.15	8.55 *	1.88 *
	High		−2.56	6.82	−23.76 *	1.69	−1311.10 **	81.75 ***	1.14	13.16 **	−1.95 *
ConCo	Low	−125.37 ***	−9.74	−295.22	350.63	1.45	−533.88	2125.34	0.59	−161.12 *	13.34
	High		8.98	297.24	−351.41	−0.75	563.44	−2116.22	−0.62	183.02 **	−13.43
GoEf	Low	−125.68 ***	11.89	−20.83	92.65 **	−1.41	1693.22	−173.03 ***	−6.70	−7.26	6.07 *
	High		−12.64 *	22.73	−93.20 **	2.02	−1663.98	184.07 ***	6.69	29.26 **	−6.14 *
RegQu	Low	−124.23 ***	2.02	−3.48	30.10 *	−1.20	534.36	−77.18 ***	−1.07	6.25	2.52 *
	High		−2.77	5.39	−30.80 *	1.84	−506.11	88.53 ***	1.05	15.51 ***	−2.60 *

This table shows the fixed-effect model for non-sensitive industries under high WGIs and low WGIs (Worldwide Governance Indicators): Ten (CEO tenure), Multip (CEO multiple directorship), Age (CEO age), Edu (CEO education), Gen (CEO gender), Nat (CEO nationality), PolSt (political stability), RuLaw (rule of law), VoAc (voice and accountability), ConCo (control of corruption), GoEf (government effectiveness), RegQu (regulatory quality), Prof (profitability), Fsize (firm size), and Lev (leverage). *, **, and *** represent significance at 0.1, 0.05, and 0.01 levels, respectively.

shows the effects of CEO characteristics on ESG performance in non-sensitive industries regarding demographic CEO characteristics; CEO age under high WGIs positively affects ESG performance only under rule of law, suggesting that strong legal frameworks can motivate older CEOs to prioritize ESG. No significant effects are observed under other high or low WGIs.

Moreover, female CEOs under low WGIs positively affect ESG performance under rule of law and voice and accountability. However, under high WGIs, their impact is negative and significant for these same indicators, possibly reflecting overregulation or cultural constraints on decision-making. No significant effects are seen for other WGIs.

Furthermore, foreign CEOs negatively impact ESG performance under low WGIs but have a positive significant effect under high WGIs, indicating that strong governance can enhance their effectiveness in promoting ESG practices.

Regarding cognitive CEO characteristics, tenure CEO negatively affects ESG performance under high WGIs, specifically government effectiveness, possibly due to reluctance to innovate, with no significant effect under all other WGIs. On the other hand, CEOs with multiple directorships do not significantly impact ESG performance under any WGIs.

Ultimately, CEOs with more educational qualifications positively influence ESG performance under low WGIs (except for political stability and control of corruption) but have a negative effect under high WGIs (except for political stability and control of corruption). This pattern aligns with prior empirical evidence suggesting that the impact of CEO attributes on ESG outcomes is contingent on governance quality and may differ across less regulated industry contexts [14,87].

Table 7. ESG combined result.

		Full Sample		Sensitive Industries		Non-Sensitive Industries
		Cognitive	Demographic	Cognitive	Demographic	Cognitive	Demographic
PolSt	Low	19.015 ***	14.092 ***	23.070 ***	14.049 ***	2.420	14.205 ***
	High	19.051 ***	17.774 ***	23.328 ***	16.593 ***	2.840	16.600 ***
RuLaw	Low	6.218	12.580 ***	8.896 **	16.586 ***	6.782 *	16.575 ***
	High	6.746 *	20.465 ***	10.314 **	22.450 ***	7.230 *	21.097 ***
VoAc	Low	8.402 **	15.569 ***	11.195 **	19.931 ***	4.127	14.619 ***
	High	9.191 **	24.157 ***	12.484 ***	26.694 ***	4.636	19.139 ***
ConCo	Low	1.397	11.873 ***	3.146	19.140 ***	1.736	1.961
	High	1.452	12.359 ***	3.728	19.290 ***	1.746	1.943
GoEf	Low	5.438	10.782 **	8.413 **	15.023 ***	6.966 *	13.737 ***
	High	5.652	14.421 ***	8.989 **	17.884 ***	7.252 *	15.185 ***
RegQu	Low	6.977 *	18.763 ***	9.618 **	23.272 ***	4.599	10.357 **
	High	7.378 *	28.687 ***	0.333 **	31.058 ***	5.060	13.437 ***

This table shows the combined result for the full sample, sensitive and non-sensitive industries, under high WGIs and low WGIs (Worldwide Governance Indicators): PolSt (political stability), RuLaw (rule of law), VoAc (voice and accountability), ConCo (control of corruption), GoEf (government effectiveness), RegQu (regulatory quality), Prof (profitability), Fsize (firm size), and Lev (leverage). *, **, and *** represent significance at 0.1, 0.05, and 0.01 levels, respectively.

presents the combined results of the effect of demographic and cognitive CEO characteristics on ESG performance across the full sample, sensitive and non-sensitive industries. In the full sample, demographic CEO characteristics significantly affect ESG performance across all WGI indicators. Cognitive characteristics also have a significant effect, but only on certain WGI dimensions—they are not significant for control of corruption or government effectiveness. This suggests that demographic characteristics have a stronger and more consistent influence on overall ESG performance than cognitive characteristics.

In sensitive industries, both demographic and cognitive characteristics have a greater impact on ESG scores. Demographic characteristics significantly affect ESG across all WGI dimensions, while cognitive characteristics are also significant for all indicators except control of corruption.

In non-sensitive industries, demographic characteristics again have a stronger influence, except for control of corruption, where there is no significant effect. Cognitive characteristics generally have no significant impact across WGI indices—except for rule of law and government effectiveness, where they do significantly influence ESG scores.

Overall, these findings reinforce prior research suggesting that demographic CEO characteristics tend to exert a more consistent influence on ESG performance, while cognitive characteristics demonstrate more context-dependent and heterogeneous effects across governance and industry settings [9,69,94].

5. Conclusions

This study examines how CEOs’ cognitive and demographic characteristics affect sustainability, under moderating WGI indicators, in STOXX Europe 600 firms for the period 1999–2023, using nonlinear PSTR with regime-switching governance (Wald tests assess heterogeneity). The findings show that CEO demographic characteristics—age, gender, and nationality—significantly influence ESG performance, with effects depending on governance context and industry sensitivity. For the full sample, under weak Worldwide Governance Indicators (WGIs), older and female CEOs tend to negatively affect ESG outcomes, whereas foreign CEOs have a positive impact. This trend reverses in strong governance contexts. Regarding cognitive CEO characteristics—tenure, multiple directorships, and educational qualifications—higher qualifications enhance ESG under weak governance but generally reduce ESG performance under strong WGIs, while CEO tenure and multiple directorships show limited or no significant effect.

Sensitive industries show results similar to the full sample regarding demographic characteristics. However, cognitive characteristics such as educational qualifications contribute negatively to sustainability performance, while tenure has a positive effect under strong governance. CEOs holding multiple directorships continue to exhibit no significant influence on sustainability outcomes.

In non-sensitive industries, the impact of CEO characteristics is more nuanced. Demographic characteristics, such as age and foreign nationality, continue to influence ESG outcomes, particularly under specific governance dimensions, while the effect of female CEOs varies depending on governance strength. Cognitive characteristics, including educational qualifications, show conditional effects: under certain high governance indicators, higher qualifications are associated with lower ESG scores, whereas under weaker governance conditions, they may enhance performance in some dimensions. CEO tenure and multiple directorships generally do not have significant impacts. These patterns underscore that in less regulated industries, demographic characteristics are often more consistently influential than cognitive traits, with the effects of education being highly context-dependent.

Overall, in the full sample, demographic characteristics significantly impact ESG performance across all WGIs, while cognitive characteristics matter only in specific dimensions. In sensitive sectors—where environmental and social scrutiny is high—both demographic and cognitive CEO characteristics significantly influence ESG across nearly all WGIs, reflecting stronger regulatory constraints and stakeholder expectations. Conversely, in non-sensitive industries—where oversight is weaker—demographics remain influential across most indicators, but cognitive characteristics only affect ESG under rule of law and government effectiveness, suggesting that softer governance cues suffice in less regulated contexts. These patterns imply that in industries exposed to strong external pressures, CEOs with both robust demographic profiles and cognitive capabilities are essential for navigating complex ESG demands, whereas in less scrutinized industries, demographic alignment alone often suffices.

This study contributes to the literature by demonstrating that the influence of CEO characteristics on ESG performance is not uniform, but varies across governance contexts and industry types. This finding is consistent with prior research emphasizing the role of CEO attributes in shaping sustainability outcomes and ESG engagement [9,10,41] while extending this literature by demonstrating that such effects are contingent on institutional governance quality, as suggested by institutional theory perspectives [19,20,27]. By combining cross-national evidence with a smooth transition regression (STR) approach, the findings highlight the importance of institutional heterogeneity and sector-specific pressures in shaping effective sustainability strategies. By leveraging variation in governance quality within the European context, this study highlights how differences in institutional environments shape the effectiveness of CEO characteristics even within a seemingly homogeneous regional setting. Taken together, the study responds to recent calls for more integrated and context-sensitive analyses of CEO characteristics and ESG performance across varying institutional environments [11,31].

This study has several limitations that offer directions for future research. First, the focus on STOXX Europe 600 firms may limit generalizability beyond developed European contexts. Future research could extend the analysis to emerging markets and conduct more granular comparisons between Eastern and Western Europe to better capture institutional heterogeneity. Second, certain CEO characteristics are measured using widely adopted proxies to ensure cross-country comparability. In particular, CEO education is proxied by the number of academic qualifications, nationality is captured as a foreign versus domestic indicator, and multiple directorships are measured on a career basis rather than contemporaneous board positions. While these measures provide consistent and comparable proxies across institutional settings, future research may consider more granular indicators to further enrich the analysis. Third, other leadership dimensions—such as behavioral traits, managerial styles, and board-level influences—remain unexplored. Future studies could incorporate these factors and examine their interaction with governance structures and ownership configurations. Finally, applying clustering or sub-sample approaches based on governance profiles may further enhance understanding of how sustainability strategies evolve across diverse institutional settings.

The results carry actionable insights for boards, investors, and policymakers. These associations suggest that aligning CEO demographic characteristics with the institutional context can enhance sustainability outcomes, older or female CEOs can bolster ESG in high-governance environments, while foreign CEOs can enhance ESG performance where governance is weaker. This aligns with prior evidence highlighting the importance of matching leadership attributes with governance structures to enhance sustainability outcomes [75,76]. Investors and ESG analysts should consider both CEO characteristics and governance quality when forming investment or engagement strategies, as the effectiveness of leadership attributes is governance-dependent. Policymakers aiming to elevate corporate sustainability should prioritize strengthening governance structures—as robust institutional environments enhance the positive impact of diverse and experienced CEOs—while in weaker regimes, incentives for leadership diversity could serve as a compensatory mechanism. This is consistent with institutional arguments that stronger regulatory and governance frameworks reinforce firms’ ESG engagement and strategic alignment with sustainability goals [85,86]. From a broader perspective, these findings also resonate with emerging views of Society 5.0, where ESG integration is central to building human-centric and sustainable systems, and where CEOs play a pivotal role in aligning corporate strategies with societal and environmental priorities [106]. This research thus advances ESG scholarship by showcasing the conditional effects of leadership within evolving governance regimes, employing a robust nonlinear empirical model, and offering nuanced guidance for governance-aware CEO selection and ESG strategy formation.