State ownership as a moderator in the relationship between board characteristics and ESG performance: evidence from Asia-Pacific markets

Environmental, social, and governance (ESG) performance has become a central benchmark for corporate accountability and long-term sustainability, driven by intensifying stakeholder scrutiny, climate-related risks, and evolving global policy frameworks (Al Kurdi et al., 2023; Helfaya et al., 2023). Boards of directors are pivotal in shaping ESG outcomes through their structure, diversity, and oversight, with attributes such as board independence, gender diversity, transparency, and CEO–chair separation frequently linked to superior ESG performance (Kim and Yang, 2025; Mazumder, 2025). However, the effectiveness of these governance mechanisms is not uniform and often varies across different institutional and ownership contexts (Abdi et al., 2022).

Ownership structure—particularly state ownership—emerges as a critical, yet underexplored, moderator in the relationship between board characteristics and ESG outcomes. State-owned enterprises (SOEs) operate under unique political objectives, social mandates, and non-market incentives, which can fundamentally alter how board practices translate into ESG performance (Eissa et al., 2024). In SOEs, governance features like transparency or gender diversity may be more compliance-driven than strategic, suggesting that conventional board-ESG relationships may not fully apply in government-linked firms (Nicolo' and Andrades-Peña, 2024). For example, the presence of independent directors or gender-diverse boards in SOEs may be shaped by regulatory requirements rather than a genuine commitment to sustainability (Sharawi et al., 2024).

The Asia-Pacific region offers a compelling context for examining these dynamics due to its institutional diversity, varying degrees of state involvement in business, and rapidly evolving ESG regulations (Arhinful et al., 2024). Yet, much of the existing research is limited to single-country studies or short timeframes, restricting our understanding of broader regional governance patterns. There is a clear need for longitudinal, multi-country analyses to capture how board practices and ownership structures jointly influence ESG outcomes over time (Villalonga et al., 2025).

While board attributes such as independence and diversity generally enhance ESG performance, their impact is often contingent on ownership type. In SOEs, politically influenced appointments or accountability to government bodies can undermine the effectiveness of governance mechanisms, challenging the assumption that board practices operate uniformly across all institutional and ownership contexts (Akhtar and Abdullah, 2025). Addressing this gap, this study will explore how state ownership moderates the influence of board characteristics on ESG performance, providing valuable insights for investors, regulators, and policymakers seeking to strengthen ESG oversight in both private and state-affiliated firms.

This study integrates four complementary theories, agency, stakeholder, resource dependence, and institutional, to explain how board characteristics affect ESG performance and how state ownership moderates this relationship. Agency theory suggests that mechanisms such as board independence and CEO and chair separation mitigate conflicts and align managerial behaviour with shareholder interests. ESG performance can signal lower agency risk and greater transparency, particularly under close regulatory or investor scrutiny (Agnese et al., 2024; Akhtar and Abdullah, 2025). Stakeholder theory emphasizes accountability to a broader constituency, with ESG outcomes reflecting both governance efficiency and responsiveness to societal demands. Board diversity and inclusion further strengthen equitable governance and stakeholder engagement (Abdelkader et al., 2024; Kampoowale et al., 2024).

Resource dependence theory highlights the board's role in accessing critical resources such as expertise, legitimacy, and networks to support sustainability. Diverse and skilled boards facilitate green financing, innovation, and regulatory compliance, especially in resource-constrained or transitional contexts (Alam et al., 2025; Arhinful et al., 2024). Institutional theory stresses external pressures, normative, coercive, and mimetic, that shape governance behaviour. In contexts with significant state ownership, firms often adopt ESG practices in response to public mandates, regulatory expectations, or reputational concerns (Eissa et al., 2024; Nicolo’ and Andrades-Peña, 2024). Together, these frameworks provide a lens for understanding ESG outcomes, emphasizing the interplay between board structure, stakeholder demands, strategic resources, and institutional constraints.

Board size influences strategic oversight, information processing, and stakeholder engagement. Larger boards offer broader expertise and access to diverse resources, enhancing their ability to manage complex ESG issues (Agnese et al., 2024; Arhinful et al., 2024). Resource dependence theory suggests that larger boards strengthen legitimacy and stakeholder engagement, supporting ESG strategy.

Excessively large boards may face coordination inefficiencies, diluted responsibility, and symbolic ESG compliance, reducing sustainability oversight (Meeprom et al., 2024; Lin and Wu, 2024). Empirical evidence indicates that board size benefits ESG performance up to an optimal threshold, with effects depending on institutional and ownership contexts, including state control. The following hypothesis is proposed:

Independent directors enhance governance by providing objective oversight, reducing managerial entrenchment, and aligning corporate practices with long-term value creation (Ji et al., 2024; Kim and Yang, 2025). Agency theory suggests that board independence limits conflicts of interest and promotes transparency, both critical for ESG governance.

From a stakeholder perspective, independent directors are more likely to address legitimacy pressures and enforce ethical, sustainable practices (Arhinful et al., 2024). Empirical evidence shows a positive link between board independence and ESG disclosure quality in both developed and emerging markets (Al Amosh and Khatib, 2022; Ding et al., 2024; Pham and Ho, 2024). Effectiveness may vary in state-influenced firms where external appointments limit autonomy. The following hypothesis is proposed:

Board gender diversity is recognized as a driver of responsible governance and stronger ESG performance. Female directors contribute diverse perspectives, greater ethical awareness, and heightened sensitivity to stakeholder needs (Abdelkader et al., 2024; Kampoowale et al., 2024; Saleh et al., 2025; Saleh et al., 2025). Stakeholder and resource dependence theories suggest that gender-diverse boards better manage social and environmental risks, engage broader constituencies, and reinforce corporate legitimacy.

Empirical evidence shows that gender diversity positively affects sustainability disclosures, CSR investment, and ESG ratings, especially when supported by inclusive board cultures (Zhu and Chen, 2025; Eissa et al., 2024; Saleh and Maigoshi, 2024). However, its impact can be limited where female participation is symbolic or lacks decision-making power. In light of these considerations:

Board transparency, reflecting the openness of disclosures on board activities, governance decisions, and ESG strategy, is a key dimension of effective corporate governance. It builds trust with investors and stakeholders while promoting accountability for ESG commitments. Agency theory suggests that transparency reduces information asymmetry, limits managerial discretion, and aligns board actions with shareholder expectations (Al Kurdi et al., 2023).

Institutional theory highlights transparency as a means for firms to meet societal norms and regulatory requirements. Transparent boards are more likely to report ESG targets, sustainability oversight, and climate-related risks, enhancing ESG ratings and stakeholder perceptions (Helfaya et al., 2023; Nicolo’ and Andrades-Peña, 2024). This is especially important in contexts where legitimacy pressures, such as state ownership or international investor expectations, shape disclosure behaviour. The following hypothesis is proposed:

CEO Chairperson duality concentrates executive and monitoring power in a single individual, potentially undermining board independence and oversight. Agency theory suggests that duality weakens internal controls, enabling managerial opportunism and reducing ESG investments (Akhtar and Abdullah, 2025; Mansour et al., 2024). ESG initiatives require long-term strategic commitment and independent evaluation, both of which may be compromised when board leadership is not independent.

Empirical evidence shows that CEO duality is linked to lower ESG scores, particularly in high-accountability or stakeholder-scrutinized firms (Agnese et al., 2024; Arhinful et al., 2024). Although duality may enable faster decision-making in dynamic industries, this benefit rarely outweighs the costs to ESG integrity. Dual leadership can dilute governance transparency, reduce stakeholder confidence, and deprioritize sustainability. The corresponding hypothesis is as follows:

Disclosing board members' competencies, including sustainability expertise, financial acumen, and ethical judgement, enhances governance transparency and signals readiness to address ESG challenges. Resource dependence theory posits that directors with relevant skills provide strategic access to external resources, legitimacy, and sector specific knowledge (Collevecchio et al., 2025). Such disclosure increases stakeholder confidence in the board's capacity to oversee ESG risks and regulatory compliance (Alam et al., 2025).

Empirical studies show that firms reporting board expertise outperform peers in ESG execution, particularly in sectors with mandatory or investor expected sustainability disclosure (Arhinful et al., 2024; Agnese et al., 2024). Disclosure also reduces perceived decision-making opacity, which is crucial in volatile institutional contexts. This leads to the following hypothesis:

State ownership adds an institutional layer that moderates how board characteristics affect ESG performance. According to institutional theory, SOEs pursue both commercial objectives and state policy mandates, often resulting in politically driven appointments that may weaken board effectiveness (Nicolo’ and Andrades-Peña, 2024; Eissa et al., 2024). Consequently, the positive impacts of board size, independence, or gender diversity can be reduced due to symbolic compliance and limited decision-making autonomy.

Conversely, state ownership can enhance the influence of transparency and board expertise. Stakeholder theory suggests that SOEs face greater public scrutiny, increasing adherence to ESG disclosure norms (Zahid et al., 2023). The bureaucratic structure of SOEs may also mitigate risks from CEO duality, as decisions are embedded in multi-tiered governance systems. The following hypotheses reflect the anticipated moderating effects of state ownership:

The study population consists of publicly listed companies in the Asia Pacific region, with the sample determined by the availability of consistent data on environmental, social, and governance (ESG) performance, corporate governance characteristics, and financial indicators. The period from 2010 to 2022 was selected as it coincides with major regulatory and institutional developments, including the International Integrated Reporting Framework (2013) and the United Nations Sustainable Development Goals (2015).

Firm level data were obtained from the Refinitiv Eikon database, which provides standardised ESG scores aligned with international frameworks such as GRI, SASB, and TCFD (Postiglione et al., 2025; Tan et al., 2023). Corporate governance variables including board structure, independence, and gender diversity were collected from Refinitiv's Corporate Governance module, while financial indicators were cross-checked with company filings and harmonised using IFRS to ensure comparability across jurisdictions.

The sample was restricted to firms headquartered in Asia Pacific countries with sufficient ESG disclosure and those with available corporate governance and financial data for at least seven years. This resulted in 1,175 firms across 14 countries, representing 15,275 firm-year observations. Market classification followed MSCI (2025) and FTSE Russell (2025), with Australia, Japan, Hong Kong, Singapore, and New Zealand defined as Developed markets, and China, India, Indonesia, Malaysia, the Philippines, South Korea, Taiwan, Thailand, and Sri Lanka as Emerging. Although Sri Lanka is officially a Frontier market, it was classified as Emerging to reflect ESG literature on transitional economies (Jayasinghe et al., 2021; Rahat and Nguyen, 2024).

As reported in Table 1, Japan contributed the largest share (31.49%), followed by Australia (13.87%) and Hong Kong (11.15%), while Sri Lanka (0.09%) and New Zealand (0.85%) were underrepresented. The distribution achieved near balance between Developed (49.02%) and Emerging (50.98%) markets. Industry level representation spanned five sectors, with the largest groups being Industrial and Manufacturing (252 firms), Environmentally Intensive (245), and Finance and Real Estate (232), followed by Consumer Facing (216) and Technology, Healthcare, and Utilities (230). To mitigate bias from outliers, continuous variables were winsorised at the 1st and 99th percentiles, consistent with prior ESG studies (Mooneeapen et al., 2022; Xie et al., 2023).

This study examines the relationship between board attributes and ESG performance, with state ownership as a moderator, using Refinitiv Eikon data aligned with GRI, SASB, and TCFD standards (Disli et al., 2022; Kim and Yang, 2025). ESG performance is measured by Refinitiv's composite score and its three dimensions—Environmental, Social, and Governance—as percentiles relative to industry peers. Disaggregating ESG allows assessment of heterogeneous board effects, such as social outcomes' sensitivity to board diversity (Kampoowale et al., 2024) and environmental outcomes' dependence on sectoral exposure (Sharawi et al., 2024).

Independent variables include six board characteristics theorised to affect ESG through monitoring, advisory, or symbolic mechanisms (Zahra and Pearce, 1989). Board size measures the total number of directors (Disli et al., 2022; Lin and Wu, 2024); board independence is the share of non-executive directors (Fama and Jensen, 1983; Ding et al., 2024; Jung, 2024); gender diversity is the percentage of female directors (Birindelli et al., 2018); CEO duality equals one if the CEO also chairs the board (Peng and Chen, 2024); board transparency captures disclosure of director attendance (Al Kurdi et al., 2023; Mahyoub et al., 2024); and board skills disclosure reflects reporting of directors' qualifications (Agnese et al., 2024; Wu et al., 2024).

State ownership is a binary moderator (government stake >50% or golden share veto) capturing institutional influence (Hayes and Montoya, 2017; Schoeneberger and Rhoads, 2024). Legitimacy theory predicts ESG promotion, while agency perspectives highlight politicisation risks; prior evidence shows contingent effects on gender inclusion and director networks (Eissa et al., 2024; Shu et al., 2024). Control variables include firm size (log of total assets) (Abdi et al., 2022; Drempetic et al., 2019; Martínez et al., 2024; Postiglione et al., 2025; Zhang and Sharon, 2023), leverage (debt to assets) (Teerapan, 2024), and firm age (years since incorporation) (Rahat and Nguyen, 2024; Zhu and Chen, 2025), reflecting resources, financial constraints, and organisational maturity affecting ESG outcomes (See Table 2).

To empirically assess the relationship between board characteristics and ESG performance, this study employs a panel data approach designed to address unobserved heterogeneity, dynamic persistence, and endogeneity. Particular attention is paid to the moderating role of state ownership. The modelling framework comprises four interdependent stages: (1) baseline estimation using fixed-effects, (2) moderation analysis through interaction terms, (3) dynamic panel estimation using system GMM, and (4) pillar-specific multivariate analysis. Each stage is supported by rigorous diagnostic and robustness procedures.

The initial empirical model adopts a two-way fixed-effects (FE) estimator to control for both firm-specific and time-specific unobservable factors. This approach allows for consistent estimation in the presence of time-invariant heterogeneity and removes potential bias stemming from omitted firm-level characteristics. The model is specified as:

Where $μi$ captures firm-specific effects, $λt$ denotes year dummies, and $εit$ is the idiosyncratic error term. The main explanatory variables include six board characteristics: board size, independence, gender diversity, CEO duality, transparency, and skills disclosure. Control variables comprise firm size, leverage, and firm age. Standard errors are clustered at the firm level to correct for potential heteroscedasticity and serial correlation.

To investigate whether the influence of board characteristics on ESG performance is moderated by the presence of state ownership, the baseline model is augmented with a binary SOE indicator and its interaction with each governance variable:

This specification allows the marginal effect of each board attribute to vary depending on state ownership status. A Wald test is conducted to evaluate the joint significance of the interaction terms, thereby determining the moderating role of state ownership.

Recognising the potential simultaneity between board characteristics and ESG performance, together with the dynamic persistence of ESG scores—we employ the two-step system Generalised Method of Moments (GMM) estimator (Arellano and Bover, 1995; Blundell and Bond, 1998). System GMM is ideally suited for panels with a large number of firms (N) and a relatively short time dimension (T), as in our unbalanced 2010–2022 sample, because it exploits internal instruments (lagged levels and first differences) to control for unobserved firm effects and reverse causality. To mitigate instrument proliferation, we collapse the instrument matrix and restrict instrument lags. Instrument validity is verified via Hansen's J-statistic, and absence of second-order serial correlation is confirmed by the Arellano–Bond AR(2) test.

We prefer system GMM to two-stage least squares (2SLS) or static IV methods because ESG scores exhibit high persistence and board structures may co-evolve with ESG outcomes. Static IV approaches require external instruments—which are difficult to find and may suffer weak-instrument bias—whereas system GMM exploits internal, lagged values of the endogenous regressors to deliver consistent estimates in dynamic panels (Baltagi, 2021; Wooldrigde, 2010). The dynamic model is written as:

Specifically, endogenous regressors (board size, independence, gender diversity, CEO duality, transparency, skills disclosure, and the lagged ESG score) are instrumented using their second and higher-order lags in levels for the differenced equation, and lagged differences for the level equation, following the collapsed-instruments approach of Roodman (2009).

To disentangle the effects of board characteristics on each ESG dimension, the model is disaggregated into three sub-components: Environmental (E), Social (S), and Governance (G). A system of seemingly unrelated regressions (SUR) is employed to account for contemporaneous correlation across these equations. The model for each pillar is given by:

Where $dϵ{E,S,G}.$ Cross-equation Wald tests assess whether the coefficients differ significantly across ESG pillars, thus revealing pillar-specific governance effects. When serial correlation is detected, a multivariate system-GMM extension is implemented.

The estimation strategy was guided by diagnostics addressing firm-specific effects, their correlation with regressors, and endogeneity due to persistent ESG performance (Baltagi, 2021; Wooldrigde, 2010). The Breusch–Pagan LM test, Hausman specification test, and Mundlak adjustment were applied to inform estimator selection. LM test statistics (Panel A–B, Table 3) decisively reject homogenous intercepts, with χ² values exceeding 30,000 in the composite ESG model and 23,000 in each pillar model (p < 0.001), while intra-class correlations (ρ = 0.63–0.74) confirm substantial firm-level heterogeneity.

Hausman tests (χ² = 180.63–864.20) and Mundlak adjustments (χ² ≥ 642) indicate significant correlation between latent effects and regressors, ruling out random-effects estimators and supporting two-way fixed effects for all static models, including the moderated specification interacting state ownership with board characteristics (Hausman χ² = 442.01; Mundlak χ² = 671.64). This ensures that firm-specific heterogeneity and potential co-evolution of board composition with government ownership are appropriately controlled.

For the dynamic specification, where ESG performance depends on its lagged values, the study employs a two-step system GMM estimator (Arellano and Bover, 1995; Blundell and Bond, 1998). This approach eliminates firm effects via first differencing and addresses endogeneity by instrumenting regressors with lagged values. Instrument validity and serial correlation are assessed using the Hansen J-statistic and Arellano–Bond AR (2) test (Table 7). The diagnostic results justify a heterogeneous estimation approach: fixed-effects models are used for static specifications to control for correlated firm-level heterogeneity, while system GMM is applied to dynamic ESG models to address persistence and simultaneity.

Table 4 presents the descriptive statistics of the variables. The ESG Score ranges from 0 to 100, with a mean of 46.301 and a standard deviation of 21.495, indicating moderate variability in ESG performance across Asia Pacific firms. The average score below the midpoint of 50 suggests that many firms fall short of international ESG benchmarks, reflecting uneven adoption of sustainability practices and underscoring the need for stronger regulatory alignment or stakeholder pressure. Among the three ESG pillars, governance has the highest mean (51.057), followed by environmental (45.315) and social (43.849). The stronger governance performance likely reflects formal institutional requirements, whereas the relatively weaker environmental and social scores may stem from sectoral asymmetries, weaker disclosure mandates, or limited activism. The wide dispersion in environmental scores (SD = 28.103) highlights heterogeneity in climate-related regulation and enforcement across the region.

Board characteristics display mixed patterns. Average board size is 10.35 members, broadly consistent with global norms, suggesting adequate representational capacity. However, board independence averages only 40.093%, far below the two-thirds threshold often associated with effective oversight, possibly reflecting concentrated ownership structures. Gender diversity remains low, with an average of 9.769% and a median of 8.33%, confirming persistent underrepresentation of women despite rising policy and investor emphasis on inclusivity.

Transparency-related governance indicators also vary. Board transparency, proxied by disclosure of director attendance, has a mean of 0.39, implying that over 60% of firms do not disclose such accountability measures. By contrast, board skills disclosure is high (mean = 0.92), indicating firms' greater willingness to report director qualifications, perhaps as a symbolic signal to investors. CEO duality is present in 29% of cases, reflecting a non-trivial concentration of leadership and oversight roles that may challenge agency theory expectations, particularly in weaker institutional environments.

Regarding ownership and firm-level controls, only 7.3% of firms are state-owned, but the institutional influence of such firms remains significant in strategic industries. The average firm age is 24.968 years, with most between 15 and 36 years, reflecting a relatively mature corporate landscape. Firm size, measured by the natural logarithm of total assets, averages 15.825 with limited dispersion, suggesting relative homogeneity in scale. In contrast, leverage averages 95.255% with a large standard deviation (131.140), highlighting stark cross-firm differences in capital structures shaped by sectoral requirements and institutional contexts. Collectively, these results underscore the uneven ESG landscape of Asia Pacific firms and reinforce the importance of examining how governance mechanisms and state ownership moderate ESG outcomes in subsequent analyses.

Table 5 presents the correlation matrix, showing all coefficients within acceptable ranges and below 0.50, indicating no multicollinearity concerns. ESG Score is positively correlated with firm size (r = 0.396), board independence (r = 0.256), gender diversity (r = 0.223), board skills disclosure (r = 0.109), board transparency (r = 0.066), and firm age (r = 0.260), all significant at the 1% level, suggesting that larger, more independent, gender-diverse, and transparent boards in mature firms achieve higher ESG performance.

State ownership (r = 0.001) and CEO duality (r = −0.013) show negligible correlations with ESG Score, indicating no direct bivariate effect. The strongest negative correlation occurs between board size and independence (r = −0.337), while positive correlations exist between board independence and gender diversity (r = 0.371) and board transparency (r = 0.419), consistent with coexisting governance mechanisms. These results confirm the suitability of including the selected variables in regression models and provide preliminary evidence of their relationships with ESG performance.

Table 6 reports the results of the multicollinearity diagnostics based on variance inflation factors (VIFs). As shown, all VIF values fall well below the conventional threshold of 10, suggesting that multicollinearity is not a significant concern in the model. The mean VIF is 1.201, and the highest observed value is 1.36 for the variable Board Independence. This value remains considerably below the level at which multicollinearity may distort coefficient estimates or inflate standard errors.

In addition, Firm Size and Board Transparency also report relatively higher—but still acceptable—VIF values of 1.32 and 1.303, respectively. Other governance variables such as Gender Diversity (VIF = 1.243), Board Skills Disclosure (VIF = 1.052), and CEO Duality (VIF = 1.014) display very low inflation factors, indicating minimal shared variance with other predictors. The same applies to the control variables Age (VIF = 1.186) and Leverage (VIF = 1.177), as well as the moderating variable State Ownership (VIF = 1.157), all of which remain comfortably within acceptable ranges.

These diagnostic results confirm that the explanatory variables used in the regression models do not suffer from severe multicollinearity. As such, the multivariate estimates presented in subsequent sections are not expected to be compromised by collinearity effects.

Panel regression models were estimated to examine the relationship between board characteristics and ESG performance, addressing heterogeneity, moderation, and endogeneity. Equation (1) uses a two-way fixed-effects model, controlling for firm- and time-specific factors. Equation (2) adds interaction terms between board variables and a state ownership dummy to test moderation effects. Equation (3) employs a two-step system-GMM approach with lagged instruments to account for simultaneity and dynamic bias. Table 7 summarizes the results.

In the baseline fixed-effects model, board independence (β = 0.091, p < 0.001), transparency (β = 3.392, p < 0.001), and skills disclosure (β = 4.195, p < 0.001) positively influence ESG performance, supporting agency and resource dependence theories. Gender diversity is positive but not significant (β = 0.023, p = 0.439), while CEO duality shows a weak negative effect (β = −0.963, p = 0.053). These results align with prior studies emphasizing the role of independent oversight, transparency, and board expertise in ESG outcomes.

Interaction terms with state ownership reveal significant moderation (Wald F = 5.73, p < 0.001). Gender diversity becomes significant (β = 0.069, p = 0.031), while board size turns negatively significant (β = −0.182, p = 0.038), indicating that political influence can weaken the benefits of larger boards and diversity. The main effect of state ownership remains insignificant (β = 3.901, p = 0.371), suggesting its impact is contingent rather than direct.

System-GMM results show that transparency (β = 0.800, p < 0.001) and gender diversity (β = 0.018, p = 0.019) remain robust, while board independence loses significance, implying that some earlier estimates reflect firm-specific traits or simultaneity bias. Control variables behave as expected: firm size is positively associated with ESG, and leverage negatively affects ESG only in dynamic models. Diagnostic tests confirm valid instruments and no second-order serial correlation (Arellano–Bond AR(2) p = 0.050; Hansen J-statistic p = 0.007).

Hypothesis testing shows consistent support for gender diversity (H3) and transparency (H4), partial support for board independence (H2) and CEO duality (H5), and limited support for board size (H1) and skills disclosure (H6). Moderation hypotheses (H7a–H7f) receive selective support, with SOE weakening the effects of board size and gender diversity in certain contexts.

Comparative evidence from Asia-Pacific and African markets confirms similar patterns: board independence, gender diversity, and skills disclosure generally enhance ESG outcomes, while board size shows mixed effects. State and institutional ownership moderate board–ESG links, but these effects vary by governance context and institutional environment. Equation (4) extends the analysis to environmental, social, and governance pillars using seemingly unrelated regressions (SUR) to account for cross-equation correlations (Table 8).

The analysis reveals substantial heterogeneity across ESG components. Board size positively affects environmental (β = 0.508, p < 0.001) and social performance (β = 0.597, p < 0.001) but negatively impacts governance (β = −0.487, p < 0.001), suggesting larger boards may enhance outward-facing ESG initiatives while reducing internal governance efficiency. Wald tests confirm significant differences across pillars (χ² = 147.1, p < 0.001).

Board independence shows consistent positive effects across environmental (β = 0.129), social (β = 0.310), and governance (β = 0.296) dimensions, with the strongest impact on social outcomes, highlighting their role in ethical practices and stakeholder engagement. Gender diversity is positively associated with all ESG pillars, particularly governance (β = 0.275), reinforcing the importance of diverse boards for transparency and ethical standards. CEO duality exhibits a mixed pattern: positive for environmental (β = 4.191, p < 0.001) and social (β = 3.259, p < 0.001) performance but negative for governance (β = −1.024, p = 0.045), suggesting centralized leadership can expedite outward-facing initiatives but undermine internal accountability.

Board transparency positively influences governance (β = 2.204, p < 0.001) but negatively affects environmental outcomes (β = −2.935, p < 0.001) and is insignificant for social scores, indicating its role may be reactive to governance pressures. Board skills disclosure strongly supports governance (β = 5.173, p < 0.001) but has limited or negative effects on environmental and social dimensions, suggesting professional expertise disclosures primarily enhance credibility rather than broader ESG impact.

State ownership (SOE) moderates these effects. SOE × board size interactions are negative across all ESG pillars, strongest for governance (β = −0.753, p < 0.001), while SOE × gender diversity reduces environmental (β = −0.241, p = 0.011) and governance (β = −0.175, p = 0.042) performance, indicating potential tokenism or politicization in state-influenced boards. SOE × transparency positively affects governance only (β = 5.131, p = 0.002), and SOE × skills disclosure shows weak negative effects in the social domain (β = −4.915, p = 0.053).

Control variables remain stable: firm size and age positively correlate with all ESG dimensions, while leverage is negatively associated, reflecting financial constraints. Wald χ² tests confirm distinct effects across ESG pillars, emphasizing the multidimensional nature of ESG and the limitations of aggregate scores. Overall, board characteristics exert domain- and context-specific effects, highlighting the need to align governance structures with targeted ESG objectives, particularly in state-owned contexts.

This study examined how state ownership moderates the relationship between board characteristics and ESG performance using data from 1,175 publicly listed Asia-Pacific firms from 2010 to 2022. Employing fixed-effects regression, interaction modelling, and pillar-specific analysis, the findings reveal that board independence, gender diversity, board transparency, and director skill disclosure generally enhance ESG outcomes, but these effects vary across ownership types. In state-owned firms, the positive impact of board size and gender diversity is weakened, likely due to politically motivated appointments, whereas board transparency is amplified, reflecting heightened reputational and public-sector accountability. The traditionally negative effect of CEO duality is also somewhat mitigated under state ownership, suggesting alignment between concentrated leadership and state-mandated sustainability goals.

Theoretically, the results highlight that board structures are institutionally contingent. Agency theory and resource dependence perspectives are extended by showing that state-owned firms operate under a hybrid governance logic where political legitimacy, social responsibility, and stakeholder accountability intersect. Mechanisms effective in private firms may function differently in state-controlled contexts, emphasizing the importance of considering ownership regimes in governance research.

For policymakers, the findings suggest that governance regulations should be tailored rather than one-size-fits-all. Enforcing director skill disclosure, mandating minimum independence ratios, and strengthening transparency can improve ESG accountability in state-influenced firms. For corporate practitioners, particularly governance committees, the evidence underscores the need to move beyond symbolic diversity targets and focus on board members' substantive contributions to sustainability strategy.

Limitations point to avenues for future research. The binary classification of state ownership does not capture the degree or type of control, which could be addressed with continuous variables or distinctions between state entities. Additionally, triangulating ESG measures with alternative databases or qualitative assessments would enhance construct validity. Longitudinal studies examining governance-ESG dynamics during privatization or regulatory reforms could further illuminate temporal aspects of board effectiveness. Overall, the study underscores that ESG outcomes are shaped by both internal board structures and broader institutional contexts, advancing theoretical understanding and practical governance design across diverse economic systems.