Board gender diversity, ESG and earnings management: evidence from the Southeast Asian firms

The increasing representation of women on corporate boards is not only a step toward gender equality but also a strategic component of improved corporate governance and financial transparency. As global stakeholders demand greater accountability and ethical conduct, board gender diversity (BGD) has emerged as a crucial mechanism influencing firms' financial behaviour and sustainability practices. Despite growing global evidence linking diverse boards to stronger governance outcomes (Bhatia and Gulati, 2021), the impact of female directors on earnings management (EM) ^[1] remains underexplored in emerging markets, especially in Southeast Asia, an institutional setting characterized by heterogeneous governance systems, concentrated ownership structures, and distinct cultural norms (Oehmichen, 2018). In particular, ESG practices and corporate governance frameworks vary markedly across Asian economies, reflecting differences in regulatory enforcement, board diversity mandates, and sustainability disclosure regimes ranging from strong, well-enforced systems in Singapore and Malaysia to more evolving frameworks in Indonesia and Thailand. Consistent with institutional theory, these cross-country differences suggest that the effectiveness of board gender diversity and ESG engagement in improving financial reporting quality is likely to be context-dependent, motivating a comparative analysis.

While prior research suggests that female participation on boards can mitigate earnings management (EM) and enhance financial reporting credibility (Srinidhi et al., 2011; Arun et al., 2015), empirical findings remain mixed and context-dependent (Gull et al., 2018; Orazalin, 2020). Much of the existing evidence is derived from developed economies with robust regulatory frameworks and mature governance mechanisms, leaving a significant gap in understanding how gender diversity operates in institutionally weaker, culturally diverse environments, such as those in Southeast Asia. This gap is particularly important because board effectiveness in emerging markets often depends not only on regulatory compliance but also on informal norms, ownership concentration, and societal expectations regarding gender roles (Nguyen et al., 2015).

To further ground the study in theory, we draw on three complementary perspectives that explain how board gender diversity may influence earnings management. First, resource dependence theory suggests that female directors enhance oversight through their diverse expertise, ethical sensitivity, and stronger stakeholder orientation, thereby constraining opportunistic managerial behaviour. Second, social role theory reinforces this view by highlighting women's tendency toward greater prudence and ethical judgement, attributes that align with more transparent financial reporting. Finally, groupthink theory cautions that these benefits may be limited in environments where female representation is symbolic or where entrenched board dynamics restrict dissent, conditions that are not uncommon in emerging markets. Together, these theories suggest that the impact of gender-diverse boards on earnings management depends not only on the presence of women but also on the organizational environment that enables or constrains their effectiveness. This perspective underpins our expectation that the influence of BGD may be strengthened in firms with authentic ESG engagement and weakened where such engagement is superficial.

Beyond gender diversity alone, Environmental, Social, and Governance (ESG) performance has become a defining feature of modern corporate governance, reflecting a firm's ethical commitment, stakeholder orientation, and long-term strategic vision (Eccles et al., 2014; Kolk, 2016). Yet, limited empirical evidence examines how ESG practices interact with board gender diversity to influence earnings management. We argue that ESG can serve as a contextual enabler, strengthening the effectiveness of female directors by embedding their ethical and stakeholder-oriented values within a supportive corporate culture. Conversely, in firms where ESG is superficial or symbolic, the benefits of board diversity may be diluted or even reversed, a phenomenon often associated with “greenwashing” or “governance washing.” Additionally, the growing emphasis on ESG practices has heightened interest in how sustainability-oriented mechanisms function in developing Asian economies, where governance systems and enforcement remain uneven. Recent evidence suggests that ESG engagement in these contexts is often driven by legitimacy pressures and evolving regulatory expectations, resulting in a combination of substantive initiatives and symbolic disclosure (Abu Afifa et al., 2024a, b). This institutional reality makes ASEAN a particularly relevant setting for examining ESG not merely as a governance outcome, but as a contextual factor that conditions how board gender diversity influences financial reporting behaviour.

Using a dataset of listed firms domiciled in Southeast Asian countries, our main findings show that board gender diversity is generally associated with lower earnings management, particularly for accrual-based manipulation. This suggests that female directors are more effective in strengthening financial reporting oversight than in constraining operational decisions. However, the moderating role of ESG engagement is highly context-dependent. Rather than uniformly reinforcing the governance role of gender-diverse boards, stronger ESG engagement often weakens the marginal monitoring effect of female directors, suggesting a substitution effect between ESG mechanisms and board-level oversight. Country-level evidence further reveals substantial heterogeneity: ESG engagement substitutes for board monitoring in Malaysia and Indonesia, has limited incremental influence in Singapore's strong governance environment, and exhibits mixed or symbolic effects in Thailand, where female board representation remains relatively low. Collectively, these results highlight that the effectiveness of board gender diversity in mitigating earnings management is contingent on institutional quality and the authenticity of ESG practices, rather than their mere presence.

This study makes significant incremental contributions to the literature in several ways. First, unlike prior studies which focus on a single country (Primacintya and Kusuma, 2025), our paper provides one of the few cross-country examinations of the BGD–EM relationship across four major Southeast Asian economies, i.e. Malaysia, Indonesia, Singapore, and Thailand, offering comparative insights into how institutional environments moderate governance effectiveness. Second, it extends prior research by examining ESG performance as a contextual moderator, showing that the influence of gender-diverse boards on earnings management depends on the authenticity of firms' sustainability engagement. This perspective clarifies that ESG is not merely a parallel governance mechanism but a condition that can strengthen or weaken the effectiveness of board diversity. Third, while prior studies (Akter et al., 2024) focus on REM, our research, by distinguishing between accrual-based and real earnings management, uncovers which forms of managerial discretion are more effectively constrained by female directors. Finally, it contributes to theory by integrating resource dependence, social role, and groupthink perspectives to explain why board diversity may yield divergent effects depending on institutional quality and ESG practices. Overall, this research provides new evidence from a region that remains underrepresented in the global governance literature, offering practical implications for regulators, investors, and firms seeking to promote genuine gender inclusivity and sustainable governance reforms in emerging markets.

The remainder of this paper is structured as follows: Section 2 provides an overview of Southeast Asia's institutional setting. Section 3 reviews relevant literature and develops our framework. Section 4 details data, sample, variables, and methodology. Section 5 shows multivariate results, including effects and robustness checks. Section 5 discusses findings, implications, and significance. Section 6 concludes with a summary and recommendations for policymakers, firms, and investors.

The institutional settings of Asian markets particularly Thailand, Malaysia, Singapore, and Indonesia are distinctive and heterogeneous, often characterized by weaker, less uniform governance infrastructures than those found in Western economies (Oehmichen, 2018). In these contexts, ESG engagement is shaped less by mature internal governance systems and more by external institutional pressures related to legitimacy, regulatory signalling, and stakeholder expectations. Recent evidence from developing Asian economies shows that ESG practices often emerge in response to external scrutiny rather than internal governance maturity, leading to a hybrid pattern that combines substantive sustainability efforts with symbolic disclosure (Abu Afifa et al., 2024a, b). This institutional environment is especially relevant for ASEAN economies, where variations in ownership concentration, regulatory quality, and enforcement intensity condition the effectiveness of governance mechanisms.

Building on this perspective, ESG engagement is conceptualized in this study not as a direct substitute for board governance but as an institutionally embedded mechanism that shapes how board attributes influence firm behaviour. In particular, ESG may either reinforce or constrain the effectiveness of board gender diversity in limiting earnings management. This approach aligns with recent evidence showing that ESG is not a monolithic governance construct but a multidimensional mechanism whose effects depend on mediating and moderating conditions (Nguyen et al., 2025). Treating ESG as an interactional force, therefore, provides a stronger theoretical foundation for examining its role in conditioning the governance impact of female board representation in developing Asian contexts.

The governance landscape across Southeast Asia further underscores the importance of contextual analysis. Malaysia has actively promoted board gender diversity, targeting at least 30% female representation (Alahdal et al., 2024), while Thailand and Indonesia are characterized by family-dominated and oligarchic ownership structures, respectively (Oehmichen, 2018). Singapore, by contrast, operates within a more advanced regulatory and market environment (Hardiningsih et al., 2024; Hermawan et al., 2025). These institutional differences imply that the governance benefits of board gender diversity and ESG engagement are unlikely to be uniform across countries. Furthermore, across ASEAN markets, female board representation is increasingly associated with stronger oversight, higher ethical standards, and more conservative financial decision-making, all of which are linked to lower earnings management (Alves, 2023; Srinidhi et al., 2011). At the same time, board gender diversity is recognized as a key driver of ESG adoption in Asian firms, with ESG performance influencing how diversity translates into governance and reporting outcomes (Alahdal et al., 2024; Hermawan et al., 2025). Given the central role of high-quality financial reporting in mitigating agency conflicts and supporting informed investment decisions (Assad et al., 2023a, b), this study argues that ESG engagement conditions rather than replace the governance role of female directors in enhancing financial reporting quality.

Prior research attributes gender differences in business behaviour to variations in ethical orientation and risk preferences, which influence managerial judgement and governance outcomes (Croson and Gneezy, 2009; Palvia et al., 2014; Abbott et al., 2012). From a resource dependence perspective, female directors represent valuable board resources, contributing diverse expertise, external linkages, and enhanced monitoring, thereby improving decision-making and financial reporting transparency (Hillman et al., 2000; Brahma et al., 2021; Dobija et al., 2022). Social role theory further suggests that women's stronger ethical sensitivity and stakeholder orientation promote integrity, prudence, and accountability in financial reporting (Eagly and Karau, 2002; Boulouta, 2013a). However, groupthink theory cautions that these benefits may be weakened when female representation is symbolic or constrained by institutional and cultural barriers, limiting their influence on board deliberations (Janis, 1972). Accordingly, the governance impact of board gender diversity depends not only on representation but also on meaningful inclusion and supportive institutional conditions.

The relationship between board gender diversity (BGD) and earnings management (EM) has garnered significant attention in corporate governance research, yet remains both theoretically and empirically debated. From a resource dependence perspective, female directors improve board effectiveness by offering diverse expertise, external connections, and independent viewpoints that enhance monitoring and limit managerial discretion in financial reporting (Hillman et al., 2000). This improved oversight is expected to restrict opportunistic accounting choices and mitigate earnings management. Complementing this view, social role theory suggests that women tend to demonstrate stronger ethical sensitivity, risk aversion, and stakeholder focus, traits that encourage transparency and conservative financial reporting (Eagly and Karau, 2002). Together, these theories imply that gender-diverse boards are better equipped to challenge managerial opportunism and curb earnings management. Empirical evidence largely backs this reasoning. Studies such as Srinidhi et al. (2011), Arun et al. (2015), and García Lara et al. (2017) show that firms with higher female board representation have better accrual quality and lower earnings manipulation, attributing these results to improved monitoring and ethical oversight. However, these positive effects are not universal and may depend on the institutional and cultural contexts in which boards operate. In emerging markets, particularly in Southeast Asia, boards often operate within concentrated ownership structures, patriarchal norms, and informal governance systems, which can either facilitate or hinder the effectiveness of female directors. When institutional support for inclusion is strong and board roles are substantive, the monitoring and ethical advantages predicted by resource dependence and social role theories are more likely to materialize. Accordingly, and consistent with this dominant theoretical perspective, we propose the following hypothesis:

Despite the arguments advanced by resource dependence and social role theories, an alternative perspective suggests that board gender diversity may not necessarily constrain earnings management in all settings. Drawing on groupthink theory (Janis, 2008), diversity can lose its governance value when board processes are dominated by consensus-seeking behaviour, hierarchical power structures, or the influence of entrenched controlling shareholders. In such environments, the monitoring advantages associated with diverse expertise and ethical orientation may be overridden by pressures to preserve harmony and avoid conflict, limiting effective scrutiny of managerial decisions. Minority board members, particularly women, may experience restricted voice, informal exclusion from key discussions, or subtle pressure to conform to dominant norms, thereby weakening their capacity to challenge managerial discretion or opportunistic financial reporting (Loh and Nguyen, 2018; Wu et al., 2024).

This perspective directly challenges the assumptions underlying H1a, which presumes that the presence of female directors translates into substantive oversight and ethical influence. Under groupthink conditions, however, gender diversity may be largely symbolic rather than functional. When appointments are driven by compliance or reputational motives rather than genuine empowerment, female directors may lack the authority or institutional support necessary to change their reporting behaviour. In such cases, diversity does not enhance monitoring but may instead coexist with, or even mask, weak governance practices. Consequently, the expected negative relationship between board gender diversity and earnings management may be attenuated, insignificant, or reversed, particularly in institutional settings where board independence and open dissent are limited. This concern is especially salient in Southeast Asian institutional environments, where corporate governance is often characterized by concentrated ownership, family control, relational ties, and culturally embedded norms regarding authority and gender roles. In these settings, female directors may be appointed to satisfy regulatory expectations or symbolic diversity objectives rather than to exercise substantive monitoring power. When gender diversity is largely tokenistic, the presence of women on boards may have little effect on earnings management or, in some cases, may even exacerbate coordination problems and reduce effective oversight. Prior empirical evidence from emerging markets, which documents insignificant or adverse effects of board gender diversity, lends support to this view (Sun et al., 2011; Gull et al., 2018; Orazalin, 2020). Accordingly, where institutional constraints limit meaningful board participation and dissent, the governance benefits of gender diversity may fail to materialize. This leads to the following competing hypothesis:

Environmental, Social, and Governance (ESG) performance has become an integral dimension of corporate governance and financial transparency, reflecting a firm's ethical culture and stakeholder orientation (Eccles et al., 2014; Kolk, 2016). Building on the discussion of board gender diversity (BGD) and earnings management (EM), this study posits that ESG performance moderates the relationship between BGD and EM by shaping the organizational context in which female directors operate. In organizations with relatively strong and substantive ESG engagement, the ethical and stakeholder-oriented values often associated with female directors may be reinforced by a more supportive corporate culture. This alignment can potentially enhance the capacity of gender-diverse boards to constrain earnings management by strengthening monitoring processes and encouraging greater transparency and more conservative financial reporting practices (Galbreath, 2018; Post et al., 2011).

However, this moderating role of ESG may not be uniform across institutional settings. In firms operating within weak governance environments, ESG engagement may be largely symbolic, driven by reputational considerations rather than genuine commitment to ethical conduct. In such cases, high ESG scores may reflect greenwashing or governance washing rather than substantive improvements in oversight, offering limited support for female directors when challenging managerial discretion. Consequently, ESG practices may fail to strengthen and may even weaken the governance benefits of board gender diversity, particularly where institutional enforcement and board empowerment are limited (Harjoto et al., 2015; Jizi et al., 2014). Taken together, these arguments suggest that ESG performance acts as a contextual moderator whose effect depends on the authenticity of sustainability engagement and the surrounding institutional environment. Accordingly, we propose the following hypothesis:

The conceptual framework summarizing the link between BGD, ESG and EM is given in Figure 1 below:

This study employs an empirical approach to investigate the intricate relationship between board gender diversity (BGD), earnings management (EM), and the moderating effect of Environmental, Social, and Governance (ESG) performance in the Southeast Asian context. The dataset comprises publicly listed non-financial firms from four major Southeast Asian economies, i.e. Malaysia, Indonesia, Singapore, and Thailand, selected purposively for the period 2012 to 2022 ^[2]. These countries were chosen due to their distinct institutional contexts, divergent corporate governance structures, and varying levels of board gender diversity. This selection enables a nuanced, cross-country analysis that captures contextual variations often overlooked in single-country studies. The sample period is particularly significant, as it coincides with heightened global emphasis on corporate governance reforms, sustainability efforts, and the promotion of gender diversity in leadership, making it an ideal period for empirical investigation in emerging markets. Financial and ESG data were sourced from Refinitiv DataStream, a financial database known for its comprehensive and reliable coverage of corporate and sustainability metrics. The initial sample included all non-financial firms listed on the stock exchanges of the selected countries. Financial institutions were excluded due to their distinct regulatory frameworks, disclosure requirements, and capital structures, which could distort comparative analysis. Further, firm-year observations with missing data on key variables were also excluded. Following this data filtering process, our final panel dataset consisted of 1,643 firm-year observations.

Accruals earnings management was estimated using discretionary working capital accruals (DisWCA), serving as a proxy for managerial manipulation of reported earnings. This proxy focuses on short-term earnings management through the manipulation of working capital. DisWCA is derived from the Modified Jones Model (Jones, 1991), a widely accepted measure for separating the discretionary and nondiscretionary components of accruals. The model for working capital accruals is expressed as follows:

where $DisWCAi,t$ is the total working capital accruals for firm i in year t, calculated as the change in non-cash current assets less the change in current liabilities (excluding short-term debt and taxes payable). $ΔREVi,t$ is the change in revenues, and $PPEi,t$ is the gross property, plant, and equipment. All variables are scaled by lagged total assets ( $Ai,t−1$⁠). α0, α1, α2 are the Regression coefficients to be estimated by the model. The discretionary portion of working capital accruals (⁠$DisWCA$⁠) is represented by the residuals (⁠$εi,t$⁠) from this regression. The absolute value of these residuals is used as the proxy, where a higher value of | $DisWCA$ | indicates a greater level of discretionary manipulation of working capital and, consequently, higher earnings management.

While the Modified Jones Model and its variants have been extensively used in prior research (Dechow et al., 1995; DeFond and Subramanyam, 1998; Kothari et al., 2005; Mouselli et al., 2012), their advantages and limitations are well-documented (Guy et al., 1996; Young, 1999; Thomas and Zhang, 2000; Lo, 2008; Dechow et al., 2010; DeFond, 2010). Despite some acknowledged shortcomings, these models remain the most prevalent and accepted approaches for estimating discretionary accruals in the absence of a universally superior alternative (Botsari and Meeks, 2008). The focus on working capital accruals is explicitly justified by their susceptibility to short-term manipulation and their relevance in capturing earnings management practices that are more likely to be influenced by board composition and oversight (McNichols and Stubben, 2018).

The key to the Modified Jones Model lies in its adjustment to the revenue variable, which accounts for managerial discretion. The discretionary portion is not just the residual of the above regression. Instead, it is calculated by first predicting the non-discretionary accruals using the coefficients from the regression, but with a crucial modification: the change in revenue is adjusted by the change in accounts receivable. The formula for predicted non-discretionary accruals is:

Where αˆ0, αˆ1, αˆ2: The estimated coefficients from the initial regression. $ΔRECi,t$⁠: The change in accounts receivable for firm i in year t. This adjustment is the “modification” that makes the model more robust by preventing managers' manipulation of revenue from being incorrectly classified as non-discretionary. The Discretionary Working Capital Accruals (⁠$DisWCA$⁠) are calculated as the difference between a firm's actual total accruals and its predicted non-discretionary accruals:

The absolute value of this number, ∣ $DisWCAi,t$∣, is often used as the final measure of the magnitude of earnings management.

To measure real earnings management (REM), we employ three proxies based on the seminal work of Roychowdhury (2006), which captures managerial manipulation of real business activities to meet earnings targets. Unlike accrual-based earnings management, REM involves actual changes in the timing or structuring of real transactions, making it more difficult to detect but potentially more costly to the firm's long-term value.

The first measure of REM captures abnormal production costs, built on the premise that managers may overproduce to spread fixed costs over a larger number of units, thereby reducing the cost of goods sold and increasing reported earnings. We estimate the normal level of production costs using the following regression model:

where $PRODit$ = $COGSi,t$ + $ΔINVi,t$⁠, representing the sum of the cost of goods sold and the change in inventory from period t−1 to t. $Salesi,t$ represents annual sales, $ΔSalesi,t$ denotes the change in sales from period t−1 to t, and $ΔSalesi,t−1$ is the change in sales from t−2 to t−1. $Assetsi,t−1$ represents lagged total assets. The residuals from this regression represent abnormal production costs (APROD). Higher values of APROD indicate overproduction relative to normal levels, suggesting upward earnings management through increased production to reduce per-unit fixed costs and improve reported profit margins.

The second proxy measures abnormal discretionary expenses, based on the premise that managers may reduce discretionary spending on research and development, advertising, and selling, general, and administrative expenses to boost short-term earnings. We estimate normal discretionary expenses using:

where $DISXi,t$ represents the sum of research and development expenses, advertising expenses, and selling, general, and administrative expenses. $Salesi,t−1$ represents lagged annual sales, and $Assetsi,t−1$ represents lagged total assets. The residuals from this regression, multiplied by −1, represent abnormal discretionary expenses (ADISX). Higher values of ADISX indicate lower than normal discretionary spending, suggesting upward earnings management through the reduction of discretionary expenses that may harm long-term firm value.

The third measure captures abnormal cash flow from operations, recognizing that managers may manipulate the timing of sales through aggressive credit terms, price discounts, or channel stuffing to accelerate revenue recognition. We estimate normal cash flow from operations using:

where $CFOit$ represents cash flow from operations, $Salesi,t$ represents annual sales, and $ΔSalesi,t$ denotes the change in sales from period t−1 to t. The residuals from this regression, multiplied by −1, represent abnormal cash flow from operations (ACFO). Higher values of ACFO indicate lower-than-normal cash flows relative to sales, suggesting potential manipulation of sales timing or terms to inflate current-period revenues.

Following prior literature (Cohen et al., 2008; Zing, 2012), we construct an aggregate measure of real earnings management by combining the three individual proxies:

This aggregate measure captures the overall extent of real earnings management activities, with higher values indicating greater manipulation of real business activities to manage earnings upward. The use of an aggregate measure is particularly important as managers may employ multiple REM techniques simultaneously, and the combined effect provides a more comprehensive assessment of real earnings management intensity.

All REM measures are industry-year adjusted to control for industry-specific and time-varying factors that may affect normal levels of production costs, discretionary expenses, and cash flows. This approach ensures that our measures capture firm-specific deviations from expected levels rather than industry or temporal patterns.

Board gender diversity (BGD) is measured as the proportion of female directors on the board (FemaleonBoard).This continuous measure is consistent with established practices in corporate governance research (Arun et al., 2015; Orazalin, 2020), providing a direct representation of female representation in the boardroom. The moderating variable, ESG engagement, is measured using the firm's aggregate ESG score sourced from Refinitiv DataStream. This score combines the environmental, social, and governance engagement indicators, providing a comprehensive assessment of a company's commitment to sustainability and ethical business conduct. Higher ESG scores reflect stronger sustainability performance. Following prior literature (e.g. Assad et al., 2023a, b; Hemmings et al., 2022), we control for omitted variable bias and capture firm-specific factors that may influence earnings management, several control variables were included in the analysis, each grounded in theoretical and empirical literature. Firm size (SIZE), measured as the natural logarithm of total assets, accounts for the complexity and resources of larger firms, which can impact financial reporting practices. Leverage (LEV), calculated as total debt over total assets, reflects financial pressure that may incentivize earnings manipulation. Investment intensity (INV), measured by capital expenditures relative to total assets, captures the potential for discretionary accruals linked to capital investments. Profitability (ROA), defined as net income divided by total assets, can affect managerial incentives for earnings management. Altman's Z-score (ZSCORE) is used to assess financial stability, with lower scores indicating higher bankruptcy risk and a greater likelihood of earnings manipulation. Growth (GROWTH), measured as the annual percentage change in total assets, reflects different earnings management incentives between high-growth and mature firms. Lastly, Board size, measured by the total number of directors, is included to account for the board's monitoring capacity and its influence on earnings management. Table 1 below provides the details of all the variables.

To examine the direct relationship between board gender diversity and earnings management (H1a), the baseline regression model is specified as follows:

Where: $EMit$ represents (i) the accrual earnings management measure (DisWCA) and (ii) real earnings management (REM_agg) for firm i in $yeart$⁠. $FemaleonBoardit$ is the percentage of female directors on the board for firm i in $yeart$⁠. $∑t=1iFirmControls$ represents the vector of control variables (SIZE, LEV, INV, ROA, ZSCORE, GROWTH, BoardSize) for firm i in $yeart$⁠. * $IndustryFixed$⁠, $Year$⁠, and $CountryFixed$ are dummy variables for industry, year, and country fixed effects, respectively, accounting for unobserved time-invariant and country-specific characteristics. * $εit$ is the error term.

To assess the moderating effect of ESG performance on the relationship between BGD and EM (H2), an interaction term $(FemaleonBoardit*ESGScoreit$⁠) is included in the model:

Where: * $ESGScoreit$ is the ESG performance score for firm i in year t. * FemaleonBoard_it × ESGScore_it is the interaction term capturing the moderating effect of ESG performance. * All other variables are as defined in the baseline model.

We employ a fixed effects panel regression model with firm fixed effects to control for unobserved time-invariant firm characteristics and country characteristics. In addition, this approach addresses potential issues of heteroscedasticity and serial correlation within firms, ensuring more reliable inference. The inclusion of fixed effects for industry, year, and country is crucial for controlling unobserved heterogeneity and isolating the specific effects of the variables of interest. Descriptive statistics and correlation matrices for all variables are presented in Table 2 and Table 3, respectively, in the Results section. Multicollinearity was assessed using the Variance Inflation Factor (VIF), with all values well below the conventional threshold of 10, indicating no serious concerns regarding multicollinearity. Furthermore, we centred the BGD and ESG variables before creating the interaction term to reduce multicollinearity and enhance the interpretability of the results. All VIF values are below 5 (well under the threshold of 10), indicating no multicollinearity issues. We also tested uncentred variables, and the results remain consistent, confirming that the conclusions are not affected by the choice of centring.

Table 2 Panel A provides a descriptive overview of the main variables. The mean absolute discretionary accruals (|DisWCA| = 0.031) and aggregate real earnings management (REM_agg = 0.204) indicate moderate levels of earnings manipulation, consistent with prior studies in emerging markets such as Orazalin (2020). The average board gender diversity (FemaleonBoard = 15.12%) indicates that women's representation remains relatively low, echoing findings by Srinidhi et al. (2011) and Brahma et al. (2021) that gender imbalance persists despite growing awareness. The mean ESG score (45.09) reflects moderate sustainability engagement, comparable to Hermawan et al. (2025), who reported similar ESG performance levels among emerging markets firms. Table 2 Panel B shows clear cross-country differences. Earnings management is higher in Malaysia and Indonesia than in Singapore and Thailand, reflecting weaker governance environments (Kyaw et al., 2015). Female board representation is highest in Singapore (19.8%) and lowest in Malaysia (8.8%), consistent with Hermawan et al. (2025), who link gender diversity to stronger institutions. Thailand's higher ESG score (54.4) compared to Singapore's lower score (42.1) suggests a varying commitment to sustainability. These patterns align with Orazalin (2020), highlighting how institutional and cultural contexts shape board diversity, ESG engagement, and earnings management across Southeast Asia. Table 2, Panel C, shows weak correlations and VIF among the variables, indicating no concerns about multicollinearity. ESGScore's positive association with firm size and board size suggests that larger firms are more engaged in ESG activities.

Table 3 Panel A shows that board gender diversity (FemaleonBoard) is negatively and significantly associated with accrual-based earnings management (AEM) for the full sample and for Malaysia, Indonesia, and Singapore, supporting H1a that greater female representation mitigates the propensity of earnings management. This finding aligns with social role theory (Eagly and Karau, 2002), which posits that women's ethical orientation and risk aversion promote more transparent reporting, and resource dependence theory (Hillman et al., 2000), which suggests that gender-diverse boards bring broader perspectives and stronger monitoring capabilities. The insignificant result for Thailand, however, supports H1b and may reflect the influence of groupthink theory (Janis, 1972), which suggests that token female presence or cultural barriers limit their effectiveness. These results are consistent with prior studies, such as those by Srinidhi et al. (2011) and Arun et al. (2015), which have found that female directors improve accrual quality. Table 4, Panel B, shows that board gender diversity has a mixed effect on real earnings management (REM), contrasting with the consistent negative relationship observed for accrual-based manipulation ^[3]. For Indonesia, the negative and significant coefficient supports H1a and aligns with social role theory and resource dependence theory, suggesting that female directors' ethical orientation and monitoring skills curb REM (Srinidhi et al., 2011; Arun et al., 2015). In Thailand, however, the positive association supports H1b and reflects groupthink theory (Janis, 1972), implying limited influence of women in less supportive institutional settings (Orazalin, 2020; Gull et al., 2018). Overall, female directors appear more effective at constraining accrual-based than real earnings management, indicating stronger oversight of accounting judgements than of operational decisions.

Table 4 reports the moderating effect of ESG engagement on the relationship between board gender diversity (BGD) and earnings management. For the full sample (Column 1), the interaction term between female board representation and ESG performance is positive and statistically significant for REM. Given that lower REM values indicate higher earnings quality and that BGD is expected to mitigate earnings management, this result suggests that stronger ESG engagement weakens the negative effect of gender-diverse boards on earnings management. In other words, higher ESG performance flattens the BGD–REM relationship, implying that the incremental monitoring role of female directors diminishes in ESG-intensive firms. This finding is consistent with a substitution view of governance mechanisms. While resource dependence and social role theories suggest that female directors enhance oversight through ethical orientation and stakeholder sensitivity, their marginal contribution is more pronounced when alternative controls are weak. In firms with strong ESG engagement, formal sustainability frameworks, enhanced disclosure, and reputational monitoring, transparency may already be institutionalized, thereby reducing reliance on board-level monitoring. Alternatively, the result aligns with symbolic governance and groupthink arguments, particularly in emerging markets, where ESG performance may be driven by disclosure intensity rather than substantive governance improvements. In such contexts, ESG engagement and board gender diversity may coexist as legitimacy-enhancing mechanisms without reinforcing effective oversight, leading to a weaker observed impact of BGD on earnings management.

Table 4 (Columns 2–5) reports country-specific results on the moderating role of ESG engagement in the relationship between board gender diversity (BGD) and earnings management, revealing notable cross-country heterogeneity. Consistent with the aggregate results, the interaction term is positive and statistically significant in Malaysia and Indonesia, indicating that higher ESG engagement weakens, rather than strengthens, the marginal monitoring effect of female directors. In these contexts, ESG appears to serve as a substitute governance mechanism, enabling formal sustainability frameworks and disclosure practices to reduce reliance on board-level oversight. This interpretation remains consistent with social role and resource dependence theories once alternative governance controls are recognized (Eagly and Karau, 2002; Hillman et al., 2000). In Singapore, the interaction term is insignificant, suggesting that strong institutional governance limits the incremental effects of both ESG engagement and board gender diversity on earnings management, consistent with prior evidence from high-governance settings (Arun et al., 2015). The interaction is also insignificant in Thailand, which may reflect the relatively low representation of women on boards, constraining their ability to influence monitoring outcomes. This finding aligns with arguments about groupthink and tokenism, whereby limited female representation weakens the effectiveness of board diversity. Overall, the country-level evidence indicates that ESG's moderating role is highly context-dependent. ESG substitutes for gender-based monitoring in Malaysia and Indonesia, has limited incremental relevance in Singapore, and yields weak or potentially symbolic effects in Thailand. These results support a nuanced interpretation of H2, showing that ESG engagement conditions, rather than uniformly enhancing, the governance role of gender-diverse boards across Southeast Asian institutional settings.

Table 5 collectively analyzes the moderating effects of the Environmental, Social, and Governance (ESG) pillars on the relationship between board gender diversity and earnings management. For the full sample, Malaysia, and Indonesia, the interaction terms between female board representation and the ESG pillars are generally positive and significant, indicating that stronger ESG engagement enhances the ability of female directors to constrain earnings management, thereby supporting H2 and aligning with social role and resource dependence theories (Eagly and Karau, 2002; Hillman et al., 2000). In contrast, interactions are largely insignificant in Singapore, consistent with a strong institutional environment that limits incremental governance effects, whereas results in Thailand are weak or mixed, particularly for the environmental and governance pillars, suggesting selective or symbolic ESG adoption. Overall, the evidence shows that ESG pillars reinforce the governance role of gender-diverse boards only where sustainability and governance practices are substantive, underscoring the context-dependent nature of ESG as a complementary, rather than uniformly effective, governance mechanism across Southeast Asian markets.

To mitigate potential common method bias and simultaneity concerns, this study adopts a temporal separation design by lagging all explanatory variables by one year. Board gender diversity, ESG engagement, and control variables are measured at t–1, while earnings management is measured at t. This approach strengthens causal inference by reducing contemporaneous alignment between governance characteristics and reporting outcomes. The models are estimated using firm fixed effects with year dummies to control for unobserved firm heterogeneity and common macroeconomic shocks. As reported in Table 6, the results from the temporal separation model are qualitatively consistent with the baseline findings.

This study examines the relationship between board gender diversity (BGD) and earnings management, and the moderating role of ESG engagement, in four major Southeast Asian economies. Consistent with prior governance arguments, board gender diversity is generally associated with lower earnings management, supporting the view that female directors enhance monitoring and ethical oversight. However, the moderating analysis reveals that ESG engagement does not uniformly reinforce this effect. Instead, the positive interaction between BGD and ESG performance indicates a substitution effect, whereby strong ESG frameworks weaken the marginal governance contribution of female directors. In Malaysia and Indonesia, ESG appears to function as an alternative governance mechanism, reducing reliance on board-level monitoring, while in Singapore, where institutional governance is already strong, neither ESG nor BGD exerts significant incremental effects. In Thailand, the mixed results suggest either limited complementarities or symbolic ESG adoption, raising concerns about governance washing in weaker institutional settings.

By integrating resource dependence, social role, and groupthink perspectives, this study reconciles mixed evidence in the BGD–earnings management literature and demonstrates that ESG engagement can both condition and constrain the effectiveness of gender-diverse boards. The findings caution against assuming that governance mechanisms are inherently complementary and underscore the importance of institutional context in shaping their interaction. For policymakers, the results suggest that gender quotas alone are insufficient without institutional frameworks that enable substantive participation and influence. For investors, the findings highlight the need to critically assess ESG scores, which may reflect symbolic compliance rather than genuine governance quality. For firms, integrating ESG and board diversity as mutually reinforcing rather than independent governance tools is essential for strengthening transparency and long-term value creation.

This study has limitations. Its focus on four Southeast Asian economies limits generalizability, and reliance on ESG scores and established earnings management proxies may not fully capture the depth of sustainability practices or reporting quality. Moreover, the analysis focuses solely on gender diversity, excluding other board attributes that may interact with ESG engagement. Future research could adopt a broader view of board diversity, explore institutional and cultural mechanisms through comparative or mixed-methods approaches, and incorporate qualitative evidence from female directors. Overall, this study shows that the governance benefits of board gender diversity are context-dependent and contingent on the authenticity of ESG engagement, emphasizing the need to move beyond symbolic compliance toward substantive governance reform in emerging markets.