Board gender diversity and ESG performance: the moderating role of earnings volatility in Australian publicly listed firms

Globally, over the past decade, investors, policymakers and other stakeholders have called for a change in the composition of corporate boards. Specifically, board reform proponents emphasize the need to increase the number of women directors on corporate boards by arguing that gender-diverse boards lead to better board performance (Higgs, 2003) and enhance corporate sustainable practices (Paolone et al., 2024).

In response to this global demand, regulatory bodies in several countries have revised their rules to increase the number of women representations on corporate boards, adopting either coercive, supporting or laissez-faire approaches (Labelle et al., 2015). For instance, European Union (EU) countries such as France, Italy, Spain and Norway have introduced mandatory quotas, often referred to as “feminization” laws, to promote women representation (De Beaufort and Summers, 2014). Literature shows that the percentage of women representation on the corporate board has been significantly increasing following the enactment of mandatory quotas (Alkhawaja et al., 2023; Deloitte, 2015). However, Australia has not adopted mandatory quotas; instead, it has relied on a comply-or-explain approach to encourage greater women representation on corporate boards (Kang et al., 2023). To support this approach, the Australian Stock Exchange (ASX) has introduced amendments to its Corporate Governance Principles and Recommendations (hereafter “ASX amended principles”). These amendments were first made in 2010 ^[1], with a focus on promoting gender diversity, followed by revisions in 2014 ^[2] to promote corporate sustainable practice, and again in 2019 ^[3] to address both gender diversity and corporate sustainable practice.

ASX principles and recommendation amendments have attracted researchers, particularly to investigate whether or not board gender diversity has improved since the recent ASX amendment principles (Nadeem et al., 2017) and whether board gender diversity leads to improved corporate environment, social and governance (ESG) performance (Yarram and Adapa, 2021). However, there is no consensus on whether board gender diversity improves or reduces ESG performance. For instance, some studies have documented a positive relationship between board gender diversity and ESG performance (Nadeem et al., 2017; Paolone et al., 2024). In contrast, other studies have reported that gender diversity has a negative or no significant effect on ESG performance (Manita et al., 2018; Abdelkader et al., 2024). These inconsistencies in the literature may stem from differences in corporate governance (CG) regulations and methodological challenges, such as failure to account for reverse causality (endogeneity) and the reliance on narrow performance proxies (Adams et al., 2015; Vafaei et al., 2015).

Critically, it remains unclear whether board gender diversity continues to enhance ESG performance during periods of high earning volatility, which imposes additional constraints on corporate decisions related to sustainable business practices (Sun and Ding, 2020). While prior research has predominantly examined the direct relationship between gender diversity and ESG outcomes, it has largely neglected the moderating role of firm earnings volatility in shaping these dynamics.

The related literature highlights that firms experiencing high earnings volatility often encounter financial pressures that drive them to prioritize short-term objectives over long-term sustainability goals (Fiori and Scoccianti, 2023; Vuong et al., 2024; Wang et al., 2023). These pressures are associated with reduced overall ESG performance (Leong and Yang, 2021).

This gap in the literature raises a critical question:

To the best of my knowledge, this study is the first to examine whether the positive impact of gender diversity on ESG performance persists under conditions of firm earnings volatility. It provides fresh insights into the effectiveness of gender diversity in promoting sustainable corporate practices, particularly in the face of firm earning volatility. This study investigates whether board gender diversity has improved and explores the moderating role of firm earning volatility on the relationship between board gender diversity and ESG performance in Australian public-listed firms for several reasons.

First, Australia differs significantly from the EU in terms of gender quota regulation. While the EU has implemented mandatory gender quotas, resulting in a substantial increase in women board representation – some countries, such as Norway, have even surpassed legal requirements, with women board representation at 39.7% (Deloitte, 2015). However, Australia has not adopted mandatory quotas (Kang et al., 2023). Consequently, the percentage of women on corporate boards in Australia remains lower than in most countries with mandatory quotas (Kang et al., 2007; Nadeem et al., 2017; Rao and Tilt, 2016; Yarram and Adapa, 2021). Evidence further suggests a stronger positive association between board gender diversity and ESG performance in countries with mandatory quotas (Alkhawaja et al., 2023). These findings highlight how institutional factors, such as CG regulations, influence the extent of board gender diversity and its impact on ESG performance (Adams et al., 2015; Aguilera et al., 2006; Peng et al., 2009; Kang et al., 2007).

Second, Australian companies are less financially resilient to unexpected financial distress than their EU counterparts (Choy et al., 2023). The literature consistently demonstrates that firms with high earnings volatility often face financial constraints, leading them to reduce long-term sustainable investments (Roche et al., 2022). As a result, findings from EU-based studies may not directly apply to the Australia context but offer a valuable setting for exploring the relationship between board gender diversity and ESG performance. Hence, the Australian market presents an interesting setting due to its voluntary CG approach, unique financial condition and the potential for distinct corporate sustainable practices.

Using a sample of 907 firm-year observations from publicly listed Australian companies, I find that board gender diversity has improved following the ASX amendment. Moreover, I identify a significant and positive association between board gender diversity and higher ESG performance, including its components, i.e. social and governance performance. These results underscore the role of women directors in fostering corporate participation in sustainable practices, including socially responsible activities and strong CG.

In addition, I find that firms with women representation on their boards are likelier to achieve better ESG performance than those with no women representation. However, my analysis reveals that high firm earning volatility weakens the positive relationship between board gender diversity and ESG performance, indicating that firms facing earnings volatility tend to prioritize short-term financial stability over long-term sustainability efforts.

The results are consistent after addressing potential endogeneity concerns and are robust across various methodologies, including (1) alternative proxies, (2) dynamic model: two-step system generalized methods of moments (GMM) estimation and (3) difference-in-differences (DID) model. These findings align with the literature, which finds a positive and significant association between board gender diversity and ESG performance (Ben-Amar et al., 2017; Fernandez‐Feijoo et al., 2014; Nadeem et al., 2017). However, my study departs from existing literature by identifying a critical nuance: the positive relationship between board gender diversity and ESG performance weakens under conditions of high earnings volatility. This moderating effect, which has been largely overlooked in prior studies, highlights that financial pressures can impair the ability of gender-diverse boards to maintain long-term ESG initiatives. This insight adds depth to the understanding of how firm earning volatility influences the effectiveness of board gender diversity in driving sustainable corporate practices.

This study builds on previous research examining the relationship between boardroom gender diversity and sustainable business practices in the Australian market, particularly expanding on studies conducted from 2010 to 2014 (Nadeem et al., 2017). Unlike these earlier studies, this research makes several key contributions. First, the study significantly extends the existing literature on the relationship between board gender diversity and ESG performance by showing the important role of women directors in enhancing ESG and its components. Secondly, covering the long period between 2010 and 2023, this study significantly contributes to understanding the role of ASX’s recommendations and principles in increasing women board representation and promoting corporate sustainability. Third, a DID estimation methodology strengthens the credibility of study findings. This methodological rigor contributes to understanding the effectiveness of regulatory interventions in increasing board gender diversity. Finally, the study offers new insights into how firm earning volatility limits the effectiveness of board gender diversity in enhancing commitment to firm sustainable practice. It reveals that earnings volatility weakens the positive impact of gender diversity on ESG performance, challenging the existing assumption in the literature that board gender diversity consistently benefits ESG performance, regardless of a firm’s financial conditions.

The remainder of the paper is organized as follows. The theories, literature review, and hypothesis development are provided in Section 2. Section 3 focuses on research methods. Section 4 discusses the results, while Section 5 concludes the paper.

This section presents theories, discusses the literature and develops hypotheses. Section 2.1 presents theories, discusses the literature on board gender diversity and ESG performance, and develops hypotheses. Section 2.2 presents theories and discusses the literature on the role of earning volatility in the relationship between board gender diversity and ESG performance and develops hypotheses.

Most earlier studies have focused on examining board characteristics and corporate disclosures primarily through the lens of shareholder interests (Jamali, 2008). However, the relationship between board gender diversity, earnings volatility, and ESG performance remains complex and requires a more nuanced approach. Therefore, this study addresses this complexity by adopting a multidimensional theoretical framework, focusing on stakeholder, resource dependency and critical mass theories. Each of these theories offers distinct insight into how board diversity impacts ESG performance.

Stakeholder theory suggests the importance of considering various stakeholders’ interests and expectations to improve organizational outcomes (Mahajan et al., 2023). From a stakeholder theory perspective, women on the board promote effective decision-making by reflecting diverse stakeholder interests and balancing board decisions, thereby enhancing overall organizational performance (Wasiuzzaman and Wan Mohammad, 2020). However, recent literature indicated that stakeholder theory might not fully explain how board gender diversity influences corporate sustainability practices, suggesting that additional theoretical frameworks are needed to understand better how board diversity contributes to corporate reporting and sustainability goals (Vafaei et al., 2015). To address this complexity, this study integrates resource dependency theory with stakeholder theory. Resource dependency theory suggests that boards of directors play a critical role in securing essential resources for the organization (Hillman and Dalziel, 2003). Women on boards, in particular, contribute unique skills, experiences and connections that enhance the board’s ability to secure valuable resources for the organization required for driving sustainability efforts (Nadeem et al., 2017).

The literature grounded in stakeholder and resource dependency theory has supported a positive relationship between board gender diversity and corporate sustainability performance (i.e. ESG performance) (Nadeem et al., 2017; Paolone et al., 2024; Shakil et al., 2021; Wasiuzzaman and Wan Mohammad, 2020). For instance, Paolone et al. (2024) examined the relationship between cultural and gender diversity and ESG performance and revealed a positive relationship between them. A similar study analyzed the influence of board composition on sustainable development disclosure, indicating that boards with higher women participation enhance the legitimacy of sustainable business reporting (Jizi, 2017). Nadeem et al. (2017) examined the relationship between boardroom gender diversity and corporate sustainability practices and found that board gender diversity positively and significantly influences corporate sustainability practices. Galbreath (2011) argued that women could manage stakeholder and firm resources due to their relationship-building ability, treating it as corporate social responsibility (CSR). Furthermore, other studies have highlighted a positive relationship between gender diversity and CSR, emphasizing that women unique skills, experiences and diverse perspectives they contribute play a crucial role in enhancing board decision-making processes that align with corporate sustainability goals (Nielsen and Huse, 2010).

While these studies highlight the positive impacts of gender diversity, literature also recognizes potential challenges. For instance, Harjoto et al. (2015) document that diverse boards may face longer decision-making processes or conflicts due to differing opinions. Another barrier to efficiently working boardroom diversity is the discrimination and serotyping challenges (Galbreath, 2010; Dawar and Singh, 2016), which may affect strategic decision-making, including corporate sustainability practices (Galbreath, 2010). However, these barriers tend to diminish over time as board members develop mutual understanding and cohesion (Harjoto et al., 2015). Therefore, I argue that the positive effect of gender diversity will outweigh the negative effect after the directors serve on the board if they work together over a more extended period. Hence, the association between gender diversity and ESG performance will remain positive. Based on these insights, the following hypothesis is proposed:

While stakeholder and resource dependency theories suggest a positive relationship between board gender diversity and ESG performance, critical mass theory offers a more nuanced perspective by focusing on the conditions under which these benefits are realized. Critical mass theory posits that small minorities within a group tend to hold only symbolic positions and lack significant influence in the organizational decision-making process (Schwartz-Ziv, 2017; Collins et al., 2010). In the context of corporate boards, a small representation of women directors may lead to tokenism, limiting their ability to drive meaningful change in corporate sustainability practices (Schwartz-Ziv, 2017; Collins et al., 2010; Dobija et al., 2022; Joecks et al., 2013).

Prior empirical literature grounded in critical mass theory revealed mixed evidence on the relationship between board gender diversity and firm sustainable performance. For instance, Joecks et al. (2013) examined the effect of gender diversity in the boardroom on firm performance and revealed that a small representation of women on the board negatively affects a firm’s sustainable performance. Similarly, Abdelkader et al. (2024) investigated the association between board gender diversity and ESG performance and revealed that board gender diversity has no significant effect on ESG performance. As evidenced in prior studies, these unfavorable results could be due to the limited proportion of women on the corporate board and their under-representation in key management positions (Konrad et al., 2008; Yadav and Prashar, 2023). However, studies indicate that once the proportion of women directors meets or exceeds a critical threshold – such as three or more women on the board – their influence on ESG performance becomes significantly positive (Alkhawaja et al., 2023; Joecks et al., 2013; Schwartz-Ziv, 2017). Based on these insights, the following hypothesis is proposed:

Agency theory suggests that conflicts of interest between managers and shareholders can lead to agency problems, exacerbating information asymmetry and creating uncertainty in decision-making (Jensen and Meckling, 2019). When agency problems intensify, asymmetric information increases earnings volatility and financial risk, complicating a firm’s ability to plan for long-term objectives like ESG investments (Ahmed and Hla, 2019; Dierkens, 1991). Resource constraint theory adds another layer of complexity, highlighting that limited financial resources, often driven by high earnings volatility, shift firms toward immediate financial stability at the expense of long-term investments like ESG initiatives (Attig, 2024; Leong and Yang, 2021). ESG initiatives are long-term investments and often require substantial resources (Henisz et al., 2019; Taglialatela et al., 2024). When firms face resource constraints due to earning volatility, corporate managers may redirect company resources from long-term ESG investment toward short-term profitable projects that are perceived as more critical for maintaining immediate financial stability (Zhang et al., 2019; Zhang, 2020; García-Quevedo et al., 2018; Vuong et al., 2024). Together, these theories explain why earnings volatility moderates the relationship between board gender diversity and ESG performance.

Prior empirical literature supports this argument, suggesting that agency problems exacerbate asymmetric information and introduce uncertainty into firm stock valuations, increasing earnings volatility (Dierkens, 1991; Ahmed and Hla, 2019). Similarly, firms experiencing high earnings volatility are more likely to encounter financial constraints, leading them to prioritize short-term performance over long-term sustainability goals (Fiori and Scoccianti, 2023; Vuong et al., 2024; Wang et al., 2023). Similar studies revealed that financial constraints can hinder firms from investing in sustainable practices due to the need to hold cash and manage risks associated with high earnings volatility (Fiori and Scoccianti, 2023; Singh et al., 2023). As a result, these financial constraints weaken the potential positive influence of board gender diversity on corporate sustainability performance (Azeem et al., 2023). Based on this literature, this study proposes the following hypothesis:

The sample consisted of public firms listed on the Australian Securities Exchange from 2010 to 2023. This period was selected because ASX has made several principles and recommendation amendments to promote gender diversity and ESG practice during these periods. ESG scores, financial information and CG data retrieved from Refinitiv-Eikon. Refinitiv was chosen for this study because it is one of the world’s most trusted and largest providers of comprehensive ESG data, offering over 450 ESG metrics with historical availability (Bătae et al., 2021; Abdelkader et al., 2024). Its strong and transparent methodology, detailed on its official website, ensures data reliability and has made it a preferred source in the literature on various CG topics, including board gender diversity and ESG performance (Morán‐Muñoz et al., 2024; Menicucci and Paolucci, 2024; Yarram and Adapa, 2021; Abdelkader et al., 2024).

The initial data set comprises 3865 firm-year observations; however, there are 1,878 missing values on key variables, resulting in a final sample of 907 firm-year observations. The study’s final sample is adequate compared to studies in a similar domain, representing a 288 final firm-year observation of the total sample (Ali et al., 2014).

ESG performance, represented by ESG Score, is the dependent variable of this study. ESG Scores obtained from Refinitiv Ekin Data Base range from 0 to 100 and indicate the level of ESG performance, with higher scores suggesting better performance (Moffitt et al., 2024). The ESG score is based on three components: environmental, social and governance scores. The study, therefore, examines the impact of independent variables on ESG and its components.

The primary variable of interest of the study is board gender diversity (BDIV), measured by the proportion (percentage) of women representation on the corporate board and used to test the relationship between the proportion of women representation on corporate boards and ESG performance. To ensure the robustness of the estimation, following previous literature, the study also employed alternative BDIV proxies (gender index) used to examine how board gender balance impacts firm ESG performance (Nadeem et al., 2017; Campbell and Mínguez-Vera, 2008).  Appendix presents the specific definition of these controlling variables.

The study controls for the firm-level, governance-level and year-specific factors to mitigate omitted variable bias. Based on prior research, the study includes firm-level variables such as firm size (FSIZE), leverage (LEV), return on assets (ROA) and market-to-book ratio (MB). In addition, the study includes governance variables like CEO duality and board size (BSIZE) (Wasiuzzaman and Wan Mohammad, 2020; Nadeem et al., 2017; Ben-Amar et al., 2017). The  Appendix presents the specific definition of these controlling variables.

The study uses a static and dynamic panel regression model to analyze the data (Nadeem et al., 2017). Moreover, the study used the ESG score and each of the three components separately to see the effect of BDIV on the ESG and its component level.

The static model is presented as follows:

The dynamic model is presented as follows:

In the above model equations, i and t stand for “firm” and “year,” respectively. The study includes further explanations for the above models. BDIV_it represents board gender diversity; ESG_i,t represents ESG score of specific firms (i) and year (t); ESGScore_(i,t−1) represents lagged ESG scores of specific firms (i) and year (t); Env_i,t represents firm specific (i) Environmental Score in a given year (t); Soc_i,t represents firm specific (i) Social Score in a given year (t); and Gov_i,t represents firm specific (i) Governance Score in a given year (t). Based on Hausman’s test suggestion, I use random-effect panel regression clustered at the firm level, including year fixed effects, and winsorize continuous variables at 1% and 99% to reduce the impact of outliers.

Following previous literature, this study used the Wooldridge test to determine the exogeneity among the variables (Nadeem et al., 2017). The statistical equation of the Wooldridge test is as follows:

where the primary variable of interest (BDIV represents % of women) and control variables are in future values/values of the next year to include in the model to test the presence of endogeneity and control for the dynamic nature of the CG–ESG relationship. The  Appendix presents the specific definition of these controlling variables.

After confirming the endogeneity using the Wooldridge test, the study used a two-step model GMM to address the endogeneity issue and control the dynamic nature of the CG–ESG relationship. Hence, following previous literature, this study used a one-year lagged ESG Score (ESGScore_{i,(t − 1)}) as an independent variable, assuming that past corporate sustainability practices significantly affect current corporate sustainability practices (Nadeem et al., 2017; Nguyen et al., 2014).

Following the recommendation of CG literature (Nguyen et al., 2014; Nadeem et al., 2017; Roodman, 2009; Wintoki et al., 2012), this study performed the GMM instrumental variables (IV) validity test. GMM instrumental variables validity test includes (i) the Arellano–Bond test for serial correlation [the AR (1), AR (2)], (ii) Hansen’s J test used to test instrument validity/overidentifying restriction, and (iii) the difference in Hansen’s J test for the GMM style/for the IV style used to test variables exogeneity (Arellano and Bond, 1991).

The literature suggests that different policy interventions can enhance or weaken a company’s efforts to increase board gender diversity (Kang et al., 2023). Therefore, a DID model is then applied to compare the impact of policy measures on the percentage of women on boards before and after policy intervention. This approach allows us to see the changes in board gender diversity during the treatment periods (post-policy) relative to the control periods (pre-policy). The DID model is designed as follows:

In the above equation, BDIV is the dependent variable, which refers to the percentage of women on the board. The study defines a dummy variable treatment group, which is an indicator variable taking the value of 1 if the percentage of women on the board is equal to or above one after the policy intervention and otherwise zero. Similarly, the study defines a dummy variable post-policy intervention, which is an indicator variable taking the value of 1 if the observation is in the post-policy-implementation period and 0 otherwise. At where X_{i, t} represents a vector of control variables and ε_{i, t} stands for the error term.

Table 1 shows that ESG scores for ASX-listed companies range from 13.39% to 71.89%, with an average of 38.59%. This result is much higher than the average ESG score of 18.67% reported between 2010 and 2014 (Nadeem et al., 2017). This result indicates a relative increase in ESG performance during the sample period (2010–2023). The mean result of the Env’t score is 25.84, the Soc score is 38.69 and the Gov’t score is 49.29. Similarly, the mean percentage of women directors is 18.52%, reflecting a notable improvement in women’s representation on corporate boards in Australia. This result is much higher compared to previous studies undertaken in Australia. For example, Rao and Tilt (2016) reported 9.3% between 2009 and 2011, while Nadeem et al. (2017) found 10.33% between 2010 and 2014, and Yarram and Adapa (2021) reported 13% between 2011 and 2016. Similarly, this result is much higher than most other developed countries like Singapore (6.90%), the UK (9.2%) and Canada (16%) (Ben-Amar et al., 2017). However, it remains lower than in Denmark (25%), Finland (29.5%) and Norway (39.7%) (Deloitte, 2015).

Moreover, the average percentage of companies with at least one women director is 72.9%. This result is much higher than most reported results earlier, for instance, Spanish (23.7%) (Campbell and Mínguez-Vera, 2008) and Australian (51%) (Nadeem et al., 2017). From the perspective of earlier findings, the incremental result of this study reflects the effectiveness of Australian policy intervention in increasing the percentage of women representation on the corporate board. The study result shows that the CEO holds the position of chairperson at approximately 9%, which shows an increment of CEO duality compared to the previous CEO duality report between 2010 and 2014 (Nadeem et al., 2017).

Table 2 results show the pairwise correlation coefficient of 0.540 (%women) shows that ESG scores significantly correlate with the percentage of women directors on the board, supporting the hypothesis that a higher percentage of women on a corporate board is positively associated with better ESG performance. Turning to control variables, MB and CEO Duality negatively and significantly correlate with ESG. In contrast, Fsize, LEV, ROA and BSIZE positively and significantly correlate with ESG. None of the included controlling variables’ correlation statistical tests exceed 0.70, suggesting no multicollinearity concerns. Figure 1 illustrates that a firm with board gender diversity within the board may have a positive relationship with a firm’s better ESG practices, aligning with the correlation results reported in Table 2.

Table 3 shows that the percentage of women directors within the firms (9.909%) is lower than across the firms (13.764%). This result is much higher than the percentage of women directors within the firms (5.69%) and across the firms (10.69%) reported between 2010 and 2014 (Nadeem et al., 2017). As this study included recent longitudinal data, the percentage of women directors may have increased over time since ASX amendments to promote board gender diversity on corporate boards, notably the study reported period between 2010 and 2023, as most firms, on average, have achieved an 18% improvement in board gender diversity. This evidence aligns with the result reported in Table 5 of the DID model, which confirms the effect of policy intervention on the percentage of board gender diversity.

Table 4 shows a t-test to capture the difference in ESG practices between firms that have women on their boards and firms that do not have women on their boards. Table 4 shows that, on average, the ESG score of firms with women on corporate boards is higher than that of firms without women, and the t-test results are statistically significant at the 1% level. Figure 1 illustrates that the ESG score of firms with women on corporate boards is higher than that of firms without women, aligning with the results reported in Table 4. These results supported the main hypothesis that the presence of higher women directors on boards significantly positively influences a firm’s ESG performance.

The proportion of women directors on the board increased from 8.37% in 2010 to 27.66% in 2023 since the ASX principles and recommendations amendment came into effect, as reported in the last column of Table 4. Similarly, Figure 2 illustrates the improvement of board gender diversity between 2010 and 2023, supporting the results reported in Table 4. Results reported in Table 5 confirm the effect of policy intervention on the percentage of board gender diversity, supporting the results reported in Table 4. More specifically, the interaction term (DID) reveals that the policy intervention led to a significant increase in board gender diversity, confirming that the policies effectively promoted gender diversity on corporate boards.

Table 6 reports static model regression results demonstrating that the proportion of women on the board positively and significantly affects ESG performance (Coeff = 0.294, p < 0.01), suggesting a positive and significant association between a higher proportion of women on a board and higher ESG performance. Specifically, the coefficient indicates that a 1% increase in the proportion of women directors is associated with a 29.4% increase in ESG performance, holding all other factors constant. Similarly, in the static model, column 3, the regression results show that the proportion of women on the board positively and significantly affects social performance (Coeff= 0.223; p < 0.01), indicating that higher women representation on the board promotes firm participation in socially responsible activities. Further, in column 4, the static model regression result shows that the proportion of women on the board positively and significantly affects governance performance (Coeff= 0.450; p < 0.01), indicating that higher women representation on the board promotes good CG practice.

Table 6 reports dynamic model regression results, incorporating a one-year-lagged ESG score, demonstrating that the proportion of women on the board positively and significantly affects ESG performance (Coeff = 0.825, p < 0.01). These findings were consistent with the results obtained from earlier studies, suggesting that past ESG practices should be considered as important variables to control for the dynamic nature of the CG–ESG relationship (Nadeem et al., 2017). Similarly, in the dynamic model, column 3, the regression results show that the proportion of women on the board positively and significantly affects social performance (Coeff = 0.831; p < 0.01), indicating that higher women representation promotes firm participation in socially responsible activities. Furthermore, in column 4, the dynamic model regression result shows that the proportion of women on the board positively and significantly affects governance performance (Coeff = 0.752; p < 0.01), indicating that higher women representation on the board promotes good CG practice. In sum, study findings lead to the acceptance of hypothesis (H1a) and are consistent with resource dependency and stakeholder theories, suggesting that women on corporate boards bring unique skills, represent diverse stakeholder interests and balance board decisions. These results corroborate existing literature that reinforces the positive influence of women on corporate board representation on ESG performance (Nadeem et al., 2017; Ben-Amar et al., 2017; Fernandez‐Feijoo et al., 2014).

Table 7 discusses the regression results with moderating variables used to indicate whether firm earning volatility moderates the relationship between board gender diversity and ESG performance. In the static model, column 1, the main effect of BDIV is positive and significant (Coeff = 0.349, p < 0.01), suggesting that, in the absence of high earnings volatility, BDIV enhances ESG performance. Conversely, the coefficient of Sd5_ROA (earnings volatility) is negative and insignificant (Coeff = −4.274, p > 0.1), indicating that there is no direct association between high earnings volatility and ESG performance. However, the interaction term (BDIV*Sd5_ROA) is negative and statistically significant (Coeff = −1.127, p < 0.01), confirming that the positive effect of BDIV on ESG performance is significantly weaker in firms with high earnings volatility.

Similarly, in the dynamic model, column 1, the results show that the main effect of BDIV is positive and significant (Coeff = 0.147, p < 0.01), indicating that BDIV enhances ESG performance in the absence of high earnings volatility. The coefficient of Sd5_ROA is positive but insignificant (Coeff = 3.02, p > 0.1), again suggesting that there is no direct association between high earnings volatility and ESG performance. The coefficient of the interaction term (BDIV*Sd5_ROA) is negative and statistically significant (Coeff = −0.471, p < 0.05), indicating consistent findings with the static model.

Finally, in the two-step GMM model, column 1, the main effect of BDIV is positive and significant (Coeff = 0.179, p < 0.01), reinforcing the finding that BDIV improves ESG performance in the absence of high earnings volatility. The coefficient of Sd5_ROA is positive and insignificant (Coeff = 2.519, p > 0.1), consistent with prior models. The coefficient of the interaction term (BDIV*Sd5_ROA) is negative and statistically significant (Coeff = −0.535, p < 0.05), indicating consistent results with the prior models.

These findings confirm that firm earnings volatility moderates the relationship between board gender diversity and ESG performance, leading to the acceptance of H2a. This result aligns with the agency and resource constraint theory and existing literature, which supports the idea that earning volatility amplifies agency problems and creates financial constraints, thereby leading firms to prioritize short-term objectives over long-term sustainable investments (Huang et al., 2018; Bin‐Feng et al., 2024; Fiori and Scoccianti, 2023; Singh et al., 2023).

Following previous literature, this study used the Wooldridge test to determine the exogeneity among the variables (Nadeem et al., 2017). The Wooldridge strict exogeneity test results documented in Table 8 show the future values of the independent variable, i.e. women % are positively significant (0.208, p < 0.01) associated with the current dependent variable (ESG), indicating strong endogeneity and dynamic nature between the relationship of board gender diversity and ESG performance and the model reject the Wooldridge’s null hypothesis, which suggests no endogeneity issues. Therefore, following previous literature, the study used a one-year lag value of the ESG score and performed a two-step GMM statistical analysis to alleviate the potential effect of endogeneity issues and control for the dynamic nature of the CG–ESG relationship (Nadeem et al., 2017; Nguyen et al., 2014).

The dynamic two-step GMM estimated result reported in the earlier Table 6, the one-year lagged ESG coefficient, is positive at the 1% significance level. These findings were consistent with the results obtained from Woolridge’s test of strict exogeneity, suggesting that past corporate sustainability practices should be considered important variables to control for the dynamic nature of the CG–ESG relationship. The instrument validity test results under the GMM model are reported in Table 6. The AR (1) test yielded a p-value of 0.000, showing that there were enough lags to control the dynamic nature of the CG-ESG relationship. The AR (2) second-order serial correlation test has yielded a p-value of 0.951, which means that a null hypothesis of no serial correlation cannot be rejected. The Hansen test result reported in table indicates a p-value of 0.625, suggesting that the null hypothesis of the instruments is valid and cannot be rejected at any conventional levels of significance. The difference-in-Hansen test result reported in Table 6, used to test exogeneity for the system GMM style and the IV style, indicate a nonsignificant p-value, suggesting that there is no statistical evidence to reject the null hypothesis suggesting that the instruments are exogenous. After using the GMM instrumental variable validity test, the results show a significant positive association between the percentage of women representation on boards and ESG performance. Therefore, the study results remain robust and the same.

Following suggestions from previous literature (Nadeem et al., 2017), the study tested the robustness of the results using alternative proxies for gender diversity, specifically a gender diversity dummy variable. The study reran the regression using static, dynamic and two-step GMM models. The results, as reported in Table 9, column 1, under the static model, indicate that the coefficient for the women dummy variable remains positive and statistically significant (Coeff = 5.726, p < 0.01). Similarly, in the dynamic model, column 1, the women dummy variable remains positively and significantly associated with ESG performance (Coeff = 2.592, p < 0.01). Furthermore, the results reported under the two-steps GMM model, column 1, indicate that the coefficient for the women dummy variable remains positive and statistically significant (Coeff = 2.848, p < 0.01). These findings confirm the robustness of the results, indicating a consistent positive association between the percentage of women representation on boards and ESG performance.

The study further tested the robustness of the moderating variable results using an alternative proxy for firm earnings volatility, specifically the difference between the maximum and minimum ROA over the previous five years (Dif_MaxMin) (Sattar et al., 2022). Table 10 presents robustness regression results using static, dynamic and two-step GMM models. In the static model, column 1, the main effect of BDIV remains positive and significant (Coeff = 0.356, p < 0.01), confirming that in the absence of high earnings volatility, BDIV increases ESG performance. The coefficient of Dif_MaxMin is negative and insignificant (Coeff = −1.158, p > 0.1), indicating that there is no direct association between high earnings volatility and ESG performance. However, the coefficient of the interaction term (BDIV*Dif_MaxMin) is negative and statistically significant (Coeff = −0.454, p < 0.01), suggesting that the positive effect of BDIV on ESG performance is significantly weaker in firms experiencing high earnings volatility.

In the dynamic model, column 1, the main effect of BDIV remains positive and significant (Coeff = 0.144, p < 0.01), reinforcing the finding that BDIV increases ESG performance in the absence of high earnings volatility. The coefficient of Dif_MaxMin is positive but insignificant (Coeff = 1.572, p > 0.1), indicating that there is no direct association between high earnings volatility and ESG performance. However, the coefficient of the interaction term (BDIV*Dif_MaxMin) is negative and statistically significant (Coeff = −0.196, p < 0.05), indicating consistent findings with the static model.

Finally, in the two-step GMM model, column 1 reports consistent results with the previous model, showing that the main effect of BDIV is positive and significant (Coeff = 0.18, p < 0.01). The coefficient of Dif_MaxMin remains positive and insignificant (Coeff = 1.416, p > 0.1), whereas the coefficient of the interaction term (BDIV*Dif_MaxMin) is negative and statistically significant (Coeff = −0.222, p < 0.05). These robustness test results confirm that the positive effect of BDIV on ESG performance is significantly weakened in firms with high earnings volatility, further supporting the study’s main findings.

This study examines the effect of board gender diversity on ESG performance. While the proportion of women on corporate boards remains relatively low, the findings show a clear positive relationship between gender diversity and ESG performance. This positive result between gender diversity and ESG performance remained significant when alternative gender diversity proxies were applied. The increasing trend in gender diversity in Australia is showing evidence of intervention and effectiveness of government policy. This means that institutional interventions can play an important role in reshaping corporate gender balance. However, firm earnings volatility weakens the positive influence of gender diversity on ESG performance. The result shows the importance of stable financial conditions for maximizing the effectiveness of gender diversity in driving sustainable corporate outcomes.

Theoretically, the findings are consistent with resource dependency and stakeholder theories, which suggest women on the board bring better connection-building skills, represent diverse stakeholder interests and contribute to balanced board decisions, ultimately leading to better ESG performance. In addition, the findings are consistent with theories that suggest agency problems and financial constraints can push firms to focus on short-term goals at the expense of long-term ESG investments. Empirically, the findings show that gender diversity is associated with high levels of ESG performance and the importance of stable financial conditions in maintaining this positive relationship. These conclusions hold even after addressing potential endogeneity concerns and conducting different robustness tests.

The empirical findings offer practical implications for investors and regulators. For investors, these findings highlight the value of focusing on companies with higher women representation on their boards and demonstrating strong financial stability, as these factors are crucial for maintaining long-term ESG investments. For regulators, the results highlight the importance of considering the time required to achieve meaningful board diversity and sufficient financial resources to meet the expectations of ASX recommendations and principles.

While this study makes significant contributions, it does have some limitations. First, the research is limited to publicly listed companies on the ASX; caution is needed when applying these findings to unlisted companies, as listed companies are typically more transparent in their structures than unlisted ones. Future research could investigate the differences between listed and unlisted firms concerning board gender diversity and ESG practices to provide a more comprehensive understanding. Moreover, this study is the first to examine the role of firm earnings volatility in Australia, and its findings may not be generalizable to other countries. Therefore, future studies could explore whether similar patterns emerge in other regions or under different regulatory frameworks.

Table A1