CSR Committee, Women on the Board, and Green Bond Issuance: Evidence from France

Abstract

This study examines the effects of internal governance mechanisms on the issuance of green bonds and investigates whether firms issuing green bonds exhibit distinct corporate governance characteristics, especially regarding board gender diversity and corporate social responsibility (CSR) committees. The analysis is based on a sample of 64 green bond announcements between 2013 and 2022. Based on the Generalized Least Squares Regression model, empirical results show that the presence of a CSR committee is positively and significantly associated with the issuance of green bonds. In other words, companies with a CSR committee are more likely to issue green bonds. In addition, companies with a lower debt ratio are more likely to issue green bonds.

1. Introduction

For several years, organizations worldwide have been implementing sustainability policies and practices to address growing environmental concerns. The principles of sustainable corporate governance enable us to meet current needs without compromising those of future generations. However, achieving this balance requires a new governance model that supports better decision-making while maintaining investor engagement. The basis of this model is a clear vision of sustainable progress, as well as the policies, practices, and tools required to achieve it. According to stakeholder theory, effective sustainable corporate governance fosters increased collaboration among stakeholders, such as investors and consumers, who demand transparency and responsible policies from the companies with which they engage. In line with Elkington’s (1998, 2018) view that social performance drives environmental success, investors are increasingly incorporating ESG factors into their selection processes. Therefore, this ensures that social controversies do not undermine environmental contributions. This approach enables the implementation of social, environmental, and ethical strategies, leading to the adoption of best practices and improved processes. While integrating the environmental dimension throughout the organization is a central objective, mobilizing both internal and external stakeholders remains a challenge. To support efforts in preventing and mitigating climate change, various innovative financial instruments have been developed. Among these instruments, green bonds play a prominent role. The ICMA defines these bonds as “any type of bond instrument where the proceeds of the issue, or an equivalent amount, are used exclusively to finance or refinance, in whole or in part, new and/or existing Green Projects.” Their success is largely attributable to the introduction of international standards such as the Green Bond Principles (GBP) and the Climate Bond Standards (CBS), which promote transparency and accountability regarding the environmental and social impact of issuing companies. According to signaling theory, companies issue bonds to declare their commitment to the environment. They seek not only to obtain financing but also to send credible signals about their environmental commitment and show their intention to reduce their carbon footprint (Yeow & Ng, 2021; Flammer, 2020).

Green bond issuance has experienced rapid growth over the past decade. The signing of the Paris Agreement in 2015 and the adoption of the United Nations’ Sustainable Development Goals (SDGs) have significantly boosted this market. Since then, issuance volumes have increased from less than $50 billion in 2015 to $452.2 billion in 2021 (Climate Bonds Initiative, 2021). This growth is further reflected in the rising number of issuing companies and countries, as reported by Cortellini and Panetta (2021). The European Investment Bank was the first to issue green bonds in 2007. However, the green bond market gained momentum only in 2013. A key factor driving this growth was the publication of the Green Bond Principles (GBP) drafted by the International Capital Market Association (ICMA) in that year.

In 2013, global green bond issuance tripled from USD 11 billion to USD 38 billion in 2014. In 2015, total issuance increased slightly to nearly USD 50 billion, and in 2016, it doubled to exceed USD 100 billion. Following this trend, in 2017, bond issuance almost doubled again to USD 200 billion. However, growth slowed down between 2017 and 2018 to only 8%. After 2018, annual growth stabilized at 50–60%, reaching a record level of USD 550 billion in 2020.

France is among the pioneering green bond markets. In 2012, local authorities issued the first green bonds to finance projects such as wastewater management, biodiversity conservation, sustainable building, and low-carbon transport (e.g., trains, bus rapid transit (BRT) systems, cycle paths, and streetcars), and renewable energy initiatives. In 2018, France ranked first in Europe and third globally, with € 37.7 billion in green bond issuance. The nation’s commitment to the 2015 Paris Agreement and its carbon-neutral policies have stimulated the growth of green bond financing.

France’s green bond market has experienced significant expansion, with companies, local authorities, and financial institutions increasingly issuing bonds to support sustainable financing. In 2021, France was the second-largest contributor to green bond issuance (after Germany), with € 42 billion issued.

Between 2013 and 2022, France emerged as a major player in the green bond market during its boom. This period is characterized by increased issuance and evolving regulatory frameworks. Notably, France issued the first sovereign green bond (Green OAT) in 2017 and established itself as a leader in climate action. Furthermore, it set a global benchmark for governmental commitment to financing the ecological transition. Over this decade, regulations such as the 2015 Energy Transition Act and the gradual integration of the European taxonomy since 2020 further bolstered the market.

This study considers that the period from 2013 to 2022 offers sufficient hindsight to analyze the impact of green bonds on financing green projects and on market evolution.

To ensure the effectiveness and transparency of these investments, internal corporate governance mechanisms such as specialized committees, sustainable investment policies, and regular environmental performance reporting are essential. By incorporating these measures, companies can align their green bond issuances with positive environmental initiatives while responsibly managing their financial resources. Effective internal corporate governance leads to solid management practices, enhanced transparency, and greater accountability to stakeholders. These factors are essential in building investor confidence that the funds raised through green bonds will be used responsibly and effectively.

A review of the literature shows an unexplored issue: the board of directors’ role in decision-making on the issuance of green bonds. Previous studies have shown that the presence of a woman on the board increases sensitivity to environmental issues (Galbreath, 2018; García & Herrero, 2020; García et al., 2023; Potapova, 2023). Furthermore, stakeholder theory argues that the presence of more women on boards improves stakeholder relations, thus improving environmental practices (Hussain et al., 2018). In 2011, based on recognition of the importance of board diversity, France adopted the Copé–Zimmermann law. This law mandated a 40% gender diversity quota on boards by 2017. In 2021, the French Parliament enacted the Rixain law, which acknowledged that progress under the previous law had been insufficient and imposed a mandatory 40% quota by 2029. This was the first law on gender diversity in a Western country.

Furthermore, a sustainability committee reporting to the board implies that the company is expected to make investment decisions for sustainable development rather than merely declaring its environmental commitment. Accordingly, this paper aims to analyze the corporate governance factors that influenced global green bond issuance during 2013 to 2022. Specifically, this study focuses on five aspects: the women on the board of directors, the board size, the board independence, the CEO duality, and the existence of a sustainability committee.

This study contributes to the literature on green bond issuance by examining the role of sustainability committees and board characteristics in shaping corporate environmental strategies. Firstly, recognizing that sustainability committees help the board in evaluating and deciding on sustainability matters, this study argues that board attributes significantly influence green bond issuance decisions. Therefore, the analysis of this study extends previous research by considering a wider range of board aspects in companies issuing green bonds. Secondly, this study analyzes 10 years of data from 64 firms, expanding the study period compared to previous studies by Flammer (2021) and Fatica and Panzica (2021). However, those studies were limited by small sample sizes. This extended timeframe captures a greater number of issuances due to the continuous growth of the green bond market. Moreover, other studies have documented the long-term impact of green bond issuances.

The remainder of this paper is organized as follows. Firstly, this study provides an in-depth theoretical exploration of green bonds, corporate governance, and their interaction, as well as a literature review. Secondly, this study presents an empirical analysis of the impact of internal corporate governance on the issuance of green bonds.

2. Literature Review and Hypotheses Development

Most empirical studies on the relationship between green bond issuance and internal corporate governance have examined various corporate governance (CG) aspects, especially the board of directors.

2.1. Board Diversity and Green Bond Issues

Board diversity is attracting growing interest due to its potential to shape board norms, attitudes, and perspectives (Amorelli & García-Sánchez, 2021). In the context of green bond issuance, board diversity can play an important role. Different boards are often better equipped to monitor and manage corporate social responsibility (CSR) performance. Therefore, this meets stakeholder expectations and promotes long-term sustainability (Xia et al., 2022).

According to Haque and Jones (2020), gender diversity is the most important attribute of boards in the CSR sector. Men and women differ in communication and leadership styles, personal characteristics, values, and ethics (Amorelli & García-Sánchez, 2020), which can provide benefits by promoting continuity of knowledge.

Using the ordinary least squares technique on a sample of 53 utility companies across 15 European Union countries, Asad et al. (2023) examine how gender diversity on boards affects environmental, social, and corporate governance (ESG) performance. They found that gender diversity positively influences ESG performance, with the effect becoming stronger in the presence of at least three female board members. Similarly, Mallidis et al. (2024) investigates the relationship between gender diversity on boards and CSR performance, focusing specifically on environmental, social, and CSR controversy dimensions. It was found that a higher proportion of female directors correlates with improved environmental and social outcomes.

Tanaka (2019) examines the relationship between female board representation and corporate performance using a sample of 2000 listed Japanese companies from 2006 to 2015. He finds that female non-executive directors are significantly and positively associated with company performance. Meanwhile, there are no significant relationships between internal female directors and corporate performance.

Jouber (2022) examines non-financial companies across 17 countries and finds that women in top management positions improve environmental and social performance. However, this positive relationship is negatively moderated by the CEO’s managerial ability and the corporate governance attribute index. Similarly, Romano et al. (2020) report a positive correlation between the presence of women on boards and higher CSR performance among listed Italian non-financial companies.

Similarly, Dodd et al. (2022) examine the relationship between board diversity and CSR performance among S&P 500 companies between 2004 and 2015. They find that visible forms of board diversity (gender, ethnic, and age diversity) may emerge endogenously through proactive diversity management. This shows that cultural diversity on the board is positively related to CSR performance.

Mohy-ud-Din (2023) compared the impact of board diversity on CSR and sustainability in the US and Australian non-financial sectors, including the S&P 1500 and the Australian Stock Exchange (ASX) from 2010 to 2020. In addition, he examined the influence of gender diversity, cultural diversity, and diversity of experience on CSR and sustainability in both countries. The results show that cultural diversity on US boards strongly supports CSR practices compared to their Australian counterparts. While the presence of women on boards positively affects local CSR initiatives, it appears to hinder global development in both countries. In addition, the experience of the US board is related to increased funding for community development rather than sustainable development. These findings align with previous research that highlights a positive relationship between board diversity and environmental, social, and CSR performance (Asad et al., 2023; Mallidis et al., 2024; Romano et al., 2020; Dodd et al., 2022).

It is plausible that board diversity may promote green bond issuance by fostering more environmentally and socially responsible practices.

H1. 

There is a positive correlation between board diversity and the issuance of green bonds.

2.2. Board Independence and Green Bond Issues

According to the theory of resource dependence (Pfeffer & Salancik, 1978), organizations are highly dependent on external resources essential for survival. In addition, their continuity depends on meeting the expectations of key resource providers. According to this analytical perspective, directors serve to manage dependency and uncertainty by creating influential connections with the external environment. An independent director is defined as one who has no direct or indirect affiliation with the company.

Directors contribute strategic resources, such as information, knowledge, experience, skills, networks, key stakeholders, and values, which are essential for a company’s survival (De Villiers et al., 2011; Pfeffer, 1972). Therefore, these resources promote strategy formulation, enhance legitimacy, facilitate access to resources, and influence decision-making, including the management of environmental issues.

According to Dalton et al. (1998), independent directors have greater expertise, which facilitates improved access to external resources. Empirical research confirms that a higher level of board independence can positively influence the level of non-financial disclosure, particularly regarding sustainability (Harjoto & Jo, 2011).

It has been shown that independent boards enhance oversight functions and are widely recognized as a critical corporate governance mechanism (De Andres & Vallelado, 2008; Khan et al., 2013; Kathy Rao et al., 2012; Said et al., 2009). By creating external connections with different stakeholder groups, they attract additional resources and bolster a company’s reputation. Many studies report a positive correlation between board independence and corporate environmental behavior. For example, Cao (2023) employed the ordinary least squares method to examine the impact of board independence on CSR practices in Chinese companies. The study concludes that effective board independence strategies can yield a high level of CSR.

Agarwala et al. (2023) aim to explore the relationship between board independence and CSR practices among Indian companies. By analyzing a sample of 76 non-financial listed companies, they find that board independence has a significant positive relationship with CSR performance.

Umar (2022) examines the relationship between board independence and CSR spending in education, health, and disaster humanitarian aid among Islamic banks in Bangladesh, using panel data from 2010 to 2020. The results indicate that board independence is positively and significantly associated with CSR spending in the education, finance, and humanitarian aid sectors. However, its association with CSR spending in the health sector is insignificant.

Rubino and Napoli (2020) have shown that boards with a higher number of independent directors are more sensitive to social issues. In addition, the presence of independent directors helps to improve the quality of social and environmental reporting (Htay et al., 2012; Van Staden & Chen, 2010; Kathy Rao et al., 2012; Jizi et al., 2014). Companies with highly independent boards are more likely to engage in CSR activities (Harjoto & Jo, 2011) by implementing strategies that are transparent and easily understood by stakeholders. Other previous studies have shown that greater board independence helps address environmental and social challenges (Agarwala et al., 2023). In contrast, several studies have found a negative relationship between board independence and the environmental behavior of companies. Michelon and Parbonetti (2012) emphasize that stronger corporate governance mechanisms can promote sustainability practices by improving relationships between companies and conflicting stakeholders. However, they note that a higher proportion of independent directors is not necessarily related to the increased disclosure of sustainability practices. Independent directors are instrumental in shaping sustainability disclosure strategies by using their influence within the community. As key influencers, they not only confer legitimacy on the company but also demand transparency regarding its social and environmental impacts. Therefore, this improves the company’s overall transparency. Based on these roles, the second hypothesis assumes that a higher proportion of independent directors on boards is positively related to companies’ commitment to issuing green bonds. Independent directors are widely recognized as a hallmark of good corporate governance (Igalens & Point, 2009) and contribute to enhanced board effectiveness (Fama & Jensen, 1983). Their objectivity and competence can positively affect the decision to issue green bonds.

H2. 

The increasing participation of independent directors on boards has a positive effect on companies’ commitment to issuing green bonds.

2.3. Effect of Duality on GB Issuance Decision

Top echelon theory (TES), developed by Hambrick and Mason (1984), is a widely used approach to understand how top management characteristics influence strategic choices and, by extension, company performance. The knowledge, skills, and values of top managers have a significant impact on strategic decisions, particularly those related to corporate sustainability (Shahab et al., 2018). Their selective focus on external environmental factors, relevant to their professional experience and cognitive framework, has been well documented (Finkelstein et al., 2009). Moreover, shared environmental protection values among top managers encourage the development of effective environmental practices (Li et al., 2017). CEO duality, where the CEO also serves as the Chairman of the Board, is a classic indicator of corporate governance quality. In addition, it is often examined in studies assessing boards’ influence on CSR (Khanchel, 2013).

Duality in a company can influence green bond decisions by making the company less transparent and less inclined to consider external stakeholder concerns, which could affect the company’s ability to engage in green bond activities. In contrast, separating the CEO and chairperson roles provide additional checks and balances. As a result, this can prevent the CEO from diverting the board’s attention away from CSR initiatives with longer payback periods (Galbreath, 2018). Several studies have examined the relationship between CEO duality and the adoption of CSR practices.

Ahmad et al. (2017) examine the influence of CEO duality on corporate social disclosure. Their study shows a negative but statistically insignificant correlation between CEO duality and CSR disclosure. Of a sample of 125 companies, only nine exhibited CEO duality, with the CEO serving as the Chairman of the Board. The study suggests that dual-role CEOs tend to adopt a more conservative approach to CSR disclosure, emphasizing positive aspects while avoiding information that might harm the company’s reputation or management credibility. Therefore, this safeguards shareholder engagement. This cautious approach may limit the extent of CSR disclosure, as the CEO’s consolidated power can influence board decisions. Although the impact of CEO duality on CSR disclosure remains negative overall, it is not statistically significant.

Tibiletti et al. (2021) argues that when available information fails to reinforce external control, CEOs who also serve as chairpersons are less likely to produce a CSR report. In a separate study, Batubara et al. (2022) examines the influence of board size and CEO duality on CSR among companies registered with the III (Indonesian Islamic Index). Their findings show that board size and CEO duality have a significant effect on CSR.

Mallin and Michelon (2011) and Naciti (2019) observe a negative association between CEO duality and CSR performance. However, Lin and Nguyen (2022) find a non-significant relationship between these variables. Other studies, such as those conducted by Cheng and Courtenay (2006) and Ho and Wong (2001), have concluded that there is no clear relationship between CEO duality and CSR disclosure. Furthermore, Gul and Leung (2004) identify a negative and significant relationship between CEO duality and the level of CSR disclosure. They argue that when the CEO also serves as chairman, CSR disclosure is lower than in companies where these roles are separated. According to Forker (1992) and Gul and Leung (2004), this study proposes the third hypothesis: CEO duality can reduce the overall accountability of companies, making them less transparent to all stakeholders. Therefore, it can be assumed that firms with CEO duality are less likely to develop solid CSR strategies and engage in CSR practices. As a result, this can undermine decisions to reduce green bond issuance.

H3. 

CEO duality may negatively influence corporate decisions to reduce green bond issuance.

2.4. CSR Committees and Green Bond Issuance

Elbardan et al. (2023) investigate how CSR disclosure, external certification, and the adoption of Global Reporting Initiative (GRI) guidelines contribute to value creation and risk reduction. They also examine the moderating effects of CSR committees and CSR-related executive remuneration. Based on 58,105 observations over 16 years (2004–2019), the study finds that CSR disclosure and external assurance are positively associated with both corporate and sector-adjusted value. However, they are negatively related to corporate value volatility. Moreover, the moderation analysis reveals that CSR committees strengthen the relationship between CSR disclosure, external assurance, and enterprise value. However, they do not influence the relationship between GRI framework adoption and the enterprise value.

Bolourian et al. (2023) examine how the CSR committee, as a component of board structure, interacts with various director characteristics (e.g., age, seniority, and experience) and other board attributes (including non-executive directors, CEO duality, and board size) to shape CSR performance. The results of their qualitative comparative analysis reveal nine board configurations that lead to high CSR performance, with six of them including a CSR committee.

Orazalin (2020) examines the impact of board sustainability committees on environmental and social performance. In addition, the researcher investigates whether CSR strategy mediates the relationship between these committees and corporate sustainability performance. Using data from UK-listed companies between 2009 and 2016, the study finds that sustainability committees improve the effectiveness of CSR strategies. As a result, companies with effective CSR strategies exhibit better environmental and social performance. Furthermore, the empirical results show that CSR strategy effectiveness explains the positive relationship between board sustainability committees and overall corporate sustainability performance.

Baraibar Diez et al. (2019) examine the role of CSR committees in overall company performance. Based on a sample of 197 listed companies from Spain, France, Germany, and the UK during 2005 to 2015, they found that companies with CSR committees achieved significantly different ESG scores compared to those without such committees. Furthermore, the presence of a CSR committee generally led to better non-financial performance in all four countries, except for economic scores in Spain. These results underline the importance of CSR committees to improving non-financial performance and corporate sustainability. Based on both theoretical arguments and empirical results, this study hypothesizes that the presence of a CSR committee is positively related to green bond issuance. In other words, an active CSR committee should encourage and facilitate the issuance of green bonds, reflecting the company’s commitment to sustainable and responsible practices.

H4. 

A CSR committee can positively influence green bond issuance within the company.

3. Data and Model

3.1. Sample Selection and Data Collection

This study’s initial sample comprised SBF 120 companies that issued green bonds between 2013 and 2022. However, it is found that green bond issuance is concentrated among the CAC 60 companies (i.e., the top 60 capitalizations in the primary and secondary markets), which collectively made 64 green bond announcements during this period. These companies operate mainly in the industrial, information technology, healthcare, consumer staples, consumer discretionary, financial services, materials, utilities, energy, telecommunications, and real estate sectors. Although the sample is limited to a specific French context, it offers an in-depth perspective on green bond issuance by major French companies. This highlights the sector dynamics and internal corporate governance practices that influence this approach. Among this study’s sample of 60 companies, the banking sector holds a dominant position. This frequent recurrence suggests the central importance of financial institutions in the issuance of green bonds. Another notable trend is the significant presence of renewable energy companies, especially independent producers of renewable energy and solar photovoltaic systems. This trend indicates a growing interest in clean, sustainable energies in line with the energy transition. This sectoral diversity shows the resilience and adaptability of economic players within the CAC 60, illustrating the French economy’s capacity for growth.

Table 1. Summary table of variables used.

Variable	Definition/Measure	Source
Issued Amount	The total value of green bonds issued by a company during a year	Eikon Refinitiv Data stream Euronext website
Issuance	The number of green bond issues made by a company in a given year.	Eikon Refinitiv Data stream Euronext website
INDP_BD	Board of Director’s independence = Number of independent members/total number of members on the board	Universal registration reports
BD_SIZE	Board size = Number of directors on the board	Universal registration reports
Com_CSR	The presence of a Corporate Social Responsibility (CSR) committee (1 for yes, 0 for no).	Universal registration reports
Duality	Duality between the roles of Chairman and Chief Executive Officer (1 for yes, 0 for no).	Universal registration reports
INST_INV	Percentage of institutional investors based on the total capital held by these investors	Orbis database and universal registration reports
GEND_DIV	Gender diversity = Number of women in the board/total number of members on the board	Universal registration reports
COMP_SIZE	Company size = Logarithm of total assets	The Orbis database
ROE	Return on Equity = (Net income/Total Equity)	The Orbis database
DEBT	Debt ratio = (Total Debt/Total Equity)	Orbis database
R&D	Expenses in Research and Development ratio in percentage: it measures company innovation	Orbis database
STOCK	Stock return = (End Stock Price − Begin Stock price − Dividends paid)/Begin Stock Price	Orbis database
CSR_perf	CSR performance Score	Thomson Reuters database

presents the variables’ definitions.

3.2. Model Specification

This study employs the Generalized Least Squares (GLS) regression to test the hypotheses. GLS is particularly useful when data exhibit non-stationarity, heteroscedasticity (non-constant error variance), or correlated errors. Unlike Ordinary Least Squares (OLS), which assumes that errors are independent and identically distributed, GLS accounts for the complex variance–covariance structure in the error terms. This makes GLS an effective method for analyzing the relationships between internal governance variables and green bond issuance. It overcomes the limitations of traditional approaches and offers a more sophisticated approach to studying the impact of governance on companies’ sustainable financial decisions. In the first regression, this study examines the impact of internal corporate governance variables and other control factors on the amount of green bonds issued by companies. The regression model is specified as follows:

Issued_Amount_i,t = β₀ + β₁ INDP_CA_i,t + β₂ Com_CSR_i,t + β₃ Duality_i,t + β₄ GEN_DIV_i,t + β₅INST_INV_i,t + β₆ COMP_SIZE_i,t + β₇ ROE_i,t + β₈ BD_SIZE_i,t + β₉ DEBT_i,t + β₁₀ R&D_i,t + β₁₁ STOCK_i,t + β₁₂ CSR_perf_i,t + ϵ_i,t

where

• Β₀ is the constant term.
• β₁, β₂, …, β₁₂ are the coefficients for each respective independent or control variable.
• ϵ is the error term.
• i: number of firms, i = 1, …, 120
• t: the issue date

In a second regression model, this study will use the number of green bond issues as the dependent variable, while all the other variables will remain as independent and control variables.

Issuance_i,t = β₀ + β₁ INDP_CA_i,t + β₂ Com_CSR_i,t + β₃ Duality_i,t + β₄ GEN_DIV_i,t + β₅INST_INV_i,t + β₆ COMP_SIZE_i,t + β₇ ROE_i,t + β₈ BD_SIZE_i,t + β₉ DEBT_i,t + β₁₀ R&D_i,t + β₁₁ STOCK_i,t + β₁₂ CSR_perf_i,t + ϵ_i,t

where

• Β₀ is the constant term.
• β₁, β₂, …, β₁₂ are the coefficients for each respective independent or control variable.
• ϵ is the error term.
• i: number of firms, i = 1, …, 120
• t: the issue date

4. Results and Discussion

4.1. Descriptive Statistics

Results of descriptive statistics are given in the

Table 2. Descriptive statistics.

Variable	Obs	Mean	Std Dev.	Min	Max
IssuedAmountUSD	460	70,921,457	2.963 × 108	0	3.128 × 109
GENDER	460	0.407	0.102	0.055	0.727
BD_SIZE	460	13.498	2.744	8	21
INDP_BD	460	0.62	0.157	0.188	1
INST_INV	460	0.366	0.192	0.019	0.862
COMP_SIZE	460	7.426	0.678	5.512	9.317
ROE	460	12.715	29.399	−514.64	130.73
Debt	460	36.218	17.613	−17.7	73.82
R&D	460	2.116	3.988	0	21.01
Stocks	460	0.118	0.185	−0.57	0.871
CSR perf	460	64.495	14.564	12.616	101.272

.

The descriptive statistics reveal that the “Issued Amount (USD)” variable has a mean value of USD 70,921,457, indicating a generally high issuance influenced by a few extremely large values. Its standard deviation of USD 296,300,000 indicates a wide dispersion around the mean. These values range from USD 0 to 3,128,000,000, with a maximum value of over USD 3 billion. For the “Gender_Div” variable, the mean is 0.407, indicating a relatively stable measure of board gender diversity. Its standard deviation of 0.102 shows a low variability. These values range from 0.055 to 0.727, with a median of 0.43. This indicates a centered distribution. Meanwhile, the average board size (BD_SIZE) is 13.498, indicating a typical moderate value. With a standard deviation of 2.744, board size exhibits moderate dispersion. These values range from 8 to 21, with a median of 14. As a result, this indicates a concentration of values within this range. For the INST_INV variable, the mean is 0.366. This indicates a generally moderate level of institutional investment. Its standard deviation of 0.192 reveals a moderate dispersion, with values ranging from 0.019 to 0.862. It has a median of 0.361, indicating a fairly balanced distribution. Regarding “INDP_BD”, the mean value is 0.62. This indicates a relatively stable level of board independence. Its standard deviation of 0.157 reveals a moderate variability, with a range from 0.188 to 1. It has a median of 0.636, showing a distribution centered around this value. For “COMP_SIZE”, the mean is 7.426, indicating the logarithm of total assets. Its standard deviation of 0.678 reveals a moderate dispersion, with values ranging from 5.512 to 9.317. It has a median of 7.471, indicating a concentrated distribution. Lastly, the descriptive statistics of the “ROE” variable show a mean of 12.754, indicating a typical return on equity. However, a standard deviation of 29.399 reveals a wide variability, with values ranging from −514.64 to 130.73. It has a median of 12.695, indicating high variability with extreme values.

For the “DEBT” variable, the mean is 36.218, indicating a typical moderate value. The standard deviation of 17.613 suggests a notable dispersion, with values ranging from −17.70 to 73.82. It has a median of 38.03, indicating a distribution with certain extreme values but generally concentrated around the mean. The “R&D” variable has a mean of 2.116, indicating a generally low value with a few high values. Its standard deviation of 3.988 reveals a moderate to high dispersion, with values ranging from 0 to 21.01. It has a median of 0.185, indicating asymmetrical distribution with some very high values. Lastly, the total stock return and the CSR_ perf variables show mean values of 0.118 and 64.495, respectively, exhibiting a typical moderate value. The standard deviations for these variables are 0.185 and 14.564, which reveal a moderate dispersion. For the total stock return, values range from −0.57 to 0.871, with a median of 0.118. This indicates a symmetrical distribution. In contrast, the “CSR_perf” variable ranges from 12,616 to 101,272, with a median of 65,290. This indicates a distribution centered around this value.

4.2. Bivariate Correlation Analysis

Results of the bivariate correlation analysis are given in

Table 3. Pearson correlation matrix.

	Issued Amount	GEND_DIV	BD_SIZE	INST_INV	INDP_BD	FIRM_SIZE	ROE	DEBT	R&D	STOCK	CSR_perf
Issued Amount	1.0000
GEND_DIV	0.0915	1.0000
BD_SIZE	0.1381 *	0.0557	1.0000
INST_INV	−0.1572 **	−0.0649	−0.2370 **	1.0000
INDP_BD	0.0374	0.1162	−0.1314	−0.0427	1.0000
FIRM_SIZE	0.2105 **	0.2643 **	0.5047 ***	−0.2260 **	0.1510	1.0000
ROE	−0.0280	0.0412	−0.0226	0.0652	−0.0051	0.0214	1.0000
DEBT	−0.1745 **	0.0154	−0.3534 ***	−0.0454	0.0589	−0.3659 ***	0.0932	1.0000
R&D	−0.0777	−0.0419	−0.1346	0.1777 **	−0.0630	−0.2096 **	−0.1016	0.2789 ***	1.0000
STOCK	0.0086	0.0732	−0.0296	−0.0055	−0.0488	−0.0957	0.0320	0.0967	0.0566	1.0000
CSR_perf	−0.0140	0.1669 **	0.0009	−0.1400	0.2275 **	0.0605	0.0835	0.1150	−0.1377	0.0627	1.0000

Note: *** p < 0.01, ** p < 0.05, * p < 0.1.

.

To quantify the linear relationship between the explanatory variables, this study calculated the Pearson correlation coefficients and associated p-values (Table 3). Notably, it can be identified that no coefficient is greater than the threshold of 0.7 (Fox & Monette, 1992), indicating that collinearity is not a concern in this study’s data. Regarding the p-values, the analysis identifies a positive correlation of 0.2105 between the issued amount of green bonds and company size. It is suggested that larger firms tend to issue higher amounts. Furthermore, higher institutional ownership is associated with lower issued amounts of green bonds, which may imply that firms with more institutional investors rely less on external funding. Therefore, companies with higher debt levels also tend to issue lower amounts, possibly indicating financial constraints.

Table 4. Variance inflation factor (VIF) test.

Variable	VIF	1/VIF
COMP_SIZE	1.75	0.571959
BD_SIZE	1.74	0.573775
DEBT	1.57	0.636778
Com_CSR	1.54	0.648309
R&D	1.24	0.803272
GEND_DIV	1.23	0.811956
INST_INV	1.20	0.831957
Duality	1.16	0.861301
INDP_BD	1.15	0.866021
CSR_perf	1.14	0.878605
ROE	1.07	0.935514
STOCK	1.03	0.967851
Mean VIF	1.32

presents the results of the VIF test, with all values well below the multicollinearity threshold of 10.

4.3. Multivariate Analysis

In panel data models, an endogeneity may arise when there is a correlation between the explanatory variables and the error term. To address this, this study performs a Hausman test to compare the fixed-effect and the random-effect models. Specifically, this study first performs a panel random-effect linear regression, followed by a panel fixed-effect linear regression for the two models (

Table 5. Random- and fixed-effect linear regressions.

	OLS(re)	OLS(fe)
GEND_DIV	−9.30 × 107 (0.812)	−1.77 × 108 (0.980)
BD_SIZE	−6,103,635 (0.616)	−1.03 × 107 (0.481)
Com_RSE	1.20 × 108 (0.097 *)	1.32 × 108 (0.084 *)
Duality	−8.16 × 107 (0.127)	−9.49 × 107 (0.252)
INST_INV	−2.68 × 108 (0.022 **)	−4.48 × 108 (0.056 *)
INDP_BD	1.31 × 108 (0.392)	3.14 × 108 (0.217)
FIRM_SIZE	3.87 × 107 (0.129)	9.30 × 107 (0.191)
ROE	−58,151.35 (0.752)	6026.978 (0.567)
DEBT	−3,578,811 (0.030 **)	−2,082,516 (0.063 *)
R&D	3,567,390 (0.659)	−8,781,428 (0.862)
STOCK	−1.78 × 107 (0.222)	2,294,754 (0.138)
CSR_perf	551,872.9 (0.592)	971,465.5 (0.820)
Constant	−1.68 × 108 (0.537)	−6.83 × 108 (0.273)

Note: ** p < 0.05, * p < 0.1.

).

This study first performed several endogeneity tests. The Hausman test yielded a p-value of 0.8472, indicating that there is no correlation between the individual effects and the independent variables (

Table 6. Hausman test.

Hausman Test
chi2 (12)	7.16
p-value	0.8472

). Therefore, the random effects model is considered appropriate.

The Breusch–Pagan/Cook–Weisberg heteroskedasticity test yielded a significant p-value, indicating that the null hypothesis of homoscedasticity is not true and the hypothesis of heteroscedasticity is true (

Table 7. Breusch–Pagan/Cook–Weisberg test.

Breusch–Pagan/Cook–Weisberg Test	ECSRC
chi2 (1)	505.92
p-value	0.0000 ***

Note: *** p < 0.01.

). This suggests that the variance of the regression residuals is not constant across the sample, which may compromise the validity of the regression results.

This study addresses heteroskedasticity by employing robust standard errors in the estimates. Specifically, it uses the “vce(robust)” option in Stata (STATA/MP 17.0), which provides a robust variance–covariance estimator to correct for non-constant error variance across observations. Therefore, this study estimates a random effects model with robust standard errors and controls for time and industry effects to account for unobservable, time-varying biases that remain constant across sectors.

In addition, it employs generalized least squares (GLS) with clustered robust standard errors to correct for both heteroskedasticity and panel dependence. While GLS is effective when there is no censoring, it may be biased in cases of a highly zero-inflated dependent variable. To verify the robustness of the results, this study performs additional Tobit regressions, which are more appropriate for modeling both the decision to issue green bonds and the amount issued when censoring is present.

4.4. Results of Regression Models

Regression model results are presented in

Table 8. Results from regression models.

	GLS(VCE)	GLS(VCE)	Tobit
GEND_DIV	−9.30 × 107 (0.582)	−2.70 × 108 (0.312)	63.31 (0.115)
BD_SIZE	−6,103,635 (0.337)	−5,426,075 (0.388)	0.09 (0.958)
Com_RSE	1.20 × 108 (0.069 *)	9.47 × 107 (0.082 *)	18.44 (0.028 **)
Duality	−8.16 × 107 (0.087 *)	−4.71 × 107 (0.251)	−12.13 (0.076 *)
INST_INV	−2.68 × 108 (0.024 **)	−2.54 × 108 (0.027 **)	−34.57 (0.079 *)
INDP_BD	1.31 × 108 (0.392)	8.15 × 107 (0.638)	32.40 (0.160)
FIRM_SIZE	3.87 × 107 (0.320)	3.60 × 107 (0.433)	12.43 (0.051 *)
ROE	−58,151.35 (0.857)	188,502.7 (0.554)	−0.12 (0.417)
DEBT	−3,578,811 (0.053 *)	−3,037,670 (0.076 *)	−0.07 (0.765)
R&D	3,567,390 (0.659)	3,546,885 (0.267)	−0.94 (0.457)
STOCK	−1.78 × 107 (0.222)	−2.91 × 107 (0.729)	−1.73 (0.918)
CSR_perf	3,567,390 (0.292)	1,302,895 (0.380)	0.29 (0.196)
Constant	3,192,283 (0.990)	2.11 × 107 (0.937)	−199.79 (0.002 ***)
R2	0.2220	0.2895	-
Year effect	NO	YES	YES
Industry effect	NO	YES	NO

Note: *** p < 0.01, ** p < 0.05, * p < 0.1.

.

We use robust standard errors in our estimates to correct for the impact of heteroskedasticity, where the variance of the error terms is not constant across observations. We can achieve this by using the “vce robust” option in Stata, which is a robust variance–covariance estimator, to obtain robust standard errors in regression analysis. In conclusion, we use a random effects model with robust standard errors (vce robust). To account for unobservable time-varying biases that remain constant across sectors, we control for the effects of time and industry.

Generalized Least Squares GLS vce cluster with robust standard errors helps correct heteroskedasticity and panel dependence. Thus, GLS works well when there is no censoring but it is biased when the dependent variable is highly zero-inflated. To check the robustness of our results we perform additional Tobit as it is more appropriate for modeling the decision to issue and the amount issued, while GLS helps when censoring is not a major concern.

The results of regression models allow us to formulate a relation between corporate governance mechanisms and green bond issuance decisions. Firms wishing to access the green bond market should pay attention to their governance mechanisms because these are beneficial to their shareholders’ wealth once they enter the green bond market.

The results show that the CSR committee variable is significant and positive in the two regressions (p < 0.05). This indicates that companies with a CSR committee are more likely to issue green bonds. The results of this study are similar to those of previous studies that found a significant relationship between the presence of a CSR committee and improved environmental, societal, and overall CSR performance (Orazalin, 2020; Baraibar Diez et al., 2019; Orazalin, 2020). They indicate that the presence of sustainability committees helps companies develop CSR strategies to improve their environmental and social performance and strengthen their sustainability and non-financial outcomes (Bolourian et al., 2023; Elbardan et al., 2023). Furthermore, Gull et al. (2023) confirm that the existence of a CSR committee demonstrates the board’s commitment and the importance it places on responsible behavior, signaling the company’s efforts to invest in building increasingly solid relationships with stakeholders. This result is in concordance with stakeholder theory, according to which the existence of a CSR committee implies that the company is not merely reporting its environmental commitment but can be expected to decide on investments aimed at sustainable development.

We also note that the presence of women in the board does not impact the decision of GB issuance, which is inconsistent with the contribution of stakeholder theory, upper echelon theory and gender socialization theory with respect to board diversity. Stakeholder theory argues that the presence of more women on boards improves stakeholder relations and can influence board members’ attitude with respect to environmental decisions, thus improving environmental practices.

According to upper echelon theory, the presence of female directors on the board encourages greater perspective diversity and enriches useful information in decision-making processes, consequently improving the board’s cognition. From the perspective of gender socialization theory, women naturally possess communal traits that are relevant for the establishment and maintenance of relationships with stakeholders and encourage boards to better understand their demands. They also have higher ethical standards than men in the workplace, which help them to play a large role in promoting discussion of ethical issues and finding solutions to these ethical issues. They can influence the board’s attitude regarding CSR and stakeholders’ interests and contribute to eco-innovation.

However, some studies (Issa & Bensalem, 2023; Ali et al., 2024) show that the relationship seems to be indirect. They also show that the relationship could be mediated by other variables like CSR strategies or green innovation. This means that female board directors have an important role in promoting CSR strategy and green innovation in a firm by stimulating the board’s cognition, and thus fostering eco-innovation.

This result can also be attributed to the specific nature of the French context, linked to the mandatory adoption of a 40% quota for women on the board of directors, which limits the variability of the diversity variable in the sample. We, therefore, conclude that it is necessary to adjust the theories on this subject to the context studied and to verify the indirect role of diversity in explaining this relationship.

The indebtedness ratio (DEBT) is significant and negative across all regressions (p < 0.05). This indicates that companies with a lower debt ratio are more likely to issue green bonds, possibly because they have a greater need for external financing. Meanwhile, a higher debt ratio could indicate greater financial stability, reducing the need to issue green bonds to attract responsible investment.

Other corporate governance variables, including CEO duality (Alabdullah et al., 2019; Ahmad et al., 2017; Khan et al., 2013; Mallin & Michelon, 2011; Naciti, 2019; Lin & Nguyen, 2022), the proportion of institutional investors (INST_INV), board independence (INDP_BD), firm size (FIRM_SIZE), return on equity (ROE), research and development (R&D) expenditures, and CSR performance (CSR_perf), do not exhibit statistically significant relationships with green bond issuance. This lack of significance is in line with previous findings reported by Dang et al. (2021), Bukair and Rahman (2015), and Tanaka (2019). Here also, we see the need to explore these relationships further in subsequent works to check the possible existence of an indirect relationship.

This study offers significant contributions for companies, researchers, and policymakers. For companies, the results highlight the importance of the CSR committee as a corporate governance mechanism in the decision to issue green bonds. In particular, the presence of a CSR committee is associated with an increased likelihood of issuing green bonds, highlighting the need to integrate sustainable and responsible practices within corporate governance. This insight can guide managers in strengthening their CSR policies to attract green and responsible investments. For researchers, this study enriches the literature on sustainable finance by clarifying the internal corporate factors that influence the issuance of green bonds. The results on the significant, negative effect of the indebtedness ratio indicate that financially less stable companies are more inclined to seek green financing. This finding paves the way for future studies on the relationship between financial stability and green financing behaviors. In addition, they underline the need to explore other potential determinants while re-evaluating traditional assumptions about governance and green finance or to check the indirect relationship between some corporate governance characteristics and green bond issuance decisions. Lastly, for regulators and policymakers, these results can help formulate policies that promote stronger corporate governance and improved sustainability practices. Therefore, this strengthens the green bond market and supports global sustainability initiatives.

5. Conclusions

This study aims to analyze the impact of corporate governance factors on the issuance of green bonds among CAC 60 companies (i.e., the 60 largest capitalizations in France’s primary and secondary markets). This theme is crucial in today’s transition to a sustainable economy, as understanding how corporate governance practices influence green bond issuance is key to promoting environmental financial strategies. Although this study’s methodology is straightforward, it offers an innovative perspective on the relationship between green bond issuance and corporate governance. To date, no previous work has applied the Generalized Least Squares (GLS) model to analyze this relationship. Using this GLS model and advanced statistical techniques, this analysis reveals a complex dynamic in which factors such as female board representation, board size, CEO duality, board independence, and the debt ratio exhibit an insignificant direct effect on green bond issuance decisions. Overall, this in-depth analysis highlights the corporate governance practices that foster the commitment of CAC 60 companies to sustainable finance and underlines the importance of a robust governance framework to encourage the issuance of green bonds. The results show that certain internal corporate governance variables have a significant impact on green bond issuance. In particular, the presence of a CSR committee shows strong statistical significance across all regressions, underlining the importance of corporate social responsibility practices. In contrast, a higher debt ratio is negatively associated with green bond issuance, possibly reflecting a reduced need for alternative financing among financially stable companies. Lastly, this study notes that the other internal corporate governance and control variables do not exhibit statistically significant relationships with green bond issuance. This suggests that other factors, perhaps more directly linked to companies’ specific sustainability practices and environmental policies, could play a more decisive role. Therefore, companies seeking to improve their green bond issuance prospects should consider strengthening their corporate governance frameworks (i.e., by creating an ESG committee).