Corporate Governance and CSR Disclosure: International Evidence for the Period 2006–2016

Abstract

In this paper, the authors examine the impact of corporate governance mechanisms on corporate social responsibility (CSR) disclosure in European and Anglo-Saxon contexts. The study is based on 324 Anglo-Saxon listed corporations and 310 European listed corporations for 11 years from 2006 to 2016 (6813 year-observations). The regression analysis shows that board gender and board age affect CSR disclosure. This study also finds that CEO duality negatively affects CSR disclosure in both contexts. Finally, the study found that the existence of a CSR committee and CSR experts positively affect CSR disclosure in both contexts.

1. Introduction

Companies have become more aware of their social and environmental responsibilities than ever. There has been an increase in attention paid globally to issues affecting communities, including environmental issues driven by stakeholders’ pressure. Companies decide to participate in CSR initiatives to boost their financial performance, as the literature has shown that corporate social responsibility (CSR) disclosure affects corporate financial performance (Budiharjo 2019; Uwuigbe and Egbide 2012; Rakhiemah and Agustia 2009), as well as satisfies stakeholders’ needs (Kim 2011). Companies provide information about their social activities. These include, but are not limited to, enhancing waste management, reducing environmental impact, abiding by environmental laws, and protecting employees.

Corporate governance (CG) has received great attention from academics and practitioners, especially after the 2008 global financial crisis. The literature shows that CG affects corporate voluntary disclosure (Al Lawati et al. 2021) in general and CSR information in particular (Albitar et al. 2022). This paper examines the influence of CG on CSR disclosure. It investigates whether CG similarly affects CSR information disclosure in two different contexts: the Anglo-Saxon and the continental European contexts. The distinction between the Anglo-Saxon model and the continental European (European) model is linked to countries’ economic, legal, and cultural characteristics using each model. In this paper, the “European” context refers to countries that belong to continental Europe, more specifically to Western Europe. These countries are found in the literature to have more quality and quantity disclosure than the East side of the continent (e.g., Hąbek 2017; KPMG 2020).

The literature has paid attention to the differences and resemblances that exist between these two contexts (e.g., Miniaoui et al. 2019; Halaoua et al. 2017; Pillay 2013; Sison 2009; Ooghe and De Langhe 2002). It was, in fact, due to their unlike legal systems, accounting systems, cultural dimensions, economic characteristics, and so on. Hence, particular CG models for companies do exist since they work in various business contexts. Looking at these contexts separately can prompt pointless ends. The goal here is not to prioritise one model over another, but to study the likenesses and contrasts in these models and practices (Ooghe and De Langhe 2002). The current state of the art offers only a handful of comparative studies based on contextual differences on the subject of CSR disclosure, hence, the originality of the valuable contribution of the present research. This paper offers an answer to the following research question: does CG likely affect CSR disclosure in Anglo-Saxon and European contexts?

The study uses a sample of an 11-year period, from 2006 to 2016. The analysis is based on 324 Anglo-Saxon listed firms and 310 European listed firms (6813 year-observations). It provides empirical evidence that gender diversity and age diversity and other CG mechanisms affect CSR disclosure.

The rest of the paper is organised as follows. Section 2 reviews relevant literature. Section 3 briefly presents the study’s contexts and major differences, while Section 4 develops the research hypotheses. Section 5 presents the empirical findings. Section 6 discusses research findings. Section 7 concludes the paper.

2. CG and CSR Disclosure

CG includes the existing controls and methods to guarantee that executives act in the best shareholders’ interest. These mechanisms may influence the levels and quality of corporate disclosure CSR disclosure was created initially for environmental issues; it activated along the rising interest in sustainability matters (García Martín and Herrero 2020; Michelon and Parbonetti 2012; Adams 2002). Companies with less relevance to environmental impacts were inclined not to publish CSR reports (Javed et al. 2020; Branco and Rodrigues 2006). However, regulatory developments have encouraged companies to exercise more governance over their CSR disclosure because the information they disclose needs to be reliable (KPMG 2005). These regulatory developments are bursting around the world strongly CSR disclosure in corporate annual reports, like the 2002 Sarbanes-Oxley Act in the USA, the Operational Financial Review (OFR) requisites in the UK, and the European Union Directive, which contribute to the increased reporting on this subject (Coates 2007).

Previous research about CG had initially embraced the perspective of agency theory (Al-Bassam et al. 2018; Al Lawati et al. 2021). Prior research shows that CG affects CSR disclosure in different contexts: for instance, Miniaoui et al. (2019) studied the impact of country-level governance and legal system on CSR engagement in Anglo-American and Euro-continental contexts; Hassan (2013) studied the relationship between CG mechanisms and extent of corporate voluntary reporting in the UAE; Peters and Romi (2012) examined the determinants of the voluntary disclosure of greenhouse gas emission information in the USA. They evaluated the relationship between GHG disclosures and two CG attributes: the existence of an environmental committee on the board and a sustainability officer. Ben Rhouma and Cormier (2007) examined the determinants of corporate environmental and social reporting of non-financial French firms listed in SBF 120. The main conclusion of these papers (and others) is that there is a significant association between CG and CSR disclosure. It is believed that companies need to consider the quality of CG when forming their CSR activities. Reporting such activities could reduce information asymmetry between insiders (managers) and outsiders (shareholders). Reducing information asymmetry would lead to positive economic consequences. The literature provides evidence that CG is an important mechanism affecting corporate disclosure practices (Al Lawati et al. 2021). Therefore, it is expected that CG would affect CSR disclosure.

3. The Contexts of Study

3.1. The Anglo-Saxon Context

The term “Anglo-Saxon” designates English-speaking countries such as Australia, Canada, the UK, the USA, and New Zealand. In the field of accounting and finance, “Anglo-Saxon” refers to countries using the same accounting model, including the Netherlands Australia, Mexico, and Hong Kong, besides those mentioned above. These countries have many similarities, such as the culture, legal system (common law), and economic model (D’Arcy 2000). Anglo-Saxon countries are countries of common law, where case law, customs, and practice are predominant (Eromonsele 2017). The laws merely state the general principles and the professional bodies responsible for establishing the accounting standards (La Porta et al. 2000). The rules can thus adapt quickly to changes in the economic environment. In Anglo-Saxon countries, shareholders are the main focus. This is a market-oriented CG model distinguished by large and more liquid stock markets, low and dispersed ownership concentration, a high level of shareholders’ minority protection, and a dominant role of institutional investors (Jabbouri and Jabbouri 2021; Thomsen 2003). It is also distinguished by a large number of listed companies, the share of which is held by a large number of (institutional) investors with distant relationships (Franks and Mayer 1994). CG reforms in countries under a market-based model are in board structures and practices that ensure that the board acts separately from management and that the board constitutes a distinct entity, capable of being objective (Lane et al. 2006). Given the significance of financial markets for Anglo-Saxon firms, investors and monetary experts’ inclinations weigh vigorously on these firms. In this specific circumstance, managers are interested in introducing more non-financial information disclosure to boost firm value and draw in additional investors.

3.2. The European Context

The European model presents the opposite characteristics of the Anglo-Saxon model. The countries of continental Europe are countries of written law. The “State” plays a leading role in the accounting normalization process. Therefore, the evolution of accounting texts is slower than in Anglo-Saxon countries. The vast majority of European countries rely upon codified rules that satisfy stakeholders’ information needs (Ben Othman and Zeghal 2006). The European CG model, also known as the stakeholder system, is common in European countries. This model considers all stakeholders (Cernat 2004). Therefore, the CG model’s framework is based on stakeholder theory. In this CG structure, corporate ownership is largely held by a small group of investors. Most publicly traded shares are controlled by shareholder groups, giving them more influence over corporate decisions. In this model, corporate ownership is concentrated among strategically oriented banks and other industrial companies. As a result, management is less affected by the market for corporate control than it would be under the Anglo-Saxon model (Becic 2011).

In contrast to the Anglo-Saxon model, employees participate in corporate strategic decision-making through works councils or union representation. The European model is based on internal control mechanisms, including director remuneration, board composition, and performance-based management rewards (Setia-Atmaja 2008 in Ishak et al. 2011). Banks greatly influence the governance process (Becic 2011). Banks play a crucial role in external governance by offering relational finance, financial services, and monitoring. Large shares in European enterprises are owned by banks and industrial companies, which actively participate in CG. Block holders rarely trade their shares, but they use their voting power to affect how firms are controlled directly (Jackson and Moerke 2005). With a more concentrated ownership structure, owners can maintain control over a corporation and make decisions that increase its profitability (Becic 2011).

Regarding CSR, non-financial information disclosure has been empowered for nearly thirty years through voluntary and, lately, mandatory regimes. The European Non-Financial Reporting Directive (NFRD) (Directive 2014/95) established the regulations relating to non-financial information and various other information that large companies in the EU must disclose in their annual report. Since 2018, this EU regulation has obliged European companies to develop a responsible approach in how they operate and manage environmental, social, human rights, and anti-corruption challenges. Then in 2017, the European Commission published non-binding guidelines on the presentation of non-financial information. These guidelines aim to help companies disclose high-quality, relevant, useful, consistent, and comparable non-financial (environmental, social, and governance) information to promote growth and strong and sustainable jobs and ensure transparency for stakeholders. These non-binding guidelines are offered as part of the reporting obligations under the directive. They aim to help companies prepare relevant, useful, and concise non-financial statements per the requirements of the directive.

Based on the above discussion, our analysis is predicated on the statement that dissimilarities in CG practices, accounting systems used, institutional determinants, and cultural dimensions affect CSR disclosure differently in Anglo-Saxon and European firms.

4. Hypotheses Development

4.1. Board of Directors

A CG system includes internal and external mechanisms designed to align the interests of managers and stakeholders (Schäuble 2019). High-quality CG improves corporate disclosure practice (Forker 1992; Al Lawati et al. 2021). Previous literature has shown that internal CG and managerial incentives can alleviate agency costs (Chung et al. 2012; Al Lawati et al. 2021), thus improving levels of corporate voluntary disclosure.

The literature has considered the characteristics of the board of directors and shows that the board of directors is an important determinant of the quality of a CG system (Bhagat and Bolton 2008; Al Lawati et al. 2021). According to Halme and Huse (1997), environmental organizations and corporate activists may request that the board increase oversight of CSR policies and disclosures to assist in making the company more socially responsible and acceptable. Governance theories hold that the efficiency of a board is judged by its composition. The board’s composition is considered one of the main drivers of CG quality (Rose 2007). The authors believe that boards’ composition is a key driver of socially responsible behaviour and strategic decision-making. (Cuadrado-Ballesteros et al. 2017). In other words, a board’s structure is a key factor in determining what each member may do.

4.1.1. Gender Diversity

Gender diversity refers to the quest for a certain level of balance in the representation of men and women on the board (Dardour et al. 2018; Al Lawati et al. 2021). Different behaviours between women and men remain a subject of debate. However, prior research showed that women and men behave differently in organizations (Ferrary 2010). They differ in many aspects but particularly in regard to their management practice (Fenwick and Neal 2001), their decision-making processes (Barber and Odean 2001), and their motivation (Kalleberg and Leicht 1991). According to agency theory, diverse boards tend to better govern management decisions since diversity increases board independence (Carter et al. 2010). The literature (Hillman and Dalziel 2003; Mallin and Michelon 2011; Haque 2017) highlight various resource-provisioning responsibilities that female directors can play in terms of human and relational capital that is necessary to improve businesses’ CSR performance, and to ease global environmental challenges such as greenhouse gas emissions and the loss of biodiversity (Haque and Jones 2018).

Female directors are crucial in promoting corporate moral legitimacy (Zhang et al. 2013) and show greater sensitivity towards societal stakeholders’ concerns (Haque and Jones 2018). Additionally, they provide their boards’ distinct expertise, experience, and skills (Sheridan and Milgate 2005). Additionally, a company with more female directors is more likely to implement a proactive and comprehensive CSR strategy (Shaukat et al. 2016). According to Khan et al. (2020), women participate in social activities more than men. They are likely to put more effort into monitoring duties (Bender et al. 2015), provide higher extent of board accountability, and are better organised during meetings (Terjesen et al. 2009). Hillman et al. (2007) stipulate that female directors dedicate significant effort while monitoring managers’ decisions. Therefore, having more women on boards may improve their understanding of CSR initiatives and disclosures (Bear et al. 2010).

Most prior research found a positive association between CSR disclosure and the presence of women members on board (Ayuso and Argandoña 2009; Giannarakis et al. 2014; Javaid Lone et al. 2016; Katmon et al. 2019; Lu and Wang 2021). Some other studies found a negative association between the two variables. These include Handajani et al. (2014), Kiliç et al. (2015), Cucari et al. (2018), and Fahad and Rahman (2020). However, very few studies found an insignificant association between gender diversity and CSR disclosure, such as Campanella et al. (2021), Zaid et al. (2019), and Galbreath (2013).

Based on the agency theory, the first hypothesis is formulated as follows:

Hypothesis H1:

CSR disclosure is likely to be influenced positively by the presence of women on the board.

4.1.2. Age Diversity

Traditionally, most board members have tended to be seniors, experienced and elderly members (Kang et al. 2007). This can be explained by the inbuilt nature of companies’ career development and management, which often considers board members, who display significant experience in other companies of the same field, as ideal directors (Gilpatrick 2000). According to Giannarakis (2014b), board members’ age is an indicator of evaluating directors’ experience. General maturity and experience are usually considered as marking features of older managers (Bantel and Jackson 1989). They have additional knowledge, the right decision-making skills, and follow a more direct and orthodox way of governing companies (Bantel and Jackson 1989). Thus, they are believed to be more oriented towards a safer way of doing business and are most likely averse to taking any business risk due to innovative business ideas, including CSR activities (Vroom and Pahl 1971). Contrarily, youthful members got their education very recently, qualifying them to be more up to date with recent management skills and orientations. They actively drive business success and prospect planning (Handajani et al. 2014). They also are more enthusiastic, loaded with dynamic energy, more risk takers, and usually attracted toward emerging innovations in the market such as technological change. As a result, they strive to include more CSR initiatives than older directors (Fahad and Rahman 2020).

The impact of board age on CSR disclosure has not been widely studied. Very few studies consider this impact. For example, Giannarakis et al. (2014) did not find any impact of board age on CSR disclosure, while Fahad and Rahman (2020) and Hashim et al. (2019) found that higher board average age reduces ESG and governance disclosure scores. Beji et al. (2021) found that age diversity is associated with higher CSR performance. In the current study, the authors believe that a more diversified board provides more richness and openness toward CSR disclosure. In other words, it benefits from the advantages of both age categories: younger and older. Thus, the second hypothesis is formulated as follows:

Hypothesis H2:

CSR disclosure is likely to be influenced positively by board age diversity.

4.2. Leadership Structure and CEO’s Compensation

4.2.1. Leadership Structure

According to the agency theory, CEO role duality could impair the board’s independence (Fama and Jensen 1983). Role duality can seriously impair boards’ monitoring function (Barako et al. 2006). The combination of the two roles makes it difficult for the board to replace an ineffective CEO (Shivdasani and Zenner 2004). Regulatory recommendations around the world stressed the necessity of separation between the two positions for better governance. For example, the Cadbury Report (1992) and The Higgs (2003) recommended that UK firms need to appoint at least three independent/non-executive directors and the roles of the chairman and CEO should be separated. In France, the law NRE of 15 May 2001, introduced the separation between management and control functions. The separation of CEO and Chairman positions or their combination and its impact on CSR disclosure has been examined in a number of studies. For example, Ananzeh (2022), Xiao and Yuan (2007), Haniffa and Cooke (2002), and Gul and Leung (2004) found a negative relationship between the combination of roles and voluntary disclosure. Contrarily, the combination of roles of CEO and Chairman of the board was found as insignificantly correlated to societal disclosure by other researchers (e.g., Michelon and Parbonetti 2012; Barako et al. 2006). Meanwhile, Pucheta-Martínez and Gallego-Álvarez (2019) found that CEO duality has a positive effect on CSR disclosure. Albeit the evidence on the relationship between CEO duality and CSR disclosure is blended, the vast majority of earlier studies uphold a negative association between CEO duality and CSR disclosure. Hence, the authors hypothesise that:

Hypothesis H3:

CSR disclosure is likely to be influenced negatively by the dual role of the CEO.

4.2.2. CEO Compensation

From a theoretical point of view, two perspectives predominate in explaining the interactions between CSR and executive compensation. The first perspective is based on agency theory (Jensen and Meckling 1976). In the principal-agent model, the quality of CSR information represents a customs clearance cost born by the agent to put the principal in confidence. CSR disclosure makes it possible, on the one hand, to reduce agency costs and, on the other hand, users’ confidence in the accuracy of the information provided by the company (Simnett et al. 2009). It also contributes to reducing information asymmetry by promoting transparency (Kolk and Perego 2010) and exempting shareholders from an individual search for reliable information. Barnea and Rubin (2010) consider the overinvestment hypothesis in social responsibility activities. These can lead to deterioration in the value of the firm. If the manager over-invests in CSR communication, they can derive private benefits such as improving their reputation as a responsible citizen. It also allows him to increase his bargaining power, market value, and career prospects from a reputation as a responsible leader. Milbourn (2003) found a positive association between the reputation of the CEO and the sensitivity of stock compensation to the creation of shareholder value. This finding motivates the authors to investigate the extent to which a CEO’s compensation and a company’s commitment to CSR disclosure are positively correlated.

The second viewpoint fits into the stakeholder theory paradigm (Freeman and Medoff 1984; Clarkson 1995). CSR disclosure is part of taking responsibility for how a company’s activities affect stakeholders (Capron and Quairel 2009). According to the second perspective, engaging in CSR activities can help to resolve conflicts between management and stakeholders. By moving beyond simple reporting, CSR disclosure allows managers to notify stakeholders of the outcomes of their decisions and actions. It is then a matter of relying on sustainable development reports to produce sustainable “balance scorecards” (Naro and Noguera 2008). Thus, to increase legitimacy and develop a positive reputation, managers make efforts, through CSR activities and social dialogue, to return control to various risk-takers (Cardebat and Cassagnard 2011), and to effectively manage corporate risks (Godfrey 2005). In the end, this increases the corporate value and its long-term economic viability. This conflict resolution hypothesis predicts that executive compensation will positively impact CSR disclosure (Cai et al. 2011; Fabrizi et al. 2014; Malik et al. 2020). This is because managers must exert significant effort to prevent potential conflicts of interest with other stakeholders, including employees and the government, and as a result, will need higher compensation than managers of less responsible companies (Suttipun 2021). A company that engages in irresponsible behaviour may experience significant issues with its stakeholders. For instance, if customers boycott a certain product brand, shareholder value may decrease. In this context, the authors assume that executive’s compensation might have an impact on CSR disclosure, and hence the study hypothesises that:

Hypothesis H4:

CSR disclosure is likely to be influenced positively by a higher compensation of the CEO.

4.3. CSR Committee and Directors’ Expertise

Another way to promote the consideration of extra-financial issues is to create board committees dedicated to CSR. A social responsibility committee (SRC) among the other board committees can be a valuable tool that motivates executives’ activities to meet all stakeholders’ expectations. Such a committee could demonstrate the board’s commitment to environmentally responsible behaviours (Monks and Minow 1995). From the agency theory perspective, forming such a committee might also be seen as a useful monitoring tool for enhancing CSR disclosure (Michelon and Parbonetti 2012). SRC aims to implement positive environmental strategies and assist companies in determining the key CSR issues that are most likely to impact their performance (Peters and Romi 2011). According to the stakeholder theory, the launch of such a committee is a sign of greater awareness and a means of addressing the interests of a wider range of stakeholders and a reaction to their expectations (Michelon and Parbonetti 2012; Peters and Romi 2011; Rupley et al. 2011).

However, the existence of an SRC does not necessarily guarantee its efficiency. Firms can put in place such committees just as camouflage for non-socially responsible activities (Aguilera and Cuervo-Cazurra 2009). Therefore, firms must assure the credibility of SRC and grant it a real and relevant function. The authors believe that having a CSR specialist on the committee is a great sign of how effective the committee is. Such a member’s expertise and experience would improve the firm’s awareness of CSR strategies and disclosures. SRC credibility is, then, further enhanced and the quality, as well as quantity of CSR disclosures, is improved. Limited researchers have undertaken studies on the impact of SRC on CSR disclosure. For example, Davis et al. (2021), Peters and Romi (2011, 2013), and Cucari et al. (2018) find that the presence of SRC and SR experts had a positive effect on CSR disclosure. Conversely, the results of Michelon and Parbonetti (2012) and McKendall et al. (1999) revealed the absence of association between the presence of SRC and the level of sustainability disclosure. It is suggested then that an effective SRC consisting of independent members with experience in CSR would enhance CSR disclosure. The authors, therefore, hypothesise that:

Hypothesis H5:

CSR disclosure is likely to be influenced positively by the existence of an SRC among the board committees.

Hypothesis H6:

CSR disclosure is likely to be influenced positively by a greater presence of directors proficient in CSR on the SRC.

5. The Empirical Analysis

5.1. Sample, Model, and Variables

The initial sample of the current study includes firms belonging to the Anglo-Saxon context that are listed on the major stock exchanges of their country of origin1 (S&P 500 for the US firms, FTSE100 for the UK firms, and S&P/TSE 300 index for the Canadian firms) (Albitar et al. 2020; Dicko et al. 2020; Manita et al. 2018). It also includes the European sample firms listed on the STOXX 600 index (Dwekat et al. 2022).

Special attention was given to British and Canadian companies. In fact, the UK was the main European country using the Anglo-Saxon accounting system (Halaoua et al. 2017). Thus, companies whose headquarters were established in the UK will be considered among the Anglo-Saxon sample. Regarding Canada, it is considered among the Anglo-Saxon countries (Ben Othman and Zeghal 2006). However, the province of Quebec has a system based on French civil law. Thus, companies with headquarter located in Quebec were avoided and deleted from the sample. The sample firms cover all non-financial companies since the financial firms are subject to different disclosure and governance regulations. The study contexts are deemed to have the highest rates of CSR disclosures worldwide (KPMG 2020; Vartiak 2016).

The study period is 11 years, starting from 2006 to 2016. The sample period ends by 2016 to exclude any direct or indirect impact of the American elections and the entry of D. Trump to the office in 2017, which may affect the research findings. Relevant literature (e.g., Klaus et al. 2022; Antonini et al. 2021; Weldon 2017) shows that the political transition in the USA after 2016 could possibly influence CG and CSR disclosure practices. In fact, companies and investors benefited from a stock market perk up in the weeks following the 2016 United States presidential election, empowered by expectations of tax cuts, increased government spending, and significant deregulation, which possibly could affect (positively or negatively) the quality and quantity of their disclosures. The Trump administration made legal and regulatory changes that affected CG and CSR in the country. The way the former US President views the climate change treaty, for instance, is a striking example of that when he announced that the US participation in the 2015 Paris Agreement on climate change mitigation is suspended. Weldon (2017) uttered that Trump Administration’s stated priorities regarding certain social issues affect both CSR and CG. These may lead a board to consider cutting back on funding CSR or sustainability initiatives. Furthermore, the findings of Antonini et al. (2021) show that companies headquartered in states that strongly supported Trump in the 2016 election have more negative changes in disclosure than other firms. Antonini et al. (2021) argued that “attention should be paid to potential differing political climates in the locales of organizations’ headquarters”, in other words, pro-Trump states versus anti-Trump states. Therefore, we limit our analysis to the financial year ending in 2016.

CG and CSR data was retrieved from the Bloomberg database. Even though Bloomberg ESG Disclosure Score was launched in 2009, it has data since 2006. Firms with missing data were deleted. Consequently, the final sample of this paper consists of 324 listed corporations from the Anglo-Saxon countries, and 310 European listed corporations, thus a total of 634 firms and 6813 firm-years observations for the analysis. The distribution of the sample firms by industry and context is provided in

Table 1. Industry summary.

N	Industry	European	Anglo-Saxon	Total
				N	%
1	Chemistry	13	19	32	5.14
2	Construction	54	56	110	17.35
3	Electricity	29	30	59	9.3
4	Food and drinks	14	16	30	4.75
5	Industrial machines	17	14	31	4.88
6	Mining	22	25	47	7.41
7	Telecommunications	63	42	105	16.56
8	Hospitality/Entertainment	43	59	102	16
9	Manufacturing, Merchandising	55	63	118	18.61
Total		310	324	634	100

.

To capture CSR disclosure (dependent variable), the present study uses the ratings from the Bloomberg database. The ESG disclosure score provided by Bloomberg’s database is used as a proxy for CSR disclosure (Giannarakis 2014a). Bloomberg database provides ESG Disclosure scores for major indexes worldwide, including FTSE100, S&P/TSE 300 index, S&P 500, and the European Region’s STOXX Europe 600.Therefore, it’s expected to ensure a wide range of data availability. Besides, Bloomberg’s ESG Disclosure scores are based on information available in annual reports, CSR reports, and firm websites (Albitar et al. 2020). According to Marquis et al. (2011), the ESG database is the most comprehensive source for evaluating companies’ environmental, social, and governance activities and outcomes. Moreover, Bloomberg’s ESG scores are based on recommendations from the Global Reporting Initiative GRI (Manita et al. 2018) which gives them credibility and reliability. In the academic literature, CSR disclosure scores published by Bloomberg have been regularly used (e.g., Albitar et al. 2020; Cucari et al. 2018; Li et al. 2018; Manita et al. 2018) thanks to Bloomberg’ wide set of CSR measures reported by publicly traded firms around the world. Accordingly, the paper relied on Bloomberg ESG ratings of CSR disclosure.

The following multiple regression model is estimated:

CSRD_(j,t) = α + β₁BD_GENDER_(j,t) + β₂BD_AGE_(j,t) + β₃CEO_DUAL_(j,t) + β₄CEO_COMP_(j,t) + β₅CSR_COM_(j,t) + β₆CSR_EXPERT_(j,t) + β₇ SECTOR + ε

where

• CSRD_(j,t) = total score received the firm _j during the year _t under CSR disclosure index;
• α = the constant, and
• ε = the error term.

Table 2. Variable summary.

Label	Description	Measures
CSRD	Corporate social Responsibility disclosure Index	CSRD is measured by ESG ratings at Bloomberg’s ESG disclosure scores for each company (Cucari et al. 2018; Manita et al. 2018).
BO_GENDER	Gender diversity on the board	The percentage of female directors on the board (Zaid et al. 2019; Cucari et al. 2018).
BD_AGE	Age diversity on the board	Herfindahl-Hirschman Indices (HHI) HHI= ∑ {(each director’s age)/(total directors age).
CEO_DUAL	Combination of roles of CEO and chairman of the board	Dummy variable 1 if the CEO is also the chairman of the board, and 0 if not (Malik et al. 2020).
CEO_COMP	Executive compensation	CEO annual compensation in US dollars (Malik et al. 2020)
CSR_COM	Existence of a CSR committee within the board’s subcommittees	Dummy variable 1 if CSR committee exists, 0 if not (Cucari et al. 2018).
CSR_EXPERT	CSR expertise of the directors within the CSR committee	Percentage of CSR expert within CSR committee (or the committee in charge of CSR issues) (Fahad and Rahman 2020).
SECTOR	The sector in which operates the company	Control Variable: Dummy variable that takes the value of 1 if the company operates in a pollutant industry, 0 otherwise (Zéghal and Dammak 2007).

shows the variables’ definitions and measurements. The chosen measures are taken from the previous literature in the field.

5.2. Descriptive Statistics

Table 3. Descriptive statistics.

Panel A: Descriptive statistics of dichotomous variable
Variable name	Yes *	No **	Unit
CEO_DUAL	51.4 (54.0)	58.6 (46.0)	%
CSR_COM	52.3 (67.4)	47.7 (32.6)	%
SECTOR	60.1 (42.9)	39.9 (57.1)	%
Panel B: Descriptive statistics for quantitative variables
Continuos Variables	Mean	Standard Deviation	Min	Max
CSRD	32.46 (31.85)	18.47 (17.93)	4.92 (3.31)	72.18 (73.42)
BD_GENDER	14.35 (13.59)	9.37 (10.88)	0 (0)	46.66% (41.66%)
BD_AGE	61.15 (54.11)	4.2 (6.88)	37.8 (41.57)	69.4 (74.5)
CEO_COMP	245.79 (307.55)	408,065.54 (664,158.35)	113,000 (142,000)	3100.000 (2845.00)
CSR_EXPERT	0.106 (0.06)	0.17 (0.24)	0 (0)	33.33% (25%)

Values between brackets () are for the European sample. ** p < 0.05, * p < 0.1.

provides descriptive statistics. With regard to total CSR disclosure scores, the statistics show a wide gap between the minimum and maximum scores, reflecting the disparity of CSR disclosure strategies adopted by different companies in both samples. The Anglo-Saxon sample has an average total CSR disclosure of 32.46, while the European sample has an average CSR disclosure of 31.85. The average presence of women on the board of Anglo-Saxon companies is 14.35 and a standard deviation of 9.37. Compared to previous research (e.g., Adams and Ferreira 2009), this number has noticeably increased, which indicates the increase in awareness of the role of women in the corporate structure.

The average CSR role duality is 51.4% in the Anglo-Saxon sample. This indicates that the most sample companies in the Anglo-Saxon adhere to CG principle of separating CEO and chairman roles. However, 54% of the European sample firms combine the positions of CEO and chairman of the board in the same person. The results also show that 52.3% (67.4%) of the Anglo-Saxon (European) companies have a CSR committee or a committee in charge of sustainability issues. The ratio of experts on the committee ranged from 0 to 33% for the Anglo-Saxon sample and from 0 to 25% for the European sample, with a mean of 10.6% (6%). Such results indicate that almost the majority of the sample companies not only support the formation of CSR committees but also hire CSR experts to assure and guarantee the committees’ activities. These results highlight the emphasis placed by companies of both samples on CSR issues.

5.3. Correlation Analysis

Table 4. Pearson correlation analysis.

	CSRD	BD_GENDER	BD_AGE	CEO_DUAL	CEO_COMP	CSR_COM	CSR_EXPERT	SECTOR
CSRD	1
BD_GENDER	0.153 ** (0.176 ***)	1
BD_AGE	−0.012 * (0.098 **)	−0.213 (−0.151)	1
CEO_DUAL	−0.103 ** (−0.451 **)	0.330 *** (0.084)	0.138 ** (0.252)	1
CEO_COMP	0.101 *** (0.075 ***)	−0.017 (−0.207)	0.341 (0.533)	0.095 (0.411)	1
CSR_COM	0.013 ** (0.341 *)	0.223* (0.663)	0.173 * (0.638)	0.128 ** (0.027)	0.216 (0.310)	1
CSR_EXPERT	0.046 * (0.042)	0.049 (0.123)	−0.131 ** (0.145)	−0.014 (0.029 *)	0.018 * (0.377)	0.076 (0.101)	1
SECTOR	−0.031 ** (−0.094 *)	0.023 (0.117)	0.014 (0.133)	−0.085 (0.210)	0.014 (0.019)	−0.122 (0.216 **)	0.021 (0.319)	1

*** p < 0.01, ** p < 0.05, * p < 0.1. Values between brackets () are for the European sample.

reports the correlation analysis. It can be seen that, regarding the Anglo-Saxon sample, the CSR disclosure score is positively associated with age diversity (r = 0.012), gender diversity (r = 0.153), the presence of a CSR committee (r = 0.013), and the presence of CSR experts on the CSR committee (r = 0.046). However, CSR disclosure is negatively correlated with CEO duality (r = −0.103). CSR disclosure is also associated with SECTOR which is negatively and significantly correlated with CSR disclosure. Similar results were found in the European context, except for the presence of CSR experts on the CSR committee which was found not significantly correlated to CSR disclosure. The highest explanatory variables’ correlation is 0.638 between board age and CSR committee existence.

5.4. Regression Analysis

The current study uses structured panel data, which combines the temporal and individual dimensions. Due to the nature of the data, biases and issues related to the analysis need to be studied. For panel data, there are three forms of regression: fixed-effects (FE) model, pooled OLS model, and random-effects (RE) model. The Breusch-Pagan and Durbin-Watson statistics were employed to determine the better suitable model. The Breusch-Pagan statistic tests the null hypothesis of homoskedasticity while the Durbin-Watson statistic tests the null hypothesis for autocorrelation. The existence of autocorrelation or/and homoskedasticity implies that FE or RE are more suitable. The results of these two statistics for both samples are provided in

Table 5. Breusch-Pagan and Durbin-Watson tests results.

MODELS	TESTS	RESULTS
Anglo-Saxon	Durbin Watson	U	2.1058
		p-value	0.013
	Breusch Pagan	Chi2	19.12
		Prob>	0.502
		Chi2
European	Durbin Watson	U	2.0872
		p-value	0.025
	Breusch Pagan	Chi2	14.31
		Prob>	0.075
		Chi2

.

Conducting the Durbin Watson and Breusch-Pagan tests shows that both models are subject to autocorrelation (p-value < 0.05) and heteroskedasticity (p-value > 0.05). Based on this result, the null hypothesis is rejected. Thus, individual effects model (FE or RE) is more suitable than pooled OLS. FE and RE models both control the unobserved characteristics that may influence CSR disclosure. Particularly, these estimators capture unobserved heterogeneity by adding specific error terms, which can be fixed over time (FE), or randomly vary over time (RE) for each firm (Baltagi 2001).

Following Torres-Reyna (2007), the research uses Hausman’s test, which allows discriminating between the fixed effects (Within estimate) and random effects (GLS estimate). The results of the Hausman test on both models are presented in

Table 6. Results of Hausman test.

Models	Chi-Square Statistic	Prob > Chi2	Result
Anglo-Saxon	12.0297	0.6732	Random effects
European	11.6215	0.0087	Fixed Effects

.

The results of the Hausman test show that the p-value (0.6732) is greater than 0.05 for the Anglo-Saxon model, hence the use of the RE model. However, for the European model, the p-value (0.0087) is below 0.05, the null hypothesis is then rejected. Accordingly, the FE model is the most suitable. Due to the presence of heteroskedasticity and autocorrelation, it has been, therefore, decided to use cluster-robust variance and covariance estimators (Wooldridge 2002) for the European model and generalised least squares (GLS) with a robust standard error method for the Anglo-Saxon context model. The robust regression has the power to overcome the problems of heteroskedasticity and autocorrelation in the data (Cameron et al. 2009).

5.5. Empirical Results

5.5.1. Anglo-Saxon Model Results

The Results of the GLS panel regression with robust standard error for the RE model, provided in

Table 7. Results of random effects regression analyses (GLS with robust standard error).

Variables	Hypothesis	Expected Sign	Coeff.	p-Value
Intercept			0.7129 *** (0.2962)	0.0001
BD_GENDER	H1	+	0.0921 ** (0.0024)	0.0213
BD_AGE	H2	+	−0.0433 * (0.0461)	0.0681
CEO_DUAL	H3	-	−0.3115 *** (0.1084)	0.0019
CEO_COMP	H4	+	0.3117 *** (0.0741)	0.0001
CSR_COM	H5	+	0.1356 ** (0.0287)	0.0395
CSR_EXPERT	H6	+	0.5237 * (0.0481)	0.0825
SECTOR		+	0.1498 ** (0.0455)	0.0307
R2		0.3564
Adjusted R2		0.3231
F statistic		10.24 ***

Figures between brackets () are standard errors robust to heteroskedasticity. * p < 0.10, ** p < 0.05, and *** p < 0.01.

, show that gender diversity (BD_GENDER) is positively and significantly associated with the extent of CSR disclosure in annual reports of companies of the Anglo-Saxon context. BD_GENDER is found positive and significant at the 5% level (coeff. = 0.0921, p = 0.0213) which confirms hypothesis H1. The results indicate that the existence of women on board plays a positive role in the extent of CSR disclosure. This result is consistent with Nour et al. (2020), Kiliç et al. (2015), and Barako and Brown (2008). Therefore, boards with more female directors of Anglo-Saxon firms may request extensive and complementary CSR information to be provided in firms’ annual reports along with the financial information. Therefore, they may induce managers to release more CSR information in the company’s annual report. This result is in line with agency theory, which suggests that boards of directors have ultimate responsibility for ensuring the firm’s financial reporting system’s reliability, integrity, and transparency (Jensen 1993). Thus, the inclusion of women improves discussions within the board and avoids the centralization of decisions in the hands of the dominant group.

Regarding board age diversity (BD_AGE), the results show a significant negative association between this variable and CSR disclosure at the 10% level (coeff. = −0.0433, p = 0.0681). This result rejects hypothesis H2. The higher the degree of age diversity on the board, the more likely it is that board members will restrain their decisions regarding CSR disclosure. This result may be because of the existence of different generations on the board. Such circumstances provoke distinct opinions and points of view between board members which leads to conflicting decisions in CSR activities and related disclosure.

The duality of the CEO (CEO_DUAL) is found to be significantly and negatively correlated with CSR disclosure (Coeff. = −0.3115, p = 0.0019). This result confirms H3 for the Anglo-Saxon context. In this regard, the separation of functions can be a favourable measure because it allows a real balance of powers to be established and thus promotes an objective assessment by the board of the quality of CSR information disclosed.

CEO compensation (CEO_COMP) is positive and significantly correlated to CSR disclosure. H4 is then accepted. Thus, a well-compensated CEO may play a decisive role in disclosing more CSR information. Moreover, this result shows that shareholders can encourage CEOs to disclose CSR information related to constructive CSR activities by changing their compensation structure. Similar results were found by Al-Shaer and Zaman (2019).

In support of H5, results show that firms with a CSR committee are more likely to disclose their CSR information. In fact, CSR_COM is found positive and significant at the level of 5%. Thus, H5 is confirmed. Therefore, the existence of a CSR committee would amplify the importance given to these particular aspects of the governance system. Accordingly, the presence of a CSR committee within Anglo-Saxon companies may encourage companies to build environmental credibility and legitimacy (Cucari et al. 2018). Similar results had been confirmed in the literature by Helfaya and Moussa (2017) and Arena et al. (2015).

Regarding the expertise of the CSR committee members (CSR_EXPERT) the results show a significant relationship with CSR disclosure. Thus, H6 is confirmed. CSR_EXPERT is positive and significant at the level of 10% for the Anglo-Saxon sample (Coeff. = 0.5237, Sig. = 0.0825). Greater expertise and the presence of CSR experts on the CSR committee increase the likelihood of CSR disclosure, especially when that expertise dominates participation. These results reveal that not only the existence of a committee in charge of CSR matters increases CSR disclosure but also the expertise of the members in charge.

Similarly, with positive coefficients and very significant p-values, the type of industry (SECTOR) of Anglo-Saxon firms positively and significantly determines the level of CSR disclosure (Coeff. = 0.1498, p = 0.0307). The industry is commonly proposed as a significant driver of CSR practices (e.g., Patten 1991; Roberts 1992; Raffournier 1995; Cormier and Magnan 2003; Haniffa and Cooke 2002, 2005). The RE model is also significant at the 1% level with good explanatory power (32.31%) Therefore, the results can be considered trustful in explaining the association between each independent variable and the dependent variable.

5.5.2. European Model Results

The robust fixed effects model is given in

Table 8. Results of fixed effects regression analyses.

Variables	Hypothesis	Expected Sign	Coeff.	p-Value
Intercept			1.0715 (1.67)	0.0421
BD_GENDER	H1	+	−0.7826 * (−2.08)	0.0973
BD_AGE	H2	+	0.1837 * (2.46)	0.0721
CEO_DUAL	H3	-	−0.4172 * (−0.96)	0.0599
CEO_COMP	H4	+	0.2493 ** (0.18)	0.0491
CSR_COM	H5	+	0.1405 *** (0.53)	0.0007
CSR_EXPERT	H6	+	0.2129 ** (0.97)	0.0325
SECTOR		+	0.9627 *** (1.04)	0.0031
R2		0.2973
Adjusted R2		0.2591
F statistic		12.620 ***

Figures between brackets () are robust t-statistics. * p < 0.10, ** p < 0.05, and *** p < 0.01.

. Contrary to the expectations, the model tends to show that gender diversity (BD_GENDER) has a negative and significant influence on CSR disclosure at the level of 10%. This result rejects H1. These results are in contrast with those of the Anglo-Saxon context. This means that the more women there are on the board of directors, the more CSR disclosure decreases in the annual report. This result is in line with Handajani et al. (2014) and Cucari et al. (2018). The authors expected that a high percentage of women on the board would improve the extent of CSR disclosure. However, this study validates that the presence of women does not necessarily support and improve CSR disclosure. The differences in the results compared to the Anglo-Saxon context may be due to the culture. Despite that the presence of women on the board has increased, it seems that their power to incite CSR disclosure is not as strong as it is on the Anglo-Saxon boards. This may be also due to the lack of knowledge that female directors have regarding CSR.

Table 8 also indicates that BD_AGE is positively and significantly correlated (Coeff. = 0.1837, Sig. = 0.0721) to the extent of CSR disclosure of European companies. This implies that boards with different age categories among members increase the likelihood of CSR disclosure. On the first hand, younger directors play a positive role in enhancing the extent of CSR information disclosed thanks to their fresh blood and awareness regarding CSR issues such as protecting the environment, global warming, and human rights. They also are noticeably mobile and are familiar with prompt change. On the other hand, older board members through their cumulative experience can significantly impact corporate disclosure. They exhibited an eminent concern for security and a yearning to avoid the risks and disasters witnessed during their early years (Ferrero-Ferrero et al. 2015). This result illustrates that older and younger directors on boards are aligned to promote the disclosure of CSR information. It is believed that the difference in the results between the two models is due to governance and cultural differences between contexts.

As predicted, the combination of the chairman of the board and CEO positions (CEO_DUAL) is negatively and significantly associated with CSR disclosure. H3 is then supported for the European context. Such a result captures a negative impact on CSR information transparency when the Chairman and CEO’s responsibilities are gathered in one person. Previous evidence for several countries reveals similar impacts. For instance, Ben Rhouma and Cormier (2007) found that the disclosure of CSR information by French companies is negatively correlated to the duality of positions. Clemente and Labat (2009) found similar results in the Spanish context.

Correspondingly to the Anglo-Saxon model results, CEO_COMP is found to be significantly and positively correlated to CSR disclosure in the European sample (Coeff. = 0.2493, Sig. = 0.0491). H4 is then supported. Thus, a well-paid CEO may play a crucial role in disclosing CSR information. This finding could imply that managers participate in CSR and disclose related information for the most part for their advantage.

According to the results, the existence of an SRC within the board’s committees (CSR_COM) improves the extent of CSR disclosure in European companies. CSR_COM is found positive and significant at the level of 1%. Thus, H5 is confirmed. Accordingly, companies that establish a CSR committee provide accountability proof and support boards’ interest in satisfying the needs of their stakeholders regarding the socially responsible behaviour of firms (Dias et al. 2017).

The expertise of SRC members is also found to have a positive effect on the extent of CSR disclosure with a positive coefficient and significant p-value (Coeff. = 0.2129, Sig. = 0.0325). These results confirm H6. Hence, the expertise of the CSR committee members is an important CG that significantly impacts the extent of CSR information disclosed in the annual reports.

Regarding the control variable, similar results to the Anglo-Saxon sample were found. In fact, INDUST is positively and significantly correlated to CSR disclosure at the level of 1%. The results show that the reputation of the industry type is among the most important factors influencing CSR disclosure practice in both contexts.

6. Discussion

CG represents a set of mechanisms that helps alleviate agency problems between managers and shareholders. The focal purpose of this study is to empirically examine and compare the relationship between firms’ CG practices and the extent of CSR disclosures in the annual reports in the European and Anglo-Saxon contexts. According to the empirical results, board gender and board age were both found significantly correlated to CSR disclosure. Thus, they affect the strategic choices of companies regarding CSR disclosure. From a managerial point of view, the results highlight the need to better consider diversity in the policy for choosing board directors. In Anglo-Saxon firms, the appointment of women on the board gives them more weight in the decision-making process, including those related to improving the quality of CSR information. However, in the European context, the results are quite the opposite. The inclusion of women on the board decreases the extent of CSR disclosure. This finding may be the outcome of social and/or regulatory pressure. In fact, the appointment of women may be motivated by a desire to improve the company’s image or to comply with the legislation rather than the quest for better board efficiency. This calls into question the legitimacy of encouraging women to act towards improving the quality of CSR disclosure. Moreover, diverging perspectives may appear within the group of female directors, which may limit the quality of disclosure. These disagreements can be linked to the desire of certain women to conform to the choices of the male directors to maintain certain functions or gain access to other more important ones within the company.

The board’s age diversity is also considered one of the most significant observable indicators of the board of directors (Kang et al. 2007). It has been demonstrated that the coexistence of different generations, and consequently of values, motivational goals, habits, experiences, and various cultural norms that affect the directors’ intuitive decision-making style, has varies effects on CSR disclosure across contexts (Cucari et al. 2018). This attribute increases the extent of CSR disclosure on Anglo-Saxon firms. It acts the opposite for European companies. In a context of strong divide between generations (Anglo-Saxon), the presence of younger directors on the board of directors can promote a vision more oriented towards the long term, particularly in the fight against climate change and global warming, thus improving the quality of CSR disclosure. Such movement has even been endorsed by some governments such as Canada. In fact, in 2016, the state of Quebec passed a law that provides that from 2021, all companies controlled by the State must have at least one person under the age of 35 on their board. Thus, board age diversity within Anglo-Saxon companies is encouraged by regulations and has been proved effective for enhancing CSR disclosure.

As anticipated, the separation of the role of CEO and chairman of the board in both contexts is related to CSR disclosure. This finding is consistent with Xiao and Yuan (2007), Gul and Leung (2004), and Haniffa and Cooke (2002). A dual management structure has then a negative impact on the extent of CSR disclosure in both contexts. Corresponding to the Agency theory, almost certainly, managers’ personal advantages will influence their obligation to CSR practices and reporting. Accordingly, the power of CEO duality might urge the managerial crew to reconsider CSR practices. Consequently, the separation of roles can be considered an indicator of effective monitoring, enhanced control, and lower risk of subjective disclosure.

CEO compensation is one of the CGP subjects of the current study. The empirical analysis suggests that higher CEO compensation is rewarded with enhanced CSR disclosure. The analysis also suggests that companies are more likely to improve the degree of CSR disclosure if they examine the CEOs’ involvement in CSR initiatives and attach it to their compensation. These results are in line with the stipulations of stakeholder agency theory stating that the implementation of an incentive-based compensation system is considered an important instrument to bring down agency conflicts between managers and stakeholders (Hill and Jones 1992). Managers of responsible companies actually put a lot of effort into preventing potential conflicts of interest with other stakeholders, such as employees and the government. In exchange, they would demand greater compensation than managers of less responsible companies.

The existence of a CSR committee and CSR expert positively affect CSR disclosure in both contexts. As a result, companies with such committees and experience are more likely to provide higher CSR disclosure. The establishment of such a committee is seen as a method of meeting the expectations and interests of stakeholders (Michelon and Parbonetti 2012). Such a committee could also ensure the quality of the information disclosed. This is because the committee members will encourage managers to demonstrate accountability by ensuring that the firm is following commonly recommended CSR reporting guidelines (Peters and Romi 2011). According to Cucari et al. (2018), effective CSR committees are supposed to increment organizations ‘transparency by increased CSR disclosure. Subsequently, they show greater commitment and greater accountability to CSR practices.

7. Conclusions

This study analyses the association between CG mechanisms and corporate social responsibility disclosure in two different contexts. It is conducted on a sample of listed companies between 2006 and 2016. This work enlarges the extent of the empirical state of the art on the connection between, on the first hand, board diversity, leadership characteristics, and CSR expertise and, on the other hand, the extent of1CSR disclosure. Most study findings are in line with those of earlier research on CSR disclosure (Lu and Wang 2021; Malik et al. 2020; Fahad and Rahman 2020; Cucari et al. 2018; Gul and Leung 2004). Although the results slightly varied depending on the context, they generally demonstrate good CG practices. The importance of demonstrating a commitment to CSR projects and activities through the provision of CSR disclosure as a means of managing their relationships with stakeholders is actually being considered by companies in both contexts. CG practices have been shown to be an effective tool for enhancing CSR disclosure. This research contributes to CG and CSR disclosure literature by offering a comparative analysis between two different contexts: Anglo-Saxon and European. Although there are no significant differences in results between the contexts, the findings are considered an addition to the literature. In fact, such results implied that the contextual differences between the Anglo-Saxon and European CG models have little effect on the extent of CSR disclosure, possibly because both contexts are well known for their quality disclosure and adhere to CG best practices. Such findings can be used as a starting point for future research into the impact of contextual differences based on other criteria.

Although the two contexts share certain similarities in the impact of CG on CSR disclosure, they differ in CG practices. Dissimilarities might be due to the managerial cultural differences and the legal, as well as institutional differences that exist between the two contexts. The present study finds that CG practices in both contexts such as the separation of board chair and CEO positions, establishing a CSR committee, and having a gender- and age-diversified board are effective CGP to reach a higher level of CSR disclosure. Despite the significant results, this paper has some limitations. However, these limitations provide opportunities for additional research. Accordingly, one of the limitations encountered during this study is the outdated data. Particularly, due to major economic changes during the Covid pandemic and post-pandemic eras, the findings and the conclusions of the study can be questioned. Perhaps future comparative research on the impact of the Covid pandemic on CSR disclosure can be considered. Reflections on ethnic origin diversity, the impact of CEOs’ compensation structure, and the optimal number of female directors on the board on CSR disclosure, are also recommended for future research.