Analysis of The Effects of Compliance with Corporate Governance Recommendations

TESIS DOCTORAL ANALYSIS OF THE EFFECTS OF COMPLIANCE WITH CORPORATE GOVERNANCE RECOMMENDATIONS Doctorando: Elena Moreno Ureba Departamento de Contabilidad y Economía Financiera, Universidad de Sevilla Director: Prof. Dr. D. Francisco Bravo Urquiza Departamento de Contabilidad y Economía Financiera, Universidad de Sevilla Programa de doctorado: Programa de Doctorado en Gestión Estratégica y Negocios Internacionales Universidad de Sevilla Sevilla, 2021 A mis padres, Francisco José y Saly. GRATITUDE En primer lugar, me gustaría expresar mi gran y sincero agradecimiento a mi director, el profesor Dr. Francisco Bravo Urquiza, por todo el esfuerzo y dedicación que me ha mostrado. Sus orientaciones, conocimientos, motivación y apoyo personal y profesional, han sido fundamentales para el desarrollo de este proyecto y de mi formación como investigadora. Quiero dar especialmente las gracias a la profesora Dra. Nuria Reguera Alvarado, por sus aportaciones, por enseñarme tanto y tan bien, y por toda la ayuda que me ha brindado. No puedo olvidarme de la Dra. Joaquina Laffarga, por integrarme en su equipo y mostrarme su apoyo desde el principio. Mi reconocimiento al Dr. Bernabé Escobar, quién me dio la primera oportunidad como alumna de doctorado, en su querida “Cátedra de Responsabilidad Social” y que junto a la Dra. María del Mar Miras, siempre han tenido unas palabras de aliento y de ánimo para mí en este duro camino. Gracias a ambos por sus buenos consejos. Mi agradecimiento a University of Edinburgh Business School, por otorgarme la oportunidad de realizar mi estancia investigadora, y en concreto a la profesora Dra. Angélica González por sus orientaciones y supervisión. A mis compañeros del Departamento de Contabilidad y Economía Financiera de la Universidad de Sevilla, por la colaboración prestada y la ayuda que de un modo u otro me han ofrecido. Mi especial agradecimiento a mi familia, por estar siempre ahí en cada una de mis múltiples caídas a lo largo de este duro camino, gracias por escucharme, por soportar mis agobios, por motivarme y apoyarme por encima de todo, y en especial, gracias a mis padres, Francisco José y Saly, por los valores que me han transmitido, y por inculcarme que la constancia y el trabajo duro son necesarios para el desarrollo personal y profesional. Por último, gracias a mi compañero de viaje, Guille, por su comprensión, apoyo y amor. No habría conseguido llegar hasta aquí sin vuestra ayuda, a todos GRACIAS. 5 INDEX OF CONTENTS CHAPTER 1: INTRODUCTION ...................................................................... 13 1.1. Research motivation ............................................................................... 13 1.2. Research objectives ................................................................................. 14 1.3. Structure of the thesis ............................................................................. 21 CHAPTER 2: CONTEXTUAL FRAMEWORK ............................................. 25 2.1. Corporate Governance .............................................................................. 25 2.2. Board of directors ..................................................................................... 29 2.3. Board gender diversity .............................................................................. 38 CHAPTER 3: LEGAL CONTEXT ................................................................... 47 3.1. Corporate Governance Codes ................................................................... 47 3.2. Board of directors ..................................................................................... 53 3.3. Board gender diversity .............................................................................. 55 CHAPTER 4: THEORETICAL FRAMEWORK ............................................ 77 4.1. Agency theory ......................................................................................... 77 4.2. Resource dependence theory................................................................... 79 4.3. Stakeholder theory .................................................................................. 81 4.4. Upper echelon theory .............................................................................. 83 4.5. Institutional theory .................................................................................. 85 4.6. Signalling theory ..................................................................................... 86 4.7. Social identity theory .............................................................................. 88 4.8. Psychological theories ............................................................................ 89 4.9. Tokenism theory ..................................................................................... 90 4.10. Group phenomena ............................................................................... 90 6 CHAPTER 5: EMPIRICAL STUDY ................................................................ 94 5.1. Does compliance with corporate governance codes help to mitigate financial distress? ............................................................................................... 94 5.1.1. Objective .............................................................................................. 94 5.1.2. Hypothesis development ...................................................................... 95 5.1.3. Research method ................................................................................ 100 5.1.3.1. Sample and data ........................................................................... 100 5.1.3.2. Variables ...................................................................................... 101 5.1.3.3. Model specification ..................................................................... 104 5.1.4. Results and discussions ...................................................................... 106 5.1.4.1. Descriptive statistics and correlations ......................................... 106 5.1.4.2. Multivariate analysis ................................................................... 111 5.1.4.3. Discussion .................................................................................... 115 5.2. Corporate governance code compliance and environmental, social and governance disclosures .................................................................................... 117 5.2.1. Objective ............................................................................................ 117 5.2.2. Hypothesis development .................................................................... 117 5.2.3. Research method ................................................................................ 121 5.2.3.1. Sample and data ........................................................................... 121 5.2.3.2. Variables ...................................................................................... 122 5.2.3.3. Model specification ..................................................................... 124 5.2.4. Results and discussions ...................................................................... 125 5.2.4.1. Descriptive statistics and correlations ......................................... 125 5.2.4.2. Regression analysis ..................................................................... 129 5.2.4.3. Discussion .................................................................................... 132 5.3. A critical approach to the true influence of female directors on environmental innovation: when are women greener? .................................... 134 7 5.3.1. Research objective.............................................................................. 134 5.3.2. Hypothesis development .................................................................... 135 5.3.3. Research method ................................................................................ 142 5.3.3.1. Sample and data ........................................................................... 142 5.3.3.2 Variables ....................................................................................... 143 5.3.3.3. Model specification ..................................................................... 145 5.3.4. Results and discussion ........................................................................ 146 5.3.4.1. Descriptive statistics and correlations ......................................... 146 5.3.4.2. Multivariate analysis ................................................................... 150 5.3.4.3. Discussion .................................................................................... 158 CHAPTER 6: FINAL CONSIDERATIONS .................................................. 163 6.1. Key findings and main contributions of the thesis ................................... 163 6.2. Implications .............................................................................................. 165 6.3. Limitations and future lines of research ................................................... 167 References ....................................................................................................... 169 Appendices ....................................................................................................... 215 Appendix A. List of Spanish CGC recommendations (Good Governance Code of Listed Companies, 2015) ............................................................................. 215 Appendix B. List of publications ..................................................................... 219 14 On an academic level, attention has focused on studying: a) ownership structure, b) the functioning of the board of directors― board size, independence, CEO duality and meetings have been analysed―, c) directors’ characteristics, such as age, diversity, tenure, and, d) CGC compliance. The vast majority of studies examine the effects of corporate governance studies on certain business strategies. One branch of studies has shown that corporate governance practices favour firm performance (Vo and Nguyen, 2014; Bhatt and Bhatt, 2017; Danoshana and Ravivathani, 2019), with previous studies having highlighted the effects of corporate governance techniques on company disclosure (Husted and de Sousa-Filho, 2019; Shan, 2019). In addition, recent research has emphasized that corporate governance, seen as the correct way to govern companies, can help to promote sustainable initiatives (Hussain et al., 2018). In particular, within corporate governance regulations, promoting board diversity has acquired a significant role in most countries. In this regard, current studies have paid increasing attention to exploring what advantages gender diversity has for boards of directors (Low et al., 2015; Kılıç and Kuzey, 2016; Li and Chen, 2018; Brahma et al., 2021). Nevertheless, the evidence remains open and more accurate and generalizable findings are required to understand the effect of compliance with corporate governance recommendations, given the direct implications this has on financial markets for governments, regulators, and firms. 1.2. Research objectives The objective of this thesis is to improve our knowledge of the relationship between corporate governance mechanisms and certain corporate outcomes. Specifically, given their relevance in financial markets, this research aims to examine the impact of corporate governance on: a) financial distress, b) environmental, social and governance (ESG) disclosure, and c) environmental innovation. Therefore, the main objective is broken down into three sub-objectives: 1. Does compliance with corporate governance codes help to mitigate financial distress? 15 2. Corporate governance code compliance and environmental, social and governance disclosures. 3. A critical approach to the true influence of female directors on environmental innovation: when are women greener? Accordingly, we analyse how corporate governance may impact three relevant and totally differentiated areas at the business level: business financial performance, disclosure of non-financial information, and environmental sustainability practices. Specifically, these objectives have led to independent empirical studies, which are summarized below and explained in detail in section 5: 1. Does compliance with corporate governance codes help to mitigate financial distress? This study aims to analyse whether compliance with CGC helps to mitigate firms’ financial distress. Three different levels of compliance are examined: overall compliance, compliance with recommendations regarding the board of directors, and compliance with recommendations on board subcommittees. This topic is timely and relevant due to the ongoing debates surrounding both the effects of CGC compliance and the determinants of financial distress. First, the adoption and repercussion of these codes has become a societal concern. Consequently, scholars, public opinion, and politicians have pushed legislators and professional bodies to reinforce governance codes in order to increase internal control and accountability (Cuomo et al., 2016). These CGC are based on recommendations about governance mechanisms and have gained increased visibility and importance in capital markets for investors and regulators (Cicon et al., 2012). Although compliance levels are generally high, most firms fail to fully comply with all the codes` recommendations since the effects of CGC compliance remain largely unexplored (Kabbach de Castro et al., 2017). Previous research has investigated whether there is an economic rationale behind compliance with CGC recommendations or whether this is merely an ethical issue that is pursued for social legitimacy reasons. In this regard, some studies suggest that CGC compliance leads to positive market reactions by increasing market value (Goncharov et al., 2006; Kaspereit et al., 2017) and corporate reputation (Hooghiemstra and van Ees, 2011). 16 Another branch of research argues that CGC compliance may positively impact major corporate strategic decisions and help to improve firm performance (Bassen et al., 2006; Stiglbauer, 2010). However, research into the effectiveness of CGC is scarce, and there are still major gaps in the literature, which calls for new research on the potential benefits of CGC compliance. In this sense, recent inquiry has highlighted the need to provide a more careful examination of code content and to further explore the relationship between code compliance and other firm financial outcomes (Cuomo et al., 2016). This research extends previous literature by filling these two research gaps. First, this study makes an in-depth analysis of CGC, since it focuses on three different levels of compliance. To that end, it considers overall CGC compliance, as do previous studies, whilst also taking into account compliance with recommendations about the board of directors and about board subcommittees, which is one of the novelties of this study. Second, the study focuses on an important firm outcome in the literature addressing business and finance; namely, the financial distress of companies, thereby providing a better insight into the effects of CGC compliance. Second, the debate surrounding the determinants of financial distress has been sparking interest in recent decades, and has been accentuated particularly after the last financial crisis, given the major consequences it has for all of a firm’s stakeholders (Mselmi et al., 2017; Boubaker et al., 2018). Previous research has highlighted the complexity of predicting firms’ financial distress situations for agency credit ratings, governments or financial creditors and has emphasized the role played by corporate governance mechanisms in preventing business failure (Manzaneque et al., 2016a). One implicit premise involved in explaining this potential association is that corporate governance is expected to have important implications for corporate decisions, especially when the business runs a high risk of failure (Dowell et al., 2011). In this sense, the nature of the link between corporate governance mechanisms and the likelihood of financial distress has been discussed among researchers and policymakers alike, and the role of corporate governance in mitigating financial distress remains a core issue today, particularly in light of the financial crisis and the financial scandals involving major companies around the world (Manzaneque et al., 2016b; Udin et al., 2017). However, the literature addressing the association between corporate governance and the 17 likelihood of firms’ financial distress is limited and has thus far failed to explore the impact of CGC compliance. In theory, the ultimate objective of CGC is to recommend best governance practices so as to mitigate agency conflicts, and thereby protect shareholder interests and ensure business prosperity. 2. Corporate governance code compliance and environmental, social and governance disclosures. This study aims to explore whether compliance with the recommendations contained in CGC contributes to environmental, social and corporate governance disclosure practices. This research is timely and relevant for different motivations that are discussed below, and fills a significant gap in the previous literature by considering active academic, professional, and regulatory debates on the role of CGC and the need to increase ESG information. Corporate governance is a key issue vis-à-vis understanding how companies work and it has become especially important for companies, regulators, and academics as well as for society as a whole. For this reason, legislators and regulators at both a national and international scale have expressed the need to manage corporate governance mechanisms properly. Notably, most developed countries have published numerous CGC. The emergence of these codes has led to intense debate at the academic, professional, and regulatory levels concerning their effectiveness. The relevance of these codes lies in listed companies’ general conviction regarding the importance of being managed properly and transparently as a fundamental element for creating value in organizations, increasing economic efficiency and boosting investor confidence (CNMV, 2015). However, what impact CGC compliance might have on companies remains an open question in the literature, and there is intense debate as to whether compliance with CGC is an economic or an ethical issue (Hooghiemstra and van Ees, 2011). CGC focus on several recommendations concerning corporate governance mechanisms and have gained significant visibility and importance in capital markets (Cicon et al., 2012). Although CGC compliance levels are generally high, as mentioned before, most companies do not follow all the recommendations, possibly because the effects of CGC compliance remain mostly unexplored, with further research being needed (Kabbach de Castro et al., 2017). Earlier research focused on whether 18 compliance with CGC had any effect on business performance and the company’s market valuation (Fernández-Rodríguez et al., 2004; Stiglbauer and Velte 2014; Kaspereit et al., 2017). Nevertheless, the results have failed to prove conclusive, and recent research has stressed the need to make further progress in studying what consequences CGC compliance might have for a company. Cuomo et al. (2016) highlight the existence of two important gaps in current research. The first relates to the need for a more detailed analysis of compliance with these codes. The second suggests the importance of studying the effect of compliance with the recommendations contained in the CGC on other types of issues. Therefore, this research aims to contribute to previous literature and to cover both matters. To this end, this study performs an in-depth analysis of CGC compliance, focusing not only on compliance at the global level, but also on compliance with recommendations specifically related to the board of directors, and with those associated with the different board committees. Accordingly, three different levels of CGC compliance are considered. Furthermore, other contributions of this study are an analysis of the relationship between the different levels of CGC compliance and the disclosure practices of ESG information. Information disclosure is a crucial aspect within corporate strategies, and several studies highlight that improved corporate governance mechanisms should lead to better oversight of companies’ disclosure processes and to a reduction in information asymmetries in capital markets (Cerbioni and Parbonetti, 2007). ESG information has acquired particular relevance in recent years for investors, entities, regulators, creditors, as well as all types of company stakeholders (Helfaya and Moussa, 2017; Dias et al., 2018; Yu et al., 2018). In particular, ESG disclosure has become a critical aspect of business strategy (Lokuwaduge and Heenetigala, 2017). One of the main reasons why companies disclose ESG information is because they are highly conditioned by existing regulations, although the absence of a mandatory and common framework leads to diversity in the reports presented, thereby complicating any comparisons (Lokuwaduge and Heenetigala, 2017). However, progress has been made in the area of regulation. European Union Directive 2014/95/EU on the disclosure of non-financial information is evidence of this. It establishes the obligation for large companies which are entities of public interest, 19 and which at the end of the financial year have over 500 employees on average, to include in the management report a non-financial statement containing information. This information must help to understand the company’s evolution, results, and situation, together with the impact of its activity, relative at least to environmental and social issues, as well as to employees, respect for human rights and the fight against corruption and bribery (European Union Directive 2014/95/EU). In this sense, there is an active debate regarding how relevant the ESG reports disclosed by companies might be to the information that interests the different economic agents in capital markets. In particular, recent studies have revealed that most investors consider that the disclosure of this type of information should be improved due to its relevance for companies and investors (PriceWaterhouseCoopers, 2016). Indeed, many studies emphasize that ESG information is vital in terms of enhancing investment decisions (Ioannou and Serafeim, 2017). In this regard, previous literature considers that the disclosure of ESG information entails significant benefits in capital markets, as it allows agency conflicts and information asymmetries to be reduced (Cheng et al., 2014). Consequently, studying the determinants of voluntary disclosure practices of this type of information is crucial for companies, regulators, and academics, and remains an issue which demands further inquiry (Galbreath, 2018). In theory, companies with better corporate governance practices are likely to supervise the information preparation process more effectively in an effort to minimize agency problems which, to this end, would also tend to improve social disclosure. Nevertheless, there is no conclusive evidence on the link between CGC compliance and ESG disclosure. 3. A critical approach to the true influence of female directors on environmental innovation: when are women greener? The objective of this study is to examine the influence of board gender diversity on environmental innovation. This topic is timely and relevant for several reasons which are commented below. It contributes to the literature related to: a) environmental innovation, b) research on gender diversity, and c) methodological approaches on gender diversity studies. 20 Firstly, recent governance failures as well as social and environmental excesses have put ever-increasing pressure on boards of directors as drivers of sustainable business strategies (Jain and Jamali, 2016; Arayssi, et al., 2020). In the current context, environmental innovation has become crucial for minimizing firms’ environmental problems and for addressing consumer and government concerns 1 as well as increasing economic outcomes through the efficient use of resources (Horbach and Jacob, 2018). Therefore, given both the economic and environmental benefits, it is essential to know how boards of directors, as the top decision-making authority in firms, can shape environmental innovation policies. Analysing the impact of board gender diversity on environmental innovation may prove particularly relevant, since female representation on boards has been at the core of political, professional and academic discussions worldwide (Cucari et al., 2018; Amorelli and García‐Sánchez, 2020). However, in terms of research on gender diversity, empirical evidence on this issue is incipient and scarce and although some studies suggest that female directors may influence environmental innovation (Galia et al., 2015; Arena et al., 2018; Nadeem et al., 2020) the existing findings are still far from complete. Specifically, the one-size-fits-all approach may be inappropriate to understand the true influence of female directors, which may be dependent on certain factors, and more research is needed to achieve more definite and generalizable findings concerning the conditions in which women in the boardroom impact environmental innovation. In this regard, recent research has emphasised the importance of exploring new methodological approaches and of taking into account the organizational context in order to ascertain the actual effect of female directors (Kirsch, 2018; Bravo and Reguera-Alvarado, 2019; Bolourian et al., 2020). Accordingly, this study extends previous research by performing an in-depth analysis of the relation between board gender diversity and environmental 1 International bodies and public representatives have strongly positioned themselves in favour of environmental innovation policies as a way to protect the environment and create a climate of social welfare (OECD (2009); European Commission (2011); United Nations Climate Change (2015); UKRI (2019)). 21 innovation in an effort to unravel what really shapes board female influence on environmental innovation. 1.3. Structure of the thesis This work is composed of several parts in order to contextualize the situation concerning corporate governance and board of directors, define the theoretical framework, and develop the empirical analysis. Specifically, this thesis is structured as follows: - Chapter 2 provides a contextual framework, focusing on the relevance of corporate governance, highlighting the importance of the board of directors as an internal corporate governance mechanism. A study of board of directors’ characteristics is provided, highlighting the significance of board gender diversity. - Chapter 3 describes the legal context and examines CGC. This chapter contains a review of Spanish CGC that have been in force recently up to the 2015 Good Governance Code of Listed Companies, compliance with which has been evaluated. - Chapter 4 discusses the theoretical framework of corporate governance, board of directors and board gender diversity. - Chapter 5 deals with the empirical study. This section is divided into three different subsections which respond to the sub-objectives of this thesis, each of which includes their own objective, hypothesis development, research method and results as well as discussion and conclusions. - The concluding chapter (Chapter 6) reports the final considerations as well as implications, limitations, and future research directions. - Finally, the references are provided. 22 23 CHAPTER 2 Contextual framework 30 boards are assumed to bring useful resources to firms, such as advice, expertise, connections to environmental contingencies, and legitimacy (Hillman and Dalziel, 2003). As a result, boards play a role in terms of strategy implementation, control, and resource provision (Madhani, 2017). Specifically, boards must be responsible for monitoring decisions in order to pinpoint and prevent financial instability (Chang, 2009; Manzaneque et al., 2016a) as well as implement the measures required to help overcome potential failure (Fich and Slezak, 2008). Due to its growing relevance in the business sphere, since the 1990s the board of directors has been a subject of debate at the academic level, and has become a topic that has spawned many scientific publications. A search in “Google Scholar” (July 2021) shows approximately 1,890,000 scientific articles dealing with the board of directors. Figure 1 represents the number of publications by time periods. As can be seen, academic activity related to the board of directors continues to grow. Figure 1. Academic response to the board of directors Source: own elaboration from the information available at Google Scholar (July 2021) Based on the fact that board of director literature is extensive, in this thesis a thorough review of the most widely researched areas of analysis in this discipline 45,500 127,500 258,500 433,500 752,500 956,500 Number of publications 31 is carried out 2 . In particular, given its link to the aims of this project, the review focuses on studies that explore board of director impact. Although inquiry into the board of directors is multidisciplinary in nature, several branches of research have arisen and become significant, some of which are summarized below. a) The board of director/financial performance linkage is the most widely explored relationship with regard to the board of directors. Some of the measures used to quantify business financial performance include Tobin’s q, market capitalization, growth opportunities, ROA and ROE. b) Board of directors and transparency/information disclosure: the composition and characteristics of the board of directors is crucial vis-à-vis corporate transparency (Torchia and Calabrò, 2016). In addition, it has an impact on both non-financial information (Cheng and Courtenay, 2006; Lim et al., 2007; Donnelly and Mulcahy, 2008) and financial disclosure (Torchia and Calabrò, 2016). In order to measure transparency, researchers commonly draw on self-constructed indices, analyst forecasts, or scores provided by analyst ratings such as those shown in the “Informe Reporta” in the Spanish setting, among others. c) Board of directors and corporate social responsibility (CSR hereinafter). Previous studies have highlighted that board composition impacts CSR practices (Fernández-Gago et al., 2016; Sundarasen et al., 2016; Godos-Díez et al., 2018). CSR activity is a broad concept that encompasses the measurement of both disclosure practices and performance. As regards CSR disclosure, content analysis of specific concerns, such as ESG, analyst assessments, Global Reporting Initiative (GRI) information measures, or carbon disclosure project (CDP) reporting are several examples of the most commonly used variables. In relation to CSR performance, the literature often draws on widely used measures provided by analysts and has included commercial databases, such as CSR scores, ESG scores, or other variables for CSR quality. 2 All references come from Google Scholar, which was consulted in July 2021. 32 d) Board of directors and corporate reputation: corporate reputation is another area of research interest linked to the board of directors. The composition and characteristics of board members have been widely analysed in order to predict corporate reputation (Brammer et al., 2009; Musteen et al., 2010; Bravo et al., 2015). When quantifying corporate reputation, previous research has tended to resort to existing databases and analyst ratings, among others. e) Board of directors and other strategies: previous studies argue that boardroom composition has an impact on other business outcomes as illustrated by innovation, internationalization, fiscal measures, dividend policies, fiscal policy, mergers, and acquisitions. In all these research areas, these types of studies have incorporated a number of different measures addressing the structure and functioning of the board as well as the characteristics of board members. As regards the main measures associated to board structure and functioning, the literature particularly highlights those based on: a) board size; b) board independence; c) board activity; d) the existence and structure of certain board subcommittees, among others. a) Board size has been traditionally measured as the number of board members (Cheng, 2008; Guest, 2008). Taking current recommendations as a reference, in some countries like Spain the desirable size could range between 5 and 15 directors. It is assumed that larger boards could lead to difficulties in decision making due to problems involving both communication as well as coordination (Jensen, 1993; Eisenberg et al., 1998), whereas a low number of directors may entail problems concerning a lack of resources or time, thereby supporting the argument of resource dependence theory (Coles et al., 2008; Adhikary et al., 2014). b) Board independence. As regards board member independence, the “Comisión Nacional del Mercado de Valores” (hereinafter, CNMV), which is the body in charge of supervising and inspecting the Spanish stock markets and the activity of those who intervene in them, proclaimed in 1998 the principal role of independent directors― directors who are not linked to the management team or to the control shareholding nucleus that most influence it (CNMV, 1998) ―. 33 In this sense, directors can belong to diverse categories (Ley de Sociedades de Capital, 2010): - executive directors― “those who perform management functions in the company or its group, whatever the legal relationship they have with it. Nevertheless, directors who are senior managers or directors of companies belonging to the group of the dominant entity of the company will have in this the consideration of proprietary directors”. - non-executive directors― “They are all the remaining directors of the company, and may be proprietary, independent or other external”. Two different categories of non-executive directors can be highlighted: • Proprietary directors― “Those who have a stake shareholding equal to or greater than that legally considered significant or that have been appointed due to their condition of shareholders, although their shareholder participation does not reach the cited amount, as well as those who represent shareholders of the above-mentioned”. • Independent directors― “directors who are selected because of their personal and professional situation and who can perform their functions without being conditioned by relationships with the company or its group, its significant shareholders or its managers”. In particular, board independence is usually measured as the proportion of independent directors among the total number of directors (Abdullah et al., 2011; Liu et al., 2015). c) Board activity. Board activity is generally quantified as the number of board meetings held in a year (Brick and Chidambaran, 2010; Chou et al., 2013). Board meetings allow directors to devote more time to analysing and addressing corporate strategy (Wincent et al., 2010). In this line, the discussion concerns whether or not meetings and board attendance are thought to be significant conduits through which directors gain firm specific information and are able to comply with their monitoring role (Ntim and Osei, 2011). In contrast, other researchers contend that board meetings are not guaranteed to prove useful because of the limited time available to directors and because more frequent board meetings may lead to less attendance (Lin 34 et al., 2014). In this sense, one branch of studies has found a negative association between board meetings and firm performance (Rodríguez- Fernández et al., 2014; Johl et al., 2015). d) Existence and structure of certain subcommittees. Audit committees (Aldamen et al., 2012; Bravo and Reguera-Alvarado, 2019) and CSR/sustainability committees (Shaukat and Trojanowski, 2016; García- Sánchez et al., 2019) have been widely studied in the literature. The audit committee is responsible for both financial and non-financial information since it should supervise both financial statements and associated documents as well as CSR disclosures (Al-Shaer and Zaman, 2018; Bravo and Reguera- Alvarado, 2019). The CSR/sustainability committee advises and encourages directors to implement the GRI-IFC strategy (García-Sánchez et al., 2019) and plays a significant role in helping managers to develop CSR strategy and examine CSR performance (Mackenzie, 2007). Apart from these characteristics based on the structure and functioning of the board of directors, previous studies have also largely employed measures regarding the personal attributes of board members. In particular, these attributes can be divided into several categories (Johnson et al., 2013); a) demographics; b) human; c) capital. a) Director demographics. Demographic characteristics include general information about groups of people which can be attributes and social characteristics. Previous research highlights that demographic characteristics shape behaviour that may influence cognition and decision making, thereby conditioning firm performance (Forbes and Milliken, 1999). Within demographic characteristics, age, educational background, gender, race and ethnicity have thus far been examined: - Age: this can be measured as how old directors are (Bilimoria and Piderit, 1994a; 1994b); as a group average (Post et al., 2011); and as a standard deviation of a group (Bohren and Strom, 2010). Johnson et al., (2013) highlight that the effect of directors’ age on firm outcomes remains controversial. On the one hand, older executives are less likely to start new directions and also tend to be more risk adverse (Platt and Platt, 2012). On 35 the other hand, they can contribute thanks to their valuable experience (Platt and Platt, 2012). Bonn et al., (2004) found a negative association between board age and firm financial performance. Nevertheless, John et al., (2020) showed that board age positively impacts firm value. - Educational background. This has been quantified as holding a degree (Singh, 2007) or belonging to a group that hold certain types of degree (Dalziel et al., 2011). Academic training conditions the decision-making capacity. John et al. (2020) highlight that educational background has a significant positive impact on firm value. Indeed, previous research has established that the educational qualifications of board members positively affect firm performance (Darmadi, 2013). In particular, the educational and functional background heterogeneity of directors is seen to enhance innovation (Sarto et al., 2019) and enrich corporate social performance (Harjoto et al., 2019). - Gender. Board gender diversity has been widely examined in the literature. Kılıç and Kuzey (2016) reported that women directors have a positive impact on financial performance. Likewise, there is empirical evidence that the inclusion of female directors promotes firm value (Carter et al., 2003; Nguyen and Faff, 2007). Furthermore, women are more sensitive and might be predisposed to take part in CSR and environmental decisions (Nielsen and Huse, 2010). Indeed, there is empirical evidence on the positive effect of female directors on CSR (Setó-Pamies, 2015; Galbreath, 2018). Given its importance, board gender diversity is discussed in a later section. - Race and ethnicity. Several measures relating to belonging to one or another category, or more generic measures of diversity, have commonly been used. Race and ethnic diversity have not been the subject of much research, although Ntim (2015) did find that ethnic diversity is a more appreciated quality than gender diversity in the South African stock market. In addition, Richard (2000) established that cultural diversity, measured as racial diversity, adds firm value and can lead to competitive advantage. 36 b) Human capital. Human capital refers to the abilities and experiences that directors can contribute to the decision-making process individually. It involves competencies acquired at the individual level, such as experience, and depends on directors’ international experience, knowledge of the company and environment, industry familiarity, CEO experience and financial know-how (Johnson et al., 2013; Volonté and Gantenbein, 2016), which could be summed up in: - Financial or industry experience: the literature assumes that directors’ human capital is expected to upgrade firm performance until it can maximize board effectiveness (Ployhart and Moliterno, 2011; Khanna et al., 2014). Volonté and Gantenbein (2016) found empirical evidence that directors’ human capital impacts firm performance, considering business strategy. - Experience in the company, which is also known as tenure. Long-serving directors can be expected to develop certain particular expertise on the company which may help them to perform their roles more efficiently. Schnake et al. (2005) evidence that longer-tenured boards improve directors’ capacity to monitor managers and so avert fraud, in addition to proving more advantageous in terms of controlling managerial opportunism when applying immoderate cash flow (Sharma, 2011). In contrast, Coles et al. (2015) highlighted that longer board service may cause directors to develop groupthink, which diminishes firm value in dynamic industries. Huang and Hilary (2018) pointed out that board tenure presents a U-shaped relation with firm value and accounting performance. Hence, additional research is required to explain these mixed results. c) Social capital. Social capital has been defined in many ways. Burt (1992) determined that social capital concerns the individual capacity to secure resources through relationships. Lin (2001) defined it as an “investment in social relations with expected returns”. Kilduff and Tsai (2003) refer to resources available to an individual which they can access through their links with others within a job network. Nevertheless, social capital can entail more costs than benefits in certain situations (Kim and Canella, 2008). For instance, 37 forming part of too many boards might lessen such benefits (Kor and Sundaramurthy, 2009) because of the need to attend to information which directors must compare as a result of the other boards they sit on (Khanna et al., 2014). Social capital theory established the need to distinguish types of knowledge between internal and external sources (Adler and Kwon, 2002). Indeed, Kim and Cannella (2008) divided social capital between interpersonal connections that involve others within the organization, and links that involve others who are outside the organization, arguing that both categorizes are based on distinct types of network links and provide the board with several resources. The former refers to internal social capital, whereas the latter applies to external social capital. - Internal social capital includes ties and connections with other people within the company, especially other directors on the board (Kim and Canella, 2008), which could be measured by surveys on board members. Previous research emphasizes that personal relationships between directors and the CEO might alter the advisory function of directors (Jones et al., 2008; Rhee and Lee, 2008). In particular, directors who develop bonds of friendship are more likely to reveal their concerns regarding company strategy (Westphal and Bednar, 2005). - External social capital refers to ties and links with outsiders: investors, customers, suppliers… (Kim and Canella, 2008). The number of connections that directors have with people outside the organization determines a director’s social capital, since more connections may imply fast access to data, ideas, and resources (Oh et al., 2006). Furthermore, outside directors sitting on several board leads to enhanced connections with other directors and executives (Beckman and Haunschild, 2002; Hillman and Dalziel, 2003). Indeed, outside directors are linked to the firm in which they provide services (Johnson et al., 2013). When the CEO of a focal company is also a board member of a director’s home firm this causes a double loop known as director interlock (Johnson et al., 2013), which allows information interchange and resources in both directions, and 38 which can both benefit or damage the firm. Moreover, across certain industry positions directors could enrich their relations with customers, suppliers, distributors (Certo, 2003), which could prove useful and valuable in obtaining legitimacy and in helping new businesses to emerge (Certo et al., 2001; Hillman and Dalziel, 2003). Additionally, director status, prestige and reputation may be key indicators to external stakeholders about the organization (Certo, 2003). Johnson et al. (2011) argued that directors who enjoy a high status tend to strive to preserve or improve their prestige. 2.3. Board gender diversity Given the special importance of board gender diversity from a legislative, professional and academic viewpoint, studying this attribute has become prominent in the literature concerning board composition, and particularly, the demographic characteristics of board members. In addition, board gender diversity plays an important role within the objectives of this project. Therefore, in this section, a specific review of board gender diversity and its importance in the professional and academic arena are provided, while the overview of the legislative perspective is provided in chapter 3. Although women account for half of the world’s population, they tend to experience more problems than men when it comes to accessing the job market. Indeed, the Organisation for Economic Co-operation and Development (OECD) calculates several employment indicators which reflect this reality and which can be seen in Figure 2. 39 Figure 2. Employment/population ratio by sex Source: OECD This inequality is extremely latent in the top management positions of companies and particularly in their boards of directors. Female directors have traditionally lacked representation in the boardroom, and although the trend is moving upwards there is still a long way to go. Figure 3 shows the current percentage of female directors on boards. Despite the fact that many countries are making a major effort to alleviate the situation, women are truly underrepresented on boards, reflecting that gender parity still poses a significant challenge. From a business point of view, there has been a strong response to regulatory measures on board gender diversity, with the vast majority of companies in different countries having increased the number of female directors. 0 10 20 30 40 50 60 70 80 1990 1995 2000 2005 2010 2015 2016 2017 2018 2019 % OECD - Average Men OECD - Average Women 46 47 CHAPTER 3: LEGAL CONTEXT 3.1. Corporate Governance Codes An increasing number of countries have joined the initiative to create CGC, which provide companies with an action guide on corporate governance that can offer diverse benefits (Acero and Alcalde, 2010). CGC include a large number of recommendations related to corporate mechanisms to help company top management to carry out their functions properly. Most countries have introduced CGC over the last few decades. Table 1 includes the main CGC from around the world. Table 1. International CGC Country Corporate governance code Australia ASX Corporate Governance Principles and Recommendations: 4th Edition Austria Austrian Code on Corporate Governance (2021) Belgium Belgian Code on Corporate Governance 2020 Brazil Código Brasileiro de Governança Corporativa – Companhias Abertas Canada Corporate Governance Guideline- 2018 Denmark Recommendations on Corporate Governance Finland Finish Corporate Governance Code 2020 France Corporate Governance Code of Listed Corporations 2020 Germany German Corporate Governance Code 2019 Hong Kong S.A.R., China. Review of Implementation of Corporate Governance Code 2018 (HKEX) Ireland Irish Stock Exchange Listing Rules applying UK Corporate Governance Code with Irish Annex Italy Italian Corporate Governance Code 2020 48 Japan Japan’s Corporate Governance Code Seeking Sustainable Corporate Growth and Increased Corporate Value over the Mid- to Long-Term 2018 Malaysia Malaysian Code on Corporate governance (2021) Netherlands Dutch Stewardship Code 2018 Norway The Norwegian Code of Practices for Corporate Governance 2018 Portugal Code of Corporate Governance 2018 (Revised in 2020) Russia Russian Code of Corporate Governance (2014) Singapore Code of Corporate Governance 2018 South Africa South Africa. King Report on Corporate Governance for South Africa - 2016 (King IV Report) Spain Good Governance Code of Listed Companies. Revised June 2020 Sweden The Swedish Corporate Governance Code (2016) Switzerland Swiss Code of Best Practice for Corporate Governance United Kingdom Co-operative Corporate Governance Code United States Corporate Governance Principle for US Listed Companies- 2017 Source: European Corporate Governance Institute (ECGI). In Spain, five CGC have been established since the late 1990s, known as the Olivencia Report (1998), Aldama Report (2003), Conthe Report, distinguished as the Unified Good Governance Code of Listed Companies (2006), and the Good Governance Code of Listed Companies (2015), which was revised in 2020. A brief summary of the main aspects of each of these codes is included below: 1. The Olivencia Report The Olivencia Report (1998) was an ethical code aimed at listed companies and whose application was voluntary, such that companies could either choose to 49 comply with it, or explain why they did not follow its recommendations. Companies were requested to inform the CNMV about fulfilment of the recommendations, although there was no obligation to do so. The main functions of the board of directors were management supervision and control. In this first Spanish Code, two board committees were specifically detailed; the audit committee and the appointments and remuneration committee, with the code being made up of 23 recommendations. According to the subsequent monitoring carried out by the CNMV in the three years immediately following publication of the report (Acero and Alcalde, 2010), the measures were not widely applied. 2. The Aldama Report A few years later, in accordance with the 2002 requirement of the Ministry of Economy, the “Comisión Especial de Promoción de la Transparencia y Seguridad en Mercados y Sociedades Cotizadas”, also known as the Aldama Commission (named after its president) was set up. The Aldama Report (2003) was implemented and followed the general lines of its predecessor, the Olivencia Report, although it also introduced other issues such as ensuring the transparency and security of financial markets, dealing with loyalty and diligence duties and the problems caused in this regard, board of director performance, general shareholder meetings, as well as listed companies’ links with other advisory and commercial companies (Banegas et al., 2006). Compliance with the recommendations included in the Aldama Report was voluntary. However, the “comply or explain” principle prevailed. The most relevant contribution of this report was to advise companies to publish an annual corporate governance report explaining the practices that corporations were carrying out in this regard. This recommendation is included in article 116 of “Ley 26/2003, por la que se modifica la Ley del Mercado de Valores y la Ley de Sociedades Anónimas, con el fin de reforzar la transparencia de las sociedades anónimas cotizadas”, which establishes that the aforementioned report should be made available to the CNMV as well as being published as a relevant event. The content and structure of the annual corporate governance report is established by the Ministry of Economy or by the CNMV, which means that this report is standardised, thus enabling comparisons to be made between reports from 50 different companies and helping stakeholders to understand it (Acero and Alcalde, 2010). 3. The Conthe Code Three years later, the government created a special working group to guide the CNMV in harmonizing and updating the recommendations of the Olivencia and Aldama Reports on the good governance of listed companies. In 2006, the aforementioned group approved the Draft Unified Code of Recommendations on Good Governance of Listed Companies. It includes both a series of recommendations aimed at listed companies and a draft of complementary recommendations aimed at Spanish financial institutions, the CNMV, and the government, and totalled 74 recommendations. The basic principles by which it is governed are voluntariness, although this is delimited by the principle of "comply or explain", and generality. It introduces new features such as board gender diversity, a commitment to women, such that recommendation 19 of the code establishes that when the number of female directors is scarce or nil, the board must explain the reasons for this and the measures taken to address it. It also includes specific recommendations on the chairman and secretary of the board of directors and ensures greater transparency in remuneration (recommendations 40-51). 4. Good Governance Code of Listed Companies The Good Governance Code of Listed Companies was published by the CNMV in 2015. Its objectives are to ensure the proper functioning of the governing and administrative bodies of Spanish companies so as to lead them to the highest levels of competitiveness; generate trust and transparency for national and foreign shareholders and investors; improve the internal control and corporate responsibility of Spanish companies, and ensure adequate segregation of functions, duties and responsibilities in companies, from a perspective of maximum professionalism and rigour. Although it follows the same line as its predecessor, it also introduces new features, and is based on a new structure that establishes the principles by which concrete and specific recommendations are created. Several recommendations from the Unified Code (2006) have since become laws, which is why they no longer appear in this code published in 2015, added to which it also incorporates a number of recommendations on CSR. 51 The predominant principle of the Code is voluntariness, subject to the principle of “comply or explain”. Similarly, this Code establishes that it is up to shareholders, investors and markets to evaluate the explanations that the different listed companies give for failing to follow the recommendations ("does not comply") or for only partially fulfilling them ("partially complies or explains"). Finally, and as its name suggests, the Code establishes that it is aimed at listed companies, which are those whose shares are traded on an official secondary stock market. 5. Good Governance Code of Listed Companies. Revised June 2020 The 2020 review process of the previous code affected, with differing degrees of intensity, recommendations 2, 4, 6, 7, 8, 14, 15, 22, 24, 37, 39, 41, 42, 45, 53, 54, 55, 59, 62 and 64. In particular, the first two modified recommendations are related to general features. Recommendations 6 to 8 refer to general shareholder meetings, while the remaining recommendations concern the board. As regards the main novelties of the review, it is worth highlighting that removed from the Code was the recommendation that “the director selection policy promotes the objective that in the year 2020, the number of female directors represents at least 30% of the total members of the board of directors” (Recommendation 14 from the Good Governance Code of Listed Companies 2015). At the end of 2020, the percentage of women directors on Spanish listed companies stood at 26.1%. Indeed, new recommendation 15 established that “the number of female directors represents at least 40% of the members of the board of directors before the end of 2022 and thereafter, previously being less than 30%”. Moreover, recommendation 37 refers to the structure of the executive committee, including as a novelty that it should be composed of at least two non-executive directors, one of whom should be independent. Additionally, the review of the code incorporates a new audit committee function, which should ensure that internal control policies are applied successfully (Recommendation 42). Finally, it is also worth mentioning recommendations on sustainability policies, which are an innovation in this latest version of the code. To sum up, Table 2 lists the main contributions of each of the Spanish CGC. 52 Table 2. Spanish CGC Code Number of recommendations Contribution Olivencia Report (1998) 23 -Board of directors’ functions: supervision and management control. -2 committees: Audit Committee and Appointments and Remuneration Committee. Aldama Report (2002) 24 - Ensure the transparency and security of the financial markets. - Publication of the Corporate Governance Annual Report. Unified Code (2006) 74 -Introduction of board gender diversity issues. -Particular recommendations about the Chairman and Secretary of the board. -Wide-ranging transparency in remuneration. Good Governance Code of Listed Companies (2015)4 64 - New structure that establishes the principles by which specific recommendations are created. -Establishment of CSR recommendations. Good Governance Code of Listed Companies. Revised June 2020 64 -Female directors should represent at least 40% of board seats by 2022. -Sustainability policy recommendations. Source: own elaboration based on the information published by the CNMV 4 Appendix A contains the recommendations contained in this Code, compliance with which has been analysed. 53 3.2. Board of directors CGC include different aspects, such as a) responsibilities; b) structure; c) composition; d) activity and functioning; and e) organization of the board of directors. Specifically, in the Spanish context, some of the recommendations included in the CGC in relation to the previous issues are highlighted below (Good Governance Code of Listed Companies. Revised June 2020): a) Regarding board responsibility, recommendation 12 establishes that boards should carry out their functions with independent judgement, and that their aim should be to pursue corporate interests, which means securing a profitable and sustainable business in the long term. b) Board structure is regulated in recommendations 13 and 14, specifying issues such as board size or gender: - Board size: between five and fifteen members. - Boards should approve a policy that guarantees an adequate composition of the board, favouring diversity of knowledge, experience, age and gender 5 . c) Recommendations 15 to 19 refer to board composition: - Proprietary and independent directors should represent the vast majority of the board of directors, and the number of executive directors should be the minimum required. - Female directors should account for at least 40% of board seats by the end of 2022, and thereafter, not previously being less than 30%. - Independent directors should account for at least 50% of the board of directors. d) With respect to board activity and functioning, there are several recommendations, ranging from 25 to 36 which are summarized in Table 3, and which include matters linked to the dedication of directors, meetings, and the existence of specific committees, among others. 5 It is considered that a measure favours gender diversity when it encourages the company to have a significant number of female managers. 54 Table 3. Recommendations on board activity and functioning Subject matter Recommendation Director dedication The Appointments Committee should ensure that nonexecutive directors have enough time to carry out their functions properly. The maximum number of boards on which a director can sit must also be established. Frequency of meetings and board attendance At least eight times a year. If a director cannot attend a meeting, which may be for an exceptional reason, he/she can delegate his/her vote. Information and advice to directors The company must provide the systems for managers to access the advice required to perform their functions. Prior to the meeting, the agenda must be known, which includes all the points to be discussed. Chairman/Chairwoman The Chair must perform the legal and statutory functions that have been attributed as well as organize the schedule of dates and matters to be discussed and the evaluation of the board. Furthermore, he/she will be the head of the board and responsible for its operation. Board of Directors Secretary The Secretary must ensure that the recommendations contained in the CGC are complied with. Board evaluation The Board should evaluate its actions once a year and adopt timely corrective measures. Source: Good Governance Code of Listed Companies. Revised June 2020 e) As regards board organization, CGC recommendations from 37 to 54 establish several aspects concerning the existence of certain board committees: e1) Executive committee. If there is an executive committee, it should contain at least two non-executive directors, one of whom should be independent. The secretary should be the same as for the board. 55 e2) Audit committee. It should be an internal audit unit. The Code also establishes audit committee functions such as monitoring and evaluating the development process and the integrity of financial and non-financial information, as well as risk control and management systems. It should ensure the independence of the department that assumes the function of internal audit, among others. e3) Risk control and management function. The risk control and management policy should identify the various financial and non-financial risks; a risk control and management model; the level of acceptable risk in accordance with the company; the measures to mitigate risk; and information systems and internal control. The Code indicates there should be an internal control and risk management function exercised by an internal department. e4) Appointments and Remuneration committee. The vast majority of the directors should be independent. These must be two separate committees if the company is highly capped. The Code sets out the remuneration committee functions, such as ensuring that any conflicts of interest do not harm the independence of external advice provided to the committee. e5) Others specialized committees. Supervision and compliance with environmental, social and corporate governance policies will be attributed to the audit, appointments, sustainability or CSR committee or to another specialized commission. In particular, within demographic characteristics, gender diversity has received a great deal of attention from diverse points of views. Because of this, and considering that one of the specific objectives of this thesis focuses on board gender diversity, the next section is dedicated to contextualising the legal framework concerning the presence of female directors on the board. 3.3. Board gender diversity The inclusion of women in the labour sphere is a reality that is still characterized by the existence of inequalities, since there are some positions, especially in senior management, which have traditionally been held by men, and which women may face various difficulties when attempting to access. Because of that, from a 62 include information on the gender quota in their annual remuneration report. If the quota is not met, companies are required to comply or to explain why, and to describe the measures that have been and will be taken to increase the numbers for the underrepresented gender (Federal Act on the Amendment of the Swiss Civil Code. Part Five: The Code of Obligations, article 734f, 2021). United Kingdom. The UK Government recommended a voluntary target of a minimum of 33 percent female representation on FTSE 350 boards by 2020 and also recommended that FTSE 100 companies should aim for a minimum of 33 percent female representation across their executive committee and direct reports to the executive committee by 2020 (FTSE Women Leaders, 2016). b2) General recommendation It refers to mentions in the CGC of each country that propose board gender diversity, without highlighting any particular percentage to be reached by female directors. This is the case of Sweden, Finland, Denmark, South Africa, Singapore, Hong Kong, Brazil, and Japan. In contrast, although they are exceptions, there are some countries whose CGC do not include any type of recommendation on board gender diversity. These countries include Canada, Ireland (Irish Annex of the UK Corporate Governance Code), the United States, and Russia. Taking into account that this thesis focuses on CGC, a summary of the recommendations previously commented on and which appear in each country’s CGC is provided, regardless of whether they are voluntary or mandatory, because they have become laws. This is shown in Table 4. 63 Table 4. Gender diversity in national CGC CGC Recommendations on board gender diversity Australia. ASX Corporate Governance Principles and Recommendations: 4th Edition Recommendation 1.5 A listed entity should: (a) have and disclose a diversity policy; (b) through its board or a committee of the board set measurable objectives for achieving gender diversity in the composition of its board, senior executives and workforce generally. If the entity was in the SandP/ASX 300 Index at the commencement of the reporting period, the measurable objective for achieving gender diversity in the composition of its board should be to have not less than 30% of its directors of each gender within a specified period. Austria. Austrian Code on Corporate Governance (2021) V. Supervisory board Qualifications of Members, Composition, and Independence of the Supervisory Board 52. When electing the members of the supervisory board, the general shareholders’ meeting shall take due care to check the expertise and personal qualifications of the supervisory board members and to ensure a balanced composition with respect to the structure and the business of the company. Furthermore, reasonable attention is to be given to the aspect of diversity of the supervisory board with respect to the representation of both genders and the age structure, and in the case of exchange-listed companies, also with a view to the internationality of the members. 64 The supervisory board shall be made up of at least 30 percent women and at least 30 percent men, provided the supervisory board consists of at least six members (shareholder representatives), and the staff representatives must consist of at least 20 percent female and male employees each. Belgium. The 2020 Belgian Code on Corporate Governance Principle 3. The company shall have an effective and balanced board The composition of the board should be determined so as to gather sufficient expertise in the company’s areas of activity as well as sufficient diversity of skills, background, age and gender. Brazil. Código Brasileiro de Governança Corporativa – Companhias Abertas 2.2. Composition of the board of directors. Recommended practices 2.2.2 The board of directors should approve an appointment policy establishing: (ii) that the board of directors should be composed taking into consideration the time availability of its members for exercise of their duties and diversity of knowledge, experiences, conducts, cultural aspects, age, and gender. 3.2. Appointment of executive management members. Principle The process of appointing and filling positions in the executive management and managerial positions should seek to form a group aligned with the principals and ethical values of the company, taking into consideration diversity, including gender diversity, seeking occupation of such positions by people with complementary abilities and qualified to face the challenges of the company. 65 Denmark. Recommendations on Corporate Governance (2017) 3. Composition and organization of the board of directors 3.1. Composition Recommendation 3.1.2. The Committee recommends that once a year the board of directors discusses the company’s activities to ensure relevant diversity6 at management levels, prepares, and adopts a policy on diversity. The policy should be published on the company website. Finland. Finnish Corporate Governance Code 2020 Recommendation 8 – Composition of the Board of directors The composition of the company’s board of directors shall reflect the requirements set by the company’s operations and development stage. A person elected as a director must have the competence required by the position and the possibility to devote a sufficient amount of time to attending to the duties. The number of directors and the composition of the board of directors shall be such that they enable the board of directors to see to its duties efficiently. Both genders shall be represented in the board of directors. 6 Comment: diversity includes age, international experience and gender. A policy on diversity should concern matters relevant to the company in relation to diversity, which promote a relevant degree of diversity, strengthen management’s qualifications and competencies and take into account the future development of the company. 66 France. Corporate Governance Code of Listed Corporations 2020 6. Membership of the board of directors: guiding principles 6.2. Each Board should consider what the desirable balance of its membership and that of the Board committees should be, particularly in terms of diversity (gender representation, nationalities, age, qualifications, professional experience, etc.). It should make public in the report on corporate governance a description of the diversity policy applied to members of the Board of directors as well as a description of the objectives of this policy, its implementation measures and the results achieved in the past financial year. 7. Gender diversity policy on the governing bodies 7.1. At the proposal of the executive management, the Board shall determine gender diversity objectives for governing bodies. The executive management shall present measures for implementing the objectives to the Board, with an action plan and the time horizon within which these actions will be carried out. The executive management shall inform the Board each year of the results achieved. 7.2. In the report on corporate governance, the Board shall describe the gender diversity policy applied to the governing bodies as well as the objectives of this policy, the implementation measures and the results achieved in the past financial year including, where applicable, the reasons why the objectives have not been achieved and the measures taken to remedy this. 67 Germany. German Corporate Governance Code 2019 C. Composition of the Supervisory Board Principle 11 The composition of the Supervisory Board has to ensure that its members collectively possess the knowledge, skills and professional expertise required to properly perform their duties; furthermore, the legal gender quota must be considered. Hong Kong S.A.R., China. Review of Implementation of Corporate Governance Code 2018 (HKEX) Chapter 1: Introduction Background To help directors carry out their role more effectively, the Exchange also published, in July 2018, “Guidance for Boards and Directors” (“Guidance”).16 The Guidance contains practical advice to boards and directors on their roles and responsibilities. It covers directors’ duties and board effectiveness, board committees, board diversity - including gender diversity - and corporate governance for WVR Issuers. Italy. Italian Corporate Governance Code 2020 Article 2. Composition of corporate bodies Principle VIII The company applies diversity criteria, including gender ones, to the composition of the board of directors, ensuring the primary objective of adequate competence and professionalism of its members. Recommendation 8 At least a third of the board of directors and the control body, where the latter is autonomous, is to be comprised of members of the less represented gender. 68 Companies adopt measures to promote equal treatment and opportunities among genders within the whole organisation, monitoring their specific implementation. Japan. Japan’s Corporate Governance Code Seeking Sustainable Corporate Growth and Increased Corporate Value over the Mid- to Long-Term 2018 Section 4. Responsibilities of the Board Principle 4.11 Preconditions for Board and Kansayaku Board Effectiveness The board should be well balanced in knowledge, experience and skills in order to fulfil its roles and responsibilities, and it should be constituted in a manner to achieve both diversity, including gender and international experience, and appropriate size. Malaysia. Malaysian Code on Corporate Governance (2021) II. Board composition Practice 5.5 Appointments of board and senior management are based on objective criteria, merit and with due regard for diversity in skills, experience, age, cultural background and gender. Guidance G5.9 All boards should comprise at least 30% women directors. Numerous studies have proven the business case for board diversity, in particular the participation of women on boards. If the composition of women on a board is less than 30%, the board should disclose the action it has or will be taking to achieve 30% or more and the timeframe to achieve this. A reasonable timeframe is one that is three years or less. 69 Netherlands. Dutch Corporate Governance Code- 2016 Chapter 2. Effective management and supervision 2.1.5 Diversity policy The supervisory board should draw up a diversity policy for the composition of the management board, the supervisory board and, if applicable, the executive committee. The policy should address the concrete targets relating to diversity and the diversity aspects relevant to the company, such as nationality, age, gender, and education and work background. Explanatory notes to the Code 2.1.5 Diversity policy & 2.1.6 Accountability about diversity The legal target figure of at least 30% male/female diversity in the management board and the supervisory board. Norway. The Norwegian Code of Practice for Corporate Governance 2018 8. Board of directors: composition and independence The composition of the board of directors as a whole should represent sufficient diversity of background and expertise to help ensure that the board carries out its work in a satisfactory manner. In this respect due attention should be paid to the balance between male and female members of the board. The board is responsible as a collegiate body for balancing the interests of various stakeholders in order to promote value creation by the company. The board should be made up of individuals who are willing and able to work as a team. 70 Portugal. Code of Corporate Governance 2018 (revised in 2020) I.2. Diversity in the composition and functioning of the company’s governing bodies I.2.1. Companies should establish standards and requirements regarding the profile of new members of their governing bodies, which are suitable according to the roles to be carried out. Besides individual attributes (such as competence, independence, integrity, availability, and experience), these profiles should take into consideration general diversity requirements, with particular attention to gender diversity, which may contribute to a better performance of the governing body and to the balance of its composition. Singapore. Code of Corporate Governance 2018 Board Composition and Guidance 2.4 The Board and board committees are of an appropriate size, and comprise directors who as a group provide the appropriate balance and mix of skills, knowledge, experience, and other aspects of diversity such as gender and age, so as to avoid groupthink and foster constructive debate. South Africa. King Report on Corporate Governance for South Africa - 2016 (King IV Report) Recommended practices. Composition 10. The governing body should promote diversity in its membership across a variety of attributes relevant for promoting better decision-making and effective governance, including field of knowledge, skills and experience as well as age, culture, race and gender. 71 Spain. Good Governance Code of Listed Companies., Revised June 2020. Recommendation 14 The board of directors should approve a policy aimed at promoting an appropriate composition of the board that: … c) favours diversity of knowledge, experience, age and gender. Therefore, measures that encourage the company to have a significant number of female senior managers are considered to favour gender diversity. To promote desirable gender diversity on the board of directors, it is recommended that female directors represent at least 40% of the total number of members by 2022. Sweden. The Swedish Corporate Governance Code (2016) 4. The size and composition of the board 4.1 The board is to have a composition appropriate to the company’s operations, phase of development and other relevant circumstances. The board members elected by the shareholders’ meeting are collectively to exhibit diversity and breadth of qualifications, experience and background. The company is to strive for gender balance on the board. Switzerland. Swiss code of best practice for corporate governance Board of Directors and Executive Board 12. Composition The Board of directors should be comprised of male and female members. They should have the necessary abilities to ensure an independent decision-making process in a critical exchange of ideas with the Executive Board. 78 be established in order to align their interests. Likewise, the company must establish supervision and control mechanisms that minimize the possibilities of using the company’s resources so that shareholders may benefit (Fama and Jensen, 1983). The precursors of the theory proposed that the key to overcome those conflicts of interest is to align the interest of owners and managers (Jensen and Meckling, 1976; Fama and Jensen 1983). Since its inception, this theory has been predominant in the corporate governance arena, especially vis-à-vis board of directors studies (Hillman et al., 2000; Nicholson and Kiel, 2007). It is in this context where the board of directors becomes particularly important. The monitoring function of boards, known as the control role, refers to directors’ responsibility to monitor managers in the interests of shareholders (Hillman and Dalziel, 2003). Prominent amongst the board’s supervisory functions are monitoring the CEO, monitoring board strategy, organising CEO turnover, as well as evaluating and establishing CEO and top manager remuneration (Hillman and Dalziel, 2003). In line with agency theory, the motivation for the board’s supervisory role are board incentives, such that when these incentives align with shareholders’ interests, directors will improve their monitoring ability and will be more efficient, which implies better performance (Jensen and Meckling, 1976; Fama, 1980). In general, agency theory offers defined implications for monitoring and for the control role of the board (Eisenhardt, 1989). Moreover, this theory may also explain the board of directors’ role in terms of company strategy. Nevertheless, its position with regard to the strategic function is not so clear. One stream of studies argues that boards play a key role in formulating and implementing company strategy (Zahra and Pearce, 1989; Useem, 2003). Agency theory declares that the board of directors has an impact on strategic decisions by avoiding opportunistic behaviour on the part of managers, protecting investors’ interests (Mizruchi, 1983), taking part in reviewing and monitoring strategic decisions (Stiles and Taylor, 2001), and thereby becoming jointly responsible for them (Sundaramurthy and Lewis, 2003). Two monitoring mechanisms which have commonly been used in agency theory are director independence and remuneration systems (Bartkus et al., 2002). 79 a) Board independence. Considering that the principal role of the board is to monitor managerial opportunism and to control managers’ actions so as to benefit shareholders, if the board is primarily made up of insider directors, then the likelihood of ineffective board monitoring increases (Dey, 2008). Furthermore, independent directors are also less committed to economic performance and may be more likely to advocate investments required for long‐term sustainability (He and Jiang, 2019). Indeed, previous studies have found that board independence is negatively and significantly associated with agency conflicts (Badu and Appiah, 2017) and can reduce the firm’s agency cost (Rashid, 2015). Jensen and Meckling (1976) define agency costs as “the sum of the monitoring expenditures by the principal; the bonding expenditures by the agent; and the residual loss”. The former refers to the cost of the supervisory role which the principal may pay. The second denotes those costs derived from ensuring that the agent will not take certain decisions which would damage the principal or that the principal would be compensated in that case. The third indicates the loss borne by the principal because of their failing to take that decision. It may therefore be a loss in wealth when making decisions that they would not have made in his place (Jensen and Meckling, 1976). b) Remuneration systems. Shen (2005) highlights the importance of rewarding incentives to non-executive directors in order to enhance board effectiveness. For many non-executive directors, the performance of the company they monitor has virtually no impact on their personal wealth since they have almost no shares in the company (Patton and Baker, 1987; Hambrick and Jackson, 2000). Accordingly, from the agency theory point of view, remuneration strategy is a key and powerful incentive system which motivates and commits managers to their supervisory functions (Shen, 2005). 4.2. Resource dependence theory Resource dependence theory is another of the most commonly cited theories in the corporate governance sphere, and is vital to the sociology, organization, and management disciplines (Hillman et al., 2009). This theory posits that the board of directors provides a crucial link in the firm and offers the necessary resources it 80 requires to maximize performance (Pfeffer and Salancik, 1978), with resources being understood as “anything that could be thought of as a strength or weakness of a given firm” (Wernerfelt, 1984). Resource dependence theory considers an organization as an open system which may be influenced by whatever occurs in the company environment. Indeed, the context in which companies undertake their activity must be taken into consideration (Pfeffer and Salancik, 1978). In particular, directors can reduce environmental uncertainty and dependence (Hillman et al., 2009). Specifically, resource dependence theory argues that board members can contribute to strategic decision-making by providing unique resources that the organization would not otherwise have access to (Pfeffer and Salancik, 1978; Hillman and Dalziel, 2003). Both the board’s capacity to bring critical resources and the specific type of resources they provide are relevant. Previous research has shown that board members can provide their companies with valuable information as well as easier access to capital, interconnections with crucial suppliers, customers and other stakeholders (Baysinger and Zardkoohi, 1986; Mizruchi and Stearns, 1988; Freeman and Evan, 1990; Banerji and Sambharya, 1996; Frooman, 1999). The board can promote continued value creation (Hillman et al., 2009), with the resources provided by directors making a contribution in (Pfeffer and Salancik, 1978): a. advising and guidance, b. legitimacy, c. providing a link between external organizations and the company by conveying important information, d. securing support from outsiders. There is theoretical evidence that boards with several connections to company environments will furnish the organization with better access to resources (Nicholson and Kiel, 2007). Therefore, the advice or resource provision assignment of board members involves real involvement in decision making and overseeing of strategic options, as well as guidance on tactical proposals (Pugliese et al., 2014). 81 4.3. Stakeholder theory Stakeholder theory is found within the framework of organization theories. According to stakeholder theory, companies must consider several stakeholder groups that either affect or are influenced by a firm’s objectives (Freeman, 1984). Company stakeholders are “individuals that contribute, either voluntarily or involuntarily, to its wealth-creating capacity and activities, and who are therefore its potential beneficiaries and/or risk bearers” (Post et al., 2002). Freeman (1984) tried to illustrate the relationship between the organization and its external environment converging with its actions within this environment. The elementary premise of stakeholder theory is that directors must manage company activity by reconciling the firm’s objectives and needs with the requirements of all the stakeholders involved such as, suppliers, customers, regulators, the environment, the local community, and the general public (Madsen and Rodgers, 2015; Yasser et al., 2017). According to Donaldson and Preston (1995), companies should consider their stakeholders for two reasons: (1) from a normative approach, their needs should be considered because it shows inherent value, and (2) from an instrumental focus, contrasting the interests of influential stakeholders could enhance company profits. Thus, stakeholders are all groups which interact with the company. They can be classified into (Clarkson, 1995): a) primary, and b) secondary. a) Primary stakeholders: those who have legitimacy vis-à-vis the company, which means maintaining formal relationships with the company (customers, suppliers, employees, shareholders…). b) Secondary stakeholders: those who have no legitimacy rights; non-formal relationships (local community, government…). In this sense, the board as the main decision-making body of the company must consider the general benefits of its decisions, not only for the company but also for its environment. Accordingly, stakeholder theory manifests the need to invest resources and time to address stakeholders’ demands (Sasse and Trahan, 2007; O’Riordan and Fairbrass, 2014) which specifically involves seeking business benefits as well as pursuing the moral responsibility the company has towards its 82 stakeholders (McKnight and Linnenluecke, 2016). In particular, previous research has highlighted that taking into account the needs of stakeholders in management favours business results through the creation of intangibles such as trust and cooperation (Zhang et al., 2014). In particular, stakeholder theory arguments are connected to CSR principles (Carroll, 1991). Several CSR perspectives can be identified (Carroll and Shabana, 2010; Wójcik, 2018): a) CSR as a marketing tool to create or improve corporate reputation (Wang and Gao, 2016). CSR practices could generate positive perceptions of the company among its stakeholders (Rothenhoefer, 2019). In this sense, the “favourable/adverse” reputation may influence or even determine present and future relationships between stakeholders and the company (Rothenhoefer, 2019). Saeed and Arshad, (2012) found that external and internal CSR activities positively impact corporate performance, generating intangible resources such as corporate reputation. b) CSR as a source to acquire competitive advantage. Prior studies have found that socially responsibly activities could produce benefits because of cost reduction, which ultimately reduces business risk (Wang and Gao, 2016; Cook et al., 2019). First, sustainable development enhances effectiveness by reducing energy costs (Lister, 2018). Second, CSR practices enrich community relations, leading to cost reduction, as illustrated by the costs associated to regulations governing business operations and gaining tax benefits (Carroll and Shabana, 2010). CSR practices may also lead to competitive advantage based on company differentiation or the firm’s market products (Carroll and Shabana, 2010). In particular, CSR activities call for board members’ ability to solve the social and environmental needs of diverse stakeholders whilst also implying innovation through the development of new technologies, corporate techniques and management systems which result in competitive differentiation (Engert et al., 2016; Cook et al., 2019). c) CSR as a signal for stakeholders. CSR practices may enable certain firm characteristics to be perceived which would not otherwise be seen, thereby possibly reducing information asymmetries between company and 83 stakeholders (Su et al., 2016). Hence, stakeholders may perceive the sustainability activities implemented by the board as a sign of transparency and commitment to social and environmental demands (Flammer, 2018). d) CSR as a tactic to achieve win-win results. Recent researchers have documented a positive association between CSR practices and financial performance (Waworuntu et al., 2014; Flammer, 2015; Cho et al., 2019; Bahta et al., 2020). Hence, firms which engage in CSR practices act responsibly towards their stakeholders while also improving corporate performance. e) CSR to manage sustainability. Social and environmental excesses have put ever-increasing pressure on boards of directors as drivers of sustainable business strategies (Jain and Jamali, 2016; Arayssi et al., 2020). In this sense, CSR is perceived as a control mechanism that can safeguard the interests of diverse stakeholders and future generations. In particular, previous studies argue that boards tend to perform CSR practices in an effort to develop a business image that ensures a more sustainable future (Farinós, 2017). 4.4. Upper echelon theory The main assumption of upper echelon theory is that board of director experience, values, and personalities significantly influence their understanding of the circumstances they are involved in and so impact their decisions (Hambrick and Mason, 1984). The predominant notion of upper echelon theory is that; a) executives take actions depending on their personal perception of the strategic situations they face, and b) these personalized interpretations are a reflection of executives’ personalities, values, and experiences (Hambrick and Mason, 1984). Cyert and March (1963) highlight that decisions made under uncertain conditions are not objective, such that they require explanation and clarification. In order to really comprehend how and why companies act, it is therefore necessary to consider the biases and dispositions of the top decision makers; the board of directors (Hambrick, 2007). Hambrick (2007) introduces two ideas of upper echelon theory. The first centres on the characteristics of the top management team, rather than merely focusing on the CEO alone, in order to obtain more suitable answers to firm 84 outcomes, with the argument being that company management is a challenging and joint task. Indeed, previous studies have indicated that board composition impacts firm performance (Carpenter and Fredickson, 2001; McIntyre et al., 2007; Rahman and Saima, 2018). The second refers to the fact that because psychological traits (experiences, values, personalities) are barely observable, the demographic attributes of top management, as reflected through educational background, industry and firm tenures, and personal associations, may be acceptable proxies of their cognitive frames and values, to predict strategic activities (Hambrick, 2007). In addition, Hambrick (2007) proposed two moderating variables of the relationship between managerial characteristics and organizational outcomes; a) managerial discretion and, b) executive job demands. The former applies to the freedom of actions that directors enjoy (Carpenter et al., 2004; Crossland and Hambrick 2011). Indeed, in external environments where there is high managerial discretion, directors’ attributes would have greater predictive power of organizational outcomes than in contexts where board members have less decision-making capacity (Hambrick, 2007). The latter reflects the degree of challenge faced by directors (Hambrick et al., 2005). Hambrick (2007) proposes that the link between managers’ characteristics and company results will be stronger when executive job demands are high because managers who face difficult challenges seem to lack the time for an in-depth analysis of decisions and consequently place greater trust in their personal experiences. Nevertheless, if the degree of managerial challenges is lower, their decisions will be more exhaustive and will rely less on their personal background. The relationship analysed will thus be weaker in this kind of situation (Hambrick, 2007). Furthermore, the attributes and strategic choices of board members are conditioned by the organization’s situation, which means that the former will vary with different industry contexts (Hambrick and Mason, 1984). The effect of the organizational environment should thus be considered as a moderator in the relationship between the top management team composition and firm outcomes (Carpenter et al., 2004; Finkelstein et al., 2009). In this sense, Yamak et al., (2014) argued that both industry and international firm contexts must be examined in order to truly understand the antecedents and repercussions of top management team composition 85 and behaviour. In particular, they identify three different upper echelon models (Yamak et al., 2014): direct, mediation, and moderation models. In direct effect models, the organizational environment has a direct impact on top management team composition and/or procedures. In mediation models, board composition, structure and methods influence the association between company environment and strategic choices and results. In moderation models, organizational environment moderates the relationship between both top management team procedures and their composition and the firm’s performance. 4.5. Institutional theory Institutional theory is applied in sociology, business sciences, economics, and politics (Gutiérrez-Rincón and Salas-Páramo, 2015). The focal point of institutional theory is that organizations are institutions that have a meaning, a value and legitimacy for their members and leaders (Selznick, 1957). Its origin dates back to the 1950s in the economics and business sciences area. In the 1970-80s, this perspective re-emerged under the name of “New Institutionalism or Neoinstitutionalism” (Meyer and Rowan, 1977; DiMaggio and Powell, 1983), which focuses on the role of institutions and their impact on business behaviour, the similarities between organizations and companies in terms of reshaping their environment (Gutiérrez-Rincón and Salas-Páramo, 2015). New institutionalism argues that organizations are influenced by pressures from their institutional environment, forcing them to modify their behaviour, methods and configuration, and moderating their economic activity (Ortas et al., 2015). In fact, institutional theory is based on the creation of socially commonly accepted rules and their application (Ortas et al., 2015). This theory assumes that companies will adopt common approaches because of their need to adjust to institutions (Qiao and Wu, 2019). In this sense, organizational institutionalism analyses the adaptions and formations of organizations due to the pressures that the institutional environment exerts on them to gain social legitimacy (DiMaggio and Powell, 1983; Scott, 2001). Institutions are norms, guidelines and regulations which condition social actors’ behaviour, thereby making social life meaningful (North, 1990; DiMaggio and 86 Powell, 1991; Scott, 2001), such that the lack thereof will hinder economic activity (Aracil, 2019). Institutional theory illustrates that institutions influence business practice, which is known as isomorphism. Meyer and Rowan (1977) established that isomorphism is the process by which formal organizations adapt to their environment in order to manage commercial and technical interdependencies. Hence, the consequences of isomorphism are the integration of factors by companies not because of how efficient these factors are, but because of their external legitimacy (Meyer and Rowan, 1977). DiMaggio and Powell (1991) identified two types of isomorphism: competitive and institutional. Competitive isomorphism applies to efficiency, whereas institutional isomorphism guides uniform behaviour and prompts similar organizational configurations. Institutional isomorphism can be coercive, normative, and mimetic (Ortas et al., 2015). Coercive isomorphism refers to pressures exerted by laws (formal) and codes of conduct (informal) on organizations. Normative isomorphism is triggered by the professionalization of decision makers (board of directors), leading to parallel socialization. It this sense, directors tend to recognize, understand and resolve issues or organizational problems in a similar manner. Mimetic isomorphism studies how companies can imitate their competitors because of uncertainty (Ortas et al., 2015). Furthermore, neo-institutional theory predicts that firms which operate in different countries establish distinct CSR preferences because their culture, as illustrated by beliefs and socially established norms, is influenced by organizational factors (Selznick, 1996), such as CSR practices (Frederick, 2006). Corporate governance mechanisms are thus influenced by the institutional environment. In this sense, Boytsun et al., (2011) postulate that the costs and benefits of corporate governance actions and board of directors’ decisions may be impacted by social norms. 4.6. Signalling theory Signalling theory deals with how the information asymmetry dilemma in a competitive environment may be addressed (Connelly et al., 2011; Mavlanova et al., 2012; Taj, 2016). Its origin can be traced back to Spence’s work (1973) applied to the labour market, to understand how employees send signals about their ability to employers, by acquiring certain levels of education. The informative value of the 87 signals is explained because employers presume there is a positive correlation between educational level and a greater ability to perform a task (Spence, 1973). The theory mainly fixes its attention on management’s aim to share information and receive signals from the market, stakeholders and society (Bae et al., 2018). Signalling theory holds that companies decrease the information asymmetry of external users through multiple signals (Connelly et al., 2011). Likewise, information asymmetry involves inherent conflicts of interests between managers and agents and the signal decreases the gap by conveying pertinent and high-quality information to both groups (Connelly et al., 2011; Taj, 2016). Signalling theory argues that the information which companies disclose is a signal designed to convey the attributes of the organization which may be hard to perceive, and which achieve organizational behaviour or behavioural intentions (Zerbini, 2017, Luffarelli and Awaysheh, 2018). Signalling theory is based on information asymmetry between two parties and how they produce signals to combat this (Luffarelli and Awaysheh, 2018). The essential components of signalling theory comprise; a) signaller, b) signal, and c) receiver (Connelly et al., 2011): a) signallers― they are insiders, such as management or executives, who had some knowledge about an individual (Spence, 1973), organization (Ross, 1977), or product (Kirmani and Rao, 2000), which outsiders do not know and which would be useful for them. b) signal― insiders acquire relevant private information which they have to communicate to outsiders. The signal includes two basics; the information itself and its sign, which can be either positive or negative (Luffarelli and Awaysheh, 2018). The theory basically focuses on the intentional disclosure of positive information, hiding negative data in the search for positive organizational attributes (Connelly et al., 2011). Indeed, organizational behaviour could be a signal of good governance practices and might be legitimate for the receiver (Groening and Kanuri, 2018). c) receivers―they are outsiders who cannot access organizational information but would like to obtain it. 94 CHAPTER 5: EMPIRICAL STUDY As mentioned before, the main purpose of this thesis is to explore current knowledge of corporate governance effects on certain firm outcomes. In particular, this general aim is divided into three more specific objectives, which look into the relation between board of directors and firm financial distress, non-financial information disclosure policy, and environmental innovation practices. Taking into consideration these objectives, the empirical analysis is composed of three independent studies: 1. Does compliance with corporate governance codes help to mitigate financial distress? 2. Corporate governance code compliance and environmental, social and governance disclosures. 3. A critical approach to the true influence of female directors on environmental innovation: when are women greener? Therefore, each of these specific empirical studies is presented separately in this chapter. Although the motivation of these empirical studies was commented on in Chapter 1, in the following sections, for each individual piece of research, the objectives, research method, and main findings are explained. 5.1. Does compliance with corporate governance codes help to mitigate financial distress? 5.1.1. Objective The succession of corporate scandals all over the world and the recent global financial crisis have brought to light deficiencies in governance mechanisms, which have had severe consequences in capital markets. Accordingly, the number of CGC has increased exponentially over the last few years. Indeed, most developed countries have recently introduced CGC and, consequently, researchers have paid a great deal of attention to investigating the effects of their compliance. As mentioned before, previous research indicates the relevance and complexity of predicting firms’ financial distress situations for agencies’ credit ratings, governments or financial creditors and has emphasized the role played by corporate governance mechanisms in preventing business failure (Manzaneque et al., 2016a). 95 The main objective of this study is to analyse whether compliance with CGC may help to mitigate the financial distress of firms. As a significant number of recommendations contained in CGC relate to boards of directors and their committees, specific measures for compliance of recommendations regarding these issues are considered, and three objectives can be highlighted: a) To analyse whether overall compliance with CGC recommendations would reduce the likelihood of financial distress. b) To examine whether compliance with CGC recommendations regarding the board of directors would reduce the likelihood of financial distress. c) To study whether compliance with CGC recommendations regarding board subcommittees would reduce the likelihood of financial distress. 5.1.2. Hypothesis development The empirical debate regarding the determinants of financial distress focused on financial and accounting information many decades ago (Altman, 1968; Beaver, 1966; Zmijewski, 1984). Nevertheless, more recent research suggests that economic and financial data alone may lack sufficient predictive power to anticipate financial distress (Chang, 2009; Fich and Slezak, 2008). In particular, a number of studies have pointed out that variables related to corporate governance structures must be taken into consideration in order to better understand the determinants of financial distress (Habib et al., 2020). Yet, the majority of these studies tend to examine specific corporate governance characteristics. In this regard, some studies find that companies with greater board independence and CEO duality are less likely to suffer financial distress (Salloum et al., 2013; Baklouti et al., 2016). Other studies have examined the relationship between board size and the likelihood of financial distress, although the evidence to emerge is mixed (Fich and Slezak, 2008; Manzaneque et al., 2016a). Parker et al. (2002) found a negative significant association between CEO replacement and the likelihood of firm survival. Moreover, Shahwan (2015) considered certain items related to shareholders’ rights and relationships with investors as potential drivers of financial distress. Several papers have also focused on ownership structure as a relevant corporate governance mechanism. For instance, the likelihood of corporate failure has been negatively associated with ownership concentration and state ownership (Li et al., 2008), with 96 institutional ownership (Manzaneque et al., 2016b; Udin et al., 2017), and with board ownership (Abdullah, 2006). Although the majority of these studies have used corporate governance characteristics individually, recent literature also calls for the need to employ composite measures, which aggregates a number of governance indicators considering the interaction between multiple corporate governance mechanisms so as to provide a better overview of the effectiveness of corporate governance structures (Brown et al., 2011; Jain and Jamali, 2016; Bravo et al., 2018). Consistent with this approach, this research focuses on aggregated measures regarding compliance with CGC recommendations in order to examine the effects of corporate governance on firms’ financial distress. CGC have several key universal principles for effective corporate governance (Aguilera and Cuervo-Cazurra, 2009) and have become widespread in the majority of developed economies. CGC are a form of soft regulations presenting a set of voluntary governance recommendations on relationships with shareholders and top management, the role and composition of the board of directors and its committees, auditing and information disclosure, and the selection, remuneration and dismissal of directors and top managers (Duh, 2017). The main objective of CGC is to strengthen internal control and maximize shareholders’ interests, which implies safeguarding business prosperity. Despite the growing diffusion of CGC and the intense academic debate about the effectiveness of these codes, empirical evidence on the impact of CGC is still far from definitive. Research into the effect of CGC compliance on firm outcomes is inconclusive and further inquiry is required (Stiglbauer and Velte, 2014). Some studies suggest that complying with CGC recommendations leads to better strategic decisions and therefore document a positive association between CGC compliance and different measures of firm performance, such as price-to-book ratio, Tobin’s q, or profitability (Fernández- Rodríguez et al., 2004; Bassen et al., 2006; Stiglbauer, 2010; Luo and Salterio, 2014; Rodríguez-Fernández, 2016). Other authors have argued that CGC compliance can be used for ethical reasons to gain social legitimacy, which will improve investors’ perceptions and lead to positive stock market reactions by improving firm value and share price (Goncharov et al., 2006; Chavez and Silva, 2009; Kaspereit et al., 2015; Kaspereit et al., 2017) as well as corporate reputation (Hooghiemstra and van Ees, 2011; McCahery et al., 2016). Nevertheless, many 97 studies fail to support the previous associations (Bassen et al., 2009; McKnight and Weir, 2009; Jain et al., 2011; Steger and Stiglbauer, 2011; Stiglbauer and Velte, 2014). Despite the increasing discussions on the determinants of financial distress and the significant debates on the effectiveness of CGC, the literature has thus far failed to analyse whether CGC compliance may have an impact on firms’ financial distress. The Spanish code, like most international CGC, includes a set of recommendations about general governance issues (such as anti-takeover mechanisms, functioning of the general meeting and the mechanics of voting), as well as specific recommendations regarding the board of directors and its committees. Therefore, this study aims to fill the previous research gap by examining three different levels of CGC compliance: (1) overall CGC compliance, (2) compliance with the recommendations regarding boards of directors, and (3) compliance with the recommendations related to board subcommittees. Since the main objective of corporate governance is to mitigate conflicts of interest between managers and shareholders (Shleifer and Vishny, 1997), the previous studies on CGC have largely relied on agency theory to explain the effects of their compliance (Cuomo et al., 2016). Following the dominant agency theory (Jensen and Meckling, 1976), CGC recommendations must aim to protect investors and reduce managerial opportunism (Zattoni and Cuomo, 2010), which is likely to guarantee company survival and therefore minimize the probability of business failure (Udin et al., 2017). In line with these premises, the potential relationship between financial distress and the different measures of CGC compliance according to the arguments of agency theory are explained. First, overall compliance with CGC recommendations should strengthen the corporate governance structures of a firm. Beyond reinforcing boards of directors and certain committees, those recommendations related with voting rights, annual meetings, and other general issues are also expected to safeguard shareholders’ interests and improve governance mechanisms (Lu et al., 2008). Agency theory suggests that strong corporate governance prevents controlling shareholders and managers from benefiting from a company at the cost of non-controlling shareholders and other stakeholders (Manzaneque et al., 2016a). This approach is 98 the one most often used in the literature, which highlights that adequate monitoring or control mechanisms of managerial decision-making processes are required in order to protect shareholders and other investors (Kiel and Nicholson, 2003; Stiglbauer and Velte, 2014). As a result, weak corporate governance may increase the probability of opportunistic behaviour by management or controlling shareholders, which could lead to an ethical conflict with shareholders and prioritize their personal aims against the overall company objective (La Porta et al., 2000). Therefore, a high degree of overall CGC compliance may decrease the wealth expropriation risk and, in turn, minimize the probability of company failure (Lee and Yeh, 2004). Additionally, a reduction in agency conflicts would lead to important benefits for firms that are likely to improve their financial situation, such as more access to capital, a reduction in cost of capital (Reddy et al., 2010), secure access to financial assets (Weber and Velte, 2011), and also would attract investment opportunities and improve capital market development (Udin et al., 2017). According to the previous arguments, if compliance with the overall recommendations included in the CGC reduces agency costs and reinforces the quality of corporate governance structures, a higher degree of compliance should reduce the likelihood of company financial distress. Hence, the following hypothesis is formulated: H1: Overall compliance with CGC will reduce the likelihood of financial distress. Moreover, boards of directors are a crucial corporate governance mechanism (Adams et al., 2010), since they are the highest decision-making authority within a firm and exert considerable power over corporate strategic actions (Galbreath, 2018). The board of directors is an essential control system which engages in decisive internal monitoring activities such as the evaluation of tasks carried out by the top management and the CEO, and the evaluation of firm strategy (Pugliese et al., 2009), which is expected to minimize the costs incurred when management pursues its own interests at the expense of the shareholders’ interests (Hillman and Dalziel, 2003). Specifically, boards must be responsible for monitoring decisions to detect and avoid financial instability (Chang, 2009; Manzaneque et al., 2016a), and to impose the necessary measures to help overcome a possible failure situation (Fich and Slezak, 2008). For this reason, the vast majority of CGC explicitly include 99 a set of recommendations concerning the board of directors (Aguilera and Cuervo- Cazurra, 2009). These recommendations aim to strengthen the degree of independence, the qualification, the diversity and the pool of resources in the board, among other characteristics. These characteristics are expected to provide the board with valuable skills and competence to improve the monitoring discipline of directors, and their effectiveness when overseeing corporate strategy (Huse and Solberg, 2006; Srinidhi et al., 2011; Field et al., 2017). Therefore, greater compliance with these recommendations must lead boards to be in a better position to monitor business decisions to minimize the risk of failure (Dowell et al., 2011) and ensure the financial situation of the firm (Simpson and Gleason, 1999). In line with the previous arguments, it could be assumed that boards with a higher compliance of CGC recommendations are in a better position to reduce the probability of financial distress. The following hypothesis is therefore formulated: H2: Compliance with CGC recommendations regarding the board of directors will reduce the likelihood of financial distress. Corporate governance mechanisms also include several board subcommittees, such as the audit committee, the appointments committee, the compensation committee, or the corporate governance committee. These committees make particular decisions and play an important role in overseeing corporate strategy to protect the interests of shareholders (Detthamrong et al., 2017). Compliance with the recommendations related to these committees should improve their quality and therefore their effectiveness when carrying out their monitoring functions. First, the audit committee has attracted great interest since it has become a key element to control and monitor management (Ruzaidah and Takiah, 2004). This committee is expected to oversee the strategic actions of a firm and, specifically, any financial or operational issues (Rahmat et al., 2009). Previous literature agrees that audit committees are needed to resolve agency conflicts (Klein, 2002) and to maintain good performance (Ainuddin and Abdullah, 2001). Therefore, the audit committee is responsible for decisions that should ensure business prosperity and mitigate the financial distress of companies (Salloum et al., 2014). On the other hand, although most of the attention paid by academics and professionals focuses on the audit committee, the other committees are also important vis-à-vis reducing agency 100 conflicts, which can result in improvements in a firm’s financial situation. For instance, the appointments committee can impact the monitoring process of the strategic actions, since this committee is likely to minimize the influence of leaders on the selection process (Shivdasani and Yermack, 1999; Baklouti et al., 2016). In addition, the compensation committee is in charge of evaluating management performance and of designing appropriate compensation packages, and hence this committee may limit agency problems by introducing incentive structures designed to align the objectives of senior management with those of shareholders (Uzun et al., 2004). Finally, the corporate governance committee focuses more specifically on the standards of directors’ qualifications, and their responsibilities, and it enhances director accountability in decision making and leads to better monitoring (Mahoney and Shuman, 2003). Therefore, the configuration of these committees remains relevant to guarantee the quality of governance structures. In theory, compliance with CGC recommendations regarding the board subcommittees should enhance the effectiveness of these committees and improve their monitoring functions, which may lead to reducing the likelihood of financial distress. Therefore, the following hypothesis is formulated: H3: Compliance with CGC recommendations regarding the board subcommittees will reduce the likelihood of financial distress. 5.1.3. Research method 5.1.3.1. Sample and data The sample of this study is composed of the 130 firms listed on the IBEX-35 for the period 2013-2016. The sample size has been proven to have sufficient statistical power in many recent studies using similar analyses (Akkermans et al., 2007; Ahmadi et al., 2018; Neifar and Jarboiu, 2018). In line with previous research, the largest firms are selected because of their representativeness (Goncharov et al., 2006; Albu and Girbina, 2015) and since they present greater agency costs and, therefore, corporate governance mechanisms are expected to be crucial (Karamanou and Vafeas, 2005). In particular, agency conflicts are likely to be especially significant in Spanish firms, due to their special characteristics in relation to corporate governance, such as concentrated ownership and control, widespread 101 ownership by outside directors, and a system based on a unitary board structure that is strongly dominated by the controlling shareholders (Acero and Alcalde, 2013; Manzaneque et al., 2016a). These characteristics make the Spanish context an interesting scenario to understand the role of corporate governance structures in safeguarding the interests of all shareholders and stakeholders (La Porta et al., 1999; Manzaneque et al., 2016b). The data needed to calculate the variables concerning CGC compliance were extracted from the Spencer Stuart Index report, which provides information about the compliance of every recommendation in the Spanish CGC. The financial data used to compute variables about financial distress and other control variables were obtained from the SABI database and companies’ annual accounts. 5.1.3.2. Variables The dependent variable is financial distress of firms. Consistent with the approach used by recent studies (Pindado et al., 2008; Manzaneque et al., 2016b), the measurement for financial distress (FD) was calculated as a binary variable that takes the value 1 if the company meets the following conditions, and 0 if not: (1) its earnings before interest and taxes depreciation and amortization (EBITDA) are lower than its financial expenses for two consecutive years; (2) a fall in its market value occurs between two consecutive periods 8 . This ex-ante approach is especially advantageous since it allows problems of ex-post business failure approaches to be overcome by considering crisis situations other than bankruptcy (Grice and Dugan, 2001). In order to increase the robustness of this empirical study, a sensitivity analysis is performed by alternatively employing a continuous variable based on the Zmijewski score (Zmscore). The Zimijeski model (Zmijewski, 1984) has been extensively used in recent studies on business failure (Tykvová and Borell, 2012; Richardson et al., 2015b; Lee et al., 2017) due to its high capacity to predict 8 Other previous studies on business failure have also employed this proxy. A major review can be found in Manzaneque (2006). 102 financial distress in comparison with older models (Husein and Pambekti, 2014). This score is calculated as follows: Zm = -4.336 - 4.513 X1 + 5.679X2 - 0.004 X3 where x1 = net income/total assets, x2 = total debt/total assets, and x3 = current assets/current liabilities. Firms with a value over 0.5 are classified as distressed companies. The main explanatory variables are related to compliance with CGC recommendations. Consistent with previous research (Campbell et al., 2009; Rodríguez-Fernández, 2016), these variables were calculated as the proportion of recommendations fulfilled by firms. Specifically, three variables were considered: (1) overall CGC compliance, measured by the proportion of total recommendations satisfied by a firm (Overall); (2) compliance with the recommendations concerning the board of directors, calculated as the proportion of recommendations related to the board of directors that had been fulfilled by a firm (Board); (3) compliance with the recommendations concerning board subcommittees, measured by the proportion of recommendations strictly related with these subcommittees that had been fulfilled by a firm (Subcommittees). Overall CGC compliance deals with a number of very different issues in addition to the recommendations concerning the board of directors and its committees, such as anti-takeover mechanisms, responsibilities of the general meeting and mechanics of voting. Recommendations about the board of directors also include a variety of topics related to the size and composition of the board, the characteristics of directors, the activity and meetings, among others. Finally, recommendations about board subcommittees include issues strictly related to the structure and functioning of the audit committee, the appointments committee, the compensation committee, and the corporate governance committee. In relation to control variables, and in line with previous literature, several variables which can influence the financial distress of firms were also considered (Parker et al., 2002; Wang and Deng, 2006; Shahwan, 2015; Udin et al., 2017): firm size, leverage, financial performance, profit margin, industry and year. The literature generally argues that the likelihood of financial distress is positively associated with firm leverage and inversely related to firm size, financial performance and profit margin. Additionally, as the financial situation of a firm may differ across industries 103 and across time periods, both variables are also included. Firm size (Size) is measured as the logarithm of total assets. Leverage (Lev) is calculated as the ratio of total debt to total assets. Financial performance (Fperf) is proxied by the ratio return on equity. Profit margin (Pmargin) is computed as net income over net sales. In order to consider the industry (Ind) in which the firm operates, dichotomous variables are created based on the classification of sectors provided by the General Index of the Madrid Stock Exchange. Finally, dummy variables are also calculated to include the years (Year) in the statistical model. Table 6 provides a summary of all the variables and their definitions. 110 Table 8. Study 1. Pearson coefficients (1) (2) (3) (4) (5) (6) (7) (8) (1) FD 1 (2) Zmscore 0.25*** 1 (3) Overall -0.09 -0.29*** 1 (4) Board -0.14 -0.17* 0.64*** 1 (5) Subcommittees -0.04 -0.22** 0.77*** 0.51*** 1 (6) Size -0.04 0.21** 0.15* 0.11 0.13 1 (7) Lev 0.22** 0.56*** -0.10 -0.07 -0.05 0.41*** 1 (8) Fperf -0.22** -0.47*** -0.04 0.03 -0.01 -0.06 -0.07 1 (9) Pmargin -0.16* -0.47*** 0.06 -0.02 0.02 -0.02 -0.26*** 0.35*** FD is a dummy which takes the value 1 if: (1) its EBITDA are lower than its financial expenses for two consecutive years; (2) a fall in its market value occurs between two consecutive periods; Zmscore refers to the Zmijewski score; Overall indicates the proportion of recommendations contained in CGC and fulfilled by firms; Board applies to the proportion of recommendations about the board of directors contained in CGC and fulfilled by firms; Subcommittees relates to the proportion of recommendations about the board subcommittees contained in CGC and fulfilled by firms; Size is calculated as the Logarithm of total assets; Lev refers to the ratio of total debt to total assets; Fperf denotes ratio return on equity; and Pmargin is calculated by net income over net sales.* p-value < 0.1; ** p-value < 0.05; *** p-value < 0.01. Source: own elaboration 111 5.1.4.2. Multivariate analysis The results from the logistic regression analysis panel are reported in Table 9. In Model 1, the relationship between overall CGC compliance (Overall) and the likelihood of financial distress (FD) is tested. In Models 2 and 3, the main explanatory variables refer to the compliance of recommendations regarding the board of directors (Board) and its subcommittees (Subcommittees), respectively. All the models are statistically significant and include the control variables commented. The results fail to find any significant relationship between the likelihood of financial distress and either overall CGC compliance or CGC compliance about subcommittee recommendations. Nonetheless, compliance with CGC recommendations in boards of directors is significantly related to the probability of financial distress. As predicted in hypothesis H2, boards that comply with CGC recommendations seem to be in a better position to carry out their monitoring function effectively and therefore to better prevent business failures. As regards the control variables, and in line with previous studies (Shahwan, 2015; Udin et al., 2017), the likelihood of financial distress is documented to increase for greater levels of leverage and to decrease for higher levels of firm performance in all the models. 112 Table 9. Study 1. Logistic regression analysis Dependent variable: FD Model 1 Model 2 Model 3 Overall -0.004 (-0.16) Board -0.05* (-1.80) Subcommittees 0.003 (0.16) Size -0.62 (-1.32) -0.69 (-1.48) -0.62 (-1.32) Lev 2.92*** (2.83) 2.96*** (2.82) 2.92*** (2.84) Fperf -0.023** (-2.46) -0.03*** (-2.57) -0.03** (-2.47) Pmargin -0.03 (-0.43) -0.06 (-0.71) -0.03 (-0.40) Ind included included included Year included included included Wald 22.36* 24.14** 22.46* FD is a dummy which takes the value 1 if: (1) its EBITDA are lower than its financial expenses for two consecutive years; (2) a fall in its market value occurs between two consecutive periods; Zmscore refers to the Zmijewski score; Overall indicates the proportion of recommendations contained in CGC and fulfilled by firms; Board applies to the proportion of recommendations about the board of directors contained in CGC and fulfilled by firms; Subcommittees relates to the proportion of recommendations about the board subcommittees contained in CGC and fulfilled by firms; Size is calculated as the Logarithm of total assets; Lev refers to the ratio of total debt to total assets; Fperf denotes ratio return on equity; and Pmargin is calculated by net income over net sales. In this table, results from logit regression analysis of the mode: FDit =  + β1Overall it + β2Board it+ β3Subcommittees it + β4 Size it+ β5Lev it+ β6 Fperf it + β7 Pmargin it + β8 Indit-1 + β9 Yearit-1 are reported. * p-value < 0.1; ** p-value < 0.05; *** p-value < 0.01. Wald refers to a statistical significance test for the model in logistic regressions. Source: own elaboration 113 In the next stage of the empirical analysis, Table 10 reports the results from the twostage procedure (2SLS), where the analysis is performed by using the instrumental variables to predict the level of CGC compliance. The instruments comply with the theoretical requirements, and the Sargan test also confirms their validity. The literature assumes that p-values over 0.10 signal that the instruments are uncorrelated with the error term (García-Castro et al., 2010). Results of the F-test show that the regression models are all statistically significant at the levels of 0.001 and 0.05. This analysis confirms that the likelihood of financial distress, measured by the Zmijewski score, is also significantly associated with compliance with CGC recommendations concerning the board of directors. In particular, the adjusted R2 for Model 2 is 0.250, suggesting that it explains an important variance in the dependent variable. Furthermore, the results again show the lack of association between financial distress and both the level of overall CGC compliance and compliance of the recommendations regarding board subcommittees. In addition to leverage and firm performance, the profit margin appears to be significantly associated with the financial distress of firms. 114 Table 10. Study 1. 2SLS analysis Dependent variable: Zmscore Model 1 Model 2 Model 3 Global -22.73 (-1.09) Board -16.37* (-1.73) Subcommittees -17.29 (-1.53) Size 109.25 (0.56) 25.09 (0.25) 5.59 (0.05) Lev -41.19 (-0.11) 281.23*** (3.16) 123.24 (0.69) Fperf -0.38* (-1.67) -0.34** (-2.27) -0.55*** (-3.23) Pmargin -8.09 (-1.17) -6.85* (-1.80) -3.08 (-1.00) Ind included included included Year included included included Sargan test 0.85 0.83 0.76 Adjusted R2 0.22 0.25 0.06 F test 2.42** 6.15** 4.76*** FD is a dummy which takes the value 1 if: (1) its EBITDA are lower than its financial expenses for two consecutive years; (2) a fall in its market value occurs between two consecutive periods; Zmscore refers to the Zmijewski score; Overall indicates the proportion of recommendations contained in CGC and fulfilled by firms; Board applies to the proportion of recommendations about the board of directors contained in CGC and fulfilled by firms; Subcommittees relates to the proportion of recommendations about the board subcommittees contained in CGC and fulfilled by firms; Size is calculated as the Logarithm of total assets; Lev refers to the ratio of total debt to total assets; Fperf denotes ratio return on equity; and Pmargin is calculated by net income over net sales. In this table, results from logit regression analysis of the mode: Zmscoreit =  + β1Overall it + β2Board it+ β3Subcommittees it + β4 Size it+ β5Lev it+ β6 Fperf it + β7 Pmargin it + β8 Indit-1 + β9 Yearit-1. are reported. * p-value < 0.1; ** p-value < 0.05; *** p-value < 0.01. Source: own elaboration 115 5.1.4.3. Discussion This study provides new empirical evidence concerning the effect of CGC compliance on the likelihood of firms’ financial distress. The relationship between corporate governance and financial distress is a matter of interest to different stakeholders, and yet there is no empirical evidence regarding the impact of CGC on this issue. This study focuses on the Spanish context, where the special characteristics of corporate governance are likely to raise serious agency conflicts. Examining the relationship between corporate governance and companies’ financial distress thus provides relevant evidence for this type of context, where overall analysis of this issue is still lacking. Unlike previous studies, three different levels of CGC compliance were examined: overall CGC compliance, compliance with recommendations about the board of directors, and compliance with recommendations about board subcommittees. The results find that the probability of financial distress may be reduced for higher levels of compliance with recommendations regarding the board of directors. However, the other measures of CGC compliance appear unrelated with the proxies employed for financial distress. This evidence specifically extends the literature on corporate governance and business failure. First, over the last few years, the proliferation of CGC has led to increasing research on the effects of their compliance. While many studies have focused on the impact that CGC compliance could have on investors’ perceptions (Hooghiemstra and van Ees, 2011; McCahery et al., 2016; Kaspereit et al., 2017), some recent studies suggest that fulfillment of CGC recommendations may also have economic and financial effects and call for further research on this issue (Stiglbauer and Velte, 2014; Haji and Mubaraq, 2015; Cuomo et al., 2016; Rose, 2016). These results complement those studies and help to explain the previous mixed evidence, thereby providing an important contribution to the existing literature. On the one hand, the results highlight that CGC compliance is not only an ethical issue, as it may also help to prevent financial distress, which is a crucial firm outcome (Boubaker et al., 2018). On the other hand, they show that analysis of CGC compliance must take into consideration the differences between the recommendations contained in CGC. Specifically, one of the novelties of this study is the design of three measures for CGC compliance based on types of 116 recommendations common to all the existing CGC around the world. In particular, overall CGC compliance fails to add value in terms of mitigating the financial distress of companies. This finding leads to support the notion that not all corporate governance recommendations might have an impact on firm strategies, which may help to explain the lack of association between CGC and firm outcomes documented in several previous studies (Bassen et al., 2009; Jain et al., 2011; Steger and Stiglbauer, 2011). Thus, hypothesis H1 cannot be accepted. Furthermore, the results fail to find a relationship between compliance with CGC recommendations about board subcommittees and financial distress, thus rejecting hypothesis H3. Only compliance with CGC recommendations regarding the board of directors leads to a reduction in the likelihood of financial distress. As a result, hypothesis H2 can be accepted. In line with previous research, the findings suggest that despite the specific functions assigned to board subcommittees, the board of directors is in charge of monitoring major strategic decisions to detect business failure (Fich and Slezak, 2008; Manzaneque et al., 2016a). Thus, these findings clarify the effects of boards. This is also an important contribution for the literature addressing boards of directors, which has traditionally highlighted the role played by boards in mitigating agency conflicts (Hillman and Dalziel, 2003). Furthermore, the study also extends previous literature on the determinants of financial distress, with only a few studies having documented that governance structures, such as the level of independence or ownership structure can influence the likelihood of financial distress (Salloum et al., 2013; Baklouti et al., 2016; Manzaneque et al., 2016b; Udin et al., 2017). This empirical research extends this branch of literature by shedding some light on the impact of CGC compliance. 117 5.2. Corporate governance code compliance and environmental, social and governance disclosures 5.2.1. Objective This study aims to contribute to the previous literature and to cover two important gaps related to research on CGC (Cuomo et al., 2016). The first relates to the need for a more detailed analysis of compliance with these codes. The second suggests the importance of studying the effect of compliance with the recommendations contained in the CGC on new and relevant corporate strategies. To this end, this study performs an in-depth analysis of CGC compliance, focusing not only on compliance at the global level, but also on compliance with recommendations specifically related to the board of directors, and with those associated with the different board committees. The above means that this research considers three different levels of CGC compliance. Furthermore, an additional contribution of this study derives from analysing the relationship between the different levels of CGC compliance and the disclosure practices of environmental, social, and corporate governance (ESG) information. As commented in Chapter 1, ESG disclosure remains important for two reasons. The first is because information disclosure is a crucial aspect within corporate strategies, and several studies highlight that improved corporate governance mechanisms should lead to better oversight of company disclosure processes and a reduction in information asymmetries in capital markets (Cerbioni and Parbonetti, 2007). In this regard, ESG disclosure has become a critical aspect of business strategy (Lokuwaduge and Heenetigala, 2017). The second reason relates to the increasing importance that ESG reporting has received from regulators all over the world. 5.2.2. Hypothesis development Corporate governance is the set of rules, principles, and procedures that regulate the structure and functioning of the governing bodies of a company. It establishes the relationships between the board of directors by setting the rules that will guide the decision-making process (CNMV, 2017). Thus, corporate governance mechanisms are crucial to ensure stakeholder rights and efficiency in capital markets (Shleifer and Vishny, 1997). For this reason, since the 1990s there has been 118 a succession of international CGC, which contain a wide range of recommendations aimed at strengthening the rights of investors and other stakeholders (Acero and Alcalde, 2010). As commented in the previous sections, several CGC have been published in Spain: the Olivencia Report (1998), the Aldama Report (2003), the Conthe Report— distinguished as the Unified Corporate Governance Code of Listed Companies (2006)—, the Good Governance Code of Listed Companies (2015), and the Good Governance Code of Listed Companies, revised in 2020. These codes comprise a number of recommendations on corporate governance mechanisms gathered in different blocks. The structure of the diverse codes is not identical, but all of them consistently contain particular recommendations related to boards of directors, as well as other specific recommendations for certain board subcommittees (mainly the audit committee). For that reason, in order to provide more precise empirical evidence, this study analyses not only compliance with the recommendations contained in the codes at a global level, but also both compliance with the recommendations of the board and those related to board subcommittees. Table 11 shows how the recommendations are structured in Spanish CGC (2015). As can be observed, the Code is made up of 64 recommendations, 11 of which relate to general aspects and questions about the General Shareholders' Meeting, with 25 focusing exclusively on the board of directors, and 28 relating to board committees. Table 11. Study 2. Block of recommendations Block Number of recommendations (64) 1. General aspects 5 2. General shareholders’ meeting 6 3. Board of directors 25 4. Board subcommittees 28 Source: own elaboration From an agency theory perspective (Jensen and Meckling, 1976), compliance with recommendations, both overall compliance and at the level of the recommendations 119 related to the board of directors and its committees, should help to protect stakeholders’ interests as well as reduce agency costs and information asymmetries. Concerning this issue, previous literature has pointed out that disclosure of ESG information is an essential tool in reducing agency problems (Cheng et al., 2014). In line with the Triple Bottom Line (Elkington, 1994), organizational performance should not be limited exclusively to financial performance but should also consider social and environmental performance (Visser et al., 2007). Consequently, ESG information has acquired great relevance in recent years, becoming a research topic, since the demand for this type of information is increasing among investors and other stakeholders (Helfaya and Moussa, 2017). In particular, numerous analysts and investors include this type of information when issuing opinions or carrying out business valuation models, which is why ESG information has become a critical tool to reduce informational asymmetries in capital markets (Ioannou and Serafeim, 2017). Previous research has investigated the association between certain corporate governance attributes and the disclosure of social information (Majumder et al., 2017; Ullah et al., 2019). Nonetheless, despite the significance and timeliness of ESG information for capital markets, previous literature has not yet explored the possible association between CGC compliance and ESG disclosure practices. In consequence, this research aims to contribute to our knowledge about the effects of corporate governance techniques, providing a new approach, analysing whether overall compliance with CGC recommendations as well as recommendation compliance at both the board of directors and board subcommittee levels may impact ESG disclosure. First, overall compliance with CGC recommendations should involve strong corporate governance structures (Cortés and Martos, 2017). In particular, companies with better corporate governance structures are likely to implement more effective strategies to protect stakeholders’ interests (Hillman and Dalziel, 2003). There is a consensus in previous literature on the positive effect that corporate governance mechanisms have on the quality of information disclosure practices (Carcello et al., 2011). In line with the above reasons, and given the relevance of ESG information in reducing agency conflicts, better corporate 126 Table 13. Study 2. Descriptive statistics Variable Mean Median Standard Dev. Min Max ESG info 23.48 24.20 4.88 3.10 32.3 C-General 91.18 93.54 9.09 50.25 100 C-Board 90.75 94.44 9.07 61.90 100 C-Committees 90.91 94.12 9.31 53.57 100 SIZE 7.05 6.85 0.81 5.67 8.70 IND 0.63 0.69 0.30 0.05 2.00 PROF 4.55 2.53 14.19 -97.05 63.10 ZMIJ 0.15 0 0.36 0 1 ESG info refers to quality of information on environmental, social, and corporate governance issues; C-General applies to the proportion of recommendations contained in CGC fulfilled by firms; C-Board indicates the proportion of recommendations about the board of directors contained in CGC fulfilled by firms; C- Committees relates the proportion of recommendations about board of director committees contained in CGC fulfilled by firms; SIZE is calculated by the total assets logarithm; IND is determined as other liabilities/ Total assets; PROF is proxied by the ratio EBIT/ Total assets; and ZMIJ refers to firm financial situation and is measured by a dichotomous variable which takes the value 1 if the company has a high probability of bankruptcy, and 0 otherwise, based on the Zmijewski score (1984). * p-value < 0.1; ** p-value < 0.05; *** p-value < 0.01. Source: own elaboration Table 14 presents the correlations between each of the variables that make up the statistical models. This table aims to show individual relationships between the variables in the study and to rule out the existence of multicollinearity problems derived from particularly highly significant relationships. First, there is a positive association between the disclosure of ESG information and CGC compliance, although it is only significant in terms of fulfilling the specific recommendations of the board of directors (significant at 10%). Consequently, compliance with the recommendations concerning the board of directors and companies’ social commitment is associated in terms of ESG information disclosure. In addition, a positive and significant association is observed between the disclosure of ESG 127 information and company size. Companies with the best disclosure practices are therefore those which may have greater visibility and come under greater social pressure because of their size. As expected, the relationship between the degrees of compliance with the different levels of recommendations is positive and significant, indicating that companies which tend to comply with more recommendations at a general level also do so with the specific recommendations regarding the board of directors and its committees. In general, none of the correlations between the independent variables exceeds 0.700 (except the association between the variables C-Committees  C-General, which would not enter jointly in any model), confirming that there are no multicollinearity problems in the econometric models used in the empirical analysis (Cooper and Schindler, 2003). 128 Table 14. Study 2. Correlation matrix (1) (2) (3) (4) (5) (6) (7) (1) ESG info 1.00 (2) C-General 0.01 1.00 (3) C-Board 0.16* 0.64** 1.00 (4) C-Committees 0.04 0.77*** 0.51*** 1.00 (5) SIZE 0.33*** 0.15* 0.11 0.13 1.00 (6) IND 0.04 -0.10 -0.07 -0.05 0.44*** 1.00 (7) PROF 0.01 0.32*** 0.22** 0.25*** -0.19** -0.43*** 1.00 (8) ZMIJ -0.13 -0.09 0.01 -0.06 -0.10 0.28*** -0.42*** ESG info refers to quality of information on environmental, social, and corporate governance issues; C-General applies to the proportion of recommendations contained in CGC fulfilled by firms; C-Board indicates the proportion of recommendations about the board of directors contained in CGC fulfilled by firms; C- Committees relates the proportion of recommendations about the board of directors committees contained in CGC fulfilled by firms; SIZE is calculated by the total assets logarithm; IND is determined as other liabilities/ Total assets; PROF is proxied by the ratio EBIT/ Total assets; and ZMIJ refers to firm financial situation and is measured by a dichotomous variable which takes the value 1 if the company has a high probability of bankruptcy, and 0 otherwise, based on the Zmijewski score (1984). * p-value < 0.1; ** p-value < 0.05; *** p-value < 0.01. Source: own elaboration 129 5.2.4.2. Regression analysis Table 15 displays the results of regression analysis using panel data and includes information for each of the models presented in the previous section. It shows the associations between the disclosure of ESG information and the independent and control variables. Model 1 includes compliance with the recommendations at the global level contained in the CGC as the main explanatory variable, while in models 2 and 3, the explanatory variables are compliance with the recommendations relating to the board of directors and its committees, respectively. First, all the models studied are statistically significant and have an adequate explanatory capacity. Column 1 offers the results of model 1, which analyses the effect of compliance with the global recommendations contained in the CGC on the voluntary disclosure of ESG information, showing that there is no relationship between the analysed variables. Therefore, compliance with the recommendations contained in the CBC at a global level does not lead to an improvement in the disclosure of this information, such that hypothesis H1 is rejected. These results are in line with other studies which indicate that CGC compliance does not have a direct impact on corporate decisions (Stiglbauer and Velte, 2014), and that compliance with CGC recommendations could be more related to moral and reputational issues (Rhode and Packel, 2014). Nevertheless, the results do not support the theoretical arguments and suggest that while compliance with CGC recommendations at the global level should ensure that the interests of different stakeholders are protected, this has no bearing on specific ESG disclosure policies. Column 2 presents the results concerning model 2. The findings show that compliance with the recommendations relating to the board of directors has a positive impact on ESG disclosure practices, which means that companies who comply with a higher percentage of recommendations related to the board of directors tend to be more committed to society and sustainable development, as far as ESG information disclosure is concerned, thereby supporting hypothesis H2. These results reinforce the evidence found by other studies on the vital role that boards of directors play in voluntary disclosure practices of this type (Helfaya and Moussa, 2017; Galbreath, 2018). Consequently, the results demonstrate that the effectiveness of CGC may depend on the type of recommendations being evaluated, 130 thereby contributing to the debate concerning the effect of CGC compliance (Cuomo et al., 2016). In line with the theoretical arguments, compliance with the recommendations regarding the board of directors does contribute to reducing agency problems by improving ESG information disclosure practices, which is probably due to the responsibility that boards have in establishing such strategies. Finally, column 3 displays the association between the fulfilment of CGC recommendations related to board subcommittees and ESG disclosure practices. The results of model 3 indicate there is no significant association between the studied variables; therefore, hypothesis H3 is rejected. Although this result is contrary to the theoretical arguments, it reinforces the idea presented above on the role played by boards of directors. This evidence makes it possible to resolve a relevant question in the literature concerning the mechanisms that can improve ESG information practices. Although board committees, specifically the audit committee, can be useful mechanisms for reducing agency conflicts and supervising the implementation of certain corporate strategies, the results show that it is the board of directors which influences policies on ESG disclosure. Regarding the effect of the control variables, it is noteworthy that only compliance with the recommendations of the board of directors and the probability of bankruptcy present a significant and negative relationship, which implies that as the risk of bankruptcy increases, the quality of ESG disclosure decreases. The absence of other relationships may be due to the specificity of the measures used. Previous studies have indicated that when specific information attributes are measured business characteristics may not be determinant of the disclosure of such information (Bravo et al., 2010). 131 Table 15. Study 2. Regression models Model 1 Model 2 Model 3 Constant 25.57*** (2.81) 10.15 (1.05) 24.53*** (2.74) C-General -0.04 (-0.74) C-Board 0.13** (2.35) C-Committees -0.03 (-0.53) SIZE 0.94 (0.83) 0.90 (0.80) 0.93 (0.82) IND -0.48 (-0.22) -0.36 (-0.17) -0.48 (-0.21) PROF 0.01 (0.23) -0.01 (-0.48) 0.00 (0.11) ZMIJ -1.39 (-1.30) -1.79* (-1.71) -1.41 (-1.32) SECT included included included YEAR included included included Adjusted R2 0.30 0.33 0.30 F test 27.53** 32.91*** 27.08** ESG info refers to quality of information on environmental, social, and corporate governance issues; C-General applies to the proportion of recommendations contained in CGC fulfilled by firms; C- Board indicates the proportion of recommendations about the board of directors contained in CGC fulfilled by firms; C- Committees relates the proportion of recommendations about the board of directors committees contained in CGC fulfilled by firms; SIZE is calculated by the total assets logarithm; IND is determined as other liabilities/ Total assets; PROF is proxied by the ratio EBIT/ Total assets; and ZMIJ refers to firm financial situation and is measured by a dichotomous variable which takes the value 1 if the company has a high probability of bankruptcy, and 0 otherwise, based on the Zmijewski score (1984). * p-value < 0.1; ** p-value < 0.05; *** p-value < 0.01. Source: own elaboration 132 5.2.4.3. Discussion According to the agency theory (Jensen and Meckling, 1976), CGC compliance should contribute to protecting stakeholder interests, as well as reducing agency costs and information asymmetries. This study seeks to explore the relationship between compliance with CGC recommendations by Spanish listed companies and the practices of voluntary disclosure of ESG information. These results represent a step forward in the literature on corporate governance, sustainability and CSR. The literature on corporate governance indicates that CGC compliance has an effect on investor perceptions, which can translate into positive market reactions to share values (Goncharov et al., 2006; Chavez and Silva, 2009; Kaspereit et al., 2015; Kaspereit et al., 2017) and an increase in corporate reputation (Hooghiemstra and van Ees, 2011; Rhode and Packel, 2014; McCahery et al., 2016). These studies assume that corporate reputation is the “increase of perception of the different participants of the degree to which the responses of the organization will satisfy the demands and expectations of stakeholders” (Pucheta, 2010), and they consider that CGC compliance improves that perception. Moreover, another branch of studies argues that CGC compliance affects the strategic decisions of companies, impacting business performance. Several studies thus reveal that a greater degree of CGC compliance positively influences different measures of financial performance, such as the market-to-book ratio, Tobin’s q or the financial profitability of companies (Fernández-Rodríguez et al., 2004; Bassen et al., 2006; Stiglbauer, 2010; Luo and Salterio, 2014; Rodríguez- Fernández, 2016). These findings therefore contribute to the literature on the role of CGC by further exploring the consequences that compliance with these codes may have on strategies related to the disclosure of ESG information. In addition, recent studies in CSR and sustainability literature indicate that the disclosure of social information, in particular ESG, may be conditioned by certain business and corporate governance factors. Lee et al. (2016) state that there is a significant negative relationship between the indebtedness of a company and ESG disclosure. They also indicate that firm size and R&D expenditure negatively impact ESG disclosure. Additionally, Arayssi et al. (2016) highlight a negative association between ESG disclosure and financial performance when there are low 133 levels of gender diversity on the board of directors. Yu et al. (2018) suggest that companies with higher levels of assets, greater investment in R&D, and good past financial results tend to be more transparent in ESG disclosure. Helfaya and Moussa (2017) find that the most independent and genderdiverse boards of directors tend to disclose higher quality ESG information. Jizi et al. (2014) highlight the existence of a positive association between board member independence and board size with CSR disclosure. Therefore, this study contributes to the literature regarding the determinants of disclosure of this type of information. To conclude, it should be highlighted that these findings provide new empirical evidence on the relationship between CGC compliance and voluntary ESG disclosure practices. The results show that only compliance with the recommendations relating to the board of directors is associated with voluntary disclosure of ESG information practices. 134 5.3. A critical approach to the true influence of female directors on environmental innovation: when are women greener? 5.3.1. Research objective Due to the environmental excesses of the last few decades, there is increasing pressure on boards of directors to drive sustainable business strategies. Environmental innovation has thus received a great deal of attention as a strategy driven by boards. Analysing the effects of board gender diversity on environmental innovation may be especially pertinent since female representation on boards has been at the core of political, professional and academic discussions worldwide (Cucari et al., 2018; Amorelli and García‐Sánchez, 2020). The objective of this study is to examine the influence of board gender diversity on environmental innovation. This particular objective is twofold. First, we explore whether the effect of board gender diversity is uniform across different levels of both female representation and environmental innovation. First, we examine whether the association between gender diversity and environmental innovation is dependent on the existence of a critical mass of women in the boardroom. This is vital to understand the dynamics of women directors given that, beyond their mere proportion, the number of women may need to reach a specific size if they are to become influential and prevail on boards in terms of fostering environmental decisions (Liu et al., 2014; Manita et al., 2018). Second, a quantile regression is performed to explore whether the effects of gender diversity are contingent upon the level of environmental innovation. This remains timely and relevant since, while linear regression analysis only estimates the conditional mean effects of a response variable, quantile regression has emerged as a key tool to determine whether the influence of directors is different across levels of organizational outcomes (Conyon and He, 2017; Chi et al., 2020). In addition, we examine how certain board characteristics moderate the influence of female directors. Specifically, since environmental innovation requires strong monitoring, this study considers that the impact of women in boards on this type of innovation depends on board characteristics that are likely to affect the monitoring ability: the existence of a CSR committee, board size, board independence, board tenure, and board meetings. Despite being theoretically evident, there is minimal 135 empirical evidence concerning whether, and how, certain board attributes might interact with one another in their links to CSR (Endrikat et al., 2020). In particular, recent research calls for a contextual approach to expand current understanding of how board gender diversity affects environmental decisions (Byron and Post, 2016; Bolourian et al., 2020). 5.3.2. Hypothesis development Environmental innovation refers to the process of creating and transforming product design, minimizing resource consumption and reducing the negative effect on the environment (Nadeem et al., 2020). The deterioration of the environment has led to worldwide attention focusing on firms’ environmental innovations (Barbi et al., 2016; Liao et al., 2019). Concurrently, board gender diversity has received a great deal of interest as a driver of environmental decisions (Bravo and Reguera- Alvarado, 2019; Naciti, 2019). The association between board gender diversity and environmental innovation has so far been examined in France (Galia et al., 2015), the United States (Nadeem et al., 2020) and the Chinese manufacturing industry (He and Jiang, 2019; Liao et al., 2019). These studies tend to argue that female directors may display both a strategic orientation towards environmental decisions and the adequate oversight to effectively develop environmental innovation. Women are likely to be more oriented to stakeholders’ interests and, particularly, more sensitive to their social and environmental concerns (Arayssi et al., 2016). Female directors also tend to be more participative than men, which would probably encourage even greater levels of debate regarding a firm’s stakeholders and the attention paid to environmental decisions (Galbreath, 2018). Moreover, women directors may improve the necessary oversight of environmental decisions since they are expected to enhance board independence and activism (Adams et al., 2011), to be generally better prepared for meetings and improved board deliberations of complex issues (Huse and Solberg, 2006), and to be more diligent as well as approach their responsibilities as directors with greater commitment (Fondas and Sassalos, 2000). Unlike previous studies, this thesis extends prior research by questioning the onesize fits all approach and by examining the specific conditions that determine the actual influence of female directors on environmental innovation. The first 142 Figure 6. Theoretical approach Source: own elaboration 5.3.3. Research method 5.3.3.1. Sample and data The study sample is made up of the firms listed on the FTSE 250 for the period 2013-2018. Selecting a single country minimizes bias in the results, since implementing environmental innovation is strongly conditioned by a country’s regulatory and institutional context (Arena et al., 2018). In this regard, the UK offers a particularly interesting setting for the specific analysis of board gender diversity because firms are subject to high environmental pressures, since the UK is one of the countries with the most environmental policies and accountability in the world (Environmental Performance Index, 2020). To guarantee the comparability and reliability of the results, the design of the variables included in the empirical analysis focuses on three recognised databases. Specifically, the data needed to calculate the dependent variable, environmental innovation, as well as the variables addressing corporate governance, are obtained from Thomson Reuters ASSET4 and Sustainalytics databases. Moreover, financial 143 data has been extracted from the Datastream database. These databases contain key information for a variety of securities markets worldwide, providing auditing data. All the necessary data from these databases are merged and the final sample is composed of an unbalanced setting of 932 firm-observations. This sample is representative, since these firms represent most of the capitalization of the UK stock exchange. 5.3.3.2 Variables Dependent variable: Environmental Innovation Similar to recent research (Arena et al., 2018; Nadeem et al., 2020), this study relies on the measure provided by Thomson Reuters ASSET4 database, which offers auditable information and minimizes the inherent subjectivity associated to other measures, thus enabling further replication and generalization of the results. Specifically, we consider the variable environmental innovation score, which reflects a company’s capacity to reduce environmental costs and burdens for its customers, and thereby create new market opportunities through new environmental technologies and processes or eco-designed products. The environmental innovation score takes values from 0 to 100; the closer to 100, the greater the firm’s environmental innovation. Independent variables and control variables The main explanatory variable, board gender diversity (BGender), is calculated as the proportion of female board members (Galia et al., 2015; Liao et al., 2018). In order to capture female critical mass, several dummy variables are designed (Liu et al., 2014; Manita et al., 2018). Specifically, FEMALE1 equals 1 when a board has at least one female director, and 0 otherwise, FEMALE2 equals 1 when a board has at least two female directors, and 0 otherwise, and FEMALE3 equals 1 when a board has at least three or more female directors, and 0 otherwise. In addition, a set of control variables, considering both board-related and firm-level variables, is also included (Lee and Min, 2015, Arena et al., 2018; Nadeem et al., 2020). As regards board-related variables, which are also the moderating variables in the empirical analysis, the variable regarding CSR Committee (CSR_Committee) is a dummy that equals 1 if the company has this committee, and 0 otherwise; board 144 size (BSize) refers to the number of directors in the boardroom; board independence (BIndep) considers the proportion of independent directors within a board; board tenure (BTenure) indicates the average number of years that directors have been on the board; board meetings (BMeetings) measures how many times the board meets a year. Furthermore, as regards firm-level characteristics, firm size (Asset) is computed as the logarithm of total assets; firm leverage (Leverage) is calculated as the ratio of total debt to total assets; firm financial performance is proxied by the ratio return on assets (ROA). Industry (Sector_CSR) is a dichotomous variable that takes the value of one if the company belongs to a sector intensive in CSR activities following the SIC CODES classification, and 0 otherwise (Alon and Vidovic, 2015; Miras-Rodríguez et al., 2020). Finally, a set of year dummies is used to control for time (Dum_Year). Table 16 summarizes the above information. 145 Table 16. Study 3. Definition of variables Variables Description Environmental Innovation Environmental innovation score. BGender The proportion of female directors on the board. Female1 Dummy variable that takes value 1 if the number of female directors is equal to 1, and 0 otherwise. Female2 Dummy variable that takes value 1 if the number of female directors is equal to 2, and 0 otherwise. Female3 Dummy variable that takes value 1 if the number of female directors is equal to 3 or more, and 0 otherwise. CSR_Committee Dummy variable that takes value 1 if the company has a specific CSR committee, and 0 otherwise. BSize Number of directors in the boardroom. BIndep Proportion of independent board members BTenure Average numbers of years that directors have been on the board BMeetings Number of board meetings per year. Asset Neperian logarithm of total assets Leverage Ratio of total debt to total assets ROA Ratio return on assets Sector_CSR CSR intensive sector Source: own elaboration 5.3.3.3. Model specification The database combines time-series and cross-sectional data to form panel data. Thus, to test the hypotheses formulated, a panel data estimation model is used for 146 the regression analysis where environmental innovation is regressed on independent and control variables. A panel data approach allows us to effectively control for possible unobserved heterogeneity, and within-firm changes are used to explain variations in the dependent variable. The Hausman test is applied to define which estimation model suits our study better: fixed effects (FE) or random effects (RE). The general model used is presented as follows: EnvironmentalInnovationit= β0 + β1BGenderit + β2CSRCommitteeit + β3BSizeit + β4BIndepit+ β5BTenureit+ β6BMeetingsit + β7Assetit + β8Leverageit + β9ROAit + β10Sector_CSRit+ β11Dum_Yearit Specifically, as regards hypothesis H1b, a quantile regression modelling is adopted. This method is a more robust technique to non-normal errors and outliers than traditional regression approaches (Gallego-Álvarez and Ortas, 2017) and permits the results of a regressor to vary from the distinct phases of the distribution (Dang et al., 2018; Chi et al., 2020). It thus enables the main relationship to be analysed depending on the degree of environmental innovation that companies undertake. This study differs from high environmentally innovative firms (those that are from quantile 50 onwards) and low environmentally innovative firms (up to the median). 5.3.4. Results and discussion 5.3.4.1. Descriptive statistics and correlations Table 17 provides the main descriptive statistics. The mean value for environmental innovation is 55.708, with a standard deviation of 26.34. Female directors average almost 22% of total board members, which shows an underrepresentation of women in boards. Although the majority of boards have at least one woman, only 27% of boards have three or more female directors. As regards the control variables, most of the firms in the sample analysed (78%) have a CSR committee. The average board of directors is composed of almost 10 directors, with around 61% of them being independent directors. With regard to tenure, directors have been part of a board for five and a half years on average. Boards meet about eight times a year. 147 Table 17. Study 3. Descriptive statistics Variable Mean Std. Dev. Q1 Median Q3 Environmental Innovation 55.71 26.34 32.85 54.90 78.76 BGender 21.78 10.09 14.29 21.43 28.57 Female 1 0.89 0.31 1 1 1 Female 2 0.53 0.49 0 1 1 Female 3 0.27 0.44 0 0 1 CSR_Committee 0.78 0.42 1 1 1 BSize 9.77 2.53 8 9 11 BIndep 61.29 11.83 53.33 61.54 70 BTenure 5.68 1.93 4.38 5.39 6.52 BMeetings 8.57 2.74 7 8 10 Asset 15.59 2.15 14.09 15.07 16.32 Leverage 0.17 0.15 0.05 0.17 0.26 ROA 0.41 0.26 0.21 0.42 0.62 Sector_CSR 0.36 0.48 0 0 1 Environmental innovation refers to the environmental innovation score; BGender represents the proportion of female directors on the board; Female 1, 2 and 3 refer to dummy variables which take value 1 if the number of female directors is equal to 1/2/3 or more, and 0 otherwise; CSR_Committee is indicated as a dummy which takes value 1 if the company has a specific CSR committee, and 0 otherwise; BSize refers to the number of directors in the boardroom; BIndep shows the proportion of independent board members; BTenure is calculated as the average numbers of years that directors have been on the board; BMeetings considers the number of board meetings a year; Asset is proxied by the neperian logarithm of total assets; Leverage is determined as the ratio of total debt to total assets; ROA is estimated as the ratio return on assets; and Sector_CSR applies to CSR intensive sectors. Source: own elaboration Table 18 reports the correlations and variance inflation factor (VIF) coefficients. Although the correlations only show univariate relations and do not allow any 148 conclusive findings to be drawn, the correlation coefficients are, in general, below 0.7, and the VIF values are all found to be below 5. Multicollinearity can thus be ruled out in this sample (Studenmund, 1997). 149 Table 18. Study 3. Correlation matrix (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (1)Environmental Innovation 1.00 (2) BGender 0.06* 1.00 (3) Female1 0.11*** 0.53*** 1.00 (4) Female2 0.11*** 0.60*** 0.38*** 1.00 (5) Female3 0.17*** 0.56*** 0.21*** 0.57*** 1.00 (6) CSR_ Committee 0.17*** 0.10*** 0.12*** 0.13*** 0.14*** 1.00 (7) BSize 0.26*** 0.04 0.12*** 0.47*** 0.47*** 0.23*** 1.00 (8) BIndep 0.12*** 0.19*** 0.14*** 0.21*** 0.24*** 0.09*** 0.18*** 1.00 (9) BTenure -0.06* -0.18*** -0.18*** -0.16*** -0.15*** -0.06* -0.14** -0.27*** 1.00 (10) BMeetings -0.01 -0.05 0.04 0.00 -0.00 0.02 0.06* 0.14*** -0.16*** 1.00 (11) Asset 0.30*** 0.12*** 0.15*** 0.38*** 0.46*** 0.35*** 0.70*** 0.39*** -0.23*** 0.24*** 1.00 (12) Leverage -0.01 0.00 0.00 0.01 -0.05 0.15*** -0.01 -0.04 -0.02 0.02 -0.05 1.00 (13) ROA -0.15*** -0.04 -0.07** -0.19*** -0.22*** -0.27*** -0.39*** -0.12*** 0.22*** -0.33*** -0.54*** -0.36*** 1.00 (14) Sector_CSR 0.04 -0.09*** -0.05 -0.16*** -0.16*** 0.11*** -0.15*** -0.14*** 0.14*** -0.12*** -0.23*** 0.01 0.29*** VIF 1.21 1.82 1.69 1.26 2.35 1.15 1.31 1.20 3.30 1.32 2.13 See Table 16 for the definition of the dependent, explanatory and control variables. * p-value< 0.1; ** p-value < 0.05; *** p-value < 0.01. Source: own elaboration 150 5.3.4.2. Multivariate analysis Main analysis With regard to hypothesis H1a, the results of the influence of a critical mass of women in boards are reported in Table 19. To that end, in addition to the proportion of female directors, the variables Female1, Female 2, and Female 3 are included in the general model as measures for gender diversity. The findings including the measure for board gender diversity (BGender) confirm that female board representation is positively associated to environmental innovation (column 1). However, female directors have a positive and significant association with environmental innovation only when they reach a critical mass of at least three women. The presence of only one or two women in the board, which is the common case for the majority of the firms, is proven to be insufficient to enhance environmental innovation. Therefore, hypothesis H1a is accepted. 151 Table 19. Study 3. Analysis of a critical mass Variables (1) (2) (3) (4) BGender 0.27*** (0.07) Female1 -2.79 (1.72) Female2 1.08 (1.33) Female3 2.11* (1.29) CSR Committee 6.94*** 8.51*** 8.52*** 8.31*** (1.96) (1.94) (1.94) (1.94) BSize -0.53 -0.62* -0.56 -0.69* (0.37) (0.37) (0.37) (0.38) BIndep 0.12* 0.13** 0.14** 0.14** (0.06) (0.06) (0.06) (0.06) BTenure 1.53*** 1.06** 1.14** 1.29** (0.52) (0.52) (0.52) (0.52) BMeetings 0.17 0.17 0.16 0.19 (0.24) (0.25) (0.24) (0.24) Asset 13.01*** 12.71*** 13.01*** 12.69*** (2.07) (2.09) (2.09) (2.10) Leverage -41.44*** -40.79*** -41.19*** -42.55*** (15.15) (15.32) (15.38) (15.28) ROA -22.87** -20.49* -21.23** -22.59** (10.57) (10.71) (10.73) (10.66) Sector_CSR Yes Yes Yes Yes Year dummies Yes Yes Yes Yes Observations 923 923 923 923 R-squared 0.20 0.19 0.19 0.19 F-test 12.13*** 11.07*** 10.89*** 11.07*** FE/RE FE FE FE FE See Table 16 for the definition of the dependent, explanatory and control variables. *** p<0.01, ** p<0.05, * p<0.1. Source: own elaboration