Amsterdam)Business)School)
Impact of corporate social responsibility on corporate governance
mechanisms
Name: Arlette van der Want Student number: 10883819
Thesis supervisor: Dr. Réka Felleg Date: 19 June 2016
Word count: 11603
MSc Accountancy & Control, specialization Accountancy Faculty of Economics and Business, University of Amsterdam
Statement of Originality
This document is written by student Arlette van der Want who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
Abstract
This paper investigates whether internal corporate governance mechanisms are influenced by CSR quality. Specifically, I argue that a low CSR quality rating in the current year will increase corporate governance quality in the next year as internal corporate governance mechanisms are capable of aligning managements’ interests with those of shareholders and other stakeholders. The mechanisms that are investigated are board size, CEO ownership, the proportion of independent directors on the board, and CEO duality. I expect that CSR quality in the current year will increase board size, decrease CEO duality, increase the proportion of independent directors, and decrease CEO duality in the next year. Findings show no significant association between CSR quality and board size and between CSR quality and the proportion of independent directors on the board. The results do show a positive association between CSR quality and CEO ownership which indicates that a low CSR quality in the current year will decrease CEO ownership in the next year. I also found, contrary to my expectations, a negative association between CSR quality and CEO duality. From an agency theory perspective, this suggests that powerful CEOs may promote CSR and CSR disclosures. The reason that drives CEOs to promote CSR and that explains differences in CSR quality may be cause by different industries because industries’ activities differ in their level of environmental and social impact. However, additional analysis shows that there are no differences between industries. This study contributes to literature as it is the first study that investigates the relation between CSR and corporate governance from the viewpoint that CSR has an impact on how internal corporate governance mechanisms are arranged.
Keywords: Corporate governance, CSR, corporate social responsibility, agency theory, stakeholder theory, board size, CEO ownership, independent directors, CEO duality
Contents
1! Introduction 5!
2! Prior research and hypothesis development 9!
2.1! Corporate social responsibility 9!
2.2! Corporate governance 10!
2.3! Hypothesis development 12!
3! Research methodology 15!
3.1! Sample 15!
3.2! Proxy for CSR quality 15!
3.3! Proxies for corporate governance quality 16!
3.4! Control variables 16!
3.5! Empirical model 18!
4! Results 22!
4.1! Descriptive statistics 22!
4.2! Results and discussion of hypothesis test 25!
4.3! Additional analysis 27!
5! Conclusion and limitations 31!
References 34!
1) Introduction
This study examines whether the pressure for more and better corporate social responsibility (CSR) disclosures influences internal corporate governance mechanisms. The motivation of this study is based on the increased interest in CSR and corporate governance from society, legislation, and research. CSR refers to the responsibility of firms towards society and the environment. Public customers and investors have put more pressure on the organizations’ environmental, social, and ethical responsibilities (Chan, Watson, & Woodliff, 2014; Harjoto & Jo, 2011) Several factor have influenced the emerge of this CSR trend. First, corporate fraud scandals, such as Enron and Worldcom. Second, the recent financial crisis. And third, globalization. Over the last decade, consumers, regulators, analysts, and markets have highlighted their concerns about the impact of organizational operations, activities, and processes on environmental aspects, for example climate change (Peters & Romi, 2013). Many organizations have expanded their businesses outside national borders which makes results in organizations having a growing impact on society and the environment (Chan et al., 2014). Because of the increase in globalization and its growing impact expectations on the responsibility and transparency of firms towards all stakeholders have increased (Chan et al., 2014). Over the last few years CSR reporting has seen a tremendous growth (KPMG, 2013). KPMG states in their report on CSR reporting (2013) that more than half of the organizations in all sectors that they investigated disclose information about CSR activities. In this report, KPMG suggest that CSR reporting has become a standard and that organizations should no long question whether or not they should publish such a report. According to Adam & Zutshi (2004) there are two main reason that caused the increase in CSR reporting: moral responsibility and business interests. However, prior research shows more reasons such as legislation enforcement (Kolk, 2004), stakeholder demand, perceived legitimacy from shareholders or other stakeholders (Adams & Zutshi, 2004), increased firm value, improved internal decision-making and cost-savings.
The increased call for transparency did not only come from this CSR angle (Kolk, 2008). Corporate scandals and he financial crisis did not only result in a deeper focus on CSR, it also highlighted the importance of corporate governance (Harjoto & Jo, 2011). Corporate fraud and the financial crisis exposed weaknesses in the way the board of directors carry out their corporate governance duties (Ferrero-Ferrero, Fernández-Izquierdo, & Muñoz-Torrez, 2013). This increased the accountability requirements for corporate governance (Kolk, 2008). The structure and effects of a good corporate governance has been well examined in prior research. It is argued that a good corporate governance system helps to mitigate agency problems. Specifically,
internal corporate governance mechanisms are designed to align the interests of management with interests of shareholders (Walsh & Seward, 1990). Corporate governance structures are not only reviewed to increase their accountability towards shareholders but also to align their practices with broader social interests. Thus, they are not only reviewed to decrease agency problems with shareholders but with all stakeholders (Chan et al., 2014). It has been argued that a good corporate governance structure does not only influences the quality of financial reporting but also that of non-financial reporting. Firms should not only take responsibility for their impact on shareholders’ wealth but also for their environmental, social, and ethical impacts. Corporate governance has been seen as the way to gear management towards this CSR view (Rossouw, 2005). Firms that pay attention to CSR may see an increase in firm value and engagement in CSR might help to improve internal control systems and decision-making processes (Chan et al., 2014). As CSR will increase firm value shareholders will also be interested in CSR (de Villiers, Naiker, & van Staden, 2011).
Because of these developments, there has been an increase in research related to CSR and corporate governance and, specifically, the relation between these two as it has been viewed that corporate governance is able to steer management towards CSR (Chan et al., 2014). Research has been done on the impact of corporate governance characteristics such as board size, CEO ownership, institutional shareholder ownership, and blockholder ownership on CSR disclosure, and on the impact of firm characteristics on CSR disclosure (see, for example, Cox, Brammer, & Millington, 2004; Jamali, Safiedinne, & Rabbath, 2008; Kolk & Pinkse, 2010; Harjoto & Jo, 2011; Jo & Harjoto, 2012; Chan et al., 2014). Prior research has found mixed evidence on the relationship between CSR and corporate governance. What motivates organizations to engage in CSR practices still is a question for future research (Holder-Webb, Cohen, Nath, & Wood, 2009).
In this paper I will respond to calls for more research on the relationship between CSR and corporate governance. I investigate whether the pressure for more transparency on firms’ environmental and social activities affects corporate governance characteristics for US listed firms under the assumption that corporate governance is used as a tool to steer organizations towards CS. The corporate governance characteristics that are used are board size Cheng & Courtenay, 2006; Buniamin, Alrazi, Johari, & Rahman, 2008; Jizi, Salama, Dixon, & Stratling, 2014; Majeed, Aziz, & Saleem, 2015), CEO ownership (Eng & Mak, 2003; Buniamin et al., 2008; Cheng & Courtenay, 2006; Donnelly & Mulcahy, 2008) proportion of independent directors on the board (Ho & Wong, 2001; Eng & Mak, 2003; Buniamin et al., 2008; Post, Rahman, & Rubow, 2011; Majeed et al., 2015) and CEO duality (Ho & Wong, 2001; Haniffa & Cooke, 2002;
Cheng and Courtenay, 2006: Donnelly & Mulcahy, 2008; Michelon & Parbonetti, 2012). By using stakeholder theory and agency theory as a framework, I propose that a low CSR quality rating in a year will increase corporate governance quality in the next year. In particular, I hypothesize that a low CSR quality in the current year increases board size, decreases CEO ownership, increases the proportion of independent directors, and decreases CEO duality in the next year.
Multiple regression analysis has been used to examine the associations between CSR quality and the corporate governance mechanisms. After controlling for the effects of firm size profitability, industry, and leverage/creditor power, findings show insignificant associations between CSR quality and board size and the proportion of independent directors on the board. The results do show a significant positive association between CSR quality and CEO ownership, and a significant negative association between CSR quality and CEO duality. The latter was not expected. A possible explanation for this finding is that powerful CEOs might promote CSR activities and disclosure.
My study contributes to literature in the following way. Despite the large amount of literature on the relation between CSR and corporate governance this relation is still undecided as previous literature shows mixed results. Most of this previous studies focus on the impact of corporate governance mechanisms or firm characteristics on CSR quality or the amount of CSR disclosure. However, no study has been done before that solely focuses on the impact of CSR quality on corporate governance characteristics. My study contributes to previous literature as I do not look at the impact of corporate governance on CSR but at whether CSR quality impacts corporate governance characteristics. To my knowledge, only Jo and Harjoto (2012) have looked at this. However, their variables to measure corporate governance are different than those used in my study. They considered mostly external corporate governance mechanisms such as blockholder ownership, number of analysts following a firm, and outside institutional ownership and find no impact of CSR quality on corporate governance mechanisms. This result is not unexpected because external corporate mechanisms are not as easily influenced because no shareholder approval is required and they are already marked-based (Walsh and Seward, 1990). My study will therefore focus on internal corporate governance mechanisms, that are organizationally based and that are designed to align the interests of shareholders and managers (Walsh and Seward, 1990). I believe that, in order to improve CSR quality and to align managements’ interests with environmental ad social interests from shareholders and other stakeholders, internal corporate governance mechanisms will have to improve. To my knowledge, this is the first study that investigates the relation between CSR and corporate governance from this viewpoint and the first that investigates the impact of CSR on internal
corporate governance mechanisms. Additionally, my study provides evidence that CSR quality is associated with some internal corporate governance mechanisms.
The remainder of this paper is organized as follows. Section two introduces CSR and corporate governance and discusses the hypothesis development. In the third section, the research methodology is explained. This section outlines the sample, the measurement of the variables, and the empirical model. The results are analyzed in the fourth section. Finally, section five contains the conclusion.
2) Prior research and hypothesis development
2.1! Corporate social responsibility
Sustainability development has been defined as the way to meet the needs of the current regeneration without jeopardizing the needs of future generations (World Commission Economics and Development, 1987). This idea has been incorporated into corporate performance which is known as the concept of corporate social responsibility (CSR) (Shrivastave & Addas, 2014). Although shareholder value maximization is still a major goal for organizations, the amount of CSR disclosure has seen an increase (Jamali et al., 2008). CSR is related to the social, environmental, and ethical activities of an organization (Harjoto & Jo, 2012). It contains a set of policies, practices, and programs to maximize the positive impacts of organizations’ activities on social and environmental aspects (Jamali et al., 2008). Organizations integrate these policies, practices, and programs in their business operations and in their interactions with stakeholders on a voluntary basis (European Commission, 2001).
Corporate scandals and the financial crisis has increased the pressures from shareholders and other non-investing stakeholders for more transparency in these social and environmental activities (Jamali et al., 2008). Additionally, due to globalization, expectation from society for organizations to act in a social responsible way has increased (Adams & Zutshi, 2004). CSR is based on the principle that organizations have a responsibility to the environment and the society as a whole and not only to shareholders. Because of these developments, also social responsible funds of organizations have significantly increased (Holder-Webb et al., 2009). Adams and Zutshi (2004) found that there are two key factors that cause the increase in CSR, namely moral responsibility, the recognition of organizations that they have wider responsibilities than merely earning money for shareholders, and business interest. However, previous research has found more reasons such as legislation enforcement (Kolk, 2004), stakeholder demand, perceived legitimacy from shareholders or other stakeholders (Adams & Zutshi, 2004), and increased firm value.
The information about social and environmental activities can be communicated through the annual or in a separate report (Adams & Zutshi, 2004). Disclosing this information is seen as a way of communicating the impact of the organization’s activities on environmental and social aspects to particular interest groups within society. KPMG (2013) defines CSR reporting as follows:
“CR reporting is the means by which a business can understand both its exposure to the risks of these changes an its potential to profit from the new commercial opportunities. CSR reporting is the process by which a company can gather and analyze the data it needs to create long term value and resilience to environmental and social change. CR reporting is essential to convince investors that your business has a future beyond the next quarter or the next year.” (p. 9).
Over the last decade, more organizations have reported or disclosed CSR information. Previous studies show that organization can obtain major benefits from demonstrating that they act in a socially responsible way (Chan et al., 2014). Disclosing CSR information is seen as essential for the long-term survival of organizations (Adam & Zuthi, 2004). It can be used to attain a competitive advantage (Majeed, et al., 2015), it can improve corporate image and relations with stakeholders, it leads to better recruitment and retention of employees, it improves internal decision-making and cost-savings, and it improves financial returns (Adams & Zutshi, 2004). Prior research has shown that the disclosure of CSR information has a positive impact on stakeholders’ perceptions of firm performance, firm value, and firm risk. CSR reporting has become a standard and that organizations should no long question whether or not they should publish such a report (KPMG, 2013).
2.2! Corporate governance
Ownership and management of organizations is usually separated. It is the management that makes the decisions in an organization and, thus, manage the organization (Rodrigue, Magnan, & Cho, 2013). The shareholders are the ones that put their investments into the organization and are the owners of the organization. To make sure management makes decisions aligned with shareholder’s interest, corporate governance exists to monitor managements’ behavior. Corporate governance can be explained as “the system by which companies are directed and controlled” (Cadbury, 2008, p. 8). This system includes processes, customs, policies, laws, and institutions to affect how organizations are directed and controlled. Corporate governance defines how the power in organization is utilized and how decisions are monitored and validated to ensure that managerial decisions are made on behalf of the organizations’ shareholders (Donnelly & Mulcahy, 2008; Jamali et al., 2008).
Literature makes a distinction between a narrow view and a broader view for corporate governance. The narrow view defines corporate governance as a system of laws and regulations that the board’s actions are subject to and focusses on management duties towards shareholders
(Jamali et al., 2008; Rodrigue et al., 2013). Due to corporate scandals caused by accounting fraud and lack of oversight, more discussions of corporate governance have emerged that has focused on accountability and transparency of the board of directors (Shrivastava & Addas, 2014). These developments have led to the broader view of corporate governance, which is related to the CSR literature and also incorporates the board’s duties towards other non-investing stakeholders and the environment (Rodrigue et al., 2013). This has changed the way corporate governance is arranged. Nowadays, corporate governance includes wider issues, such as business ethics in the value chains, human rights, bribery and corruption, and climate change (Elkington, 2006). The board of directors is therefore not only responsible for maximizing shareholder wealth but they also have a responsibility towards important employees, suppliers, customers, and communities who are also affected by the organization’s actions (Jamali et al., 2008).
Contributing to previous literature, this study focuses on the broader view of corporate governance. Stakeholder theory and agency theory provide a framework for finding the link between CSR and corporate governance. Stakeholder theory tells us that the board of directors not only has an obligation towards shareholder but also to other important stakeholder groups (Holder-Webb et al., 2009). Agency theory charges the directors of an organization with the responsibility of representing shareholders’ interests (Hillman et al., 2001). The theory explains that managers may act in their own best interest and not in the best interest of their shareholders (Hillman et al., 2001). As managers have more private information than shareholders or other stakeholders, they might abuse their power to misuse the investments from shareholders and stakeholders (Jizi et al., 2014). In general, corporate governance exists to ensure that organizations run according to law and regulation and to align managers’ interests with interests from shareholders. An effective management process is necessary to enhance corporate performance and accountability and not only shareholder value has to be incorporated but also stakeholder value (Jamali et al., 2008). In the perfect world without agency problems, corporate governance will not be necessary. However, because it is the management that makes the decisions in an organization, they might abuse their power to exploit the organization’s shareholders and other stakeholders. Disclosure of CSR information contributes to the decrease of information asymmetry between managers and shareholders as well as with other stakeholders (Jizi et al., 2014). Mandated (financial) disclosure can help firms to communicate with their shareholders but they, according to Healy and Palepu (1995), do not completely solve the problems associated with information asymmetry. A good corporate governance structure helps to establish and reach the objectives and strategies and is able to align the decisions managers make with the interests from shareholders and other stakeholders. When CSR activities are
viewed as a part of meeting shareholder and other stakeholder demands, it can be expected that there is a positive relationship between their power and CSR performance (Chan et al., 2014). 2.3! Hypothesis development
Corporate scandals and the financial crisis have raised questions about the effectiveness of corporate governance mechanism (Raphaelson & Wahlen, 2004). Recent years have seen attempts to strengthen not only corporate governance mechanisms like board behavior, auditor independence, risk management, and remuneration, but also voluntary responsibilities towards all stakeholders in terms of environmental and social aspects have seen an increase in importance (Kolk & Pinkse, 2010). CSR disclosure is one of these mechanisms and has been regarded as a good corporate governance mechanism (Buniamin et al., 2008). Many of the prior research on the impact of the characteristics of the board of directors on the quality of CSR reporting has found that good corporate governance has a positive impact on the quality of CSR reporting. Organizations with a better corporate governance structure seem to be more social and environmental responsible than organizations with poor corporate governance (Chan et al., 2014). However, there has also been research that found no relation between corporate governance and CSR. Additionally, previous literature has also found that corporate governance related to social and environmental activities has a symbolic role (Rodrigue et al., 2013). Also Hillman, et al. (2001) found results that are consistent with this and conclude that CSR reporting is a way of gaining legitimacy.
Even though there is a lot of research that focuses on CSR, the reason why firms engage in CSR and disclose CSR information is still unclear and no agreed-upon rational has been found on the relationship between CSR and corporate governance (Majeed et al., 2015; Jo & Harjoto, 2012). This is because mixed evidence has been found in previous literature regarding the relationship between CSR and corporate governance. Jo and Harjoto (2012) find that corporate governance variables such as ownership concentration, board structure, takeover pressures, institutional investors and security analysts, affect the extent of CSR engagement and that this positively affects the financial performance of firms. Also evidence found by Harjoto and Jo (2011) show that CSR positively influences corporate governance characteristics such as board independence, institutional ownership, and analysts following the firm. Additionally, Cox, Brammer, & Millington (2004) find positive evidence that long term institutional investors influence CSR engagement. However, Donnelly and Mulchay (2008) and Eng and Mak (2003) find evidence contrary to these results. Donnelly and Mulchay (2008) examine the relationship between outside, non-executive directors on the board, CEO/chairman duality, proportion of shares held
in block by institutional investors, managerial ownership, and the extent of voluntary disclosure. They only find evidence that the number of non-executive directors is positively associated with the amount of voluntary disclosure but do not find evidence that shares held by institutional investors is associated with institutional shareholders. They also find that voluntary disclosure is even greater when the chairman is a non-executive, independent director. Eng and Mak (2003), on the other hand, found evidence that an increase in the proportion of independent directors on the board results in a decrease in CSR. Also Buniamin et al. (2008) find different results. They look at similar corporate governance characteristics within Malaysian firms, namely board independence, CEO duality, management ownership and board size and investigate which governance attributes are essential for firms to decide on disclosing CSR information. They find that only board size has a significant relationship with CSR reporting. Ferrero-Ferrero et al. (2015) take a different approach by looking at generational diversity as a corporate governance characteristic and whether this influences CSR performance. Their results suggest that generational diversity positively influences the adoption of sustainable approaches by firms. The main results also show that more generational diversity improves engagement with stakeholders. Also Chan et al. (2008) take a different approach as they base their corporate governance score not on different corporate governance characteristics but on the WHK Horwath report. Their findings indicate that firms with a higher corporate governance rating provide more CSR disclosure. Hillman et al. (2001) look at the link between stakeholder representation on the board and CSR quality. They expected that stakeholder directors would positively affect CSR. However, they did not find significant evidence to support this. Additionally, previous studies also examine the extent of CSR disclosure with firm characteristics like firm size and industry (Holder-Webb et al., 2009; Brammer & Pavelin., 2006).
As explained in the above paragraph, previous research has found mixed evidence on the relation between corporate governance and CSR. The reason why organization engage in CSR activities and why organizations disclose CSR information remains unclear (Holder-Webb et al., 2009). CSR has been of increased interest and importance to both shareholders and other stakeholders. This increased focus could have had an impact on the way corporate governance is arranged. Previous research has focused on the possible influence of corporate governance mechanisms on CSR reporting, but, to my knowledge, no research has been done on the impact of CSR quality on internal corporate governance mechanisms. I believe more research on this relationship is necessary as it remains a question how the board of directors can commit to pressure for any CSR activities when it is a voluntary decision. Also, internal corporate governance mechanisms are able to align managements’ interest with the environmental and
social interests form shareholders. Thus, I propose that CSR quality will influence corporate governance quality because to steer management towards CSR disclosure and to increase CSR quality, corporate governance has to change.
Although previous research has found mixed evidence on which corporate governance mechanisms are related to CSR reporting and quality, where relations where found between the two, they were positive. Therefore, I propose the following hypothesis:
Hypothesis 1: A low quality CSR score in one year will lead to a higher corporate governance quality in the next year.
3) Research methodology
3.1! Sample
To test my hypothesis, I obtain quantitative data from public databases. I extract data from the Compustat database, the MSCI ESG KLD STATS database (formerly KLD and GMI), and the Institutional Shareholders Services database (ISS, formerly called RiskMetrics Governance and Directors database). I use a sample of US firms over the years 2007 to 2014. I used this time period because the minimum year available using the ISS database is 2007 and the maximum data year available in the MSCI ESG KLD STATS database is 2014. I started with gathering data for US firms over the year 2007 to 2014 period using Compustat. This resulted in 100,537 firm year observation from different industries. I decided to include the financial industry in my sample. Although some studies choose not to include this industry in their sample, inclusion of the financial sector will give a more generic result (Ho & Wong, 2001; Eng & Mak, 2003). The sample firm has to be available from the ISS database and CSR information has to be available from the MSCI ESG KLD STATS database. After removing all firms with insufficient data, my final sample is made up of 6,436 firm year observations.
3.2! Proxy for CSR quality
To measure the quality of CSR disclosure I use data obtained from the MSCI ESG KLD STATS database because this is the most widely used database in CSR research for CSR quality (Hillman et al., 2001; Harjoto & Jo 2011; Jo & Harjoto, 2012). It is also regarded as the most appropriate measure of CSR quality (Hillman et al., 2001). This database contains more than 3,000 organizations with various CSR characteristics. The social CSR rating is based on different strengths and concerns over six categories namely community relations, diversity issues, employee relations, environmental performance, human rights, and product issues. In addition to these six categories, the database also has some exclusionary screens (Jo & Harjoto, 2012). These exclusionary screens deal with sociopolitical issues, such as production of alcohol or tobacco, involvement in gambling, military contracting, and nuclear power (Hillman et al., 2001). Following Jo and Harjoto (2012) and Hillman et al. (2001), I do not include these screens in my main test because only concerns ratings and no strength rating are assigned to firms. Thus, I only include those items that are based on their direct relationship to primary stakeholders. These categories include all important stakeholder groups like employees, customers, society, and suppliers. Also not included, again following Jo and Harjoto (2012), is the corporate governance category in the MSCI ESG KLD STATS database because the definition of corporate
governance differs from that of corporate governance in finance. An overview of the strengths and concerns per category can be found in Appendix 1.
In each category, firms are rated on a Likert scale from -2 to +2, where -2 ratings are major concerns, -1 ratings are concerns, 0 ratings are neutral, +1 ratings are strengths, and +2 ratings are major strengths (Hillman et al., 2001). Following Hillman and Kein (2001), I give equal importance to each category and sum up the score of the dimensions to create my variable for CSR. I will calculate the CSR score as the total number of strengths minus the total number of concerns. Additionally, because I want to investigate whether the CSR score in the current year is associated with corporate quality in the next year, I have lagged the CSR score by firm year.
3.3! Proxies for corporate governance quality
To measure corporate governance quality, I use four internal corporate governance characteristics. The following variables are used: board size (Cheng & Courtenay, 2006; Buniamin et al., 2008; Gue; Jizi et al., 2014; Majeed et al., 2015), CEO ownership (Eng & Mak, 2003; Buniamin et al., 2008; Cheng & Courtenay, 2006; Donnelly & Mulcahy, 2008) proportion of independent directors on the board (Ho & Wong, 2001; Eng & Mak, 2003; Buniamin et al., 2008; Post, Rahman, & Rubow, 2011; Majeed et al., 2015) and CEO duality (Ho & Wong, 2001; Haniffa & Cooke, 2002; Cheng and Courtenay, 2006: Donnelly & Mulcahy, 2008; Michelon & Parbonetti, 2012).
All data for these corporate governance variables is gathered from the ISS database. Board size is measured by the number of board members on each firm’s board in the respective year. CEO ownership is measured as the proportion of shares held by the CEO and executive directors. The proportion of independent directors is measured by dividing the total number of independent directors with the total board size. CEO duality is measured with a dummy variable where 1 = CEO duality, the CEO is also the chairman and 0 = no CEO duality, the CEO is not the chairman.
3.4! Control variables
I also included four control variables, of which I expect may be of influence on CSR or corporate governance and which have been widely used in previous research. The first control variable is firms size that often impacts firm behavior in general (Ho & Wong, 2001; Cormier & Magnan, 2003; Eng & Mak, 2003; Buniamin et al., 2008, Donnelly & Mulcahy, 2008; Kolk &
Pinkse, 2010; Harjoto & Jo, 2011), the second control variable is profitability (Ho & Wong, 2001; Eng & Mak, 2003; Kolk & Pinkse, 2010; Harjoto & Jo, 2011), the third control variable is industry (Kolk, 2005; Buniamin et al., 2008; Chan et al., 2014; Harjoto & Jo, 2011), and the fourth control variable is leverage (Jo & Harjoto, 2012; Michelon & Parbonetti, 2012; Jizi et al., 2014; Peters & Romi, 2014).
Firm size is measured by the natural log of total assets and ROA is used as a measurement for profitability. ROA is calculated by dividing net income with total assets. Firms that are bigger become more visible and will therefore feel more pressure to disclose CSR information (Kolk & Pinkse, 2010). Additionally, disclosing CSR information is relatively less costly for large firms than for small firms, and small firms may be less willing to disclose CSR information as their annual report is the main source of information for their competitors (Donelly & Mulcahy, 2008)
Industry is included because in some sectors the CSR trend started somewhat later such as in the financial industry (Kolk, 2005). Some industries have higher environmental and social impacts, in these industries pressures from all stakeholders concerning more transparency on environmental and social activities started earlier. This led that firms in these industries started disclosing CSR information earlier than in other industries (Kolk, 2005). Following Chan et al. (2014), I measured industry based on a 1 to 3 scale, where 1 = low environmental and social impact, 2 = medium environmental and social impact, and 3 = high environmental and social impact. The low environmental and social impact groups include firms that are classified as being active in the financial sector. Firms that are categorized as having a high environmental and social impact are involved in the material sector such as agriculture, forestry, fishing, mining, construction, and manufacturing. All other firms are classified in the second group and include transportation and public utilities, wholesale trade, retail, services, and public administration.
Leverage is measured by dividing total liabilities with shareholders’ equity (Michelon & Parbonetti, 2012). Besides shareholders, creditors are an important financial resource (Chan et al., 2014). Leverage is included because highly leveraged firms may have closer relationships with their creditors and may have more pressure from creditors to disclose CSR information (Khan et al., 2013; Chan et al., 2014; Jizi et al., 2014). Additionally, previous research based on stakeholder theory has found that capital structure is part of the stakeholder strategy (Chan et al., 2014). Creditors can be seen as important stakeholders of a firm.
3.5! Empirical model
The empirical models used to test my hypothesis are as follows:
CORPORATE GOVERNANCEt= CSR_SCOREt-1+ SIZE + ROA + INDUSTRY +
LEV
Where:
CORPORATE GOVERNANCEt= corporate governance quality in period t, where: BOARD_SIZEt = The total directors on the board of directors
CEO_OWNt = CEOownership, proportion of shares held by the CEO INDEPENDENTt = Proportion of independent directors on the board CEO_DUALITYt = CEO and chairman are the same person
CSR_SCOREt-1 = CSR quality score in period t-1
SIZE = Control variable for firm size measured by the natural log of total assets ROA = Control variable for profitability
INDUSTRY = Control variable for industry
LEV = Control variable for leverage/creditor power
In Table 1 an overview is provided with all the variables including description and measurement method. In subsection 2.3 I explained my hypotheses. I expect that a low CSR quality in the current year will increase corporate governance quality in the next year. Corporate governance quality is measured by board size, CEO ownership, the proportion of independent directors and CEO duality. I expect that a low CSR score in the current year will increase board size, decrease CEO ownership, increase the proportion of independent directors, and decrease CEO duality in the next year. The association between CSR are these different proxies are explained below.
A board of directors should be effective in monitoring management and reducing information asymmetry. Smaller groups are often expected to be more effective at this than larger boards (Jizi et al., 2014). However, larger boards have lower workloads than smaller boards, which could increase the effectiveness of larger boards. Additionally, larger boards have a greater number of directors with expertise and is probably also more diverse (Guest, 2009).
Greater group diversity can improve the quality of group decisions. A greater board size is therefore expected to be a better corporate governance quality (Jizi et al., 2014). Thus, it is expected that a low CSR quality score in this year will increase board size in the next year.
CEO ownership is the proportion of ordinary shares held by the CEO and executive directors (Eng & Mak, 2003). Agency problems are higher when CEO ownership is low because managers have less incentives to maximize their job performance but more incentives to act in their own best interests (Eng & Mak, 2003). When CEO ownership decreases, managers will have increased incentives to expand their own benefits (Donnelly & Mulcahy, 2008). It is assumed that CEO ownership motivates managers to behave like shareholders because their incentives are aligned (Donnelly & Mulcahy, 2008). However, Ruland et al. (1990, cited by Donnelly & Mulcahy, 2008) find that CEO ownership is negatively associated with earnings forecast. Additionally, Eng and Mak (2003) also argue this point. They argue and find evidence that CEO ownership is negatively associated with CSR quality as CSR disclosure is seen as a substitute for outside monitoring. It is then expected that a low CSR score this year will decrease CEO ownership.
The proportion of independent directors is measured by dividing the total number of independent directors on the board with the total board size. Independent directors are the directors that are not affiliated or employed by the organization in any way (Majeed et al., 2015). The not affiliated requirement is met when the director does not hold any significant ownership and does not hold an executive position in the company (Buniamin et al., 2008). Values of independent directors are different than those of inside directors as independent directors are less attached to the economic performance of the firm and represent more the interests of shareholders (Post et al., 2011). Independent directors are considered as a tool for monitoring management because they can influence the behavior of managers and take actions when management acts opportunistically (Ho & Wong, 2001; Post et al., 2011). Therefore, they are an important corporate governance tool. The effectiveness of monitoring management behavior will increase, when the proportion of independent directors increases (Ho & Wong, 2001). A higher percentage of independent directors increases financial disclosure quality and reduces information asymmetry at these directors represent the interest of stakeholders (Forker, 1992; Buniamin et al., 2008). It is then expected that CSR quality is negatively associated with the proportion of independent directors on the board as a low CSR quality this year will increase the proportion of independent directors the next year.
CEO duality means that the CEO of the firm is the same person as the chairman of the board of directors. When the CEO also holds the chairman position on the board of directors, accountability of the board will decrease because CEO duality constrains the independence of the board (Michelon & Parbonetti, 2012). CEO duality compromises the monitoring role of the board of directors. CEOs may hide information as this is easier and they may influence board decisions towards their own interests (Haniffa & Cooke, 2002). Previous research has found evidence the monitoring activities of the board are negatively affected by CEO duality and the quality of CSR disclosure (Donnelly & Mulcahy, 2008). Therefore, it is expected that a low CSR score this year will decrease CEO duality the next year and, thus, CSR quality is positively associated with CEO duality.
Table 1
Overview of variables
Variable Definition Measurement
BOARD_SIZEt Total number of board members on the board of directors in period t
Number of board members
CEO_OWNt Percentage of ordinary shares held by the CEO and executive directors in period t
Total shares held by the CEO and executive directors divided by total equity outstanding
INDEPENDENTt Total number of independent directors on the board of directors in period t
Total number of independent directors divided by total number of directors on board
CEO_DUALITYt CEO is also the chairman on the board of directors in period t
1 if CEO is the chairman, 0 if CEO is not the chairman CSR_SCOREt-1 Total CSR quality in period t-1 Total strengths minus total
concerns
SIZE Firm size Natural log of total assets
ROA Profitability Net income divided by total assets
INDUSTRY Industry 1 if low environmental and social
impact, 2 if medium environmental and social impact, 3 if high
environmental and social impact
LEV Leverage/creditor power Total liabilities divided by
4) Results
4.1! Descriptive statistics
Table 2 provides the descriptive statistics for all the dependent, independent, and control variables used in the main test. Note that have been winsorized to exclude extreme outliers that may influence the data and results. Additionally, time dummies have been created to fill in the gap years for each firm. The descriptive statistics provide the following insights. The independent variable, CSR quality rating (CSR_SCORE), has a mean score 0.003 with a minimum rating of -3 and a maximum rating of 5. It can be concluded that the firms included in the sample do not have a high CSR quality. Which may indicate that management interests are not aligned with interests from shareholders and other stakeholders. The mean of board size is 9.5 and the mean of independent directors is 0.780 which indicates that the average board size consists of around 10 members with an average of 78% who are independent directors. This is consistent with previous research. For example, Post et al. (2011) findings show board size with a mean of 9.5 for board size and a mean of 0.59 for the proportion of independent outside directors. The mean of CEO ownership (CEO_OWN) is 0.008 indicating that on average 8% of the common outstanding shares of a firm is in hands of the CEO. The mean of CEO duality is 0.292.
All dependent variables, except CEO duality which is a dummy variable with either a value of “zero” or “one”, the independent variable, and the control variables are checked for normality. I did this by using the skewness and kurtosis test for normality, the sktest. This test conducts one normality test based on skewness and another based on kurtosis. The null-hypothesis of this test is that the variable is normally distributed and is rejected when the p-value exceeds the alpha level of 0.10. The results of this test show p-values lower than the alpha level for board size, independent directors, firm size, and industry. However, this is not regarded as a problem because the multiple regression analysis is not very sensitive to non-normality. Variance inflation factor (VIF) was calculated to detect any multi-collinearity problems. All variables have a VIF value that is lower than 2, which means that multi-collinearity is not an issue in the multiple regression analysis.
Table 3 shows the Pearson correlations between the different variables. The Pearson correlation calculates the separate correlations between the variables used in the main test. Correlations show that CEO ownership (CEO_OWN) is negatively correlated with board size (BOARD_SIZE) at a significance level of 5%. Boards with more members are considered to be
more effective at monitoring management than smaller board sizes (Jizi et al., 2014). This may decrease the proportion of shares held by the CEO. Additionally, a smaller board size may increase CEO ownership to incentive managers not to behave in an opportunistic way as smaller boards are less effective at monitoring management. The proportion of independent directors (INDEPENDENT) is positively correlated with board size at a significance level of 1% which indicates that a higher board size increases the proportion of independent directors on the board. The proportion of independent directors is also negatively correlated with CEO ownership at a significance level of 1%. The same conclusion can be drawn from this correlation as from the correlation between board size and CEO ownership. More independent members on the board of directors is seen as more effective and thus CEO ownership decreases (Ho & Wong, 2001). Furthermore, CEO duality (CEO_DUALITY) is negatively correlated with CEO ownership and positively correlated with the proportion of independent directors, both at a significance level of 1%. CEO duality decreases the effectiveness of the board of directors (Donnelly & Mulcahy, 2008). When the CEO is the same person as the chairman, more independent directors are needed to make sure that the monitoring activities of the board stay effective and that the CEO does not influence board decisions. The quality rating of CSR (CSR_SCORE) is positively correlated with CEO ownership and negatively correlated with CEO duality, both at a significance level of 10%. The first supports the expectation that CSR quality has a positive association with CEO ownership. However, the significant negative correlation between CSR quality and CEO duality was not expected. Overall, the correlations between the variables are not considered to be a problem as they are all relatively weak. However, it can be noted that the correlation between ROA and firm size is relatively high (0.588 at a significance level of 1%). This correlation is explained by the fact that the main element of both variables is total assets.
Table 2
Descriptive Statistics of Selected Variables
n Mean Std. Dev. Min Max
Dependent variables BOARD_SIZE 5,556 9.530 2.137 6.000 14.000 CEO_OWN 5,556 0.008 0.011 0.000 0.055 INDEPENDENT 5,556 0.780 0.107 0.571 0.909 CEO_DUALITY 5,556 0.292 0.459 0.000 1.000 Independent variables CSR_SCORE 6,436 0.003 2.064 -3.000 5.000 Control variables SIZE 6,436 5.373 2.587 0.147 9.683 ROA 6,436 -0.163 0.469 -1.800 0.171 INDUSTRY 6,436 2.253 0.796 1.000 3.000 LEV 6,436 2.486 4.099 0.023 15.183
Note: Variable definitions can be found in Table 1.
Table 3
Correlations between variables
1 2 3 4 5 6 7 8 9 1 BOARD_SIZE 1.000 2 CEO_OWN -0.037** 1.000 3 INDEPENDENT 0.157*** -0.312*** 1.000 4 CEO_DUALITY -0.017 -0.043*** 0.150*** 1.000 5 CSR_SCORE 0.001 0.025* -0.019 -0.031* 1.000 6 SIZE -0.003 -0.037*** -0.004 0.002 -0.029* 1.000 7 ROA -0.010 -0.0023 0.000 0.0173 -0.014 0.588*** 1.000 8 INDUSTRY 0.001 -0.001 -0.011 -0.014 0.017 -0.276*** -0.181*** 1.000 9 LEV -0.001 0.009 0.002 0.011 -0.007 0.213*** 0.051*** -0.484*** 1.000
Note: Variable definitions can be found in Table 1. This table shows Pearson correlations. *** indicates a significance at the 10% significant level, ** indicates a significance at the 5% significant level and *** indicates a significance at the 1% significant level.
4.2! Results and discussion of hypothesis test
Table 4 contains the results of a multivariate regression analysis on the effect of the CSR quality score in the current year on corporate governance quality in the next year using linear regression. Model 1 shows the beta-coefficients and t-statistics of CSR score as the independent variable, board size as the dependent variable and the control variables. The results show a positive association between CSR and board size, which indicates that a lower CSR quality in the current year decreases board size in the next year. However, this association is not significant which is consistent with findings from Cheng and Courtenay (2006) who also find that CSR and board size are not associated. This finding is not consistent with Jizi et al. (2014), Majeed et al. (2014), and Buniamin et al. (2012) who did find significant associations between CSR and board size. Nonetheless, this finding was not expected as I argued that the two variables have a negative association. This would mean that a low CSR score in the current year would decrease the board size in the next year. Previous research has argued that smaller board sizes are more effective because their communication and coordination is more efficient, and board members are more committed. John and Senbet (1998) argue that, the higher monitoring effectiveness of a larger board may not exceed the incremental costs of poorer communication and decision making. Thus, boards that are too large may decrease monitoring effectiveness because of different opinions and incoherent viewpoints (Chen & Courtenay, 2006). This could be a possible explanation for this result.
Model 2 shows there is a significant positive association between the CSR quality and CEO ownership at a significance level of 10% which indicates that a lower quality of CSR is the current year decreases the proportion of total common shares hold by CEO or executive directors. This is consistent with expectations as it is expected that higher CEO ownership reduces the incentives for managers to disclose CSR information and thus is positively associated with CSR quality. This finding is consistent with Eng and Mak (2003) but not consistent with Donnelly et al. (2008) who do not find significant evidence.
In model 3, the proportion of independent directors is regressed on CSR quality and the control variables. The results show a negative relationship between CSR and the proportion of independent directors. This negative association is expected, however, it is not significant. This finding is not consistent with the finding from Post et al. (2011) who did find that CSR quality and independent directors are associated. The finding is consistent with Majeed et al. (2015), Haniffa and Cooke (2002), and Buniamin et al. (2008) who all did not find an association between CSR and the proportion of independent directors.
Lastly, model 4 examines the effect of CSR quality on CEO duality. The results show that CSR is significantly negatively related to CEO duality at a significant level of 5%. This suggest that a low CSR quality in the current year increases CEO duality in the next year. However, this association is not expected. It was expected that CEO duality is positively associated with CSR. This finding is not consistent with results from Ho and Wong (2001), Chen and Courtenay (2006), and Michelon and Parbonetti (2008) who did not find an association between CEO duality and CSR. My finding is also not consistent with Gul and Leung (2004). They find that lower CEO duality is associated with higher CSR, suggesting that CSR quality is better when CSR duality is low. However, my findings are in line with findings from Jizi et al. (2014) and Bear, Rahman, and Post (2010). One possible explanation for this finding comes from stewardship theory where it is perceived that managers’ interests are aligned with those of shareholders and will thus promote CSR and CSR disclosure. In previous literature associated with stewardship theory it has been argued that CEO duality may improve the board of directors’ efficiency because the CEO will able to steer the firm into the right direction to achieve objectives as there is less interference (Haniffa & Cooke, 2002). Another possible explanation may be that powerful CEOs will promote to become more successful and to enhance their own reputation (Jizi et al., 2014). These CEOs may have the incentive to increase CSR disclosure and quality. Additionally, agency theory argues that when markets are competitive corporate governance mechanisms are endogenous (Hermalin & Weisback, 2003). If shareholders and other stakeholders are aware of the powerful position of a CEO they may choose to appoint this CEO as a chairman. From this viewpoint, CEO duality might be used to improve monitoring effectiveness. According to Jizi et al. (2014) it is important to know what drives powerful CEOs to promote CSR because the rationale for disclosing this information is different for CEOs and stakeholders. The CEO’s reason to disclose CSR information may be caused by personal interests, risk management, or by competition and public pressures. As I have controlled for industry and the mean of industry is around 2 indicating that most of the firms in my sample have medium to high environmental and social impact, my finding will probably point towards the competition and public pressures. However, with my study and data I will not be able to give a conclusive answer to this.
In subsection 2.3 I explained my hypothesis. I proposed that a low CSR quality rating in the current year will increase the corporate governance quality in the next year. Specifically, I expected a negative association between CSR quality and board size, a positive association between CSR quality and CEO ownership, a negative association between CSR quality and independent directors, and a positive association between CSR quality and CEO duality. My
findings show contrary association between CSR quality and board board size and between CSR quality and CEO duality and only significant associations between CSR quality and CEO ownership and CEO duality. Therefore, I conclude that H1 is not supported. However, even though I did not find associations between CSR quality and some internal corporate governance mechanisms and even though I found some assocations contrary to my expectations, my findings do suggest that there are associations between CSR quality and corporate governance which implies that CSR does impact the way corporate governance mechanisms are structured, in particular internal corporate governance mechanisms.
Table 4
Multiple regressions of hypothesis test Model 1: CSR on
board size Model 2: CSR on CEO ownership Model 3: CSR on independent
directors Model 4: CSR on CEO duality CSR_SCOREW 0.000 0.000*** -0.001 -0.007** (0.01) (1.76) (-1.42) (-2.36) SIZEW 0.004 -0.000*** -0.000 -0.003 (0.27) (-3.54) (-0.56) (1.56) ROAW -0.064 0.001* 0.0001 0.025 (-0.85) (1.83) (0.20) (1.56) INDUSTRYW -0.013 -0.000 -0.002 -0.005 (-0.29) (-0.11) (-0.83) (-0.60) LEVW -0.004 0.000 -0.000 0.001 (-0.54) (1.12) (-0.16) (0.54) Constant 9.538*** 0.009*** 0.786*** 0.323*** (63.58) (9.78) (103.33) (10.20)
Note: This table shows the regressions between CSR quality and corporate governance variables. Variable definitions can be found in Table 1. * indicates a significance at the 10% significant level, ** indicates a significance at the 5% significant level, and *** indicates a significance at the 1% significant level.
4.3! Additional analysis
In the previous subsection I found a negative association between CSR quality and CEO duality. I argued here that industry may have had an impact on this results and that powerful CEOs are driven by personal interests, risk management, or by competition and public pressures to promote CSR activities and disclosure. In developing my hypothesis, I also suggested that
differences between may impact CSR quality as different industries differ in their environmental and social impacts. Therefore, I expect that firms operating in industries with high environmental and social impact would be more likely to have a higher CSR quality and thus impact on CSR quality on corporate governance quality will be less than for industries with low to medium environmental and social impact.
To examine these different industries, I split up the industry measure in three categories by level of environmental and social impact. This resulted in the following regression models:
CORPORATE GOVERNANCEt= CSR_SCOREt-1+ SIZE + ROA + INDUSTRY1
+ LEV
CORPORATE GOVERNANCEt= CSR_SCOREt-1+ SIZE + ROA + INDUSTRY2
+ LEV
CORPORATE GOVERNANCEt= CSR_SCOREt-1+ SIZE + ROA + INDUSTRY3
+ LEV Where:
INDUSTRY1 = industries with low social and environmental impact INDUSTRY2 = industries with medium social and environmental impact INDUSTRY3 = industries with high social and environmental impact.
Table 5 provides the descriptive statistics of this additional test. These show that the mean of industries categorized as having a low environmental and social impact (INDUSTRY1) is 0.213, the mean of industries characterized as having medium environmental and social impacts (INDUSTRY2) is 0.304, and the industries categorized as having a high environmental and social impact (INDUSTRY3) has a mean of 0.484. This indicates that most firms have medium to high environmental and social impacts.
Table 6 shows the regression results. My main test showed a significant positive association between CSR quality and CEO ownership at a 10% significance level and a significant negative association between CSR quality and CEO duality at a 5% significance level. This additional analysis also shows significant associations between these variables. The regression analysis shows that no large differences between the different industries. These results may suggest that CSR pressure on firms does not differ between industries anymore while it was expected that firms operating in industries with high environmental and social impact would be more likely to disclose information about their environmental and social impacts and thus would
maybe have a higher CSR quality and therefore show different results on the relation between CSR quality and corporate governance quality. For the discovered association between CSR quality and CEO duality this means that it is still not clear what may drive powerful CEOs to promote CSR. Overall, these additional findings confirm again that is still unclear what the relation is between CSR quality and corporate governance quality.
Table 5
Descriptive Statistics of Selected Variables
n Mean Std. Dev. Min Max
Dependent variables BOARD_SIZE 5,556 9.530 2.137 6.000 14.000 CEO_OWN 5,556 0.008 0.011 0.000 0.055 INDEPENDENT 5,556 0.780 0.107 0.571 0.909 CEO_DUALITY 5,556 0.292 0.459 0.000 1.000 Independent variables CSR_SCORE 6,436 0.003 2.064 -3.000 5.000 Control variables SIZE 6,436 5.373 2.587 0.147 9.683 ROA 6,436 -0.163 0.469 -1.800 0.171 INDUSTRY1 6,436 0.213 0.409 0.000 1.000 INDUSTRY2 6,436 0.304 0.460 0.000 1.000 INDUSTRY3 6,436 0.484 0.500 0.000 1.000 LEV 6,436 2.486 4.099 0.023 15.183
Table 6
Multiple regressions of additional analysis Model 1: CSR quality on board
size
Model 2: CSR quality on CEO ownership
Model 3: CSR quality on independent directors
Model 4: CSR quality on CEO duality CSR_SCORE 0.000 0.000 0.000 0.000* 0.000* 0.000* -0.001 -0.001 -0.001 -0.006** -0.007** -0.006** (-0.01) (0.00) (-0.00) (-1.76) (-1.74) (-1.74) (-1.40) (-1.43) (-1.40) (-2.35) (-2.37) (-2.35) SIZE 0.004 0.004 0.004 -0.000*** 0.000*** -0.000*** -0.000 -0.000 -0.000 -0.003 -0.003 -0.003 (-0.30) (-0.30) (-0.31) (-3.52) (-3.50) (-3.51) (-0.47) (-0.47) (-0.49) (-0.97) (-0.97) (-0.98) ROA -0.061 -0.062 -0.063 0.001* 0.001* 0.001* 0.001 0.001 0.001 0.025 0.026 0.026 (-0.81) (-0.82) (-0.84) (-1.83) (-1.78) (-1.82) (-0.24) (-0.27) (-0.31) (-1.58) (-1.62) (-1.64) INDUSTRY1 0.057 -0.000 -0.008 -0.023 (-0.45) (-0.23) (-1.22) (-1.30) INDUSTRY2 -0.003 0.001 0.001 -0.002 (-0.03) (-0.87) (-0.10) (-0.11) INDUSTRY3 -0.035 -0.000 0.005 0.017 (-0.37) (-0.61) (-0.96) (-1.17) LEV -0.003 -0.003 -0.003 0.000 0.000 0.000 0.000 0.000 0.000 0.001 0.001 0.001 (-0.46) (-0.47) (-0.46) (-1.34) (-1.37) (-1.37) (-0.24) (-0.27) (-0.23) (-0.91) (-0.92) (-0.89) Constant 9.492*** 9.505*** 9.520*** 0.009*** 0.009*** 0.010*** 0.782*** 0.781*** 0.779*** 0.313*** 0.309*** 0.300*** (-96.24) (-93.13) (-88.7) (-13.7) (-13.27) (-13.17) (-160.28) (-153.42) (-146.79) (-15.67) (-15.06) (-14.5)
Note: This table shows the regressions between CSR quality and corporate governance variables for the additional test. Variable definitions can be found in Table 1. * indicates a significance at the 10% significant level, ** indicates a significance at the 5% significant level, and *** indicates a significance at the 1% significant level.
5! Conclusion and limitations
Research in corporate social responsibility (CSR) and corporate governance and the association between these has seen an increase. This has resulted from the increased interest in CSR and corporate governance due to financial scandals, the financial crisis, and globalization. Previous research shows mixed results as to how CSR and corporate governance are associated and thus it still remains a question as to why organizations engage in CSR disclosure. This study examines the effect of CSR quality on corporate governance quality. More specifically, I investigate the influence of CSR quality in the current year on corporate governance quality in the next year. In my hypothesis I predict that a low CSR quality in the current year will increase corporate governance quality in the next year. I use four internal corporate governance mechanisms to measure corporate governance as internal corporate governance mechanisms help to align managements’ interests with shareholders and other stakeholders’ interests. The mechanisms used are board size, CEO ownership, the proportion of independent directors, and CEO duality. Specifically, I predict that a low CSR quality in the current year will increase board size in the next year, will decrease CEO ownership in the next year, will increase the proportion of independent directors in the next year and will decrease CEO duality in the next year.
To measure CSR quality, I used a total score of strengths and concerns over the six categories diversity, employee relations, environment, human rights, product, and community from the MSCI ESG KLD STATS database. Based on a sample of US listed firms and after controlling for firm size, profitability, industry and leverage/creditor power, I found the following results. First, I examine the association between CSR quality and board size and find a positive insignificant association between these variables. This implies that a low CSR score in the current year decreases board size in the next year which was not the association that was expected. One possible explanation for this finding is that the diversity in a larger board size does not improve monitoring effectiveness but may even decrease monitoring effectiveness. Larger board sizes may not exceed the costs of poorer communication and decision making. Smaller board sizes are more effective in their communication and coordination is more efficient which implies that they may be better able at communicating CSR activities and disclosure. Next, I examine the association between CSR quality and CEO ownership. I find that CSR quality and CEO ownership have a significant positive association at a significance level of 10%. This is consistent with expectations that a low CSR quality in the current year decreases CEO ownership. This finding is consistent with findings from Eng and Mak (2003). Third, I examine the association between CSR quality and the proportion of independent directors on the board
of directors. I predicted that a low CSR quality in the current year will increase the proportion of independent directors. Findings show a negative association which is consistent with this expectation. However, this finding is insignificant. Lastly, I examine the association between CSR quality and CEO duality. Findings suggest that CSR quality is significantly negatively association with CEO duality at a 5% significance level. As was predicted that a low CSR quality in the current year will decrease CEO duality in the next year, this result is not what I expected. A possible explanation for this finding is that powerful CEOs are driven to promote CSR and CSR disclosure. However, I am not able to establish what drives these CEOs to promote CSR. Future research could explore the reasons of this negative association and explore what drives powerful CEOs to promote CSR and CSR disclosure.
In the additional analysis, I examined the differences between industries as this may explain the found significant negative association between CSR quality and CEO duality, and may explain differences between industries. However, the results of this multiple regression analysis does not show differences between the different industries. This finding may indicate that in the last years, importance given to CSR and pressures from shareholders and stakeholders is not different between the different industries. However, it was expected that the industries with higher environmental and social impact would influence CSR quality more as in these industries pressures for more transparency on environmental and social activities starter earlier.
Based on my predictions on the association between CSR quality and the four internal corporate governance mechanisms, I argue that my hypothesis is not supported because I am not able to find evidence for my predictions. However, as also argued in section 4, it does seem that CSR quality influences corporate governance quality. Hence, future research would be helpful in order to gain a better understanding on the impact of CSR quality on corporate governance mechanisms. Additionally, future research could also help in gaining better understanding on which corporate governance mechanisms are helpful or harmful in protecting all stakeholders’ interests. This could help in providing a better understanding in how corporate governance mechanisms will help aligning management’s interests with those of shareholders and other stakeholders.
While my paper provides additional research into the still unclear relation between CSR and corporate governance and provides insights into why organizations choose to disclose CSR information, it is subject to various limitations. First, the findings are based on a single-country investigation and only US firms are included in the main test. This may imply that results cannot be generalized. Another implication of my study is the sample period. It may be possible that
corporate governance quality is already high as the overhaul of corporate governance systems started in the beginning of the 21st century while my sample period starts in 2007.
