Master Thesis Accountancy
Corporate Social Responsibility
The influence of board composition on CSR decoupling
Fleur Hoogers S2371561 f.hoogers@student.rug.nl University of Groningen June 24, 2019 Supervisor: N. Hussain Word count: 9,187
Abstract: In this thesis, the influence of board composition on firms’ decoupling their CSR disclosure from CSR performance will be examined in light of the agency theory perspective. The required data have been extracted from the Thomson Reuters ASSET4 and Bloomberg database. By using a sample of 15,790 observations from 67 different countries spread over 7 years (2011-2017) the absolute CSR decoupling score has been calculated as the difference between the CSR disclosure score and the CSR performance score. Three corporate governance mechanisms were included in the study, namely board size, board independence and CEO duality. Board size and CEO duality were expected to positively influence CSR decoupling and a negative relation was expected between board independence and CSR decoupling. Through a time fixed regression analysis, it has become apparent that no significant relation between these three corporate governance mechanisms and CSR decoupling could be detected. A possible reason for this finding could be that in the researched data two different forms (overstating and understating) of CSR decoupling are included.
Table of contents
1. Introduction
3
2. Theory & Hypothesis development
7
2.1. CSR decoupling from an agency theory perspective
7
2.2. Hypothesis development
10
3. Methodology
15
4. Results
20
5. Conclusion and discussion
27
1. Introduction
Over the past decades, corporate social responsibility (CSR) received significant interest from investors and the general public because of the growing amount of money involved and especially since corporate scandals like Enron (2004), BP (2010) and Volkswagen (2015). These scandals have given rise to the awareness of corporations to not only satisfy the needs of its shareholders, but also of its stakeholders (Freeman and Reed, 1983). The growing realization that stakeholders expect firms to behave in an ethical and social responsible way, has made firms invest more time and money on CSR activities in order to respond to these external pressures (Hawn and Ioannou, 2016) and to legitimize themselves to their stakeholders (Schons and Steinmeier, 2016). They take on external actions (e.g. branding and disclosure), and internal actions (e.g. providing training and guidelines, forming board committees) to integrate stakeholders’ expectations into firm’s operations and processes (Hawn and Ioannou, 2016; Crilly, Zollo and Hansen, 2012).
Furthermore, due to technology improvements, companies are disclosing CSR information via new channels (e.g. social media), which provides them more opportunities to engage with external audiences and improve transparency (Castillo and Vial, 2016). This development contributes to the reduction of information asymmetry in the CSR field (Du, Bhattacharya and Sen, 2010). Although this seems to be a good development, the added value of and motives behind CSR activities are questionable (Hawn and Ioannou, 2016). Managers could use CSR as a tool to pursuit self-interest by diverting attention from their misbehaving activities (Hemingway and Maclagan, 2004) and might disclose insincere CSR claims to improve their competitive standing (Pope and Waeraas, 2016). Therefore, there are a lot of uncertainties regarding the impact of the policies that enterprises deploy and the realization of their CSR goals.
When taking a closer look at the above mentioned corporate scandals, it is striking that these scandals occurred as a result of companies hiding information and making false claims in their CSR reports (Marquis, Toffel and Zhou, 2016). According to Tashman, Marano and Kostova (2018), most of the CSR reports do not mirror the company’s actual CSR performance, which means that firms engage in “CSR decoupling”. This phenomenon indicates the gap between CSR disclosure and CSR performance. This gap, caused by incomplete and non-transparent CSR reports, is misleading towards stakeholders. In the literature about CSR decoupling, a
distinction is made between “walking” and “talking” CSR (Tashman et al, 2018). Crilly et al. (2012) describe two ways in which CSR decoupling can take place. On the one hand, the intentional overstating of CSR performance as “faking it” and on the other hand, CSR understating as a “muddling through” process, where decoupling emerges as a result of variation within the firm when implementing the corporate-wide policy (Crilly et al., 2012). Marquis, Toffel and Zhou (2016) explored “how, when and why” decoupling occurs, however, in spite of this, CSR decoupling is still underexplored and is seen as a black box. Studying the conditions that lead to decoupling is warranted because stakeholders need to have a realistic view of the firms’ CSR behavior and the reliability of CSR reporting. Furthermore, high-quality CSR reporting will improve CSR performance, because it improves the effectiveness of policy implementation as it makes inconsistencies within the firm visible. Therefore, high-quality reporting will induce firms to behave in a more socially responsible way and CSR performance will consequently improve (Hussain et al., 2018).
Previous studies showed the relationship between the quality of information provided and the various factors that differentiate companies. For example, Greenwood and Hinings (1996) found that there is a higher chance of decoupling in loosely coupled organization fields in which compliance monitoring is low. This finding gave rise to questions about how corporations are governed and how governance mechanisms influence corporate social behavior. The main reason for the lack of transparency in the field of CSR is that the information is provided on a voluntary base by the board of directors (Hussain et al., 2018). Since the directors have a lot of freedom regarding their decision in what information to disclose about their CSR-performance, the effectiveness of the board of directors plays an important role in CSR reporting (Chen and Jaggi, 2000). If the board effectively executes its monitoring function, namely on the disclosure quality, the transparency will be increased (Fama and Jensen, 1983). According to several researchers (e.g. Lim et al, 2007; Andres et al., 2005), the efficiency of the board depends largely on the board composition. However, findings regarding board composition and CSR are conflicting, especially when looking at the difference between CSR disclosure and CSR performance (Abdullah et al., 2011; Bear et al.,2010; Williams, 2003; Hussain et al., 2018). For example, Walls et al. (2012) and Lim and Chow (2007) studied the associations between CSR performance and board independence and found contradicting results. Lim and Chow (2007) reported a positive relation, whereas Walls et al. (2012) found a negative association. These outcomes conflict and, therefore, result in unclear relations between
corporate governance mechanisms and CSR, which limits theory development on this issue. In addition, prior literature shows variation in the used theories, which means that similar variables are tested based on opposing theories. Some studies apply the stakeholder theory, while others use the agency theory, which are contradictory theories (Hussain et al., 2018). This study will use the agency theory as the foundation in answering the research question. The inconsistent results in prior studies prove that there is still much room to explore the relationship between corporate governance mechanisms and CSR (Walls et al. 2012). The board is, from any point of view, an important corporate governance mechanism and its monitoring role has a big impact on organization’s well-being (Zang, 2010).
More detailed studies are required to fill the research gap and therefore, the research question of this study will be: “How are corporate governance mechanisms associated with CSR decoupling?”.
This study complements the limited number of studies on CSR decoupling and contributes to the increase of insight into the relations between different board characteristics and the extent to which companies decouple their CSR disclosure from actual CSR performance. The objective of this study is to obtain associations between CSR decoupling and the following three board characteristics: board size, board independence and CEO duality. Board size is included in the study as an independent variable, since this is a measure of many different characteristics of board composition. The size of the board could provide an indication on the diversity and skills within the board. As board size is included as an overall dimension, it might be interesting to also include one positive dimension and one negative dimension. Therefore, board independence is selected, because existing literature indicates a positive relation between board independence and transparency (Lim et al., 2007; Hussain et al., 2018). CEO duality is added as the third independent variable, because this is considered as impairing CSR disclosure (Giannarakis et al., 2014; Walls et al., 2012).
In order to test the theory, a sample of 15,790 observations of 4,838 firms around the world, covering a period from 2011 – 2017, is used to conduct a regression analysis. The analysis reveals that a larger board has a positive impact on CSR decoupling, which means that the gap between CSR disclosure and CSR performance will increase (defined as hypothesis 1). The association between board independence and CSR decoupling is found negative (hypothesis 2),
which is in line with the positive association of CEO duality and CSR decoupling (hypothesis 3). Although, the results do not show significance, this is still a step forward in exploring the black box “CSR decoupling”.
This study makes four key contributions to existing academic literature. Firstly, most of the research regarding board composition focuses on financial reporting. When it comes to non-financial information like corporate social responsibility reporting, there is still too little knowledge available, which is surprising because board composition can be of great importance for CSR disclosure (Giannarakis, 2014). Secondly, this study contributes to the agency theory by aiming to provide evidence about the influence of board characteristics on management actions, which has influence on the information asymmetry and the agency costs.
Thirdly, previous studies regarding this topic have been conducted in only one or a few countries (Hafsi and Turgut, 2013; Mallin & Michelon, 2011; Villiers et al., 2011). The results of those studies are limited with regard to generalization and therefore more research regarding CSR decoupling and board characteristics in a broader range of countries needs to be performed. To conclude, most literature regarding CSR decoupling only focuses on the environmental dimension of CSR (Arena et al., 2014; Walker and Wan, 2012) and on overstating performance. However, stakeholders are also misinformed through the decoupling of other CSR dimensions, namely environmental, social and governance, or as a result of companies understating their CSR performance (Schons and Steinmeier, 2016; Westphal and Zajac, 1998). Diminishing this gap in the CSR literature could be of great importance, because if CSR does not have a positive impact on society, the whole concept is useless.
In this study, the theoretical background will be discussed, which contains a review of the agency theory and the hypotheses development. Subsequently, the research methodology and variables will be described. The outcomes of the study will be discussed in the results chapter. The study will be concluded with a summary, conclusions, and a discussion of this study’s limitations. Recommendations for future research are also included.
2. Theory & Hypotheses development
2.1 CSR decoupling from an Agency theory perspective
The review of existing literature on CSR decoupling reveals that this phenomenon of decoupling CSR disclosure from actual CSR performance is mostly explained through the legitimacy theory (Boiral, 2013; Cho et al., 2015, Christmann and Taylor, 2006; Kim and Lyon, 2015) and the institutional theory (Crilly et al., 2012; Graafland and Smid, 2016; Jamali et al., 2017; Marquis et al., 2016) as showed in table 1 (page 9). The extant literature focuses mainly on CSR decoupling caused by external stakeholder pressure and the effects of CSR decoupling on the firm’s market value. This study, however, aims to explain the factors that underlie the phenomenon of CSR decoupling. Since CSR decoupling is described in terms of CSR performance and disclosure, it is essential to study these concepts in order to understand the drivers of CSR decoupling. In research regarding CSR performance and disclosure, CSR is primarily considered as an agency problem (Dhaliwal, Li, Zhang, and Yang, 2011; Wang and Coffey, 1992; Lim and Chow, 2007; Walls et al., 2012). According to Jensen and Meckling (1976), the agency theory describes the conflict of interest between management (agent) and shareholders (principal) who each want to pursue their own interests, which leads to agency problems (Fama and Jensen, 1983). The actions of the agent are not always in the best interest of the principal, therefore board members are appointed by the shareholders in order to supervise the management (Jensen and Meckling, 1976).
A key aspect of the agency theory is the lack of full transparency from the agent towards the principal, which leads to information asymmetry (Jensen and Meckling, 1976). In this situation the agent has more information than the principal, which results in an advantage for the agent (An et al., 2011). This concept is applicable to CSR, because the management has more access to information about the CSR activities of a firm than their stakeholders, the outsiders (Cui et al., 2018). For example, information regarding a firm’s sustainability activities can be difficult for an outsider to obtain (Hahn and Kühnen, 2013). The agent could use this information advantage to maximize his own interest, which could be at the principal’s expense. In that case, the agent is willingly withholding information to maximize his own interests (Jensen and Meckling, 1976). As shareholders do not have resources to decently control the managers, they suffer from this information disadvantage.
Corporate governance mechanisms can help resolve agency problems by reducing information asymmetry (Fama and Jensen, 1983). To reduce information asymmetry, firms can communicate CSR performance with their stakeholders by disclosing information. Engaging in CSR reporting will lead to a reduction of the information asymmetry, since CSR reporting is a form of voluntary disclosure where more information is disclosed (Michaels and Grünig, 2017). The board of directors aims to reduce the information asymmetry to ensure that the consequences of management decisions are in line with the expected value creation for the stakeholders.
Currently, firms are also using social media as a communication tool to reach a greater public and to appear more transparent and trustworthy (Du et al., 2010; Hawn and Ioannou, 2016). However, these platforms could also be used by managers to manipulate the stakeholders’ perceptions instead of increasing the firm’s actual performance (Crilly et al., 2012). Firms are recognizing opportunities of marketing and communication CSR strategies, but without actually undertaking CSR actions (Baldassarre and Campo, 2016). Baldassarre and Campo (2016) speak of ‘opaque companies’ when companies ‘appear to be’ socially and environmentally responsible instead of ‘actually being’. The other side of decoupling is called “understating” (Tashman et al, 2019). This is, for example, the situation where managers are deploying their resources on CSR in an inefficient manner, and therefore want to hide the investments for their shareholders to avoid criticism. Furthermore, corporations do not only decouple for intentional reasons. It could also be an outcome of uncoordinated response to the rapidly changing pressure from stakeholders (Weaver, Trevino and Cochran, 1999). When understating CSR performance, firms are undertaking actions regarding CSR, but they are not fully aware of CSR as a marketing tool, which also results in a gap between performance and disclosure (Du et al., 2016). Concluding, both types of CSR decoupling lead to growing information asymmetry and may be perceived as a lack of credibility and transparency (Brown, 2008). In order to increase transparency, companies report about their CSR activities (Cormier et al., 2011). Through more openness in reporting, the conflict of interest between agent and principal can be reduced (Cai et al., 2006). When shareholders have more information, they will have better opportunities to monitor and, if needed, correct managers (Hooghiemstra, 2015). As a result, agency costs can be reduced, due to the fact that shareholders have to exert less effort in monitoring the company (Healy and Palepu, 2001). The disclosure of social and sustainability information turns private information into public information. A higher quality of reporting reduces the information asymmetry and thus the agency problem.
Since CSR decoupling increases information asymmetry, the corporate governance functions are of a large necessity in order to reduce this problem. Through its function to monitor the actions of the directors and to ensure that the interests of the shareholders are pursued, a board can be seen as an essential governance mechanism. A board that monitors closely reduces the possibilities of selfish and opportunistic behavior by the management and will therefore resolve the agency problem (Fama and Jensen, 1983). The board should monitor the agent in a way that the agent will be transparent, human and responsible towards not only the shareholders, but also towards the stakeholders. Furthermore, the board must act in an effective way to hold agents accountable for their behavior (Li et al., 2008).
Table 1. Review of prior research
Study CSR Decoupling Data source Theory applied Country
Arena, Bozzolan and
Michelon (2015) Relationship between optimistic language in environmental disclosure and future environmental performance KLD database Companies 10-K filings Self-serving communication tactics; Resource dependence theory US
Boiral (2013) Sustainability reports as simulacra Content analysis Voluntary disclosure theory; Legitimacy theory
Global
Cho, Laine, Roberts and Rodrigue (2015)
Gap between corporate sustainability talk and practice
Case study Legitimacy theory US
Christmann and Taylor (2006)
Symbolic and substantive implementation of standards
Survey Legitimacy theory China
Crilly, Zollo and Hansen (2012)
Decoupling policy from practice Interviews; Rating agencies Institutional theory Global
Graafland and Smid (2016)
Quality of CSR implementation Rating agency Sustainalytics
Institutional theory 24 countries
Hawn and Ioannou (2016)
Gap between external and internal actions
Thomson Reuters (ASSET4)
Neo-institutional theory 33 countries
Hyatt and Berente (2017)
Symbolic environmental strategies Survey Impression management; Institutional theory
US
Jamali, Lund-Thomsen and Khara (2017)
Selective decoupling in relation to core humanitarian an labor rights issues
Interviews Institutional theory India
Kim and Lyon (2015) Deviation between reported and actual emissions reductions
DOE’s Voluntary Reporting of Greenhouse Gases Program; FERC Form 1
Legitimacy theory US
Marquis, Toffel and
Zhou (2016) Selective disclosure in environmental reporting Trucost Institutional theory, strategic management, information asymmetry
45 countries
Pope and Waeraas (2015)
False CSR claims to deceive consumers Literature review Global
Schons and Steinmeier (2016)
Gap between symbolic and substantive CSR actions, the social dimension
Thomson Reuters (ASSET4)
Neo-institutional theory and stakeholder theory
Cross-country Tashman, Marano
and Kostova (2016)
Conditions that lead EM-MNS to CSR decoupling
MSCI IVA database; Content analysis
Neo-institutional theory 15 countries
2.2 Hypothesis development
According to various researchers, the effectiveness of the board monitoring and the degree of disclosure depends, among other things, on the characteristics of the board (e.g. Madhani, 2015; Carter et al., 2003). Bamber et al. (2010) concluded that board members have a large influence on the disclosure style and therefore on the reduction of information asymmetry. The characteristics of the board of directors influence the effectiveness of the individual members to reduce the agency problems between the owners and the management (Healy and Palepu, 2001). In the literature, the most widely studied characteristics are those regarding board independence and board composition (Hussain et al., 2018). As for this study, three independent variables previously found relevant to CSR were selected, namely board size, board independence and CEO duality. Hypotheses for these three corporate governance characteristics are introduced in the sub-sections below.
Board size
Researchers who investigate the relationship between board characteristics and the degree of voluntary disclosure often regard the size of a board as an important characteristic, as this has a major influence on how effectively the board could perform its monitoring task (Madhani, 2015) and is therefore seen as a determining factor of the company’s performance. However, whether a large board size increases overall performance is still debated. According to Lipton and Lorsch (1992) the agency problems get bigger if there are more members on the board of directors. This leads to more internal conflicts between board members, to slower decision-making, poorer communication and therefore to less effective monitoring (Prado-Lorenzo and Garcia-Sanchez, 2010; De Andres, Azofra, and Lopez, 2005). A smaller board size could make both internal and external communication more efficient, which results in increased transparency and accountability (Ahmed et al., 2006).
On the other hand, Pfeffer (1972) argues that a larger board can offer the organization more access to important resources and therefore improves business performance. Furthermore, a larger board results in a lower workload per board member, which enables them to monitor and control effectively (John and Senbet, 1998). Considering each director as a resource, a larger board results in a greater diversity of expertise, opinions and backgrounds, which would lead to tighter monitoring and the reduction of information asymmetry (Adams et al., 2005; Zahra
and Pearce, 1989). A larger board has is more likely to have directors on the board with a greater knowledge of CSR. Therefore, they can provide management with better advise on CSR (De Andres et al., 2005). Besides that, a larger board has a larger network to gain access to external resources. In addition, Brown et al. (2006) found that companies with a larger board donate more money to charities. The paper of Akhtaruddin et al. (2009) elaborates further on the positive relationship, as they find that large boards have more capability to make managers engage in voluntary disclosure. In addition, larger boards have a better resistance towards the executive board, and could therefore pay more attention to meeting the stakeholders’ expectations (Abraham and Cox, 2007). However, the potential benefits of large board may be outweighed by the costs associated with larger groups (Jensen and Meckling, 1993).
Following the perspective of agency theory and the arguments of Prado-Lorenzo and Garcia-Sanchez (2010) and De Andres et al. (2005) a larger board has a negative impact on governance efficiency and consequently decreases the quality of monitoring. Therefore, the first hypothesis is formulated:
Hypothesis 1: There is a positive relationship between board size and CSR decoupling. Board independence
Board independence, a majority of independent board members, is considered as a main corporate governance characteristic regarding CSR performance and disclosure. It is suggested by the agency theory that board independence can improve quality of the boards’ monitoring and controlling function, and therefore can enhance higher transparency (Jensen and Meckling, 1976; Hussain et al., 2018). This includes better quality of disclosure, which helps stakeholders in making informed decisions (Fama and Jensen, 1983).
A more independent board appears to promote CSR activities and is more interested in demonstrating compliance with CSR standards (Johnson & Greening, 1999). Furthermore, independent directors tend to pursuit long-term success, whereas inside directors seem to be more concerned with short-term economic performance (Wang and Coffey, 1992). Since CSR investments might not be visible in the short term (Burke and Logsdon, 1996), CSR activities will be less interesting to inside directors. The independent directors will exert more influence to participate in CSR activities and they will intervene when managers are making
disadvantageous decisions regarding the firms’ CSR performance (Mallin and Michelon, 2011). Due to greater responsibility and higher reputational costs, independent directors will not only protect the interest of shareholders, but also those of the stakeholders (Prado-Lorenzo and Garcia-Sanchez, 2010; Lim et al., 2007). In addition, Ibrahim and Angelidis (1995) state that independent board members are more sensitive towards stakeholders’ social expectations. A possible explanation could be that independent director do not experience pressure from competition (Sonnenfeld, 1981), and therefore will be more likely to behave in a social responsible way (Freeman & Reed, 1983). Walls et al. (2012) confirm this by stating that firms with more independent boards show better environmental performance.
However, there are some competing results in the literature on the relation between CSR and board independence. Eng and Mak (2003), for example, report a negative association between the number of independent directors and disclosure. As an increase of independent directors could be considered a substitute for voluntary disclosure as a mechanism to monitor the managers, this results in a decrease of CSR disclosure (Eng and Mak, 2003). Michelon and Parbonetti (2012), Coffey and Wang (1998) and Cormier et al. (2011) found no significant relationship between board independence and CSR activities.
Following the agency theory and prior empirical research, a larger share of independent directors on the board is expected to have a negative effect on CSR decoupling. Therefore, the second hypothesis is as follows:
Hypothesis 2: There is a negative relationship between board independence and CSR decoupling.
CEO duality
CEO duality connotes that the CEO of a company is also the Chairman of the board (Baliga et al., 1996). This situation results in a blurred line between management and control (Fama and Jensen, 1983). When a person is both the CEO and the chairman of the board, he can exert a greater influence over the board (Cannella and Shen, 2001). Several studies found that CEO duality gives the CEO greater power, which will enable him or her to make decisions that are in line with the CEO’s agenda and not with the interests of stakeholders (Finkelstein and
D’Aveni, 1994; Muttakin et al., 2018). Additionally, as a result of the concentration of managerial power, CEOs who serve as chairpersons can set the agenda for board meetings and could therefore control the issues brought up in those meetings (Imhoff, 2003). This provides the opportunity for CEOs to steer away attention from selfish behavior and irresponsible actions. In addition to that, the duality power could affect the selection of board candidates, which results in reduced board independence (Muttakin et al., 2018). According to Mallin and Michelon (2011), CEO duality reduces the ability of a corporation to be more involved in CSR and to implement the right strategy regarding CSR reporting. When the CEO is a manager, the company’s strategy always remains focused on the short-term financial performance (Walls et al., 2012).
The CEO’s dual role results in increased agency problems since CEO duality upsurges the managerial power concentration and decreases the board independence (Boyd, 1994). From the agency theory perspective, concentration of managerial power, caused by CEO duality, results in an increased risk of the CEOs implementing strategies that benefit his or her personal interests and may harm firm value (Jensen and Meckling, 1976). The agency theory indicates the boards as a mechanism to monitor the actions and decisions of the agent to protect the rights of their shareholders (Jensen and Meckling, 1976). When a single person occupies both roles, CEO and Chairman of the board, this compromises the system of checks and balances, and therefore might result in weak monitoring (Rechner and Dalton, 1991). This weakening in the monitoring function and independence of the board could lead to reduced involvement in CSR activities and thus affect the related reporting. The board’s ability to monitor the CEO is compromised, which reduces the accountability and transparency of the corporation (Hussain et al., 2018).
Although most research on CEO duality suggest a negative association between CEO duality and transparency, there are also studies stating a positive link between CEO duality and transparency regarding CSR (Arena et al., 2014; Mallin et al., 2013; Lam and Lee, 2008). Arena et al. (2014), report a positive association between environmental performance and CEO duality. In addition, Mallin et al. (2013) found the CEO’s dual role associated with higher voluntary reporting. Furthermore, from the stewardship theory perspective, Lam and Lee (2008) argue that managers do not want to risk their reputation, therefore they will not participate in self-serving and opportunistic behaviour.
Ensuing the agency theory perspective and results of prior studies (Ahmad et al., 2017; Muttakin et al., 2018; Walls et al., 2012), CEO duality is found to weaken the monitoring function and board independence, which lead to reduces accountability and transparency. This leads to the third hypothesis:
3. Methodology
In this chapter the research method, data collection methods and the used measures will be described. The first section elaborates on the sample determination and the data collection. The second section explains how the dependent, independent and control variables are measured. Lastly, the research method will be further explained.
For this research, a quantitative research method is used to investigate relations between the independent variables and dependent variable. The sample is based on several databases. The data regarding CSR is obtained from the Thomson Reuters Asset4 ESG Database, which provides objective CSR information. Prior CSR literature also made use of this database (e.g. Cheng et al., 2014; Hawn and Ioannou, 2016; Visser, 2014). The data are based on three aspects: environmental, social and corporate governance data (ESG). Data regarding CSR disclosure are obtained from the Bloomberg database. This database contains an annually rating of companies based on the disclosure of ESG data. ESG data for more than 10.000 publicly listed companies are included in this database.
Furthermore, data are available for the period 2011 – 2017. To be able to make statistically reliable and representative statements, a large data set is used. The observations of companies that did not receive a score in a certain year were deleted. After deleting incomplete observations, a sample of 15,790 firm-year observations for 4,838 firms remained. In this research panel data are used. Panel data are a combination of cross-sectional data and time series data. The panel data contain observations on multiple entities, where each entity is observed at two or more points in time. The entities are surveyed over time and the data are pooled over space as well as time. In this study the panel is unbalanced, which means that some observations are missing. The use of panel data can alleviate possible multicollinearity. Since the absolute variance increases overtime, it is less likely that the explanatory variables are highly correlated when they vary over two dimensions (N and t).
In order to make the research generalizable, the sample includes firms from 67 different countries, resulting in a strong external validity. In addition, the firms in the sample are from 115 different industries, which makes the results also generalizable for different sectors. To prevent certain outliers from influencing the research, some variables are corrected for extreme values.
Dependent variable
In line with previous studies, CSR decoupling can be considered as a measure for the gap between CSR performance and CSR disclosure (e.g. Tashman et al., 2019; Crilly et al., 2012). The CSR decoupling score is based on the difference between the Bloomberg ESG disclosure score and the Asset4 ESG performance score. Both scores were already standardized since they were obtained from databases. The performance score is measured with a one-year lack compared to the disclosure score (taken into consideration that CSR actions take at least one year to be legitimized). The calculated decoupling score measures to what extent the disclosed information is in line with the performance. The higher the score, the higher the CSR decoupling. Since this study is concerned with the absolute gap, it is not of importance whether the decoupling is positive or negative, only the absolute score is taken into consideration.
Independent variables
The independent variables are board size, board independence and CEO duality. The size of the board of directors is determined based on the total number of non-executive and executive directors. This measurement method is used in a large number of studies, including the studies by Prado-Lorenzo and Garcia-Sanchez (2010) and Michelon and Parbonetti (2012). The board independence is defined as the percentage of non-executive directors (Andres et al., 2005). CEO duality will be taken into account as a dummy variable, with a value of 0 if the CEO is not chairman of the board and 1 if the CEO and the board chairman are the same person (Hussain et al., 2018).
Control variables
In order to be able to accurately test the different hypotheses, there are five control variables included in this study in addition to the dependent and independent variables. The selection of the control variables, based on academic literature, will be explained in this section.
The first control variable is size of an organization, as it has become apparent through other studies that this influences the extent of disclosure (Brick & Chidambaran, 2010; Gamerschlag et al., 2011; Gao et al., 2005). Larger organizations are known to have greater transparency (Galani et al., 2011). This may be due to the fact that large companies experience a greater
pressure from external stakeholders. Moreover, a large company is often controlled more precisely by regulatory authorities (Perez-Batres et al., 2012). In addition to that, larger organizations are more likely to have the resources to collect, generate and publish information (Owusu-Ansah, 1998). The firm size is measured using the log of total assets to reduce excessive positive kurtosis.
Leverage is included as a second control variable to control for firms that may be more hazardous than other firms (Hussain et al., 2018). In accordance with the agency theory, Ahmend and Courtis (1999) expect that companies with a great amount of borrowed capital have higher monitoring costs and will therefore be more transparent to reduce the monitoring costs. Also, long-term debt providers demand more information from organizations to reduce their risk (Chau & Gray, 2002). The leverage ratio is measured by dividing total liabilities by total equity.
As the third control variable, Return on Assets (ROA) is included to measure financial performance. Many scientific studies found that a positive association between profitability and non-financial reporting (Haniffa and Cooke, 2005; Eng and Mak, 2003). When a company is more profitable, more financial resources are available to invest in CSR activities and disclosure (McWilliams and Siegel, 2006). Therefore, it is expected that profitable companies will release more CSR information. The ROA is calculated by dividing the net income by total assets (Frias-Aceituno et al., 2013).
The fourth control variable is the research and development (R&D) intensity since innovative companies may be more able to improve CSR performance (McWilliams and Siegel, 2000). However, when firms invest a great amount of money in R&D and they might spend less on CSR (Hussain et al., 2018). The R&D intensity is measured as the proportion of R&D expenses to sales. Missing data are replaced by the industry average in the certain year (Strike, Gao and Bansal, 2006).
Lastly, the sales growth is included in the research. Sales growth opens more opportunities for investing, this will include the operating performance in the research (Hawn and Ioannou, 2016).
Table 2. Measurement of dependent, independent and control variables
Name of Variable
Mnemonics Role Measurement Source
CSR Decoupling
CSR_DEC Dependent CSR disclosure score minus CSR performance score (one year lack)
ASSET4, Bloomberg Board Size BSIZE Independent Total number of directors on governance board ASSET4 Board
Independence BINDP Independent Percentage of independent directors to total directors Bloomberg CEO Duality CEOD Independent Binary variable which takes value 1 if the CEO of
the company is also the chairperson of the governance board and 0 otherwise
Bloomberg
Profitability ROA Control Calculated as ratio of operating income and total assets
ASSET4
Firm Size SIZE Control Log of total assets of the firm ASSET4 Capital
structure
D/E Control Ratio between total debts to shareholders’ equity ASSET4
Sales Growth SGROW Control Percentage change in total sales with respect to previous year
ASSET4
R&D Intensity RDINT Control Ratio of total R&D expenditure to total sales ASSET4
Because panel data are used in the sample, suitable regression models for data of entities across time will be used. When dealing with this data, the simplest way to do this would be to estimate a single, pooled regression on all the observations together. However, with panel data, a pooled ordinary least squares (OLS) regression is not the best estimation method because there has to be dealt with company-specific characteristics that do not change over time and therefore do not form independent observations in the data set.
Two models can be used to take panel structure into account: the random effects model and the fixed effects model. When deciding between fixed or random effects, a Hausman test is executed. In this test it is examined whether the random effects model should be preferred over the fixed effects model or vice versa. The null hypothesis in this test is that random effects is the preferred model versus the fixed effects model (Green, 2008). The Hausman test showed that the null hypothesis must be rejected and that the fixed effects model must be chosen. The fixed effects regression applies demeaning; for each variable, the average is determined over time and subtracted from the variables “center over time”. Fixed effects regression only uses information that varies over time.
Within the fixed effects regression, it is assumed that each company can be characterized by means of a specific business effect that is assumed to be constant for that company. No assumptions are necessary about the coherence of business effects between companies and between the business effects and the explanatory variables. To test whether there are time fixed effects when running an fixed effects regression, an F-test is executed. This examines whether the year dummies that display time fixed effects are equal to 0 or not. The F-test showed that there are time fixed effects, because p (0.0000) is lower than alpha (significance level =0.05). Therefore, time fixed effects are used in this study. The following regression model is used: CSR decoupling = b0 + b1BSIZE + b2BINDP + b3CEOD+ b4SIZE + b5ROA + b6DE + b7RDINT + b8SGROW + e
4. Results
In this chapter the results of the research are presented and discussed. This chapter only deals with the statistic results, the interpretation follows in the next chapter. Firstly, the descriptive statistics and the result of Pearson correlation are presented (Table 3) and explained. Subsequently, the outcomes of the regression analysis are presented (Table 4) and analyzed. Furthermore, two additional analyses will be executed to produce more insight on the interplay between the three different CSR dimensions, namely environmental, social and governance (Table 7). In the last section the sample is divided into two subsamples to test for positive and negative CSR decoupling separately (Table 6).
Table 3 presents the descriptive statistics for the examined variables. The descriptive statistic is a summary of the results of data collection and does not yet provide an interpretation for the regression model. Table 3 sets out the mean and the standard deviation, of each variable. The sample (N) consists of 15,790 observations. The order that is used in the table concerns dependent variables, independent variables and finally the control variables. The dependent variable CSR decoupling, based on the difference between the Bloomberg disclosure score and the ASSET4 performance score, has an average of 39.142. The average board size is 10.073, with an average of 49.491% independent board members. The CEO duality average is 0.645, which means that in 64,5% of the observations the CEO and the board chairman are the same person. All the variables are winsorized in order to deal with outliers.
Furthermore, table 3 shows the outcomes of the Pearson correlation analysis. This matrix is constructed to control for potential multi-collinearity problems among the variables. Correlations between the variables used in the analysis are indicated, correlations higher than (-).5 are a sign of multi-collinearity in the study (Hair et al., 2005). There is no evidence on multicollinearity, none of the correlation values are above the specified threshold. For example, we see that the ‘Board size’ and ‘CEO duality’ positively correlate with ‘CSR decoupling’, since CSR decoupling is used as dependent variable in this study, we expect those correlations. As mentioned, the variables do not exceed the limit for correlation, therefore it is not necessary to perform more detailed correlation tests.
Table 3. Descriptive statistics and Pearson correlation Variable Mean SD (1) (2) (3) (4) (5) (6) (7) (8) (9) 1. CSRDEC 39.142 27.101 1 2. BSIZE 10.073 3.450 0.130*** 1 3. BINDP 49.491 30.862 0.0372*** -0.133*** 1 4. CEOD 0.645 0.479 0.00697 -0.0722*** -0.185*** 1 5. SIZE 16.731 2.928 0.0554*** 0.176*** -0.231*** -0.0287** 1 6. ROA 0.044 0.093 0.0290*** 0.0144 0.0215* -0.0310*** 0.0270** 1 7. D/E 98.569 168.556 0.0189* 0.0765*** -0.0227** 0.0238** 0.216*** -0.0599*** 1 8. RDINT 0.044 0.062 -0.0358*** -0.0278*** 0.0666*** -0.0132 -0.113*** -0.0669*** -0.0325*** 1 9. SGROW 9.159 17.46787 -0.0468*** -0.0313*** 0.0151 -0.0170 -0.0956*** 0.0185* -0.0226* 0.101*** 1
Table 4 contains stepwise fixed effect regression results in different models. The different regression analyses were computed in order to determine the effect of board size, board independence and CEO duality on CSR decoupling. The first regression (1) includes the dependent variable and all the control variables. Models 2 to 4 test the influence of the individual governance variables and controls. The last regression analysis runs a regression with the dependent variable, the independent variables and the control variables. All models are a result of panel regression and use time fixed effects, to decrease autocorrelation and heteroscedasticity problems (Greene, 2013).
In the first regression we see that the values for the five control variables, firm size, profitability, leverage, R&D intensity and sales growth are respectively, 0.117, 0.057, 0.009, 0.059 and -0.052. We see that firm size has a positive relation with CSR decoupling, with a significance level of 0.01, which indicates a 90% confidence level. R&D intensity also shows a significant positive relation with CSR decoupling, namely on a 0.05 level. Profitability and leverage also show a positive association, however these results are insignificant. In this regression, sales growth is the only control variable which shows a negative relation with CSR decoupling. Model 2 shows the interaction coefficient (0.002) of board size on CSR decoupling. This coefficient indicates a positive relation between the size of the board and the engagement in CSR decoupling. However, the relation is not significant. This means that hypothesis 1, which is formulated as There is a positive relationship between board size and CSR decoupling, cannot be accepted.
The interaction coefficient of board independence (-0.030) is not significant as well (Model 2). This coefficient indicates a negative relationship between the percentage of independent board members and CSR decoupling. Due to insignificance, hypothesis 2, formulated as There is a negative relationship between board independence and CSR decoupling cannot be accepted. Model 4 shows the interaction coefficient (0.002) between CEO duality and CSR decoupling. This coefficient implies a positive relation, which is an insignificant positive result of CEO duality. Therefore, hypothesis 3, which is formulated as There is a positive relationship between CEO duality and CSR decoupling, must also be rejected.
To conclude, the results do not support the influence of board size, board independence and CEO duality on CSR decoupling. The low R-squared could indicate that there is more to CSR decoupling than the three corporate governance mechanisms tested in this study. According to Cohen (1992), the size of an effect is considered as large if it is higher than 0.50, medium when it is 0.30 and small when it is 0.10. Given that none of the interaction coefficients is larger than 0.30, none of the tested corporate governance mechanisms seems to be closely related with CSR decoupling.
Table 4. Fixed-effects regression on CSR decoupling
Variable (1) (2) (3) (4) (5) BSIZE 0.002 0.002 (0.004) (0.004) BINDP -0.030 -0.029 (0.050) (0.050) CEOD 0.002 0.001 (0.028) (0.028) SIZE 0.117*** 0.117*** 0.116*** 0.117*** 0.116*** (0.033) (0.033) (0.033) (0.033) (0.033) ROA 0.057 0.056 0.058 0.057 0.057 (0.163) (0.163) (0.163) (0.163) (0.163) DE 0.009 0.009 0.009 0.009 0.009 (0.009) (0.009) (0.009) (0.009) (0.009) RDINT 0.059** 0.059** 0.059** 0.059** 0.059** (0.028) (0.028) (0.028) (0.028) (0.028) SGROW -0.052 -0.053 -0.052 -0.052 -0.053 (0.061) (0.061) (0.061) (0.061) (0.061) Constant 1.491*** 1.465** 1.512*** 1.491*** 1.486*** (0.569) (0.571) (0.570) (0.569) (0.572) Observations 15,790 15,790 15,790 15,790 15,790 R-squared 0.002 0.002 0.002 0.002 0.002 Number of id 4,383 4,383 4,383 4,383 4,383
Notes: *** p<0.01, ** p<0.05, * p<0.1; This table shows the results of the time fixed effects panel regression. Values in the parentheses contain robust standard errors
In addition to the latter regression, additional analyses are performed to ascertain the robustness of the results and to obtain a more detailed insight on CSR decoupling. Several interesting results emerged from the additional analyses. Table 5 shows the first additional analysis, which is a fixed effects regression analysis on the three different dimensions of CSR, namely environmental, social and governance. In line with the above-mentioned results of the overall
measure of CSR decoupling, the corporate governance mechanism ‘Board size’ shows a positive insignificant relation with environmental decoupling, social decoupling and governance decoupling. Board independence and CEO duality, on the contrary, show different results when testing the CSR dimensions individually. The environmental and social CSR dimensions show a positive association between board independence and decoupling, whereas before, a negative association was found. Furthermore, the additional regression results on CEO duality also contradict the earlier found results. The interaction coefficient between CEO duality and CSR decoupling implied a positive relation, whereas the results in the additional regression analysis show a negative influence of CEO duality on environmental and governance decoupling.
Table 5. Additional regression analysis CSR dimensions
(1) (2) (3)
Variable ENV_DEC SOC_DEC GOV_DEC
BSIZE 0.001 0.003 0.004 (0.005) (0.004) (0.004) BINDP 0.042 0.054 -0.041 (0.059) (0.048) (0.046) CEOD -0.015 0.043 -0.017 (0.033) (0.026) (0.025) SIZE 0.168*** 0.095*** 0.000 (0.039) (0.031) (0.030) ROA 0.160 0.096 0.177 (0.193) (0.154) (0.149) DE 0.000 0.004 0.002 (0.010) (0.008) (0.008) RDINT 0.021 -0.006 0.003 (0.033) (0.027) (0.026) SGROW -0.204*** -0.083 -0.049 (0.072) (0.058) (0.056) Constant 0.497 1.564*** 3.129*** (0.674) (0.540) (0.520) Observations 15,790 15,790 15,790 R-squared 0.002 0.001 0.000 Number of id 4,383 4,383 4,383
Notes: *** p<0.01, ** p<0.05, * p<0.1; This table shows the results of the time fixed effects panel regression. Values in the parentheses contain robust standard errors
The second additional analysis contains fixed effects regressions concerning the effect of both CSR decoupling gaps, namely overstating CSR performance and understating CSR performance. On the one hand, the subsample ‘overstating CSR performance’ concerns the observations where the CSR disclosure score is higher than the CSR performance score. On the other hand, the subsample ‘understating CSR performance’ concerns the observations where the CSR disclosure score is lower than the CSR performance score. In this way, a more detailed insight into the misalignment of CSR disclosure and CSR performance is provided. As it should be, the sum of both subsamples is equal to the number of observations in the full sample (2,896 + 12,894 = 15,790).
The analyses results, earlier performed on the full sample, showed that board size and CEO duality have a positive impact on CSR decoupling. Furthermore, the results show a negative relation between board independence and CSR decoupling. The results of the additional analysis on the overstating CSR performance subsample show contradicting associations between CSR decoupling and the corporate governance mechanisms. Whilst a positive relation is expected between board size and CSR decoupling, the statistics show opposite results. This contradicting result is also found on the relation between CEO duality and CSR decoupling. The findings of board independence are consistent with the earlier results, however there is still no significance. The results of the second subsample, understating CSR performance, are in line with the relations found in the full sample on CSR decoupling. Board size has a positive effect on CSR decoupling and board independence is negatively associated with CSR decoupling, although these results are still not significant. The second subsample, however, does indicate a significant positive relation between CEO duality and firms understating CSR performance. CEO duality seems to be significant at a 90% confidence level. The coefficient indicates that one standard deviation increase in CEO duality is associated with a 0.050 standard deviation increase in CSR decoupling (understating CSR performance).
Table 6. Additional regression analysis sub samples (Dependent variable: CSR Decoupling) Variable Overstating subsample (1) (2) (3) (4) (5) Understating subsample (6) (7) (8) (9) (10) BSIZE -0.015 -0.014 0.001 0.001 (0.012) (0.012) (0.004) (0.004) BINDP -0.126 -0.124 -0.039 -0.028 (0.153) (0.153) (0.049) (0.049) CEOD -0.067 -0.068 0.050* 0.049* (0.080) (0.080) (0.027) (0.027) SIZE -0.200** -0.200** -0.203** -0.197** -0.199** 0.112*** 0.113*** 0.112*** 0.110*** 0.110*** (0.079) (0.079) (0.079) (0.079) (0.079) (0.034) (0.034) (0.034) (0.034) (0.034) ROA 1.183** 1.185** 1.196** 1.203** 1.218** -0.176 -0.176 -0.175 -0.179 -0.179 (0.513) (0.513) (0.513) (0.513) (0.513) (0.160) (0.160) (0.160) (0.160) (0.160) DE -0.004 -0.005 -0.003 -0.006 -0.005 0.005 0.005 0.005 0.005 0.005 (0.031) (0.031) (0.031) (0.031) (0.031) (0.008) (0.008) (0.008) (0.008) (0.008) RDINT 0.102 0.102 0.101 0.101 0.099 0.059** 0.059** 0.059** 0.060** 0.060** (0.086) (0.086) (0.086) (0.086) (0.086) (0.027) (0.027) (0.027) (0.027) (0.027) SGROW 0.297* 0.295* 0.295* 0.298* 0.293* -0.100 -0.101* -0.100 -0.099 -0.100 (0.169) (0.169) (0.169) (0.169) (0.169) (0.061) (0.061) (0.061) (0.061) (0.061) Constant 5.700*** 5.837*** 5.788*** 5.676*** 5.897*** 1.860*** 1.848*** 1.892*** 1.867*** 1.874*** (1.410) (1.413) (1.414) (1.410) (1.418) (0.572) (0.573) (0.573) (0.572) (0.575) Observations 2,896 2,896 2,896 2,896 2,896 12,894 12,894 12,894 12,894 12,894 R-squared 0.014 0.015 0.014 0.014 0.017 0.002 0.002 0.002 0.003 0.003 Number of id 1,809 1,809 1,809 1,809 1,809 4,461 4,461 4,461 4,461 4,461
5. Conclusion and discussion
For a long time, researchers have been investigating the value of CSR, and its effect on companies. This study contributes to the current literature by exploring the relatively new concept “CSR decoupling” and by addressing the recommendations from previous studies to explore governance mechanisms that impact the engagement in CSR decoupling (i.e., Delmas and Burbano, 2011). Decoupling as a symbolic adoption of CSR reporting is already researched by several studies, which explored certain factors affecting the reporting quality (e.g., Luo et al., 2017; Marquis et al., 2016; Marquis and Qian, 2014). However, those studies measured decoupling as a CSR reporting property, while the gap between performance and reporting is not considered in those studies. This research used the direct measurement of decoupling as the absolute gap between CSR reporting and performance and explored which factors contribute to the likelihood of firms to engage in decoupling practices. The objective of this study is to analyze relationships between CSR decoupling and certain board characteristics, namely, board size, board independence and CEO duality.
Firstly, the board seems to play an important role regarding CSR activity. Board size and CEO duality are both positive predictors of the gap between CSR performance and disclosure, indicated as CSR decoupling. Board independence is expected to affect CSR decoupling in a negative way. Although, contrary to the expectations, the results of the fixed effects regression analysis do not show significant results for these three governance mechanisms. A possible reason for this finding could be that in this research both forms of CSR decoupling, namely understating and overstating, are included, which could lead to conflicting results. The additional analysis on the overstating and understating subsamples only partly certifies the conflicting results within the full sample regarding CEO duality. Board size and board independence still show insignificant results. In addition, Yang, Wen & Yung (2009) state that the results strongly depend on the environment in which the company is located. In line with this, Aquilera et al. (2006) suggest that the relationship could be influenced by geographical settings. The institutional theory states that corporations are subject to the beliefs, norms and values of the society (Eccles, Ioannou and Serafeim, 2014). Therefore, different societies define different rules and expectations firms need to take into account, which influences the actions of firms differently. Board size showed a positive relation with CSR decoupling, but this association was not significant. It could be that a high number of board members is positively associated with CSR decoupling, but that there is a maximum to this number after which there
will be a negative association (Veronica, Siregar and Bachtiar, 2010). Salmon (1993) agrees, as he argues that the board must have sufficient expertise and knowledge, but should not exceed a certain number of members as this prohibits the board from efficient decision-making. The number of outsider directors seems to have a positive influence on the CSR decoupling, which means that the decoupling gap is bigger when there are more outsider directors. This could be explained by the fact that CSR activities can be industry-specific and insider board members probably possess more relevant expertise regarding this area. The results on the association between CEO duality and CSR decoupling are consistent with the existing literature (Wang and Coffey, 1992; Lim and Chow, 2007; Walls et al., 2012; Frias-Aceituno et al., 2013) and with the agency theory. From the agency theory perspective, the key role of the governing board is to monitor the agents’ decisions. This monitoring process will not be effective if the CEO is also chairman of the board (Allegrini and Greco, 2013).
Furthermore, this study supports the agency theory regarding the monitoring role of the board in delivering high-quality CSR disclosure that mirror the actual CSR performance of the company. The results assume that effective governance mechanisms help firms to gain fair advantages from their CSR activities and reporting. Managers should incorporate CSR performance and reporting as part of their strategy and as one of the main goals to reduce the chance of inconsistencies or misalignment. The practical implication of this study is to better understand the relation between corporate governance and CSR decoupling, which will enable firms to utilize governance mechanisms in a more effective way. This could result in an improvement of their CSR performance and their CSR disclosure. Within this study, no hard evidence is found on the relation between board size, board independence and CEO duality as corporate governance mechanisms and CSR decoupling. Therefore, the question whether there really is a link between corporate governance and CSR remains (Walls et al., 2012), and further research is needed.
Limitations and future research directions
This study has certain limitations. First of all, CSR is a multidimensional construct. Companies may treat environmental and social issues in a different way (Bansal and Gao, 2008), which may lead to different results. To get deeper into CSR decoupling, it could be insightful to make a distinction between environmental and social performance and disclosure. Future research could also take into account other corporate governance mechanisms. The inclusion of diversity
within the board may also uncover interesting results. Furthermore other management variables can be included, for example the backgrounds of top management or the presence of certain functions or committees.
The results in the relationship between corporate governance mechanisms and CSR are restricted to large companies. This limits the results, since large companies generally have more resources available to invest in CSR activities and usually have better governance mechanisms than the smaller firms (Hussain et al., 2018). In addition, larger firms experience more visibility, which has a big impact on the way they behave and the manner in which they present themselves. Taking a sample of small and medium-sized companies may result in different outcomes. The degree of CSR decoupling does not only differ between firm sizes, but also across industries, countries and economies. Firms experience different pressures to align reporting and performance. Thus, further research could improve the understanding of decoupling practices among different environments.
The results of this research are based on archival measures. All the data used is collected from databases. Researchers are warned not to collect all their data from databases, especially when making use of archival data (Fogarty, 2006). It is advised to hand collect at least one variable in order to make the results not fully dependent on the reliability of databases. Also, to create more depth in the relationships to CSR, underlying mechanisms must be identified and measured, which could be done by performing for example case studies or survey methods. Those methods can probably better capture sociological and behavioral characteristics of firms and board members. Further research could create a more qualitative setting.
The research on corporate governance and CSR is still fragmented. Future researchers can contribute by exploring possible explanations for the mixed results. The findings regarding corporate governance and CSR can stimulate future theoretical development. From an agency perspective, the board is considered as a mechanism to protect the interests of shareholders. Other theoretical frameworks could explore this relationship further. It will be interesting to explore directors’ behavior, motivation, social norms, power and moral obligations from a more social perspective. In addition, a stakeholder lens could be useful (Freeman, 1984), as other researchers recommended a more non-economic perspective on corporate governance (Jones, 1992).
Finally, further research could focus on the effect of positive decoupling (muddling through) and negative decoupling (faking it) separately to gain better insight about the effects of CSR actions and the best way of reporting about it (Crilly et al., 2012). Realizing that managers sometimes muddle through, rather than intentionally deceive their stakeholders, is a good starting point for exploring this side of the gap, which could result in better implementation and compliance with CSR.
References
Abdullah, S. N., Mohamad, N. R., Mokhtar, M. Z. (2011). Board independence, ownership and CSR of Malaysian Large Firms. Corporate Ownership & Control, 8(3): 417-431
Abraham, S., Cox, P. (2007). Analyzing the determinants of narrative risk information in UK FTSE 100 annual reports. The British Accounting Review, 39(3): 227–248
Adams, R. B., Almeida, H., Ferreira, D. (2005). Powerful CEOs and their impact on corporate performance. Review of financial studies, 18(4): 1403-1432
Aguilera, R. V., Williams, C.A., Conley, J.M., Rupp, D.E. (2006). Corporate governance and social responsibility: A comparative analysis of the UK and the US. Corporate Governance: An International Review, 14(3): 147-158
Ahmed, K., Courtis, J. K. (1999). Association between corporate characteristic and disclosure levels in annual reports: a meta-analysis. British Accounting Review, 31(1): 35–61
Ahmed, K., Hossain, M., Adams, M. B. (2006). The effects of board composition and board size on the informativeness of annual accounting earnings. Corporate Governance: An International Review, 14: 418–431
Akhtaruddin, M., Hossain, M. A., Yao, L. (2009). Corporate governance and voluntary disclosure in corporate annual reports of Malaysian listed firms. Journal of Applied Management Accounting Research, 7(1): 1-19
Allegrini, M., Greco, G. (2013). Corporate boards, audit committees and voluntary disclosure: evidence from Italian Listed Companies. Journal of Management & Governance, 17(1): 187-216
An, Y., Davey, H., Eggleton, I.R.C (2011). Towards a comprehensive theoretical framework for voluntary IC disclosure. Journal of Intellectual Capital, 12(4): 571-585
Andres, P., Azofra, V., Lopez, F. (2005). Corporate boards in OECD countries: Size, composition, functioning and effectiveness. Corporate Governance: An International Review, 13(2): 197-210
Aras, G., Crowther, D. (2008). Governance and sustainability: An investigation into the relationship between corporate governance and corporate sustainability. Management Decision, 46(3): 433-448.
Arena, C., Bozzolan, S., Michelon, G. (2014). Environmental reporting: Transparency to stakeholders or stakeholder manipulation? An analysis of disclosure tone and the role of the board of directors.
Baldassarre, F., Campo, R. (2016). Sustainability as a marketing tool: To be or to appear to be? Business Horizons, 59(4): 421-429.
Baliga, B. R., Moyer, R.C., Rao, R.S. (1996). CEO duality and firm performance: what’s the fuss? Strategic Management, 17: 41-53
Bamber, L.S., Jiang, J., Wang, I.Y. (2010). What's my style? The influence of top managers on voluntary corporate financial disclosure. The Accounting Review, 85(4): 1131-1162
Bansal, P. (2005). Evolving sustainability: a longitudinal study of corporate sustainable development. Strategic Management Journal, 26: 197-218
Bear, S., Rahman, N., Post, C. (2010). The impact of board diversity and gender composition on corporate social responsibility and firm reputation. Journal of Business Ethics, 97(2): 207-221
Boiral, O. (2013). Sustainability reports as simulacra? A counter-account of A and A+ GRI reports. Accounting, Auditing & Accountability journal, 26(7): 1036-1071
Boyd, B. (1994). Board control and CEO compensation. Strategic Management Journal, 15: 335-344
Brick, I.E., Chidambaran, N.K. (2010). Board Meetings, Committee Structure, and Firm Value. Journal of Corporate Finance, 16(4): 533-553
Brown, L. (2008). Creating credibility: Legitimacy and accountability for transnational civil society. Kumarian Press: Sterling, VA.
Burke, L., Logsdon, J. (1996). How corporate social responsibility pays off. Long Range Planning, 29: 495-502
Cai, C.X. Keasey, K., Short, H. (2006). Corporate Governance and Information Efficiency in Security Markets. European Financial Management, 12(5): 763-787
Cannella, A.A., Shen, W. (2001). So close and yet so far: Promotion versus exit for CEO heirs apparent. The Academy of Management Journal, 44(2): 252-270
Carter, D. A., Simkins, B. J., Simpson, W. G. (2003). Corporate governance, board diversity, and firm value. The Financial Review, 38: 33-51
Castillo, M., Vial, V. (2016). Exploring post-financial crisis CSR digital communications by MNEs in Mexico.
Chau, G.K., Gray, S.J. (2002), Ownership structure and corporate voluntary disclosure in Hong Kong and Singapore. The International Journal of Accounting, 36(2): 247-265
Chen, C.J.P., Jaggi, B. (2000). Association between independent non-executive directors, family control and financial disclosures in Hong Kong. Journal of Accounting and Public Policy, 19 (4-5): 285-310
Cheng, B., Ioannou, I., Serafeim, G. (2014). Corporate social responsibility and access to finance. Strategic Management Journal, 35(1):1–23
Cho, H., Laine, M., Roberts, W., Rodrigue, M. (2015). Accounting, Organizations and Society, 20: 78-94
Christmann, P., Taylor, G. (2006). Firm self-regulation through international certifiable standards: determinants of symbolic versus substantive implementation. Journal of International Business Studies, 37(6): 863-878.
Cormier, D., Ledoux, M. J., Magnan, M. (2011). The informational contribution of social and environmental disclosures for investors. Management Decision, 49(8): 1276-1304
Crilly, D., Zollo, M., Hansen, M. T. (2012). Faking it or muddling through? Understanding decoupling in response to stakeholder pressures. Academy of Management Journal, 55(6): 1429-1448
Cui, J., Jo, H., Na, H. (2018). Does corporate social responsibility affect information asymmetry? Journal of Business Ethics, 148(3): 549-572
