The influence of corporate social responsibility on investor activism

(1)

Amsterdam Business School

The Influence of Corporate Social

Responsibility on Investor Activism

Name: Sjors Smeulders Student number: 10204148

First supervisor: Prof. dr. B.G.D. O'Dwyer Second supervisor: Dr. G. Georgakopoulos Date: 16 June 2016

Word count: 11.559

MSc Accountancy & Control, specialization Accountancy Amsterdam Business School

(2)

2

Statement of Originality

This document is written by student Sjors Smeulders 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.

(3)

3

Abstract

Increasing attention on corporate social responsibility performance to stay legitimate can give rise to investor pressure on firms policies. This study search with a quantitative database research for the answer if corporate social responsibility influences investor activism. Investor activism is measured with the quantity of shareholder proposals on a corporate social responsibility- and a total level. By using the GRI sustainability report and the KLD social rating score as indicators of corporate social responsibility, the effect of CSR on investor activism is examined. A sample containing 3,722 firm-year observations from 523 firms listed in the S&P 500 during the period 2006-2013 is used. Results show that a the production of a GRI sustainability report increases investor activism. The GRI report

decreases information asymmetry with investors, but the increased transparency and trust in firms will not result in less investor activism. Possible clarifications could be; that firms with a GRI report are more open for conversations, beforehand unobservable information give new insights and firms producing a GRI report have already more closely monitoring investors. In contrast with the KLD score findings, which indicate that higher KLD scores could already give investors a more trustworthy feeling and decreases investor activism.

(4)

4

4.2 Control Variables ... 24 4.2 Model Building ... 24 5. Results ... 27 5.1 Descriptive Statistics ... 27 5.2 Correlation Matrix ... 29 5.3 Regression Analysis ... 31 6. Discussion ... 35 6.1 Results Discussion ... 35 6.2 Limitations ... 37 6.3 Future Research ... 38 References ... 39 Appendices ... 43

(5)

5

1. Introduction

After centuries of solely focus on the financial performance of firms. There’s now a shift more and more to the sustainability subject (Viller, Rinaldi and Unerman, 2014). This shift results in disclosures of firms social and environmental topics. Lodhia (2015) mention that present forms of reporting have developed two major strand of reporting. On the one hand financial reporting and on the other hand sustainability reporting.

Current sustainability reporting of multinationals incorporates corporate governance aspects. This reporting is used to create an organizational report on economic, social and environmental performance, the sustainability report (Kolk, 2008). Sometimes this is labelled as ‘corporate social responsibility’ (CSR) or ‘triple bottom line’ (People, Planet, Profit)

reporting (Kolk, 2008). CSR reporting not just focuses on the inclusion of corporate

governance sections in reports. It also, and more importantly, examines whether companies pay attention to board supervision in relation to sustainability, codes of ethics and

complaints procedures, and auditor involvement in assessing the reliability of sustainability reporting (Kolk, 2008).

Akisik and Gal (2014) argued that ﬁrms that don’t engage in socially responsible behavior may risk their legitimacy and reputations with customers, workers and other stakeholders. Some other researchers like Borkowski, Welsh and Wentze (2010) say that CSR reporting is just a symbolic way of giving information to stakeholders. Nowadays, there is a transition to more integrated reporting. A new way of reporting, where financial and corporate social responsibility reporting is one. Before firms take a big steps to integrated reporting, we first need identify the added value of the corporate sustainability reporting.

In a research of EY (2014) managers think that introducing CSR reporting gives them a competitive advantage over others who don’t. For example Akisik and Gal (2014) conclude that there is a direct relationship between reviewed CSR reporting and financial

performance of the firm, like an effect on sales, leverage and growth. Besides these direct financial links of CSR reporting there could be some other effects, like the findings of Fernandez, Romero and Ruiz (2014) suggest that CSR reporting is a communication tool companies use to convey a transparent image. With the production of a sustainability report, firms give more information about their actions to their stakeholders. Also Gwendolen (2016) mention that CSR reporting provides a more transparent view of

(6)

6

companies environmental, economic and social impacts. In this way stakeholders can use more information to make their decisions and information asymmetry between firms and shareholder is reduced.

Jaggi and Freedman (1992) mention that the legislative activities, disclosure requirements and social, political and economic pressures were expected to encourage management to become socially responsible and pay greater attention to environmental and social consequences of corporate activities. So external pressure encourages managers to focus on CSR and this is extended by Mahoney and Thorne (2005) with the findings that companies introducing a long-term compensation to focus executives’ efforts on optimizing the longer term. Actions taken or decisions made by firms that maximize longer term firm performance tend to be socially responsible (Mahapatra, 1984; Mahoney & Thorne, 2005). By the growing CSR governance a shift is caused from a short term focus to a more longer term firm focus.

This longer term attitude is also noticed by firms’ shareholders and explained by Monks, Miller and Cook (2004) who tell us that more responsive, accountable leadership is seen by investors as less risky investments in the long run. Firms that use sustainable business practices are rewarded by stake- and shareholders, including consumers and employees (Monks et al., 2004). Above statements suggest that management with CSR reporting take decisions more in line with their stake- and shareholders. The long term focus is accompanied by the findings of Cai, Jo and Pan (2011) who mention that CSR executive compensation of socially responsible firms, on average, care more for their employees and have better self-control rather than follow the path of greed. García-Beneau, Siera-Garcia and Zorio (2013) agree that CSR is an effective management tool which offers confidence to stakeholders as the company is perceived as responsible and trustworthy. This research is build up on the theory that CSR reporting lead to increased transparency and

trustworthiness of the firm. With this firms can ,as Clark and Crawford (2012) mention, manage its public policy pressure by responding on shareholder resolutions and increase corporate transparency. Disclosing information in a response to a shareholder request about one’s climate change strategy is a political activity designed to signal management’s

reevaluation of its practices and policies and provide details about future policy changes (Clark and Crawford, 2012). On the other side Oded and Wang (2010) suggest that better

(7)

7

exogenous governance mechanisms, leads to higher valuation through more frequent but less intense activism. This research looks for the deeper understanding of the influence of a sustainability reporting on the pressure a company encounter on their corporate governance policies. Looked is for the answer on the following research question; Does corporate social responsibility influence investor activism?

This is a contribution to the academic literature, because there has been a lot of research about CSR reporting, but most of this literature focuses on the financial effect and the reliance of information produced by firms, instead of the aspect of investor activism. New insights in the pressure on corporate governance policies can give rise to the significant importance of CSR reporting. Another point is that if CSR reporting creates less short term terrorism, it influences also other important subjects in the firm, like CEO compensation and different investments decisions in the short and long run.

To answer the research question a quantitative database research is used. Firstly, the concepts of information asymmetry, stakeholder theory, corporate social responsibility reporting, investor activism and shareholder proposals are described in paragraph two. Subsequently, the hypothesis are developed and explained. Paragraph four gives a clarification about the research method with the sample and models used to test the hypothesis in this study. Afterwards, there follows an analysis of the results found. Finally, there will be a discussion about the results and an answer on the research question.

(8)

8

2. Literature Review

In this part the concepts of information asymmetry, stakeholder theory and corporate social responsibility are clarified. Subsequently, the concepts of investor activism and shareholder proposals are described.

2.1 Information Asymmetry

Cornier and Ledoux (2011) mention that firms stakeholders, like regulators and investors, have fundamental concerns about the extent of the information asymmetry between firm’s management and themselves. This information asymmetry concept can be described with two important theories often used in studies, the agency theory, described by Jensen and Meckling (1976), and the market for lemons theory, described by Akerlof (1970). The agency theory describes the relationship between a principal and an agent. Jensen & Meckling (1976) describe the principal as the person that tries to monitor another person, which is the agent. The agent perform tasks on behalf of the principal. An example of an agency relationship is the relationship between the shareholders (principal) and firm managers (agent). Jensen and Meckling (1976) describe that it’s likely that managers won’t always act in the best interest of the shareholder. The manager know what he is doing, but it’s hard for the shareholder to monitor the effort the manager gives. The agency problem introduces the monitoring- and bonding costs for the principal (Jensen & Meckling, 1976). With payouts and incentives the principal tries to influence agents behavior to work in the optimal way for the principal, which are the bonding costs. (Jensen & Meckling, 1976). Monitoring costs are incurred when the principal is trying to get the right information about the agent, so when shareholders are searching for the right information to make their investment decisions (Jensen & Meckling, 1976).

The information asymmetry is also mentioned by Akerlof (1970) in his lemons theory in which information about the quality of the product is known by the seller, but unknown by the buyer. Akerlof (1970) mention that the difficulty of distinguishing good quality from bad quality is inherent in the business world. When shareholders try to distinguish good quality companies from bad quality companies, the same information asymmetry problem arise. Healy and Palepu (2001) argue that it’s an important challenge to resolve the

(9)

9

information and agency problems to come to the optimal allocation of savings to investment opportunities. Akerlof (1970) conclude that this problems are about trust. To overcome this trust problems there are numerous institutions that try to resolve the information

asymmetry (Akerlof, 1970). For example, information asymmetry and trust problems could be partly resolved by giving more information with the production of a firm report with financial and/or corporate social responsibility related information.

The research of Cornier and Ledoux (2011) shows how social and environmental disclosures affect information asymmetry on the stock market. Their results show that social and environmental disclosures reduce stock market asymmetries. Also Gwendolen (2016) mention that the production of a sustainability report gives the stakeholder more

information to base their decisions on and this will reduce the information asymmetry. Coming back on the monitoring cost of Jensen and Meckling (1976) we can say that the monitoring cost can be reduced by the production of a corporate social responsibility reporting. On the other hand is mentioned that despite the growing CSR disclosure, Cornier and Ledoux (2011) assert that there is some degree of information asymmetry between companies and stakeholders beyond stock markets, but the disclosure of CSR reporting can help to reduce the information asymmetry between stakeholders and the managers of the company.

2.2. Stakeholder Theory

An important consideration for firms nowadays is the question to be accountable to which people. Shareholders and investors are important, but also environmental activist or work unions can have significance influence on the company. Deegan and Blomquist (2006) introduce for this accountability problem of firms two theories; the legitimacy and

stakeholder theory. The legitimacy theory explains that firms have one general social contract with the society to act in a way society expects the organizations should behave (Deegan and Blomquist, 2006). Companies behaviors should be in line with the norms and values of the society. Failure of this compliance can have a significant negative impact on the continuation of firm’s business. The growing attention on sustainability subjects causes a change in perceptions how corporations and industries should interact and report. Deegan and Blomquist (2006) discover the growing importance of information disclosure about firms

(10)

10

activities. It’s not only important that an organization or industry act in the way society expects, but there is also a need to inform society about their activities and changes.

The stakeholder theory, first introduced by Freeman(1984), is in line with the legitimacy theory, but the important difference is that the stakeholder theory recognizes that there are different social contracts with particular groups in the society (Deegan and Blomquist, 2006). Obviously, most organizations first react to the demands of those

stakeholders groups that control the most important resources, ‘the powerful stakeholders’, but also the less powerful stakeholders becoming more important. Deegan and Blomquist (2006) conclude that the disclosure of types of information can be useful for maintaining the support of particular stakeholders. An important disclosure opportunity are the

sustainability reports for particular environmental and societal stakeholders. García-Beneau et al. (2013) mention that growing CSR awareness has led companies to disclose CSR reports that extend the traditional financial information provided to shareholders to a wider range of stakeholders.

Rinaldi, Unerman and Tilt (2014) mention that motives for firms to engage in CSR and sustainability reporting could be of holistic concern to be responsible and accountable for firms impact, and on all those impacted by their actions. In the book of Rinaldi et al. (2014) is argued that there could be couples of other reasons to take into account stakeholder

opinions and their information demands in the reporting process, besides the growing demand for wider reporting. For example, Taking stakeholders into account leads to a more equitable and sustainable development of reports and enables understanding of firms complex operating environments (Rinaldi et al., 2004). Rinaldi et al. (2004) also suggest that it enables learning form stakeholders resulting in product and process improvement and lead to better management of risk and reputation. The process of identifying all stakeholders and their opinions is a complex process, but could lead to positive effects that can increase shareholder value.

(11)

11

2.3 The Concept of Corporate Social Responsibility

Various research focused on various dimensions of corporate social responsibility. García-Beneau et al. (2013) explain that CSR can be understood as an evolving concept, by which firms integrate economic, environmental and social concerns into their strategy and decision making process. Others like Garriga and Mele (2004) identified instrumental, political, integrative and ethical theories as corporate social responsibility dimensions. Whereas, Windsor (2006) come to the dimensions of ethical, economic and corporate citizenship. In figure 2 the Three-domain model of corporate social responsibility introduced by Schwartz and Carrol (2003) is presented with the three domains; economic, legal and ethical.

Figure 1; Three- Domain Model of Corporate Social Responsibility (Schwartz and Carrol, 2003)

Firstly, the economic domain refers to those activities maximizing profits and/or maximizing share value. Secondly, in the legal domain the focus is on firms responsiveness to legal expectations by society and their compliance, avoidance and anticipation of the law

(Schwartz and Carrol, 2003). Thirdly, the ethical domain refers to ethical responsibilities of firms expected by relevant stakeholders. Firms can report purely economic, legal or ethical, but with growing demand for information on corporate behavior firms report on a

(12)

12

disclosure is especially important for efficiently functioning of capital markets. These corporate disclosures could have several forms, such as; reports, footnotes, management discussions, analysis and forecasts, regulatory filings, press releases, internet sites and intermediaries who provide information (Healy and Palepu, 2001).

Corporate social responsibility can be executed, measured and reported in different ways. In our research we look to two variables which are common in CSR literature; firstly, the production of a sustainability report under the Global Reporting Initiative (GRI, 2016A) rules and secondly, the rating agency KLD Social Rating. To begin, the Global Reporting Initiative (GRI, 2015), seen as the pioneer in sustainability reporting since the late 1990s, builds on the three-domain model from Schwartz and Carrol (2003) and reforms the social corporate responsibility activities idea to an independent sustainability report, describes as:

The sustainability report is a report published by a company or organization about the economic, environmental and social impacts caused by its everyday activities (GRI, 2015).

In other words sustainability reporting is also known as corporate social responsibility reporting or triple bottom line reporting. GRI (2015) mention that the sustainability report presents organization’s values and governance model and demonstrates the link between its strategy and commitment to a sustainable global economy. Nowadays, GRI’s (2015)

sustainability reporting standards G4 are voluntary adopted by thousands of reporters and has become world’s most widely used standard on sustainability reporting and disclosure. The information disclosed in the sustainability report is according to Rinaldi et al. (2014) most commonly used by the stakeholders; shareholders and investors, insurers and banks, customers and suppliers, employees and trade unions, NGO’s, and the media.

Secondly, in the literature CSR can be measured with the Kinder, Lydenberg, Domini Research & Analytics, but is most common measured by the KLD Social Rating (Chatterji, Levine and Toffel, 2009; Cai, Jo and Pan, 2012). The KLD social rating agency measure firms strengths and weaknesses on different social responsibility aspects. Firms are rated on variables like community, diversity, employee relations, environment and product quality to evaluate firms social responsibility (Cai et al., 2012). Other exclusionary variables measured in KLD scores are alcohol, gambling, military, nuclear power, and tobacco (Cai et al., 2012).

(13)

13

Investors can use the KLD Social Rating for a more transparent view that contributes to identifying environmental responsible firms (Chatterji et al., 2009). The investor information needs for a combination of past environmental outcomes and recent firm policies for the predictable future are all accounted in the KLD rating, which give this rating an important value to investors (Chatterji et al., 2009).

2.4. The Concept of Investor Activism

Goranova and Ryan (2014) define investor activism as actions taken by shareholders with the explicit intention of influencing corporations’ policies and practices, rather than as latent intentions implicit in ownership stakes or trading behavior. Investor activists attempt to influence corporate behavior, but they aren’t in a quest for managing firms by themselves and take responsibility for executive decision (Goranova and Ryan, 2014). Corporate

behavior which is not in line with shareholders view is trying to be targeted and influenced. Also Gillan and Starks (2007) explain investor activists as shareholders who are dissatisfied with some aspect of a company’s management or operations and try to bring change within the company without a change in control. This investor activism process isn’t new, but takes place for decades (O’Rourke, 2003). For example, Brancato (1997) claims that US

shareholders have been actively pressuring companies to address social issues through the proxy voting process for more than 60 years. Nowadays, increasingly environmental and social activists turn into active lobbyist to influence firms behavior (O’Rourke, 2003). The practices of investor activism is currently used to address a wide range of issues, from executive compensation, to labor rights, to environmental responsibilities (Slater, 2007). Goranova and Ryan (2014) build a model to organize previous investor activism research (figure 2).

(14)

14

Figure 2. Investor Activism: antecedents, processes and outcomes (Goranova and Ryan, 2014).

Our research is especially focused on the antecedents and processes parts of investor activism research and not on the outcome aspect. The antecedents are classified by

Goranova and Ryan (2014) in levels of firm, activist and environment. The firm level research show us that larger firms, which are more susceptible to agency problems, are more prone to be targeted by investor activists (Cai and Walkling 2011). Major firms are hard to monitor and likely to be targeted. In line with the agency problem firm performance have the impact of a better operating performance leads to a decreasing investor activism (Ertimur, Ferri and Muslu, 2011). Better performance will satisfy shareholders by generating a profit. Another point researched is the alignment of managers and shareholders interest. Bizjak and Marquette (1998) found that managers holding higher shares are less likely to be targeted by shareholder activism. Owning shares reduce the agency problems and increase

alignment of managers and shareholders. The activist level is constructed on research with components of influence shareholder activists and their activities. Ertimur et al. (2011) mention that the cost and resource are of major importance for activism. High investment could raise barriers to entry in investor activism. Another factor is the ability to gain support by other stake- or shareholders (Gifford, 2010). To generate a powerful sound, the topic

(15)

15

needs to be supported by other groups to enlarge pressure executed on firms management. Besides, activist can be triggered and motivated by their social identity (Rowley and

Moldeveanu, 2003). Holistic views could be the driver for executing investor activism. On the environmental level of Goranova and Ryan’s (2014) research are mentioned the macro and task environment of firms. Activism on other firms could affect the firm itself in the long run. For staying legitimate in the market firms have to comply with stakeholder demands.

The process literature about shareholder activism can be classified in managerial actions, shareholder actions and activist demands (Goranova and Ryan, 2014). The managerial actions are the likely results when the majority of shareholders approve the shareholder proposal (Ertimur et al., 2011). Increased pressure on management, needs a change or adjustment of managements agenda. Another important point is the shareholder actions aspect. Traditionally, shareholders choose between the actions hold, by staying loyal to the firm, or trade and exit the firm as owner. Nowadays, clarifies O’Rourke (2003) the concept of investor activism with the sentence that investor activists try to influence corporate behavior by exercising ownership rights.

2.5 Shareholder Proposal

Investors can execute activism by either preparing and/or voting on shareholder proposals or by entering into dialogue with companies directly on issues (O’Rourke, 2003). Under the Securities Exchange Act Rule 14a-8 of 1934;

‘’Each investor that continuously holding shares worth $2,000 or by having a 1% market value of the equity for at least one year is allowed to include one (only one) proposal with a 500-word supporting statement in the proxy distributed by the company for its annual meeting. The requests for proposals have to be submitted at least 120 days before the proxy is mailed to investors’’ (Ertimur et al., 2011).

SEC Rule 14a-8 requires proxy materials to be disseminated on the expense of companies and be discussed on annual meetings (O’Rourke, 2003). These shareholder proposals proxies should also be included in their proxy statements (Goranova and Ryan, 2014; O’Rourke, 2003). O’Rourke (2003) describes a shareholder proposal as a document recommendation that a shareholder formally submits to a publicly traded company advocating the company

(16)

16

to take a specific course of action. With this a shareholder tries to get a point of voting on the annual general meeting (AGM). Ertimur et al. (2011) mention that firms could avoid voting processes by convincing proponents to withdraw their proposal, by making

concessions or agreeing to proxies. Firms could also avoid voting by asking the Securities and Exchange Commission (SEC) to exclude proposals if proposals violate some conditions, like ‘ordinary business matters’ or if it’s considered as ‘improper under the state of laws’ (Ertimur et al., 2011). If the proposal isn’t excluded or withdrawn, the voting comes to the annual meeting. By using an explanation statement provided by firms board, which include the issue on hand and boards opinion, shareholders can make their decision on the voting contest (Ertimur et al., 2011). Important to see is that proposals that indicate binding effects on firms behavior, are indicated as ‘improper’, because it limits a boards ability to exercise business judgment and its fiduciary role (Ertimur et al., 2011) This results in the circumstance that shareholder proposals are mostly non-binding recommendations to firms board, but non-binding doesn’t mean no influence. Brancato (1997) conclude that investors actively pressuring firm to pay attention to issues through the proxy voting process. Firms feeling this pressure and are adapting, not always, to the investors sound.

(17)

17

3. Hypothesis Development

In this section the hypothesis are developed. Consecutively, the concepts of corporate social responsibility related proposals and total proposals are explained.

3.1 Corporate Social Responsibility Related Shareholder Proposals

Borkowski, Welsh and Wentel (2011) mention that in the past 10 years there has been a growing demand for information on corporate behavior, including environmental impact, employee health and safety initiatives, community development, supply chain practices and a lot of other topics. KPMG (2013) published an international survey with the conclusion that in 2013 86% of the largest corporations in the USA published some form of social or environmental report, distinct from their financial report. Sustainability reporting remains voluntary, but research suggests two distinct theories on why firms are embracing: the ethical vs. the economic view (Borkowski et al., 2010). The ethical view is described by Borkowski et al. (2010) in the way that firms hold societal and moral obligations to engage in socially responsible activities and report on these activities, because it’s the “right thing to do”. This means that firms and management have a feeling to be responsible for cost induced by firms activities to society out of their own. Secondly, the economic view is clarified by Borkowski et al. (2010) with the argument that CSR reporting creates better corporate reputations that, in turn, create shareholder wealth via increased profits. This view is seen as staying legitimate by acting how society wants you to behave and not acting for ethical reasons on how a firm thinks it should behave. In the late 1990’s trends investors linking their environmental and social issues to financial performance, as well as to risks faced by the company, were rising (O’Rourke, 2003). Investors argue that environmental and social issues have increasing effects on firms value. A call for more disclosure requirements and pressure from society were seen as factor to encourage management members to keep an eye on environmental and social consequences of firms actions, as being socially

responsible (Jaggi and Freedman, 1992). The growing demand for corporate social responsibility information is caused by the idea of Porter and Miles (2013) that today’s shareholders and society is granting a license to operate to business imposes and

(18)

18

causes uncompensated cost on communities and the natural environment. Firms should be aware, but most of the time also need to compensate and report to remain legitimate. The stakeholder theory indicates that more disclosure of information should accomplish a

legitimate firm. CSR can be used as a tool, that could show organization objectives and social performance to investors and regain trust (Prado-Lorenzo, García-Sánchez and Gallego-Álvarez, 2012). Firms can manage public policy pressure by replying on shareholder proposals and increase transparency by corporate sustainability reporting (Clark and Crawford, 2012). Clark and Crawford (2012) mention that responding on shareholder proposals and disclosing information about one’s climate change strategy is a way to signal management’s changes in practices and future plans, which leads to decreasing investors pressure. CSR is influencing the information asymmetry of Akerlof (1970) in a way that it’s provide a more transparent image of firm policies and performance, because investors can access increasingly more important information (Gwendolen, 2016; Borkowski et al, 2010; Fernandez et al., 2014). If investors are uniformed or don’t agree with firms policies they can set up proposals and influence corporate behavior by exercising ownership rights (O’Rourke, 2003). The increasing transparence and information availability with a GRI sustainability report, decreases investor pressure (Clark and Crawford, 2012). This decreasing pressure could be resulting in a decrease in sustainability shareholder proposals. The reasoning of a growing shareholder demand for corporate social responsibility reporting and an increased reporting role of firms under the GRI sustainability reports, which covers most common topics of corporate sustainability reporting on social issues information, resulted in the following hypothesis.

H1a: Providing a GRI sustainability report decreases the quantity of corporate social responsibility related shareholder proposals submitted.

Besides the production of a GRI sustainability report, is argued that a corporate social responsibility measures, like the KLD Social Rating, could influence investors willingness to submit proposals. Firms that are really aware of their corporate social responsibility could set up their policies to align their actions to remain legitimate for their investors (Cornier and Ledoux, 2011). The KLD Social Rating measure corporate social responsibility on different

(19)

19

aspects and investors can access this data for their judgment about firms CSR performance (Parthiban, Bloom and Hillman, 2007). A higher score on the KLD index indicate that a firm is performing better on the corporate social responsibility aspects. The following hypothesis of this study looks to the KLD Social Rating index and the effects and reactions of investors, in the trend of shareholder proposals. The following hypothesis is made up:

H1b: A higher KLD Social Rating decreases the quantity of corporate social responsibility related shareholder proposals submitted.

3.2 Total Shareholder Proposals

There are couples of reasons why firms voluntary embrace sustainability reporting. Firstly, an overview of the CSR reporting literature clarifies the change in accountability with the introduction of CSR reporting. Haigh (2009) mentioned that investors are looking for more information from companies through for example sustainability reports. CSR reporting can help organizations to measure, understand and communicate their economic,

environmental, social and governance performance (GRI, 2016B). With this firms can set goals and manage change more effectively. Also Pleon (2005) say that recent international survey showed that most investors, financial research and rating agencies found these reports useful for their professional work. This demonstrates that the demand by society for increased accountability of firms can be met with the introduction of a sustainability report. Also Cooper and Owen (2007) conclude that the degree of administrative reform with a sustainability report has certainly been substantial in terms of the production of new accounting which heighten levels of organisational transparency and potentially in the interests of improved accountability. Secondly, CSR reporting influences the transparency of a company. With a larger variety of stakeholders calling for greater accountability and transparency, it’s obvious that increasing focus on corporate responsibility reporting and disclosure can give firms various advantages (Pleon, 2005). This is accompanied by the statement of Gwendolen (2016) who mention that sustainability reports provide a more transparent view of a company’s environmental, economic and social impacts. Stakeholders can access increasing data to make decision, which result in a decreased information

(20)

20

sustainability reports symbolize an ever growing demand by stakeholders for more transparency. With corporate social responsibility reporting, value is added to the transparency aspect. More information about firms actions is given by presenting

information that isn’t provided in the financial reports. Borowksi et al. (2010) conclude that corporate social responsibility reporting presents the non-financial factors that have

potential impact on aspects like; future performance, income generation and value

preservation. Aspects that where beforehand unknown by stakeholders are now observable. The effect that will be measured under hypothesis 1a/b could be not only concentrated on the social responsibility proposals, but may have a much bigger effect on the company as a whole. Ertimur et al. (2011) found growing evidence that shareholder proposals impact governance practices, increase the likelihood of declassification of firms board and the most important finding is that boards becoming more responsive to shareholder proposals. Monks et al. (2004) add to these argumentation that firms are rewarded by investors to incorporate corporate social responsibility in their policies. Investors see these longer term firm attitudes as less risky investments in the long run (Monks et al., 2004). The increased investor

confidence in firms board creates a perception that the company is responsible and

trustworthy (García-Beneau et al., 2013). Firms should take investors into account otherwise they will try to influence corporate behaviour, which can be accomplished by shareholder proposals. CSR tooling could decrease this from happening. Managers should consider how decision making affect society and try to achieve an adequate return on investment, while considering social and environmental impact of their actions (Cornier and Ledoux, 2011). Behaviour which isn’t aligned with investor idea’s will be targeted, by for instance

shareholder proposals. The CSR’s increasing accountability, transparency and trust leads to a more overall aligned way of corporate control between investors and management. By above reasoning the following hypothesis is set up:

H2a: Providing a GRI sustainability report decreases the quantity of total shareholder proposals submitted.

(21)

21

In line with hypothesis 1 is argued that a corporate social responsibility measures, like the KLD Social Rating, could influence investors willingness to submit proposals. KLD’s scoring on environmental performance and environmental management activities increase firms

transparency to investors, just like credit ratings enhance transparency and efficiency in debt capital markets by the reduction of information asymmetry (Chatterji et al., 2009). Firms paying extensive attention to their corporate social responsibility could increase their KLD social index scoring. On the other hand poor social, environmental and cultural performance on the KLD social index could harm firm performance and reputation (Chatterji et al., 2009). Increasing investor confidence in firms corporate policies by an increase in corporate social responsibility results is seen as more responsible and trustworthy (García-Beneau et al., 2013). There can be argued that a higher KLD score could lead to more trust by investors and a decreasing willingness to influence corporate behavior, because investor and

managements ideas are more aligned. With above reasoning the following hypothesis is constructed.

H2b: A higher KLD Social Rating decreases the quantity of total shareholder proposals submitted.

(22)

22

4. Research Method

In this paragraph is clarified how this research is conducted. Firstly, the used sample is described and the independent and dependent variables are introduced. Afterwards, the control variables are presented. Lastly, the introduced variables are used to build up four different models.

4.1 Sample, Variables and Method

This study search for the answer on the question if corporate social responsibility influence investor activism (figure 3). To get an answer on this research question a quantitative database research method is used. There is a focus on the S&P 500 firms, because these major firms are significantly targeted by investor activism and mostly produce some form of CSR reporting. For the period 2006 – 2013, the time of rising CSR importance, all data about GRI sustainability reports (GRI, 2016A), KLD Social Ratings and shareholder proposals are collected. S&P 500 firms are selected which creates a sample of 5549 observations. After the elimination of inactive market participants, KLD missing values, negative sales and missing values for board independence, the final data sample consists of 3722 firm-year observations at 523 firms.

Figure 3: Influence of Corporate Social Responsibility on Investor Activism

To specify corporate social responsibility two different approaches are used, the GRI sustainability report and the KLD social rating (figure 4). The first approach measures CSR only by the production of a GRI sustainability report, to see differences in investor activism when a GRI sustainability report is produced. After correspondence with the Global

Reporting Initiative (GRI, 2016A) access is obtained to all the GRI Sustainability Reports produced. For all firms listed in the S&P 500 there is a dummy variable created to divide the

(23)

23

firms in a group which have a GRI sustainability Report over a year; 1, and 0; otherwise, a group that doesn’t provide a GRI report over the year.

Secondly, the KLD Social Rating, that has been extensively used in various studies (Parthiban et al., 2007; Cai et al., 2012). In the study of Parthiban et al. (2007) investor activism is matched with CSR by using the KLD Social Rating to calculate firms CSR performance. In this study the same approach is used to measure corporate social responsibility performance. Firms KLD Social rating is made up out of the variables; community, diversity, employee relations, environment and product quality ratings. The scores given to each company are collected from COMPUSTAT database. Each year strengths are added and weaknesses are subtracted to composite a number, the KLD Social Rating number (Parthiban et al., 2007).

The dependent variable investor activism is in the literature extensively measured with the shareholder proposals proxy, which is mandatory to report under the SEC rule 14-a8 (Goranova and Ryan, 2014; O’Rourke, 2003; Parthiban, Hitt and Gimeno., 2001) (figure 4). Firms are required to disclose information like, shareholder proposals, subject, outcomes and sponsors. We obtain this variable from the COMPUSTAT database.

(24)

24

4.2 Control Variables

Consistent with prior research is argued that investor activists could create more value by targeting large firms, because these firms are hard to monitor and more prone to agency problems (Clark and Crawford, 2012; Goranova and Ryan, 2014; Jensen and Meckling, 1976). Subsequently, in the literature is mentioned that it’s more attractive for activist, because the greater visibility (Goranova and Ryan, 2014). By above reasoning is concluded that firm size has a positive effect on investor activism and in this study is controlled for size by the log of sales from COMPUSTAT (Parthiban et al., 2007).

In the literature is economic performance seen as an important control variable (Ertimur et al., 2011; Parthiban et al., 2001; Parthiban et al., 2007). Reasoned is that, consistent with the agency theory, companies which have an better economic performance tend to be less attractive to activists, because of good performance (Ertimur et al., 2011). Concluded is that economic performance is negative related to investor activism and controlled with the measure ROA (Ertimur et al., 2011; Parthiban et al., 2007).

Ertimur et al. (2011) mention that independent board members play an important role. Subsequently, Bizjak and Marquette (1998) find results that it’s less likely to be

targeted by investor activism when managers holding higher rates of shares. Also Kesner and Johnson (1990) are reasoning that more investor activism is carried out when the

percentage of outside board members is higher, because outside members are, in line with the agency theory, seen as less aligned with firms performance, instead of inside board members. Board independence have a positive influence on investor activism. This research uses the measure for director independence on the COMPUSTAT-ISS database, which

classifies board members in various groups; E-employee/insider, I-Independent, L-Linked and NA-notascertainable. The number of independent directors is calculated by dividing the quantity of independent board members at the annual meeting after year ends by the total number of board members (Kesner and Jonson, 1990; Ertimur et al., 2011).

4.2 Model Building

The dependent, independent and control variables are summarized in table 1 and are used to formulate four OLS models. To use an OLS regression the sample of 3722 firm-year observations have to meet several assumptions. Checked and confirmed is that all the

(25)

25

variables are measured continuous or categorical. By using scatterplots of the dependent, independent and control variables the linear relationship is confirmed. To check and exclude significant outliers, the variables CIASR, TCIA, KLDsr, LogTSales, ROA and BoardInd are

winsorized. The GRIreport variable isn’t windsorized, because this is a dummy variable which only can indicate a 1; if a firm produce a GRI report and, 0; otherwise. The independence of observations is tested and confirmed with a Durbin Watson test. There is also checked for residuals, homoscedasticity, skewness and kurtosis with the following three tests; a plot, a Cameron and Trivedi’s decomposition of M-test and a Breusch-Paga/ Cook-Weisberg test, no problems are found. After performing a Kernel density estimation, checking for normally distributions and, if necessary, gladder and adjust for non-normal distributions the criteria for an OLS regression are met.

For hypothesis 1a and 1b the effect of sustainability reports and KLD Social Ratings on the corporate social responsibility aspect of investor activism is examined. A cumulative count variable, named Cumulative Investor Activism on Social Responsibility related issues (CIASR) (Parthiban et al., 2007) is used. This variable is build up by counting the times of CIASR shareholder proposals, called SRI in COMPUSTAT, under SEC rule 14-a8 each year from 2006 till 2013 for each firm. Hypothesis 1a, where the effect of GRI sustainability reports is examined on the social responsibility aspect of investor activism, is tested with the following OLS model:

Model 1; CIASR = α0 + ϐ1 GRIreportt=0 + ϐ2 ROAt=0 + ϐ3 logTSalest=0 + ϐ4 BoardIndt=0 + ε.

Secondly, the effect of the KLD Social Ratings is examined on the social responsibility aspect of investor activism. In line with above model, hypothesis 1b is tested with the following OLS model,

Model 2; CIASR = α0 + ϐ1 KLDsrt=0 + ϐ2 ROAt=0 + ϐ3 logTSalest=0 + ϐ4 BoardIndt=0 + ε.

For hypothesis 2a and 2b the total effect of GRI sustainability reports and KLD Social Ratings on investor activism is measured. A cumulative count variable, named Total Cumulative Investor Activism (TCIA) is used (Parthiban et al., 2007). This variable is build up by counting the times of total shareholder proposals under SEC rule 14-a8 each year from 2006 till 2013.

(26)

26

The OLS model for hypothesis 2a, where the effect of GRI sustainability reports on total investor activism is tested, is designed:

Model 3; TCIA = α0 + ϐ1 GRIreportt=0 + ϐ2 ROAt=0 + ϐ3 logTSalest=0 + ϐ4 BoardIndt=0 + ε.

The hypothesis 2b test the effect of KLD Social Ratings on total investor activism and following OLS model is developed:

Model 4; TCIA = α0 + ϐ1 KLDsrt=0 + ϐ2 ROAt=0 + ϐ3 logTSalest=0 + ϐ4 BoardIndt=0 + ε.

Table 1. List of Variables.

Variables Descriptions

CIASR Cumulative Investor Activism on Social Responsibility Issues TCIA Total Cumulative Investor Activism

GRIreport GRI Sustainability Report, measured by a dummy variable; 1 if a firm produce a GRI Sustainability Report and; 0 otherwise

KLDsr KLD Social Rating

ROA Return On Assets

LogTSales Log Total Sales BoardInd Board Independence

(27)

27

5. Results

In this paragraph results of the quantitative database research are presented. Firstly the descriptive statistics of the sample are described. Afterward the Pearson and Spearman’s Rho correlation matrixes are presented and lastly the regression result are analyzed.

5.1 Descriptive Statistics

The sample contains 3,722 firm-year observations from 523 firms listed in the S&P 500 during the timeframe 2006-2013. Table 2 provides the descriptive statistics overview of the variables used in this study. Shown are the total observations, mean, standard deviation, minimum, maximum, first quartile, median and third quartile of the used data. The CIASR variable show a mean of almost 0.5 which means that each firm have on average a half proposal subjected to social responsibility issues, with a maximum of 5 each year. For the TCIAT there is logically a higher mean, because social responsibility proposals are part of the total proposals measured under TCIAT, approximately 1.34 on average and a maximum of 11 proposals each year. The GRIreport variable is measured by a dummy variable, 1 if a firm produce a GRI report and, 0 otherwise. The statistics show that on average 19.5 % of the 523 firms produced a GRI sustainability report from 2006 till 2013. In table 3 an overview is given about the quantity of CSR reports produced under the GRI standards each year. These results are far below the result of the 86 % in the year 2013 of KPMG’s (2013) research, but a rising trend from 6.38 % in 2006 to 34,12% in 2013 is ascertained. This difference is caused by the fact that KPMG (2013) researched only the biggest USA firms and not the whole S&P 500. Another point is that KPMG (2013) measure corporate social responsibility reporting and not specific on sustainability reporting under the GRI standards, also other forms are included. This study argue that GRI is the most important one and most useful for this research. The KLD social rating score, which is a score consisting on the community, diversity, employee relations, environment and product quality of firm, shows a mean of approximately 0.48 and only a very small standard deviation, which suggest little differences between firms. These results are consistent with previous research of Cai et al. (2012) and Parthiban et al. (2007). The control variable log of the total sales aren’t consistent with Parthiban et al. (2011) research, because now the years from 2006 till 2013 are used and

(28)

28

back than the years 1992 till 1998. The variable ROA provides a 6.32 % growth of the return on assets each year. These results are in line with findings of Ertimur et al. (2011) and Cai et al. (2012). Board independence is on average 0.8146 which means that 81,46 % of the board members are independent. Another point shown in table 2 is that there is never; a fully independent board (maximum is approximately 93%) or a totally none independent board (should be 0%), but the minimum is 50 % of independent board members. The founded results are consistent with the findings of Parthiban et al. (2007) and Ertimur et al. (2011).

Table 2. Descriptive Statistics.

Variables Observations Mean Std. Dev. Min. Max. First

Quartile Median Third Quartile CIASR 3,722 0.48979 0.951965 0 5 0 0 1 TCIAT 3,722 1.33853 2.073978 0 11 0 1 2 GRIreport 3,722 0.19506 0.396297 0 1 0 0 0 KLDsr 3,722 0.48401 0.069500 0.338811 0.687413 0.434965 0.471329 0.524301 LogTSales 3,722 3.89733 0.528813 2.777429 5.202197 3.519695 3.870611 4.218457 ROA 3,722 0.06317 0.069931 -0.192778 0.274671 0.023965 0.056642 0.099929 BoardInd 3,722 0.81463 0.100267 0.5 0.928571 0.750000 0.833333 0.900000

Notes: Table 2 provides the descriptive statistic for the sample with 3,722 firm-year observations at 523 firms listed in the S&P 500 during

the time period 2006-2013. CIASR = Cumulative Investor Activism on Social Responsibility issues ; TCIA = Total Cumulative Investor Activism; GRIreport = GRI Sustainability Report, measured by a dummy variable, 1 if a firm produces a GRI report and, 0 otherwise; KLDsr = KLD Social Rating; ROA = Return On Assets; LogTSales = Log Total Sales; BoardInd = Board Independence.

Table 3. GRI Sustainability Report Overview.

Year Total GRI reports provided Total companies in sample Percentage Firms producing a GRIreport

2006 27 423 6.38 % 2007 40 445 8.99 % 2008 56 450 12.44 % 2009 65 458 14.19 % 2010 83 469 17.70 % 2011 130 478 27.20 % 2012 152 492 30.89 % 2013 173 507 34.12%

Notes: table 3 show the total sustainability reports produced under GRI guidelines, resulting in the percentage of firms with

(29)

29

5.2 Correlation Matrix

To check for multicollinearity between the variables a Pearson correlation matrix is created in table 4. For the Pearson correlation all the variables are checked to be normally

distributed and if necessary a gladder for the non-normally distributed variables is executed. By using a gladder the BoardInd variable is transformed in an square_BoardInd variable which is normally distributed and used in the correlation test in Table 4.

The matrix shows the strength and direction of the association between the variables used in this research. Multicollinearity can occur when variables correlate too much with each other, which could create strange results for the regression. Multicollinearity can be a large concern when the correlation matrixes show a value higher than 0.70 (Neter,

Wasserman and Kutner, 1987; Cohen, 1988). In the Pearson correlation matrix presented in table 4 there aren´t any figures that cause concerns. Only the 0.7990 relation between CIASR and TCIAT can be concerning, but isn´t relevant because these variables aren´t used in the same regression model. The relationship is logically high, because CIASR is part of the TCIAT variable. Also the variables KLDsr – GRIreport aren´t used in the same model and the

correlation isn´t relevant. Overall the correlations are quite low, but most concerning is the 0.5493 association between LogTSales and TCIAT. This number is clarified by the fact that literature states that investors could create more value by attacking larger firms, because of greater visibility and agency problems (Clark and Crawford, 2012; Goranova and Ryan, 2014). Multicollinearity isn’t found in this study. Also the Spearman’s Rho Correlation matrix in table 5 don’t show significant differences with the Person correlation matrix and any concerns for multicollinearity. To check for interrelated data a variance inflation factor (vif) table is created for each model. These vif figures show low values and no indication of interrelated data.

(30)

30

Table 4. Pearson Correlation Matrix

Variables CIASR TCIAT GRIreport KLDsr LogTSales ROA BoardInd

CIASR 1.0000 TCIAT 0.7990*** 1.0000 GRIreport 0.2147*** 0.2103*** 1.0000 KLDsr 0.0249 0.03292** 0.3308*** 1.0000 LogTSales 0.4859*** 0.5493*** 0.3108*** 0.1928*** 1.0000 ROA 0.0224 -0.0398** 0.0323** 0.0843*** -0.0148 1.0000 Squarert_ BoardInd 0.1104*** 0.1421*** 0.1648*** 0.1543*** 0.2225*** -0.0637*** 1.0000

Notes: Table 4 provides the Pearson correlation matrix for the sample with 3,722 firm-year observations at 523 firms listed in the S&P 500

during the time period 2006-2013. *; ** and *** show the significance at the 10, 5 and 1% level. CIASR = Cumulative Investor Activism on Social Responsibility issues ; TCIA = Total Cumulative Investor Activism; GRIreport = GRI Sustainability Report, measured by a dummy variable, 1 if a firm produces a GRI report and, 0 otherwise; KLDsr = KLD Social Rating; ROA = Return On Assets; LogTSales = Log Total Sales; BoardInd = Board Independence. The correlations between TCIAT – CIASRT and KLDsr – GRIreport aren’t relevant, because these variables aren’t used in the same model.

Table 5. Spearman’s Rho Correlation Matrix

Variables CIASR TCIAT GRIreport KLDsr LogTSales ROA BoardInd

CIASR 1.0000 TCIAT 0.7223*** 1.0000 GRIreport 0.2030*** 0.2059*** 1.0000 KLDsr 0.0335** 0.0441*** 0.3035*** 1.0000 LogTSales 0.4283*** 0.4945*** 0.2976*** 0.1866*** 1.0000 ROA 0.081 -0.0478*** 0.0342** 0.0604*** -0.0288* 1.0000 Squarert_ BoardInd 0.1382*** 0.1761*** 0.1860*** 0.1611*** 0.2717*** -0.0991*** 1.0000

Notes: Table 5 provides the Spearman’s Rho Correlation matrix for the sample with 3,722 firm-year observations at 523 firms listed in the

S&P 500 during the time period 2006-2013. *; ** and *** show the significance at the 10, 5 and 1% level. CIASR = Cumulative Investor Activism on Social Responsibility issues ; TCIA = Total Cumulative Investor Activism; GRIreport = GRI Sustainability Report, measured by a dummy variable, 1 if a firm produces a GRI report and, 0 otherwise; KLDsr = KLD Social Rating; ROA = Return On Assets; LogTSales = Log Total Sales; BoardInd = Board Independence. The correlations between TCIAT – CIASRT and KLDsr – GRIreport aren’t relevant, because these variables aren’t used in the same model.

(31)

31

5.3 Regression Analysis

In table 6 the results of the OLS regression models 1 and 2 are presented. Model 1 identify the association between the quantity of social responsibility related shareholder proposals and the production of a GRI sustainability report, with 1; if a firm produces a GRI

sustainability report and 0; otherwise. Model 1’s adjusted R-square indicates that 24.05 % of the quantity of social responsibility related shareholder proposals is explained by the

independent variables. The regression results show a positive effect if a firm produce a GRIreport and a significant outcome at the 1% level. It indicates that the production of a GRIreport result in a 16.76 % rise of social responsibility related shareholder proposals. This finding is contrary the findings of Clark and Crawford (2012) who implicate that an increase in transparency by corporate sustainability reporting result in decreasing investor pressure. The result could be explained by the fact that more information availability raises question about the information provided. This was explained by Oded and Wang (2010) who suggest that exogenous governance mechanisms leads to more frequent, but less intense investor activism. Oded and Wang (2010) built on the argumentation that when quality of

governance mechanisms increases the returns on investor activism are decreasing, but the high quality of exogenous mechanism suggest that investor activism is more effectively. It seems that investors expect to heard by the firm. Concluded is that a GRI sustainability report is associated with the quantity of sustainability related shareholder proposals, however this effect isn’t a decrease of the quantity, like H1a suggest, but an increase of proposals which result in failing to support hypothesis 1a. Controlling for the size of companies results in a positive outcome at a significance level of 1 %. This indicates that firms with higher sales are increasingly charged with social responsibility related shareholder proposals. This is in line with the findings that larger firms are more prone to attract investor activism by the research of Clark and Crawford (2012) and Goranova and Ryan (2014). Another control is executed by the economic performance of firms which is measured with the return on assets variable (ROA). A positive effect at a significance level of 10 % is found. This indicates that an extra social responsibility related shareholder proposal increases the return on assets by approximately 36.55 %. A better performing company will attract more CSR related proposals. This is in contrast with findings of Ertimur et al. (2011) that good

(32)

32

performance leads to less attractiveness for investor activists. This can be related to the size aspect of the company when sales are growing, a firms size is growing as well and this attract more proposals. The high p-value of the board independence variable suggest there is no significant linear dependence between board independence and the social responsibility related shareholder proposals, which was indeed found by the study of Kesner and Johnson (1990) and Ertimur et al. (2011).

Model 2 identifies the quantity of social responsibility related shareholder proposals and the association with the KLD social rating score. The adjusted R-square indicates that 24.17 % of the quantity of social responsibility related shareholder proposals is explained by the independent variables of model 2, almost the same as model 1. Hypothesis 1b, suggest by the reasoning that firms could set up CSR policies to align their actions to remain

legitimate to investors. The increased CSR performance would causes decreasing investor pressure, resulting in decreasing social responsibility related shareholder proposals submitted (Cornier and Ledoux, 2011; Clark and Crawford, 2012). In line with above

reasoning the results in table 6 show that firm with more CSR related proposal have alower KLD social rating score at a significance level of 1 %. Also could be concluded that a higher KLD social rating leads to less social responsibility related shareholder proposals which supports hypothesis 1b. Controlling for size and performance significant results are found, in line with model 1. In other words, bigger in size and better performing firms are more prone to conceive social responsibility related proposals. Same as in model 1, insignificant results are found for the association between board independence and CSR proposals.

Table 6. Model 1&2; CSR Shareholder Proposals and GRI Reports/KLD score.

Model (1) Model (2)

CIASR Coefficient Std. Error t-statistic P-value Coefficient Std. Error t-statistic P-value

GRIreport 0.167644*** 0.036344 4.61 0.000 KLDsr - 1.04581*** 0.201588 -5.19 0.000 LogTSales 0.837884*** 0.027538 30.43 0.000 0.896897*** 0.026745 33.54 0.000 ROA 0.365479* 0.195104 1.87 0.061 0.505415*** 0.195678 2.58 0.010 BoardInd - 0.33884 0.140328 - 0.24 0.809 0.125104 0.140470 0.89 0.373 N 3722 3722 R² 24.14 % 24.25 % Adjusted R² 24.05 % 24.17 %

Notes: Table 6 provides the outcomes of the regression for model 1 and 2 for the sample with 3,722 firm-year observations at 523 firms

listed in the S&P 500 during the time period 2006-2013. *; ** and *** show the significance at the 10, 5 and 1% level. CIASR = Cumulative Investor Activism on Social Responsibility issues ; TCIA = Total Cumulative Investor Activism; GRIreport = GRI Sustainability Report,

(33)

33

measured by a dummy variable, 1 if a firm produces a GRI report and, 0 otherwise; KLDsr = KLD Social Rating; ROA = Return On Assets; LogTSales = Log Total Sales; BoardInd = Board Independence.

For the models 3 and 4, presented in table 7, the focus is changing from the social responsibility related proposals to the total shareholder proposals submitted. Model 3 focusses on the association between the production of a GRI sustainability report and the quantity of total shareholder proposals. The adjusted R-square is logically higher, because the dependent variable sustainability related shareholder proposals is part of the dependent total shareholder proposals variable of model 3 and 4. This result in an explanation increase to 30.77 %. The results, in table 7, show a significant association on a 1% level between the total shareholder proposals and the production of a GRI sustainability report. This result is a positive effect, in contrast with hypothesis 2a, which suggested a negative effect. An

increase in total shareholder proposals when firms producing a GRI sustainability report can again be clarified with the reasoning of Oded and Wang (2010) that more information availability raises more questions about the information disclosed and open new ways of thinking by investors. Managers take into account the social and environmental impact of decision making (Cornier and Ledoux, 2011). Decision which aren’t aligned with investors will be attacked. The GRI report discloses more information and rises the accountability and transparence of an organization, but this doesn’t mean that investors agree with the

presented information. The disclosed information in a GRI sustainability report could be not in investors’ interest, which could be a reason for the increasing shareholder proposals when a GRI report is produced. After controlling, significant results are found for the variables size and performance. Size is, in line with statements of Goranova and Ryan (2014), positive associated with the total shareholder proposals. Investors could create more value by targeting larger firms, because bigger firms have greater visibility (Goranova and Ryan, 2014). In contrast with model 1 and 2 a negative association for firm performance is found. This suggest that better performing firms have less total shareholder proposals. This is in line with findings of Ertimur et al. (2011) that good performance results in less investor

intervene, because investors are pleased with firms result. Just like the results in model 1 and 2, no association is found between shareholder proposals and board independence in model 3.

Lastly, model 4 search for the association between the quantity of total shareholder proposals and the KLD social rating score. Table 7 shows the results with an adjusted

(34)

R-34

square of 30.70 %, which is the percentage model 4 clarifies of the variances in total shareholder proposals. The adjusted R-square is in line with model 3 and logically higher than model 1 and 2, because the CSR proposals are part of the total proposals. Hypothesis 2b is built on the reasoning of García-Beneau et al. (2013) who argue that an increasing KLD social rating score lead to more investor trust and a decreasing willingness to influence corporate behavior, a higher KLD score decreases the quantity of total shareholder proposals submitted. This statement is supported with significant results on a 1% level in model 4. With these results is argued that paying attention to firms own corporate social

responsibility to increase the KLD score decreases investors willingness to influence firms behavior (Chatterji et al., 2009). Model 4’s control variables size and performance are significant and indicate that bigger firms performing worse are more attacked by investors with shareholder proposals. Good performance indicates better alignment of investors and firm policies (Parthiban et al., 2007). For the first time board independence have a significant association at a 10% level with the dependent variable, total shareholder proposals. In line with the reasoning by Kesner and Johnson (1990) a positive association is found which indicates that extra shareholder proposals increase the quantity of independent board members. More outside board members increase the independence of the board. Outside board members are less aligned with firms performance and this agency problem does increase the quantity of shareholder proposals submitted (Kesner and Johnson, 1990). Inside board members are seen as more aligned by investors and create logically less agency

problems, which result in a lower quantity of shareholder proposals.

Table 7. Model 3&4; Total Shareholder Proposals and GRI Reports/KLD score

Model (3) Model (4)

TCIA Coefficient Std. Error t-statistic P-value Coefficient Std. Error t-statistic P-value

GRIreport 0.228670*** 0.075799 3.02 0.003 KLDsr -2.096377*** 0.419857 -4.99 0.000 LogTSales 2.087156*** 0.057432 36.34 0.000 2.182296*** 0.055702 39.18 0.000 ROA -0.961860** 0.406910 -2.36 0.018 -0.707808* 0.407548 -1.74 0.083 BoardInd 0.282400 0.292669 -23.72 0.335 0.556521* 0.292564 1.90 0.057 N 3722 3722 R² 30.48 % 30.40 % Adjusted R² 30.77 % 30.70 %

Notes: Table 7 provides the outcomes of the regression for model 3 and 4 for the sample with 3,722 firm-year observations at 523 firms

listed in the S&P 500 during the time period 2006-2013. *; ** and *** show the significance at the 10, 5 and 1% level. CIASR = Cumulative Investor Activism on Social Responsibility issues ; TCIA = Total Cumulative Investor Activism; GRIreport = GRI Sustainability Report,

(35)

35

measured by a dummy variable, 1 if a firm produces a GRI report and, 0 otherwise; KLDsr = KLD Social Rating; ROA = Return On Assets; LogTSales = Log Total Sales; BoardInd = Board Independence.

6. Discussion

In this section the results are being analyzed and discussed. After in depth reasoning, an answer on the research question will be drawn up. Subsequently, the limitations of the research will be discussed and afterwards ideas for new research are presented

6.1 Results Discussion

This study search for the question if there is an association between corporate social responsibility and investor activism. To find answers, four hypothesis and models are

developed. Investor activism is measured at a corporate social responsibility level and a total level. This study uses the GRI sustainability report and the KLD social rating score as

indicators of corporate social responsibility. The trend of increasing demand, pressure and attention on environmental and social aspects of firms behavior described by Clark and Crawford (2012) and Deegan and Blomquist (2006) and Haigh (2009) is also found in this study. With an increase of GRI sustainability reports from 27 (6,38 %) in 2006 to 173 (34.12 %) in 2013, a lower result than was found in the research of the KPMG (2013) (86%), a rising trend is shown. Firms are adapting to the norms and values of the society to stay legitimate, because failure could harm the firm. This is in line with the stakeholder theory which suggest an increasing accountability to more stakeholders to stay legitimate. To stay legitimate the corporate social responsibility actions should not only be implemented, but of great importance is the need to inform society about their activities and changes (Deegan and Blomquist, 2006). The information asymmetry between investors and firms could be reduced by corporate social responsibility reporting. By reasoning of Clark and Crawford (2012) was expected that public policy pressure could be managed by responding with the production of a GRI sustainability report, which increase corporate transparency and

accountability. This study find results that there is an association between the production of a GRI sustainability report and investor activism. Firstly, the quantity of corporate social responsibility related shareholder proposals is, in contrast with hypothesis 1a, positive associated with the GRI report which means that the production of a GRI sustainability

The influence of corporate social responsibility on investor activism

Amsterdam Business School