Is CEO compensation related to the level of CSR reporting? Maaike Knijn
10668187
Master thesis accountancy and control, specialisation control Date final version: 17 June 2015
Word count: 13058
MSc Accountancy and control, variant Control Amsterdam Business School
Faculty of Economics and Business, University of Amsterdam Supervisor: George Georgakopoulos
1.1 Background 2
Statement of originality
This document is written by student Maaike Knijn who declares to take full responsibility for the contents of this document.
I declare that the text and 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.
1.1 Background 3 Abstract
In the past few years, executive bonuses and corporate incentives have been strongly
criticized. The interest in CSR is growing. In Europe 71 percent of the companies report CSR. The main goal of CSR is to create value for stakeholders. In this thesis I want to research the relationship between the adherence level of CSR reporting and the amount of incentives of CEO. For the incentives of the CEO there will made a distinction in salary, bonuses and stock-options. For the dependent variable adherence level the GRI level of application will be used. In this research I focus on Europe in 2011 until 2013. In total, there are 130 firm year observations. The data is collected with CapitalIQ, Amadeus and Datastream. With this research there is show that there is a relationship between adherence level and salary and the adherence level and equity incentives. Contradictory there is no relationship between
adherence level and bonuses. It can conclude that the hypothese of equity incentives (H3) is rejected and the hypotheses of salary (H1) and bonuses (H2) can be accepted.
Keywords: CSR reporting, salary, bonuses, equity incentives, adherence level, CEO, voluntary disclosure
Data availability: All data incorporated in the examination of this research is available Through CapitalIQ, Amadeus and Orbis databases via University of Amsterdam.
1.1 Background 4
Inhoud
1. Introduction ... 6 1.1 Background ... 6 1.2 Research question ... 8 1.3 Motivation ... 9 1.4 Structure ... 112. Literature review and hypotheses ... 12
2.1 CSR ... 12
2.1.1 Why would a CEO report corporate social responsibility (CSR)? ... 13
2.1.2. Employee-organization CSR. ... 15
2.1.3 Measures instruments for CSR ... 15
2.2 Parts of incentives compensations ... 18
2.3 Incentives CEO ... 19
2.4 Empirical evidence on compensation incentives and CSR ... 20
2.5 Hypothesis ... 22 3. Research methodology ... 23 3.1 Data ... 23 3.2 Independent variables ... 25 3.3 Dependent variables ... 27 3.4 Control variables ... 27 3.5 Assumptions ... 29 4 Results ... 30 4.1 Descriptive statistics ... 31 4.2 Correlations coefficients... 31 4.3 Empirical results ... 33 4.4 Sensitivity analysis ... 34
5. Discussion and conclusion ... 36
5.1 Discussion ... 36
5.2 Contribution to academic literature ... 37
5.3 Important limitations ... 37
5.4 Directions for further research ... 37
Appendix: Reference ... 39
Appendix Variables ... 43
Appendix: Overview researches ... 44
1.1 Background 5
1.1 Background 6 1. Introduction
1.1 Background
“In the past few years, executive bonuses and corporate incentives have been strongly
criticised for encouraging excessive risk-taking and cost-cutting that created the conditions for the global financial crisis an subsequent economic recession” (Lorsch and Khurana, 2010). For example, In the Netherlands there was a debate with minister of Finance and diverse organisations that the directors get too much bonus. A recent example is ABN AMRO, ING bank and AEGON where the directors get too much salary and bonuses. (NRC, 2015)
In that context bonuses started to be reconsidered, with the emergence of so-called ‘sustainable bonuses’ as a new trend that might signal a move towards corporate
responsibility. (WBCSD, 2010; Conference Board, 2012). While pay-for-performance is not a new phenomenon, in recent years there has been a growing pressure to consider sustainability performance as a part of the executive compensation formula. An increasing number of multinationals, such as Intel, Alcoa, Group Danone, National Grid and Xcel Energy, have reportedly incorporated sustainability intro their bonus structure (Kolk and Perego, 2014). However sustainability will be more important the coming years. (KPMG, 2011)
Regulated financial reports and disclosures are informative to investors and add value to the capital market. According to Healy and Papelu (2001) disclosure is associated with the stock price performance, bid-ask spread, analysts’ following and institutional ownership. Furthermore, strategic alliances are significantly affecting the nature of financial reporting and disclosure.
Graham et al. (2005) suggest that managers make voluntary disclosures to promote a reputation for transparent reporting, reduce the information risks and boost stock price but at the same time, try to avoid setting disclosure precedents that will be difficult to maintain. Graham et al. (2005) also suggest that firms make voluntary disclosures for three main reasons: to promote a reputation for transparent reporting, to reduce the information risk assigned to the firm’s stock and to address the deficiencies of mandatory reports. The authors don’t find strong support that CFOs manage financial reporting practices in private firms to influence the bonus hypothesis, exercisable stock options, and exercisable stock options held by managers (Graham et al., 2005).
The study of Jones et al. (2009) in Australia suggest a strong statistical relationship between sustainability disclosure and operating cash flows, cash positions, working capital,
1.1 Background 7
asset backing, retained earnings, capital expenditure, debt servicing, capital structure and price-to-book ratio. Such a finding is consistent with earlier research from Gray et al. (2001) that found positive associations between firm characteristics and level of social and
environmental reporting. Accordingly to Jones et al. (2009) suggests that larger firms tend to reveal a statistical higher level of sustainability disclosure. Jones and Ratnatunga (1997) argue that the firm size hypothesis would predict that decision-relevant account information
generally would have greater economic consequences in the external reporting environments of larger firms given the dependence on a wider range of internal and external users. (Jones et al., 2009).
What is exactly the difference between corporate social responsibility (CSR) disclosure and sustainability reporting (SR)? Montiel (2008) shows in his analysis that the conceptualizations and measures of CSR and corporate sustainability (CS) seem to be
converging, even though there are points of difference between the two constructs. Out of the 112 U.S. companies enrolled in the GRI, Montiel (2008) find all kind of terminologies used to refer to the companies’ social and environmental reports. Companies have used titles as varied as “Global Citizenship Report,” “Corporate Responsibility Report,” “Environmental Sustainability Report,” “Sustainability Report,” and “Environmental & Social Responsibility Report” to refer to their annual document summarizing their social and environmental
initiatives. In this thesis the term CSR will be used because it is the most used term in articles. According to Orlitzky (2003) CSR can be measured by four broad strategies: 1) CSR
disclosures, 2) CSR reputation ratings, 3) social audits, processes and observable outcomes, and 4) managerial CSR principles and values. So sustainability reporting is a part of CSR.
Since the 1990s, several large companies have been publishing non-financial
performance reports. Focusing initially on the physical environment, these reports evolved to consider social relations, as well as data on the firm’s economic performance (Perez and Sanchez, 2009). Sustainability developments focus on the three P’s namely: profit, people and planet. In the S&P 500 Index 57 percent of the companies in America published a
sustainability report in May 2012, compared with 20 percent in America in 2006-2010 (Clark and Master, 2012). According to KPMG (2011) the traditional CSR reporting nations in Europe (71%) continue to see the highest rates, followed by America that are quickly follow by 69 percent. There are different standards for the sustainability reports such as GRI,
1.2 Research question 8
2.1.3. According to KPMG (2011) the GRI guideline is most used for the sustainability reports with 69 percent.
Rank 1 2 3 4 5 Countries UK France Spain Finland Netherlands Percentage 100% 94% 88% 85% 84% Rank 1 2 3 4 5 Industry
Forestry, Pulp and Paper Mining
Automotive
Communication and Media Utilities Percentage 84% 84% 78% 74% 71% Table 1: Countries and industries that report CSR (KPMG, 2011)
In table 1 there is show in which countries and industries in Europe where there report the most CSR. In this research there is also used a control variable for the industries divided in sensitive industries such as mining, metal and automotive and non-sensitive industries. In table 1 you can see that the highest reporting rates are in the sensitive industries (forestry, pulp and paper, mining, automotive and utilities).
The reason that has motivated companies to voluntarily disclose social, environmental and economic information has been researched in many different studies (GRI, 2013;
Christoffi, 2012; Searcy and Buslovich, 2013; Christofi et al, 2012). Advantages of
sustainability reporting are driving the need for more transparency and accountability (GRI, 2013). The advantages are to improve the internal process, engage stakeholders and persuade investors (Christoffi, 2012). Because there is a general lack of regulation, sustainability reporting is primarily voluntary in nature.
Many companies work with some kind of incentive system. By giving monetary incentives, companies seek to stimulate the behaviour of employees within the company, which then leads to certain organizational outputs and competitive advantages (Barney & Zajac, 1994).
1.2 Research question
This paper will explore the incentives that companies seek to create in order to stimulate the behaviour of their employees, have a positive influence on the level of CSR disclosure. Every company that publishes a CSR disclosure has an adherence level by GRI, the company gets a rate from A+ till C. See appendix: application level criteria for an explanation of the
1.3 Motivation 9
research will check when the CEO get a higher amount of incentive the adherence level of GRI will be higher. The research question will be as followed:
“Is there an association between the amount of incentives and CSR reporting?”
1.3 Motivation
There is a lot of research between the relationship of sustainability reports and financial performance. There are positive, negative and neutral relationships. Jones et al. (2009) finds a strong statistical relationship between sustainability disclosure and various performance indicators such as ratios and measures based on cash position, operating and free cash flow, working capital, profitability and earnings performance, turnover, financial structure, debt servicing capacity, capital expenditure, and a number of valuation multiples such as market-to- book value and the price-earnings ratio.
Belu and Manescu (2013) find a neutral relationship between the strategic CSR index and the economic performance as measured by return on assets (ROA) and Tobin’s Q.
Lo and Sheu (2007) find no significant change in the research which not supports that corporate sustainability increases market value. Whereas the main purpose of the paper of Lo and Sheu (2007) is linking corporate sustainability to firm value. For further research, one might extend this topic by incorporating other corporate governance variables (e.g.
managerial ownership, size of the board of directors, the proportion of outsiders on the board, CEO compensations, etc.). (Lo and Sheu, 2007)
Jizi et al. (2014), Michelon and Parbonetti (2010) and Shrivastava and Addas (2014) and more researchers focus on the subject sustainability and corporate governance. So there is already a lot of information about this subject but the writers didn’t research the variable CEO compensation. CEO compensation is also an incentive and makes an interesting variable for this research. CEO compensation is an interesting variable because the board compositions, media and public affects executive pay have been the subject of debate for decade (Gurthrie et al., 2012; Croci et al, 2012). Also after the accounting scandals to the Sarbanes-Oxley Act of 2002, NYSE and NASDAQ revised their listing standards to improve their corporate
governance in the companies (Croci et al, 2012). A recent example in the Netherlands is the dispute of bonuses of director between the ministry and companies as ABN AMRO, AEGON and ING Bank (NRC,2015)
1.3 Motivation 10
For this research there are three articles about CSR and incentives (Fabrizi et al., 2014), (McGuire et al., 2003) and (Mahoney, L and Thorne, L, 2015) where the writers focus on US and Canada. This research focuses on Europe from 2011 until 2013. I choose this years because this are recent years and there is also a dynamic of the bonus culture in the public sectors this years. The years before were in the financial crisis so that is irrelevant for this study. In “Appendix: overview researches” and paragraph 2.4. There is make a distinction of the three articles.
Fabrizi et al. (2014) splits between monetary and non-monetary incentives. This paper focuses on the difference between short-term incentives (salary and bonus) and long-term incentives (equity holdings) in Europe.
This paper brings a new scientific perspective in sustainability reporting and incentives in Europe by proxy and distinguishing between short-term and long-term
incentives. According KPMG (2011) Europe published more CSR reports than America. The independent variables that I use are variables of incentives. There is a lot of information about sustainability and financial performance, but the results are positive, negative. So, I want to research if the variables of short term or long term incentives enhance or weaken the effect of the level of CSR disclosures.
Through this it will contribute to the social point of view by expressing the importance the best incentives choice by CSR disclosures. The stakeholders of the company can use the information to make better decisions about sustainability reports. Furthermore, it enhances the understanding between level of sustainability reporting and incentives in EU.
In figure 1 there is show a Libby box. According Libby (1981) a Libby box is “a
framework that captures the researches concept and illustrate the research process”. The
Libby boxes includes five boxes. There are five variables that are independent, dependent, conceptual, operational and control variables. The first arrow from incentives to sustainability reporting shows the predicted effect. The arrow to incentives to short-term incentives and long-term incentives and the arrow from sustainability reporting to voluntary disclosure shows the construct validity. If the construct validly is higher the better the research is estimate (Macintosh and Quattrone, 2010). The arrow from short-term incentives and long-term incentives to voluntary disclosure describes relationship between incentives and sustainability reporting. This arrow is for the internal validity (Macintosh and Quattrone, 2010). Finally, the last arrow reflect the effect of other factors (control variables) on the outcome of sustainability reporting. The six control variables are: independent on the board,
1.4 Structure 11
leverage, ROA, log total assets, financial slack and industry. This control variables can increase the internal validity of this study. This because higher internal validity results in a better association between incentives and sustainability reporting. (Libby, 1981)
Figure 1: Libby boxes (Libby, 1981)
1.4 Structure
This research describes the relationship between the adherence (application) level of CSR reporting and the amount of incentives of CEOs. See appendix for the explanation of the application level. The second chapter there will be explained the CSR and incentives and the relationship between the CSR and incentives. The third chapter will describe the research methodology of this research. Furthermore the results of this research will be discussed. Finally, the conclusion and the limitations of this study will be developed.
Incentives (independent variable) Sustainability reporting (dependent variable) Short-term incentives and long-term incentives Voluntary disclosure
Independent on the board, leverage, ROA, log total assets, financial slack
and industry (control variables)
Conceptual variables
2.1 CSR 12 2. Literature review and hypotheses
This chapter describes as first the definition of CSR reporting. Next there will be explained why the CEO would report CSR, the relationship between employees and CSR and the measures instruments for CSR. After that I will describe the parts of the incentives of compensation. In 2.3 there will discuss about the influence of the incentives of CEO.
Furthermore the relationship between the incentives and the CSR reporting will be discussed. Finally, based on the prior literature in this chapter the hypothesis for this study will be developed.
2.1 CSR
According to KPMG (2011) more companies are publishing sustainability reports. The interest for CSR becomes greater because they get more value for CSR. There are different terms for CSR disclosures such as ‘‘sustainability,’’ ‘‘corporate social responsibility,’’ and ‘‘corporate citizenship’’ reports. The key is that the report, whatever its title, contains qualitative and quantitative information of interest to stakeholders on the company’s key sustainability issues and initiatives. (GRI, 2013)
The definition for sustainable development can be defined as “Development which meets the needs of present generations without compromising the ability of future generations to meet their own needs” (WCED, 1987, p.3). CSR can be defined as “In general, corporate sustainability and CSR refer to company activities – voluntary by definition – demonstrating the inclusion of social and environmental concerns in business operations and in interactions with stakeholders” (Van Marrewijk, 2003).
CSR reporting is sometimes mandatory for companies. For example in Sweden all state-owned companies need to report the CSR activities and in Denmark there is a reporting requirement for all listed companies (KPMG, 2011).
With increasing social and environmental concerns, CSR gives insight in a company's contribution on the environment and society as a whole. The term stakeholders is important with CSR. Stakeholders are the organisations and persons that concerned by an organisation. According to Falck and Heblich (2007) short-term action such as donating money for social purposes or sponsoring popular events are not the most effective means of attaining an effective management strategy. Effective CSR should be a long-term action. CSR should be seen as an investment for the company’s future. It is a way of actively contributing to the
2.1 CSR 13
society’s basic order and for enhancing the company’s reputation. A good reputation is necessary to attract, retain and motivate quality employees. So, a good reputation increases the value of the brand which will increase the company’s goodwill. (Falck and Heblick, 2007)
There are not many continental studies on this subject while 71 percent of the companies in Europe publish sustainability reports and therefore CSR become a more vital issue in this part of the world. Multinational enterprises face a multitude of different
institutional conditions. Civil society, reacts with hostility to such not sustainable behaviour, scandals such as Nike’s sweatshops are illustrative. Therefore, cost savings in the short run need to be balanced against the potential risk to the company’s goodwill in the long run. It will be in their own self-interest if companies would consider filling this regulatory gap (Falck and Heblish, 2005). Furthermore, according Cooper and Owen (2007) also in the view of the fall-out from Enron and similar affairs, reputation building appears to provide a primary motivating factor for companies going down the CSR path.
2.1.1 Why would a CEO report corporate social responsibility (CSR)?
In the previous paragraph it was suggested that CSR is not mandatory in all companies. For the CEO and the company, there are many reasons to publish CSR reporting. The following reasons for the importance of CSR will be described in this paragraph.
The first reason is based on the stakeholder theory. According to Deegan and
Unerman (2006), while legitimacy theory discusses more about the expectations of the society in general, stakeholders are discussing more about the particular groups within the society. Today, it is commonly accepted that under certain conditions the satisfaction of social
interests contributes to maximizing shareholder value and most large companies pay attention to CSR, particularly in considering the interests of people with a stake in the firm (Melé, 2008). Also Freeman and Velamuri (2006) have suggested that the main goal of CSR is to create value for stakeholders fulfilling the firm's responsibilities to them, without separating business from ethics.
The second reason that is based on agency theory. Agency theory describes why a company chooses a particular method, such as CSR disclosure. According to Jensen and Mecking (1976) the agency theory is defined as “a contract under which one or more persons
(the principals engage other person (agent)) to perform some service on their behalf which involves delegating some decision making authority to agent”. Building upon the behavioural
2.1 CSR 14
agency model (Wiseman and Gomez-Mejia, 1998) it could be argued that ownership stake may lead to more conservative strategies as managers seek to preserve their wealth. Such wealth preservation motives may lead managers to avoid potentially costly socially dubious behaviour.
According to Jiraporn and Chintrakarn (2013) the agency view suggest that CSR investments are made to enhance the private benefits of managers. Therefore, when the CEO is more powerful, he/she expects to engage more in CSR (to enhance his own private
benefits). Nevertheless, when CEO power reaches a certain threshold, the CEO reduces CSR investment significantly. However, when the CEO consolidates his power to a certain point, he is so entrenched and invulnerable that he/she no longer views CSR favourably and
therefore reduces CSR investment significantly.
The last reason that I put forward is institutional theory. According to DiMaggio and Powell (1983) institutional theory can be defined as “Organizations adopting innovations not
because they are efficient but to legitimize themselves within their cultural and institutional context”. Institutional theory defines that a company feels pressure from several stakeholders.
This pressure affects the way of thinking about CSR reporting. There can be a pressure from society that the company needs to implement CSR disclosure. According to Falck and Heblich (2007), if CSR is understood as a voluntary corporate commitment that helps
establish social trends and institutional demands, the differences in its level of importance for countries becomes evident. As long as society's basic order binds companies to comply with social demands, there is no necessity for CSR. Not until society's basic order is unable to represent social trends in an appropriate way will CSR come into play.
Some important benefits that CEO want to implement CSR are, according to Christofi et al. (2012), which CSR increase transparency because the CEO provide information that would otherwise not have been disclosed. Second, they can change the internal management practices. Third, the CEO can reinforce the relationship between firms and local communities by explaining the contribution of companies to these communities. According to KPMG (2011) the growth of this green products market is directly related to many of these drivers. For example, their availability of CSR increase brand reputation.
2.1 CSR 15 2.1.2. Employee-organization CSR.
Discussed in section 2.1.1. CSR create value for stakeholder and engage stakeholders with CSR. But what is exactly the effect of CSR on employees. The employees need to see the advantages of CSR, otherwise the employees are not motivated to implement CSR reporting in the organization. A success implementation of CSR reports depends on the interest of the managers and employees of the organization. Employees are a key stakeholders group. The CEO cannot implement CSR in his own, so the employees need to be motivated. In the next paragraph there will explain the effect of CSR on the (prospective) employees.
Stakeholders get more value for CSR that is the reason that directors want to implement CSR reporting. To motivate the CEO to implement CSR they can get bonuses when the CEO implement the CSR objectives well.
There are advantages to implement CSR for the employees’ work-life. Singhapakdi et al. (2015) mentioned that an organization that use CSR can be another way for people to behave in a socially responsible manner. It is possible that people are more likely to join an organization with CSR. Singhapakdi et al (2015) found a strong evidence that CSR is a strong predictor of work life of employees. They mentioned that CSR is not only good for strategic motives but also it has a positive impact on the quality of work life of the employees of the organization.
Farooq et al. (2014) mentioned that the research have significant implications for CSR strategies of a firm. As a mentioned before CSR helps create a competitive advantages. The paper of Farooq et al. (2014) gives also evidence that CSR helps to create a competitive advantage. They suggest that firms with high CSR practices attain high productivity because of the motivation and knowledge sharing among their employees. This motivation and knowledge sharing among their employees also reduce the absenteeism, greater extra-role behaviour and lowered turnover costs.
2.1.3 Measures instruments for CSR
There are different reporting frameworks for CSR that are the Global Reporting Initiative (GRI), AccountAbility (AA), Fédération Des Experts Comptables Européens (FEE), International Standard for Assurance Engagements (ISAE 3000) and Royal Dutch Nivra
2.1 CSR 16
3410. The principles of this standard are inclusivity, materiality and responsiveness. This can be used by non-accountants as well as accountants. (AA1000, 2008) Second, there is the FEE a non-profit organization, which use non-audit assurance standards (FEE, -). Third, there is the ISAE 3000 and fourth there is the Royal Dutch Nivra 3410.
The most important standard for CSR is GRI. GRI is a network-based organization that produces a comprehensive sustainability reporting framework that is widely used around the world (KPMG, 2011). GRI is an important driver of CSR disclosure, 80 percent of G250 and 69 percent of N100 companies adhere to GRI Sustainability reporting guidelines. This reporting guideline published the principles and standard disclosures organisations can use this guidelines to report their economic, environmental and social performance and impacts. The guidelines assist in the preparation of sustainability report by all types of organisations. The guidelines are developed through a global multi-stakeholder process which involving representatives such business as well the auditors, experts and governmental agencies (GRI, 2013). The latest development of GRI are the GRI Reporting Guidelines (G4), developed in 2013.The substantial divergence observed between companies in various regions has raised an urgent need for users to improve the quality of CSR reporting by adopting a reliable universal standard, with the main aim of improving the robustness, reliability, credibility and
consistency of reporting practices (Toppinen et al. , 2012). Example of the application level criteria are in the appendix “Application level criteria”.
“GRI's mission is to make sustainability reporting standard practice for all companies and organizations. Its Framework is a reporting system that provides metrics and methods for measuring and reporting sustainability related impacts and performance.” (GRI, 2013).
The most recent version of GRI framework is the G4 version that introduced an Application Level procedure to “demonstrate a pathway for incrementally developing, expanding, and deepening approaches to reporting over successive cycles” (GRI, 2013). The procedure helps companies to gauge their maturity levels in sustainability reporting. It guides organizations to self-declare their reporting level (A+, A, B+, B, C+, or C), or hire an external organization to verify their self-declaration. Every type organisation (big/small, different types of industries) can use the GRI application
With this Application Level I want to show what kind of level the CSR reports are in the company. A+ is advanced level of CSR reporting and C is entry-level. To indicate that is GRI-based the company should declare the level to which they have applied the GRI
2.1 CSR 17
advanced reporters and those somewhere is between there are three levels in system. See appendix for further explanation of the application level criteria. I want to use this with the research with the Likert scale, 1 = C and 6 = A+ (GRI, 2013)
Follow GRI (2013), CSR reporting needs to include six dimensions; economic, environmental, labour practices and decent work, human right, society and product responsibility. The item below is an explanation of the six dimensions.
The economic dimension of sustainability concerns the organization’s impacts on the economic conditions of its stakeholders and on economic systems at local, national, and global levels.
The environmental dimension of sustainability concerns an organization’s impacts on living and non-living natural systems, including ecosystems, land, air, and water. Environmental Indicators cover performance related to inputs (e.g., material, energy, water) and outputs (e.g., emissions, effluents and waste).
The specific aspects under the category of Labour Practices are based on
internationally recognized universal standards. There is growing global consensus that organizations have the responsibility to respect human rights, see below.
Human rights Performance Indicators require organizations to report on the extent to which processes have been implemented, on incidents of human rights violations and on changes in the stakeholders’ ability to enjoy and exercise their human rights, occurring during the reporting period. Among the human rights issues included are non-discrimination, gender equality, freedom of association, collective bargaining, child labour, forced and compulsory labour, and indigenous rights.
Society Performance Indicators focus attention on the impacts organizations have on the local communities in which they operate, and disclosing how the risks that may arise from interactions with other social institutions are managed and mediated. Product Responsibility Performance Indicators address the aspects of a reporting
organization’s products and services that directly affect customers example in the way they can effect are: “health and safety, information and labelling, marketing, and privacy.”
2.2 Parts of incentives compensations 18
2.2 Parts of incentives compensations
Based on the agency theory, managers should have a compensation package which is in line with the goals of the shareholders (Jensen & Meckling, 1976). The major elements of executive compensation are fixed compensation (salary), variable pay (bonus and stock options) (which include stock options, other forms of market based compensation, and non-market based long term incentive plans) (Murphy, 1999). The bonuses and long-term incentives are variables pays.
Base salary is an example of a fixed payment what the manager get every month. This payment is not dependent on the performance of the manager, this payment is a short-term incentive. According to Deckop et al. (2006) the payout for maximizing short-term earnings is more salient for CEOs when their pay plan offers substantial rewards based on short-term performance and/or a lack of disincentive to focus on the short term. Empirical research shows that managers are more likely to manipulate earnings to show larger short-term gains when bonuses are based on annual earnings and this tendency decreases under pay plans tied to long-term performance (Deckop et al, 2006)
One example of variable pay is annual bonus. Annual bonus is a short-term incentive. According Murphy (2001) executive bonus plans can be categorized in terms of three basic components: performance measures, performance standards, and the relation between pay and performance. Performance measures are measured with a variety of financial and
non-financial measures. Examples of performance measures are revenues, net income, pre-tax income, EBIT or accounting-based ratios such as return on assets. The most performance standards for accounting-profit performance measures are based on a single criterion.
Examples of performance standards are budgets, prior-year performance, board discretion and cost of capital. Pay-performance are the effects of performance standards on executive
behaviour and compensation. The better the performance of the employee, the higher the bonus will be.
An example of long-term incentives are stock-options. Managers get the right to buy shares from their company at the exercise price. According to Hayes et al. (2012)
shareholders can potentially reduce the risk-related agency problem by structuring
compensation to be a convex function of firm performance for example through the use of stock options. Compared to the bonus, stock-options are long-term incentives, because the stock-options are vested for a couple of years. Long term incentives are good for the company
2.3 Incentives CEO 19
because managers will to perform better in the future because of the stock-option. So the managers need to be long-term oriented.
According to Deckop et al. (2006) attention to strength and concern areas of CSR reports are more likely to longer term positive payouts and consequences and less likely to have positive short-term financial effects. Also investments in CSR can have short-term effects in the companies effect and thus affect the stock-price. Short-term CEO pay is based heavily on firm accounting performance as opposed to stock price. However, executives that have a particular financial disincentive to engage in CSR when their pay focuses on short-term incentives. The negative financial effect of actions that result in CSR concerns such as problems in areas of environmental neglect, employee and product safety are not likely to be caught immediately and are unlikely to show up in short-term financial performance. The actions of the long term are more likely to affect the firm and provide little disincentive to the CEO in the short term (Deckop et al., 2006).
2.3 Incentives CEO
The major elements of executive compensation are fixed compensation (salary), short-term incentives, long-term incentives (which include stock options, other forms of market based compensation, and non-market based long term incentive plans) (Murphy, 1999).
McGuire et al. (2003) suggests the performance pressures implied by long term incentives may encourage CEOs to focus on achievement of financial objectives to the detriment of the interests of other stakeholders. In essence, the long term benefits of building stakeholder relations may be less visible than strategies (such as CSR reports) that may boost firm market performance, at least in the short term. Stock options have been increasingly criticized for encouraging earnings manipulation and other strategies aimed at improving market performance without significantly building stakeholders value. This is illustrated for example by Enron and Worldcom bankruptcy. (McGuire et al., 2003)
Furthermore, according to McGuire et al. (2003), executive incentives are a visible and potentially important mechanism through which owners and the board of directors can direct managerial attention to specific objectives having both financial and social
implications. Compensation can therefore be a potentially important mechanism for directing managerial attention to social objectives.
2.4 Empirical evidence on compensation incentives and CSR 20
McGuire et al. (2003) also suggests that in developing executive compensation plans, and awarding executive incentives, the board of directors can encourage management to pursue social, as well as financial objectives. Thus, compensation can be an important mechanism to promote the implementation of the firm’s social objectives.
I want to verify this with the agency theory, according to Jensen and Meckling (1976) in this theory they describe the relationship between principal and agent. The latter bear fiduciary duties towards the former, and are generally subject to strong incentives in order to align their economic interests with those of the owners, and with the maximization of
shareholder value. (Melé, 2009) So, the principal gives the agent incentives which will maximize the principles welfare. Where the strong incentives for individuals will minimize the agency costs. CSR disclosures are made to decrease the agency costs and reduce the information asymmetry of the company by giving the stakeholder information about the company (Jensen and Meckling, 1976).
According Fabrizi (2013) CEO’s compensation scheme is not the only driver of managerial attention to specific objectives. Other incentives that might affecting CSR-related decisions are career concerns, turnover and power and entrenchment. Although compensation is a visible and fundamental mechanism that directs managerial attention to specific
objectives. Career concern, turnover and power and entrenchment are example for non-monetary incentives what is used in the paper of Fabrizi (2013). With this research I want to focus on the short-term incentives (salary and bonus) and the long-term incentives (stock-options).
2.4 Empirical evidence on compensation incentives and CSR
Prior research shows that there can be a positive relationship between CEO compensation incentives and CSR (Fabrizi, 2013), a negative relationship between long-term compensation and total CSR weakness (Mahoney and Thorne, 2005) and salary and long-term incentives have a positive association with weak social performance (Mcguire et al., 2003). Fabrizi (2013) bases the analysis on a sample of 597 US firms over the period 2005–2009 and split between monetary (based on both bonus compensation and changes in the value of the CEO’s portfolio of stocks and options) and non-monetary (career concerns, incoming/departing CEOs, and power and entrenchment), in relation to corporate social responsibility (CSR). Mahoney and Thorne (2005) examines the association between bonus and long-term compensation and corporate social responsibility for 90 publicly traded Canadian firms.
2.4 Empirical evidence on compensation incentives and CSR 21
Mahoney and Thorne also measure the salary, bonus and stock-options. McGuire et al. (2003) bases the analysis on a sample of more the 300 US firms and also used the measures salary, bonus and stock-options.
Fabrizi (2013) suggests that the empirical analysis provides evidence of a negative and significant relationship between CSR and the CEO’s equity incentives and annual bonus. Overall, the evidence suggests that the CEO’s decisions on CSR are not driven only by monetary incentives but also by personal non-monetary incentives that relate to the role of the executive within the firm (incoming, powerful and entrenched CEOs) and her career concerns (age).
Mahoney and Thorne (2005) suggests that the results of this study suggest that certain aspects of CSR are related to long-term compensation of CEO’s for the sample of firms. He found that the composite measure of total CSR is marginally positively related to higher levels of longer term compensation. The results suggest that the use of long-term
compensation may discourage executives from making decisions that are risky to the firm and their own compensation, which in turn may benefit society. In other words, rather than
serving to be more socially responsible, long term compensation may serve as a deterrent to potentially damaging actions.
McGuire et al. (2003) suggests that with using the KLD database they find that incentives have no significant relationship with strong social performance. Salary and long-term incentives have a positive association with weak social performance.
Today, companies are increasingly demonstrating that CSR reporting provides financial value and drives innovation, whereby the incentives will be higher. (KPMG, 2011) In 2011, 95 percent of the 250 largest company’s now report on sustainability, so when you didn’t report you get the pressure to also report.
With this research I want to focus on Europe because there are not so much research papers about Europe with the subject CSR reporting and incentives. But traditional CSR reporting nations in Europe continue to see the highest reporting rates. (KPMG, 2011)
I can say that are two kind of incentives, short-term and long-term. Short-term will consist of salary and bonuses and long term which include stock options, other forms of market based compensation, and non-market based long term incentive plans.
From prior research you can see that there can be a (no) relationship between CEO incentives and CSR reports. So, I want to research if the variables of short term and/or long term incentives enhance or weaken the effect of the level of CSR reports in Europe. Above
2.5 Hypothesis 22
text on the literature gives a mixed results regarding the research. The results on relevant previous researches have been summarized in Appendix “Overview researches”
2.5 Hypothesis
I want to research if higher short-term and long-term incentives enhance or weaken the effect of the level of CSR reports. My research will be based on the study of Fabrizi (2013) and McGuire et al. (2003) My first question will focus on the short-term incentives, the short-term incentives will be consistent with salary and bonus. Furthermore, I will make a distinction in salary and bonuses.
What earlier mentioned in paragraph 2.1 the term stakeholder is an important term with CSR. Effective CSR is usually a long-term action. CSR is also an investment for the company, a good reputation increase the value of the brand which will increase the company’s value. Furthermore McGuire et al. (2003) mentioned that compensation can be a potentially important mechanism to promote the implementation of the firm’s social objectives. Also the agency theory mentioned that the incentive will maximize the principal welfare. Where the strong incentives for individuals will minimize the agency costs. CSR disclosures are made to decrease the agency costs and reduce the information asymmetry of the company through give the stakeholder information about the company. According the above paragraphs, incentives will positive and negative influence the CSR reporting.
I think that there is no significant relationship between salary and CSR reports. The salary is fixed, and will not show large fluctuations. Furthermore, it is short-term whereby as suggested in paragraph 2.1. CSR reports will mostly have impact on the long-term factors. I propose the following hypothese:
H1: There is a no relationship between salary and CSR reports
Bonuses are short-term incentives. This has an important implication for CSR reporting. CSR is from nature a long-term action. Accordingly, I anticipate a negative relationship between the bonuses and CSR reporting. I propose the following hypothese: H2: There is a negative relationship between bonuses and CSR reports
I expect a positive relationship between long-term incentives and CSR reports. Stock-options are a form of long-term incentive compensation and the manager is responsible of the
3.1 Data 23
value of the stock in the future. With stock options the executive focus on increase the future stock value, as the executive wealth will increase if the stock price increase. According Mahoney and Thorne (2005) is that executives that receive stock options are more likely to take actions consistent with maximizing the interest of the firm in longer term such as CSR. According Falck and Heblich (2005) CSR reports are long-term actions and when an owner gives term incentives the owner will motivate employee to do a good job in the long-term. I expect a positive relationship between long-term incentive and the CSR reporting. I propose the following hypothese:
H3: There is a positive relationship between stock options and CSR reports.
3. Research methodology
This chapter describes how the research is executed. This research is a quantitative research.
The section will explains how the data is collected. After that it will be explained which variables will be used to test the hypotheses. There will be a distinction in independent, dependent and control variables.
3.1 Data
For collecting data I used the database ‘The Global Report List’ in 2011, 2012 and 2013. The Global Report List gives an overview of all the companies over the world that report CSR with the GRI. In this list there is also an overview of adherence level of the CSR reports. The adherence level explains the level of the CSR report of the company. The database provides also an overview where the company is settled and in which sector it operates.
An important point in this thesis is that the stock-options need to be included. In Europe the data of stock-option of the CEO is limited. First, I filtered in the Global Report List five sectors and five countries. However, the amount of data I found was too small. Therefore, I have downloaded the list with companies with stock-options on the database ‘CapitalIQ’ and looked through ‘Global Reporting List’ which companies make a CSR report. Mostly, when the stock-option of the CEO is available also the salary and bonus of the CEO is available. So I choose for this method to obtain a bigger sample.
The sample period in this research is the period 2011-2013. I choose for this period because in this period most data is available. A lot of the financial reports of 2014 are not
3.1 Data 24
finished yet, so the data is limited. I choose for different years because the sample is bigger what results in more validity. The data in this research will be shown in €.
For the financial data I use Datastream and Amadeus. This database has data about Europe. The advantages of the Amadeus database is that it includes standardized financial statements data for companies in Europe.
Through the above selection there is a sample of 130 firm year observations and 47 unique firms see table 1. In the beginning there were 359 firm years observations. Some companies the salary, bonus and stock-options were not published, this data has been removed from the list. Also some companies were removed because in some specific year the
adherence level is not known. Further, some control variables of the companies are not published, this data is also deleted this is the reason that there are 130 firm year observations now, see table 1. In paragraph 3.5 there I want to tell more about the assumptions of the regression model.
TABLE 1
Sample Selection Firms Firm-years 2011 through 2013 data 124 359 Less: unknown adherence level 32 96 Less: missing incentives 16 50 Less: control variables 29 83
Final sample 47 130
In the sample I make also a distinction in years. In 2011, 2012 and 2013 there are
3.2 Independent variables 25
Table 2 Sample selection and distribution Total unique firms: 47
Total firm years: 130
Years Freq. Percentage
2011 42 33,09%
2012 43 33,09%
2013 45 33,81%
Total 130 100%
The regression model is:
Adherence level = β0 +β1 *Salary + β2 * BONUS + β3 * Equity incentives + β4 * D_IND + β5 * LEV + β6 * ROA + β7 * Log total assets + β8 * IND + β9 * financial slack + ε
The variables will be explained in the next paragraphs. In the table below you can see an overview of the variables that is used.
Adherence level of CSR report (Adh_level) 6 point Likert scale variable
1= C 2= C+ 3= B 4= B+ 5= A 6= A+
Salary Total salary
Bonus Bonus / total compensation
Equity_incentive (eq_inc) CEO_ONEPCT/ (CEO_ONEPCT +
CEO_CASHCOMP)
Independent on the board (D_IND) 1 = higher than 26 %
0 = lower than 26%
Leverage (LEV) Total debt divided by total assets
ROA (ROA) Total net income/ total assets
Total assets (SIZE) Logarithm of the total assets
Financial slack Earnings before tax and interest / interest
expense
Industry (IND) 1 = sensitive industry (mining, metal, oil, gas,
chemical, pharmaceutical, paper, alcohol, defense and utilities)
0 = other industries
Table 3: Variables
3.2 Independent variables
The independent variable in this research is the compensation of the CEOs. The database that will be used is CapitalIQ. For this research is chosen the short-term incentives (salary and
3.2 Independent variables 26
bonus) and long-term incentives (stock-options). First I was in trouble with the long-term incentives because it is hard to find the data in the databases of Europe but as I mentioned earlier that stock-options can be an important mechanism to promote the implementation of the firm’s social objectives. Also CSR will be more important to the stakeholders. According Falck and Heblich (2005) long-term incentives have a risk for a longer time so this will be trigger the CEO to do a good job with the company. According to McGuire et al. (2003) the relationship between social performance and long-term incentives revolves the issue of whether socially responsible actions build the firm’s long-term potential.
Salary (Salary) will be measured with the salary as reported in CapitalIQ in € in 2011, 2012 and 2013 (the same as Mcguire et al., 2003).
Bonus (Bonus) will be measured with the bonus/ total compensation as disclosure in CapitalIQ in € in 2011, 2012 and 2013 (the same as Fabrizi, 2013).
Stock-options (Equity incentive) will be measured as the incentive ratio computed as in Bergstresser and Philippon (2006). First, I start by computing the euro change in the value of the CEO stock and options holding that would come from a one-percentage point increase in the company stock price (CEO_ONEPTC) using the Capital IQ data for the options and the Amadeus database for the stock-option price the following formula:
CEO_ONEPCT = 0,01 * PRICE * (CEO_SHARES + CEO_DELTA * CEO_OPTIONS).
(1)
PRICE is the fiscal year-end company share price and the CEO_OPTIONS is the number of options held by the CEO as at the fiscal year-end. CEO_SHARES is the number of shares held by the CEO. CEO_DELTA is an estimate of the CEO options portfolio. Secondly, the measure of incentive computed in (1) is standardized by the amount of cash compensation (salary and bonus in €) received by the CEO during the fiscal year as in CapitalIQ. Therefore, we compute the CEO’s incentive ratio as follows:
EQUITY_INCENTIVE = CEO_ONEPCT/ (CEO_ONEPCT + CEO_CASHCOMP)
(2)
According Bergstresser and Philippon (2006) the measures above are based on the implicit assumption that the “delta” of the options in the CEO’s portfolio is one, i.e. a dollar increase in the price of a firm’s share translates one-for-one to the value of an option. While
3.3 Dependent variables 27
this is approximately true for options that are deep in the money, it is a less accurate
assumption for options that are out of the money. It is impossible for my research to calculate the delta of the stock-option. The amounts that I need to use are not available in the databases. This is a limitation for this research because I use for CEO_DELTA one. In the research of Fabrizi they follow the Core and Guay (2002) methodology for estimating the delta of
executives’ option portfolio. They calculate the sensitivity of individual stock options to stock price.
McGuire use also the variable activist institutional investment as variable. There they calculate the percentage of shares held by the ‘activist’ pension funds identified by Wahal reported in our sample of firms (McGuire, 2003). I don’t use this variable. The amount of the shares held by the ‘activist’ pension funds are not in the database.
3.3 Dependent variables
The dependent variable that will be used for this research is the adherence level of CSR reports. It was not possible to use KLD or CSID as previous researches. KLD has only data until 2011 and covered not data from Europe. For this research I will use the website
www.globalreporting.org. The adherence level is to which dimensions of the GRI are applied by compose the report. The levels are C, B and A. The criteria in every level gives more elaborate application or coverage of the GRI reporting framework. An organization give a “+” when extern verification is included. A+ is the highest (Likert scale variable 6) and C (likert scale 1). The likert scale variable adherence level taking the value 6 when the adherence level of the CSR report is A+ and 1 when it is an C. In previous researches they use the database KLD or EIRIS, this is not data for Europe so it is not useful for my research.
3.4 Control variables
Based on prior research, I include some firm characteristics as control variables which might affect the adherence level of CSR reporting. For the determination of the control variables were analysed the variables of Mahoney and Thorne, McGuire et al. and Fabrizi.
Mahoney and Thorne (2005) used the variables firm ownership, size and industry. The variable firm ownership refers to shareholders with more than 20% ownership as having significant influence over a firm’s strategies and policies. They found that a firm’s
3.4 Control variables 28
compensation plans are related to firm ownership concentration. Mahoney uses also the variable size they use sales as a control variable and industry since the risk and opportunities regarding social performance may vary with the industry.
McGuire et al. (2003) used the variables firm size, industry, profitability and financial slack. Profitability and financial slack have be linked to social performance. They also likely influence salary and incentive compensation. They use therefore the ROA and the log transformation of the number of employees as control variables. Financial slack is measured by leverage and the times interest earned ratio. McGuire et al. (2003) also suggest for further research to add some variables with the management characteristics, as well as that of the board of directors, may help shape the firm’s social agenda. Therefore I add the variable independent on the board.
Fabrizi used the variables industry, independent non-executives on the board, size, ROA, leverage and the book-to-market ratio. Industry, independent non-executive on the board (firm ownership), ROA and leverage is the same as describe above. Size is the natural logarithm of the firm’s total assets.
For this research five control variables will be used. These control variables are selected based on the previous studies in this area of research (Fabrizi, 2012, McGuire et al., 2003 and Mahoney and Thorne, 2005). The six control variables will be explained below.
First, I include amount of independent members on the board as control variable which might affect the adherence level of CSR reports. Independent on the board (D_IND) will be used as a dummy variable, 1 if the proportion of independent non-executives on the board is more than 26% reported in Orbis and 0 when it is less. Mahoney and Thorne (2005) suggest that Craighead et al. 2004 found that firm’s compensation plans are related to firm’s ownership concentration. According Mackazie (2007) company boards are key participants in ensuring companies meet CSR standards.
Second, I added leverage by total debt divided by total assets (LEV). Leverage can be used as a proxy of a company’s risk tolerance. Firms with low risk tolerance are more likely to use disclosure on their CSR activities as a tool to manage risks (Cormier et al., 2005). In general, a high debt-to-equity ratio indicates that a company may not be able to generate enough cash to satisfy its debt obligations.
3.5 Assumptions 29
Third, I include ROA (ROA) is the yearly return on assets. The ROA is calculated by total assets/ total net income. Mcguire et al. and Fabrizi also use this variable.
Fourth, I add the natural logarithm of the total assets (SIZE). According to Fabrizi is expected to affect CSR disclosure. Size is an important variable in this research. In bigger company
Fifth, I add the variable times interest earned ratio. The times interest earned ratio is calculated by earned before interest and taxes divided by interest expenses.
Finally, I include industry dummy (IND) to control. Industry can be used as proxy because of the risk and opportunities regarding social performance may vary in the industry. It will be used the same as Fabrizi. Fabrizi uses a dummy variable 1 when the firm belongs to either an environmentally or a socially sensitive industry (mining, metal, oil, gas, chemical, pharmaceutical, paper, alcohol, defense and utilities).
So the regression model will be
Adherence level = β0 +β1 *Salary + β2 * BONUS + β3 * Stock-option + β4 * D_IND + β5 * LEV + β6 * ROA + β7 * Log total assets + β8 * IND + β9 * Financial slack + ε
3.5 Assumptions
In this paragraph will explain the assumptions of the regression model. The first assumptions is that there should be no outliners. In my opinion is that inconsistent, sometimes for example the bonus is € 0 but that is than an outliner but it is an important value in this test. I did a regression test with the outliners and without the outliners. I found out that with the regression test with the outliners there are came better results and the fit (r2) doesn’t change so much (adjusted r2 of respectively 0.59 and 0.62). So I choose for a test without remove the outliners.
One another assumption is that there is no multicollinearity. Multicollinearity means that two or more predictors variables in a multiple regression are highly correlated. When the VIF is equal to one than is there no multicollinearity, the multicollinearity gives really a problem when it is higher than 5. To check whether the multicollinearity problem affect the sample, the VIF values of the variables are calculated. In this test I add for every independent variable two formulas and with the highest multicollinearity I removed. The variables in parentheses will not use in the regression test. With the variable bonus I decided to choose the value with the lowest VIF that is bonus/total compensation this variable is also used in the paper of McGuire. (table 4)
3.5 Assumptions 30 Table 4
Testing of the multicollinearity VIF
Variables Multicollinearity (1) Multicollinearity (2)
Salary 2.61 2.60 Salary/comp 9.74 (-) LNbonus 2.98 (2.17) Bonus/comp (omitted) 1.89 Eq_inc 3.14 1.92 Stock/comp 9.44 (-) IND 1.45 1.45 D_IND 1.32 1.24 SIZE 2.71 2.70 LEV 1.10 1.09 ROA 1.39 1.31
This table provides the VIF of the tests. (1) Is with all the variables
(2) Is without the variables that are too high in model 1 (VIF is higher than 5)
Conclude, I use a test with the outliners because that give better results and use the independent variables salary, bonus/compensation and the equity incentives.
4 Results
This chapter shows the empirical results. First, there will represents the descriptive statistics. Second, the correlation will be explained. Finally, the results on the hypotheses will be explained.
4.1 Descriptive statistics 31
4.1 Descriptive statistics
Table 5 Descriptive statistics
Variable n Mean Std. Dev. Min. Max.
Adherence level a 132 4.17 1.88 1 6 Salary b 132 900783.70 394947.90 278.41 2.100.000 Bonusb 132 0.19 0.18 0 0.63 EQ_INC b 132 0.059 0.12 0 0.51 D_IND 132 0.26 0.44 0 1 Leverage 132 0.70 0.63 0.29 6.78 SIZE 132 7.13 0.79 5.01 8.42 ROA 132 0.05 0.07 -0.21 0.49 Sector 132 0,46 0.50 0 1 Financial slack 130 9.28 13.40 -39.34 74.85 N = number of observations a Dependent variable ᵇ Independent variable
This initial sample consists of 130 firm-year observation for which data for CSR adherence level, independent variables and control variables
Appendix Variables provides detailed information on all the variables
Table 5 presented descriptive statistics for the variables in analysis. The mean of the adherence level is around 4 which means that the mean of the adherence level is B. The average salary is € 900783.7, the bonus/ total compensation is 0.1958 and the average of the equity incentive is 5%. The mean of the independence on the board is 0.26, leverage is 0.70, log total assets is 7,13 and ROA is 0.050. The sector has an average of 0.46 which means that there are almost as many sensitive sectors as non-sensitive sectors.. The average of the salary is € 911.109, whereby the minimum is € 278.411 and the maximum is € 2.100.000.
4.2 Correlations coefficients
Table 6 represents a correlation model matrix for all the variables discussed in chapter 3.1. The correlation matrix is a measure of linear association between two variables. A perfect negative correlation is represented by the value -1 while +1 indicates a perfect positive correlation.
4.2 Correlations coefficients 32
Table 6
Pearson correlations
Adh_level Salary LNBonus Eq_inc IND D_IND LEV SIZE TIMESIN
Salary 0.5929* 0.00 Bonus/comp 0.1972** 0.1665* 0.0234 0.0097 Eq_inc -0.23* -0.0031 -0.5105* 0.0080 0.9717 0.0000 IND 0.3701* 0.1966** 0.3150* -0.3199* 0.0000 0.0239 0.0002 0.0002 D_IND 0.2715* 0.3204* 0.1099 -0.0662 0.2532* 0.0016 0.0002 0.0582 0.4505 0.0034 LEV -0.0674 -0.0091 -0.0038 -0.0505 -0.0405 0.0976 0.4428 0.9176 0.9652 0.5653 0.6450 0.1952 SIZE 0.6716* 0.7120** 0.2313* 0.0562 0.2292* 0.2771* - 0.0914 0.0000 0.0000 0.0076 0.5219 0.0082 0.0013 0.2972 ROA -0.1276 0.0820 -0.0418 0.0337 -0.0553 -0.1419 -0.0437 0.0756 0.1448 0.3501 0.6339 0.7009 0.5286 0.1046 0.6188 0.3891 Financial slack 0.1251 0.1832** -0336 0.0393 0.1435 -0.056 -0.1510 0.0530 0.4174* 0.1562 0.0369 0.7046 0.6574 0.1033 0.5271 0.0863 0.5491 0.0000
* Denotes significance at 0.01 levels ** Denotes significance at 0.05 levels Pearson correlation matrix
To test the correlation of all the variables there is made a correlation matrix. A
correlation matrix is a matrix where there is show how much the variables correlate with each other. The values are between the 0 and the 1. Zero is when there is no connection between the variables and 1 when there is a connection between the variables. When the variables is minus 1 that means that there is a negative connection between the variables and plus when there is a positive connection between the variables.
In table 6 there is show the Pearson Correlation matrix. On basis of this matrix is show that there is a positive correlation between adherence level (0.59) and salary and bonuses (0.20) and a negative correlation between adherence level and equity incentives (-0.23).
On basis of this matrix there is show that there are positive relationship is between adherence level and size (0.67). That means when the company is bigger that the adherence level of the company is higher. So, CSR reporting is important for the company because they have a higher adherence level.
An another point is when the salary (0.71), bonus (0.23) and equity incentives (0.06) is higher the company is also larger (SIZE).
4.3 Empirical results 33
4.3 Empirical results
In this paragraph will be test the regression model. Table 7 Full sample
Model Beta T statistics Significance
1 (constant) -6.76 -6.21 0.000 IND 0.69 0.25 0.006 D_IND 0.12 0.40 0.687 LEV 0.03 0.15 0.885 SIZE 1.49 9.74 0.000 ROA -6.15 -3.33 0.001 Financial slack 0.02 -6.21 0.019 2 (constant) -5.89 -4.68 0.000 Salary 7.96 1.74 0.085 Bonus -1.73 -2.37 0.019 Eq_inc -4.87 -4.41 0.000 IND 0.43 0.24 0.084 D_IND 0.05 0.18 0.855 LEV -0.04 -0.27 0.784 SIZE 1.39 6.43 0.000 ROA -6.31 -3.69 0.000 Financial slack 0.02 2.31 0.023
Model 1 (controls) F ratio: 24.97; Significance: 0.000; Adjusted r2: 0.5271 Model 2 (full model) F ratio 22.28; Significance: 0.000; Adjusted r2: 0.5976
This table provides a multiple regression test.
There are 130 firm years observations and 47 companies. Variables are shown in the appendix Variables
In table 7 there is shown the results of hypotheses 1 through 4. The model has an adjusted r2 of 0.5271 and 0.5976. That means that 52.71% and 59.76% of the variables is explained by the model. This means that the overall fit of the model is well. In model 1 there is show that the control variables industry, size, ROA and financial slack has an influence on the adherence level of CSR reporting.
The first hypotheses tests whether the adherence level of CSR reporting has an effect on the amount salary of the CEO. Thus, to not reject the H1 the significance level need to be p > 0.05 for my test. Consistent with the predictions, the results shows not a significant result. In model 2 there is show
4.4 Sensitivity analysis 34
that the significance is 0.085 that means there is no significant relationship between salary of the CEO and the adherence level of CSR reporting, so this hypotheses is accepted. The second hypotheses tests whether the adherence level of CSR has an association with the amount of bonuses of the CEO. I expected that there is a negative relationship. The significance level is 0.019 and the β is -1.73 this means that there is a negative significant relationship between bonuses and the adherence level. So, the hypothese 2 is accepted. The third hypotheses argues that the adherence level is associated amount of stock options of the CEO. The. For hypothese 3 I predicted a positive significant relationship between equity incentives and the adherence level of CSR reporting. The significance level is 0.000 and the β is - 4.87 this means that there is a negative significant relationship between equity incentives and the adherence level. So hypothese 3 is rejected.
4.4 Sensitivity analysis
In this section, I show the sensitivity analysis. By far the best option if you have irksome data is to use test that is robust to violations of assumptions and outliners. (Field, 2009)
4.4 Sensitivity analysis 35 Table 8 Sensitivity analysis Adh_level (n = 130) Intercept -5.8912 (0.000)* Salary 0.0000 (0.074)*** Bonus 1.7334 (0.011)** Equity incentives -4.8758 (0.000)* Control variables IND 0.4251 (0.125) D_IND 0.0493 (0.859) LEV -0.0493 (0.763) ROA -6.3164 (0.000)* SIZE 1.3947 (0.000)* Financial slack 0.0231 (0.027)** The data is based on 2011 through 2013
Adh_level is based on adherence level of CSR reporting (n = 130) */**/*** represent a significance level of respectively 1 / 5 / 10%
The r-square of the model is 0.6256, which is consistent with the regression model in table 6. The number of observations in this model is 130 over 3 years namely 2011, 2012 and 2013.
The results in table 8 demonstrates that the relationship between adherence level and salary is not significant, which is congruent with the previous regression. Furthermore the relationship between bonus and the adherence level is significant. However, the relationship has changed from
