The influence of CSR performance on CEO compensation

The Influence of CSR Performance on

CEO Compensation

Name: Maarten Spruijt Student number: 10194223 Date: June 20, 2016 Word count: 19,844

Supervisor: dhr. dr. S.W. (Sanjay) Bissessur MSc Accountancy & Control, both variants

Amsterdam Business School

1 Statement of Originality

This document is written by student Maarten Spruijt who declares to take full responsibility for the contents of this document.

I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.

The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

Abstract

This study investigates whether Corporate Social Responsibility (CSR) has an influence on CEO compensation. For this matter, an extensive sample including S&P 1500 firms over a sample period from 1996 until 2014 is used. Hypotheses are developed using the stakeholder theory and agency theory and tested by using a regression-based approach. This study finds that CSR performance negatively affects short-term CEO compensation comprised of base salary and cash bonus. Furthermore, this study finds that CSR performance is unrelated to total CEO compensation. Finally, long-term CEO compensation is only significantly positively affected by CSR performance for the period from 2010 until 2014. These results contribute to the limited scientific knowledge on the direct influence of CSR performance on CEO compensation, since prior literature mainly studied the inverse or indirect relationship.

2 Acknowledgements

I would like to thank my supervisor, dr. Sanjay Bissessur, for his continuous support and for sharing his knowledge with me. Although the sentence “this is something you have to think about…” was not always the feedback I wanted to hear, it steepened my learning curve greatly. Additionally, I am thankful for his clear explanations with regard to his so-called “hobby” Stata.

3 Table of Contents Abstract 1 1 Introduction 6 1.1 Relevance 6 1.2 Background 7 1.3 Research question 8 1.4 Results 9 1.5 Contribution 9 1.6 Structure 10 2 Literature review 11

2.1. Main accounting theories tested 11

2.2. CEO compensation 14

2.2.1. Base salary 15

2.2.2. Cash bonus 16

2.2.3. Equity compensation 16

2.3. Corporate social responsibility (CSR) performance and reporting 17

2.4. Prior literature on this topic 18

2.4.1. Prior literature on the inverse relationship 19

2.4.2. Prior literature on influence of CSR performance on financial performance 19 2.4.3. Prior literature on influence of corporate financial performance on CEO

compensation 22

2.4.4. Prior literature on studied relationship 23

2.4.5. Prior literature on other (economic) determinants of CEO compensation 26

3 Development of hypotheses and research methodology 32

3.1. Development of hypotheses 32

4

3.2.1. The model and predictive validity framework 36

3.2.2. The sample 37

3.2.3. Measurement of variables 37

3.3. Assumptions for using OLS regressions 47

4 Empirical results 50

4.1. Descriptive statistics and correlations 50

4.2 Results from the regressions 54

4.3 Examining endogeneity in the model 56

4.4 Additional analyses 58

5 Discussion 64

5.1. Short-term compensation model 64

5.2. Long-term compensation model 66

5.3. Total compensation model 68

5.4. Additional analyses 70

6 Concluding remarks 71

References 73

5 List of tables

Table Title Page

1 Chronical presentation of scientific articles used in studying relationship in prior literature 30

2 Dependent and independent variables of this study 43

3 Control variables included in the regression model 47

4 Descriptive statistics for the samples 51

5 Pearson correlation matrix 54

6 Regression results for main analyses 55

7 Regression results for main analyses with different outlier management 60 8 Results for the main analyses divided in two sample periods based on CSR performance

measurement

61 9 Regression for the effect of CSR performance on long-term compensation 62 A1 Appendix 1: Variables included in used categories for measuring CSR performance, using

MSCI ESG Database variable descriptions.

6 1 Introduction

Relevance 1.1

Recent years have witnessed an increase in the portion of companies or organizations reporting on the economic, environmental and social impacts caused by its everyday activities. According to a recent study of EY and Boston College Center for Corporate Citizenship (2013) sustainability disclosure was once the province of a few unusually green or community-oriented companies, while today it is a practice employed by a variety of companies across the world. A report of KPMG (2015) supports this finding by stating that the portion of Global 250 companies that report on its corporate responsibility is more or less stable at 92 percent. This number has increased significantly in the last two decades, since in 1999 only 35 percent of the Global 250 companies reported on their corporate responsibility. Given this development it is perhaps not surprising that this relatively new form of reporting has come to figure highly in recent studies. On the other hand, executive compensation has been of heightened interest in recent years, since it has risen sharply (Callan & Thomas, 2014; Murphy, 2012). Especially, in the context of business failures and government bail-outs this has led to public outrage. Callan and Thomas (2014) make this rise concrete by stating that the average annual wage of Chief Executive Officers (CEOs) have increased six-fold over the past 20 years, while several crises have occurred during this period. This growth coincides with the shift towards more performance-based compensation, such as stock options and restricted stock grants (Callan & Thomas, 2014; Murphy, 2012). Callan and Thomas (2014) and Benson and Davidson (2010) indicate that the performances on which compensation is based can roughly be divided into financial performance and a broader set of measuring performance, such as incorporating corporate social performance.

Taken the abovementioned developments together, McGuire, Dow and Argheyd (2003) state that executive compensation is a tool to direct managerial attention to specific objectives. These objectives could include financial ones or more broader defined objectives, such as Corporate Social Responsible (CSR) objectives. The incentives associated with different types of executive compensation and the criteria on which compensation is based are visible and

7 potentially important mechanisms through which this directing is made possible. The concept of executive compensation may therefore be potentially important in directing managerial attention to CSR objectives (Deckop, Merriman, & Gupta, 2006; McGuire, Dow, & Argheyd, 2003). This mechanism of directing managerial attention towards certain objectives is expected to be most useful for the CEO of a firm, since he/she is clearly accountable for the overall (social) performance of the firm (Mahoney & Thorn, 2006). In addition, CEOs are perceived to have direct control over decisions that directly impact all stakeholder groups (Coombs & Gilley, 2005).

Background 1.2

Prior studies have mainly studied the effectiveness of executive compensation as a tool to encourage CSR performance (Deckop et al., 2006; Mahoney & Thorn, 2006; McGuire et al., 2003; Merriman & Sen, 2012; Stanwick & Stanwick, 2001). A significantly smaller amount of published studies test for and statistically identify the inverse relationship in which CSR performance is a determinant of executive pay. The overwhelming majority of these CSR empirical studies are focused on the link between CSR performance and corporate financial performance (Perrini, Russo, Tencati, & Vurro, 2011), while on the other hand a large extent of research has focused on the link between corporate financial performance and executive compensation (Callan & Thomas, 2014).

Nevertheless, prior literature does include some papers that study the relationship between CSR performance and CEO compensation directly. This literature includes a study from Coombs and Gilley (2005), in which they investigated the effect of five indicators for the extent of stakeholder management on the various aspects of CEO compensation. Additionally, Benson and Davidson (2010) studied whether CEOs are compensated based on shareholder maximization or stakeholder maximization. As such, the authors proposed that firms will compensate CEOs based on the goals of the firm, being either shareholder or stakeholder maximization or both. In other words, being either maximizing financial performance or more

8 broadly defined goals related to stakeholder management (Benson & Davidson, 2010). Next, a study from Cai, Jo and Pan (2011) studied whether firms that are more socially responsible pay their CEOs less. Lastly, studies from Callan and Thomas from 2011 and 2014 have tested the influence of CSR on CEO pay, while including the determinant of corporate financial performance.

Additionally, prior literature identified other (economic) determinants of CEO compensation. First of all, Core, Holthausen and Larcker (1998) and Armstrong, Gow and Larcker (2013) focus on the influence of corporate governance on CEO compensation by studying board-of-director characteristics, ownership structures, and shareholder voting outcomes. Secondly, studies by Agarwal (1981) and Finkelstein and Hambrick (1989) develop models including the determinants of CEO compensation. Agarwal (1981) uses three determinants: job complexity, the employer’s ability to pay and human capital, while Finkelstein and Hambrick (1989) extent this model by deviating between market forces, such as firm size, corporate performance, corporate complexity and human capital, and political forces, such as board vigilance and CEO power. Furthermore, Ittner, Larcker and Rajan (1997) focus on the choice between financial and non-financial performance metrics, such as CSR performance, in determining annual bonuses. Additionally, Graham, Li and Qui (2012) add to prior literature that unobservable determinants, such as firm and managerial fixed effects, may cause the majority of variety in CEO compensation.

Research question 1.3

Prior literature has found that compensation levels have risen sharply in the last two decades, but the objective of this study is to investigate whether this rise is justifiable and on which performance indicators, financial or more broadly defined ones, this is justifiable. Additionally, the effect of CSR performance, thus the extent of being socially responsible, on the compensation of the CEO becomes of interest, since this has largely been overlooked in prior literature. Therefore, the aim of this paper is to enhance scientific knowledge with regard to the

9 association between the degree of CSR performance and the composition of the compensation of the CEO. More specifically put: how does CSR performance influence the compensation of the CEO?

Results 1.4

This research question is tested by using an OLS regression including a proxy for CSR performance and various control variables and corrected for fixed industry and year effects and two-dimensional clusters for firms and years. With regard to the effect of CSR performance on short-term compensation, this study found a significant negative relationship. This means that, with regard to the short-term compensation part, CEOs are given disincentives to enhance CSR performance. Additionally, this study found that CSR performance is not significantly related to total CEO compensation. The interpretation for this non-significant relationship is that CEOs are not compensated for enhancing CSR performance. Furthermore, using additional and robustness analyses, this study found that CSR performance only has a positive significant effect on long-term compensation for the period from 2010 until 2014.

Contribution 1.5

This study contributes to scientific knowledge by extending limited prior research in analyzing the effect of CSR practices on of CEO compensation by using a large sample over an extensive period of time. Following prior literature, this study investigates the effect of performance on CEO compensation and introduces the determinant of CSR performance directly. Combining the two performance concepts, this study includes whether CEOs that outperform other CEOs with respect to CSR receive higher compensation even if these CEOs already outperform other CEOs based on corporate financial performance. Given the increasing public awareness of sustainability issues and the fact that sustainability strategies may become a key factor in competitive advantage, understanding the relationship between CSR performance and CEO

10 compensation is crucial to successfully managing companies. Therefore, the contributions of this paper are not merely limited to theoretical knowledge, but also to practical content.

Structure 1.6

The remainder of this paper is structured as follows. The next section provides the theoretical background for the study. In this section the possible forms of CEO compensation are set out together with the associated incentives and the concept of CSR reporting is explained. Additionally, prior literature with regard to the determinants of CEO compensation is discussed, whereby the concepts of corporate financial performance and CSR performance are highlighted. The third section details the research method used to address the research question. The hypotheses are also developed in this section. The fourth section contains the results of the empirical analysis. The fifth section provides the discussion and analysis of the empirical results. The final section provides concluding remarks and recommendations for future research.

11 2 Literature review

This section firstly explains the theories that are drawn upon in this paper, namely the agency theory and the stakeholder theory. In order to analyze the influence of CSR performance, and more specifically the degree of CSR performance, on the composition of CEO compensation, the concept of CEO compensation and the concept of CSR performance are explained. Finally, prior literature that studied the relationship between CSR performance and CEO compensation is discussed.

2.1. Main accounting theories tested

The concept of CEO compensation and the concept of CSR performance have been widely studied in literature. Despite this, there is no common theory on the relationship between the two concepts (Cai, Jo, & Pan, 2011). A reason for this is that empirical investigations of CEO compensation and the effectiveness of organizations have been limited by researchers’ reliance on financial measures instead of more broadly defined measures (Coombs & Gilley, 2005). Another reason is that the main dependent variable included in studies with regard to the influence of CSR performance is the corporate financial performance of a firm (Callan & Thomas, 2011, 2014). In order to analyze the relationship between the two concepts, the reasoning behind performing CSR practices is discussed by using two competing, but explanatory theories. Both theories are rooted in the agency theory (Callan & Thomas, 2014), but they have different implications for the studied relationship. The theories that are used are the agency theory with regard to shareholder wealth maximization and the stakeholder theory. Cai et al. (2011) studied both theories in their study concerning whether firms that are socially responsible pay their CEOs less.

The agency theory tries to explain the relationship between the principal(s) and the agent in the metaphor of a contract (Jensen & Meckling, 1976). Jensen and Meckling (1976) assume that both parties of the relationship (the principal(s) and the agent) are utility maximizers and that there is therefore good reason to believe that the agent will not always act in the best interest

12 of the principal. An important influential factor in the contract is that information asymmetry exists between the agent and the principal, which leads to the fact that the principal is not able to completely monitor the actions of the agent (Eisenhardt, 1989). The principal can reduce the divergences from his utility function by closely monitoring the actions of the agent, and thereby incurring so-called monitoring costs, or creating incentives for the agent to act in alignment with his utility function (Jensen & Meckling, 1976). The latter method is discussed in this study, since the compensation scheme is one of the ways to align these incentives.

Firstly, consider the agency theory with regard to the contract between shareholders and the CEO. In this contract the principal is the shareholder and the agent is the CEO. Most of prior research using this agency theory has focused on the complex link between executive compensation and the firm’s financial performance (Callan & Thomas, 2014). Prior research thus focused on whether firms have found ways to compensate their executives in such a way that the agency theory is circumvented. The extent to which executive compensation is based on financial performance is the key issue in these studies. However, this current study mainly studies the extent to which executives are compensated for their CSR performance. The agency theory with regard to shareholder wealth maximization therefore suggests that CEOs would perform CSR practices if doing so provides private benefits, even if this would lead to costs for shareholders (Cai et al., 2011). This automatically leads to the need to compensate executives in such a way that performing CSR practices is beneficial for those executives, assuming that shareholders value CSR performance. However, since the agency problem with regard to shareholders generally assumes that shareholders value financial performance primary (Callan & Thomas, 2014; Eisenhardt, 1989; Jensen & Murphy, 1990), this theory is mainly applicable for these studies.

On the other hand, the stakeholder theory requires managers to serve other non-investing stakeholders, such as bondholders, employees, NGOs, social activists and government, as well (Cai et al., 2011). Stakeholders are defined as “any group or individual who can affect or is affected by the achievement of the firm’s objectives” (R. W. Roberts, 1992). This theory therefore assumes contracts with a wider range of implicit and explicit stakeholders having

13 legitimate expectations, urgent claims, and/or power regarding the firm (McGuire et al., 2003). This approach requires that the interests of these stakeholders are considered in managerial decision-making. The expectations from non-shareholder stakeholders are generally seen to be less focused on financial results and more towards sustainability principles than those of shareholders (Cai et al., 2011). McGuire et al. (2003) state that the concept of corporate social performance is an evaluation of how well firms have met the expectations of stakeholder and environmental concerns.

Merriman and Sen (2012) argue in favor of the agency theory perspective. They state that even when managers and organizations consider sustainability issues to be of strategic importance, they may withhold support if these efforts are perceived to be contradicting with the economic utility of their own compensation, and vice versa. This implies that the importance that managers tie to sustainability issues can be explained from an agency theory viewpoint. Firms may assess the costs and benefits of enhancing CSR performance in order to determine the optimal level of investment (Waldman, Siegel, & Javidan, 2006). Additionally, Waldman et al. (2006) state that agency theory would suggest that CEOs can promote increasing CSR performance for their own benefit.

Coombs and Gilley (2005) take both theories together and argue that “stakeholder-agency theory” would provide a useful foundation for studying the effect of stakeholder management on the various types of CEO compensation. The principal-agent relationship between shareholders and managers that is studied by Jensen and Meckling (1976) is then, by introducing the stakeholder theory, seen as one of a series of relationships that form an entity (Coombs & Gilley, 2005). CEOs can then be seen as stakeholders’ agents, instead of merely shareholders’ agents.

As mentioned above, since both theories are rooted in the agency theory, an important factor that is of influence in both theories is the existence of asymmetry in information that is accessible for stakeholders, including shareholders, and firms (Cho, Lee, & Pfeiffer, 2013; Jensen & Meckling, 1976; Jensen & Murphy, 1990). Cho et al. (2013) focus on the relationship between CSR performance and information asymmetry. In their study, they hypothesize and find that CSR

14 performance, both positive and negative, reduce information asymmetry. Based upon past research the authors argue that CSR performance drives transparency, since good- and bad-performing firms are eager to signal their CSR quality. Good-bad-performing firms want to signal their quality by voluntarily disclosing good CSR performance, while poor-performing firms want to explain their poor performance (Cho et al., 2013). Additionally, CSR performance may also be a driver of earnings quality, since firms with a good CSR performance are less likely to engage in earnings management (Kim, Park, & Wier, 2012). This increased transparency and earnings quality reduce information asymmetry (Cho et al., 2013). The figure below is a visual representation of the link between CSR performance and information asymmetry that was discussed above.

Figure 1. Link between CSR Performance and Information asymmetry (Cho et al., 2013).

The concept of sustainability and the increased presence in current society implies that companies are required to structure their activities in such a way that takes into account the consequences of their actions on shareholders and non-shareholder stakeholders. The next section explains the various types of CEO compensation and the implied effect on the extent to which the actions of the CEO are focused on the long- and short-term. This knowledge is then used when the hypotheses of this study are formulated in subsequent sections.

2.2. CEO compensation

Large US companies have seen significant changes to executive compensation in recent times (Murphy, 2012). For example, scandals, such as accounting fraud, led to new pay-related laws and tax, accounting and disclosure rules in order to mitigate perceived abuses in executive

15 compensation (Murphy, 2012). During the 1990s the median total pay for CEOs in the S&P 500 more than tripled, while in the early 2000s this number relatively stagnated. Callan and Thomas (2014) even stated that the median pay for CEOs has increased six fold in the last two decades. In order to analyze to what extent the composition of the CEO compensation may be affected by the CSR performance of a firm, a distinction is made in the type of compensation. Cash compensation, including base salary and cash bonus, is separated from equity compensation. Total compensation is then the summation of these two categories of compensation and all other compensation. In addition, the incentives that theory associates with these different types of compensation are set out and discussed in depth.

2.2.1. Base salary

First of all, a base salary is a fixed amount that is contractually agreed upon between the firm and the CEO (Murphy, 1999). This type of compensation is generally considered to be paid out in cash. This type of compensation is generally known to have no incentive effect for increasing future performance, since it is a relatively fixed contractually agreed amount and only to a relatively small extent based on past performance. Therefore, this type of compensation is referred to as short-term pay in this study.

Theory suggests that making payment schemes based less on fixed amounts and more based on variable and achievement-dependent of specific individual and organizational goals, financial and more broadly defined ones, makes it more likely that these goals will be realized (Gerhart & Milkovich, 1990). However, base salary may have an incentive effect to increase future performance if the change of base salary from year to year is dependent on performance (Jensen & Murphy, 1990). However, Jensen and Murphy (1990) found that the implied effect of this possibility is significantly low, as was discussed above.

16 2.2.2. Cash bonus

Secondly, in general, cash bonuses are expected to have a short-term incentive effect. Cash bonus payments reward executives for achieving short-term targets, rather than for building longer term performance potential (McGuire et al., 2003). Performance targets that have to be achieved to receive the cash bonus are to a great extent financial ones and based on a single-year’s performance (Murphy, 1999). Due to the pay-performance structure in cash bonuses, executives are inclined to withhold effort if performance will exceed the performance required to achieve the bonus cap (Healy, 1985), which implies a short-term focus. Additionally, Healy (1985) states that bonus contracts specify annual rather than long-term objectives. Furthermore, Murphy (1999) acknowledges that financial performance measures, mostly accounting numbers, are backward-looking. This type is therefore, just as base salary, regarded as short-term pay in this study.

2.2.3. Equity compensation

Thirdly, equity compensation is generally expected to have a long-term incentive effect. Equity compensation includes, among other things, the rewarding of stock options and restricted stock (Murphy, 1999). This type of compensation is contingent on the value of the company in the future (Murphy, 1999). Equity incentives are therefore defined as the sensitivity of the executive’s wealth to stock price changes (Core, Guay, & Larcker, 2003). According to Hall (1998), if performance is measured by changes in shareholder value, stock and stock option-based payments account for approximately 98 percent of the pay-for-performance relationship. Therefore, equity compensation will incentivize CEOs to focus their effort on increasing firm value in the future, since their compensation is positively associated with this value (Mahoney & Thorn, 2006). Therefore, equity compensation is assumed to increase the long-term focus of the CEO and thus the firm.

17 2.3. Corporate social responsibility (CSR) performance and reporting

Our Common Future (hereafter: the Brundtland Report) issued by The Brundtland Commission is generally seen as one of the first important developments that introduced environmental concerns to the formal political agenda. The concept of a sustainable development path was hereby introduced (Brundtland et al., 1987) and led to a major shift in the awareness of company activities and its consequences. This sustainability concept implies that society must use no more of resource than can be regenerated (Aras & Crowther, 2009). In addition, Aras and Crowther (2009) quote from the Brundtland Report that “Sustainability development is development that meets the needs of the present without compromising the ability of future generations to meet their own needs”. The rise of CSR practices is often assumed to be directly linked to globalization and expectations that firms will fill gaps that are left behind by global governance failures (Frynas, 2005). Concerns of sustaining life are thus pertinent at the macro level of society, but are equally relevant at the micro level of the corporation. The quote from the Brundtland Report above implies that the concept of sustainability triggers companies, and thus the micro level, to be more long-term focused than before. Firms have been pressurized to act on the social and environmental impact of their activities by stakeholders. In addition, social and environmental initiatives may have a potentially negative effect on short-term financial outcomes (Deckop et al., 2006; Merriman & Sen, 2012). Deckop et al. (2006) and Merriman and Sen (2012) explain this by giving the example that sustainability initiatives may require a significant up-front investment, which will impact short-term financial performance negatively.

Focusing on this micro level of the corporation, CSR practices are broadly defined as the active and sometimes voluntary contribution of enterprise to environmental, social and economic improvement (Garay & Font, 2012). Garay and Font (2012) go deeper into the reasoning for implementing sustainability practices by stating that the most widely quoted argument for business engagement is “the increased competitive advantage due to the development of some valuable capabilities”. This may include capabilities such as stakeholder integration, continuous innovation or higher-order learning. The implementation of CSR practices may still be aligned with the classic objective of maximizing returns and obtaining competitive advantages through

18 cost reduction, sales increases, new market opportunities and enhanced company image (Garay & Font, 2012). This argument is supported by Waddock and Graves (1997) as well, since they argue that a competitive advantage can be achieved when better CSR performance allows companies to attract better personnel more easily, avoid battles with governmental officials or communities or pay fines for environmental problems.

The reporting on these CSR practices is then a vital tool for organizations to provide transparent communication with their stakeholders, especially about organizations’ social and environmental performance (Junior, Best, & Cotter, 2014). Garay and Font (2012) state that the main consequences of such a CSR report is to consider the role and influence of stakeholders, such as shareholders but, also employees, customers and other actors. The Global Reporting Initiative (GRI) describes a sustainability report as a report published by a company or organization about the economic, environmental and social impacts caused by its everyday activities. They go on to state that such a report also presents the organization's values and governance model, and demonstrates the link between its strategy and its commitment to a sustainable global economy. Possible topics that may be addressed in CSR reports include internal and external employee issues, community, economic, and environmental issues (Fortanier, Kolk, & Pinkse, 2011; Garay & Font, 2012; McGuire et al., 2003). As discussed above, CSR performance motivates firms’ voluntary CSR disclosure, irrespective of whether this CSR performance is positive or negative (Cho et al., 2013).

2.4. Prior literature on this topic

As discussed above, both concepts have been studied to a large extent, but a common theory on the relationship between both concepts has not been established. However, the relationship between the concepts have been studied in prior studies and this is discussed in this section. Studies that investigated the direct relationship mainly focused on the influence of CEO compensation on CSR performance. On the other hand, studies that focused on the inverse relationship mainly focused on the influence of CSR performance on corporate financial

19 performance and the influence of corporate financial performance on CEO compensation. Some studies, however, investigated the relationship between CSR performance on CEO compensation.

2.4.1. Prior literature on the inverse relationship

As stated above, the major part of prior literature that studied the relationship between CSR performance and CEO compensation is aimed at studying the effect of CEO compensation on CSR performance (Berrone & Gomez-Mejia, 2009; Deckop et al., 2006; Mahoney & Thorn, 2006; McGuire et al., 2003; Merriman & Sen, 2012; Stanwick & Stanwick, 2001). These studies found that there exists a link between the two concepts, but that the relationship may vary if the institutional environment differs. For example, McGuire et al. (2003) found that high levels of base salary and long-term incentives are related to poor social performance, while Mahoney and Thorn (2006) found that long-term incentives are related to strong social performance. A possible reason for this is that McGuire et al. (2003) analyzed the studied relationship in firms in the U.S., while Mahoney and Thorn (2006) performed their analysis on firms located in Canada. On the other hand, Stanwick and Stanwick (2001) found that there exists an inverse relationship between total CEO compensation and environmental reputation. Deckop et al. (2006) found that the more firms focus on short-term CEO pay, the lower the firm’s corporate social performance, while the more firms focus on long-term CEO pay, the higher the firm’s corporate social performance.

2.4.2. Prior literature on influence of CSR performance on financial performance

Prior literature that used CSR performance as the independent variable used the corporate financial performance often as the dependent variable (Callan & Thomas, 2011). Margolis and Walsh (2003) present a review of this literature and conclude that this relationship has been studied 127 times in published studies between 1972 and 2002. The seminal work is generally credited to Bragdon and Marlin (1972), and Moskowitz (1972) (Callan & Thomas, 2014). Almost

20 half of the papers (54) reviewed by Margolis and Walsh (2003) point to a positive relationship between CSR performance and financial performance, while only seven found a negative relationship and 28 found no significant relationship. A clear signal that a positive association exists between company’s social performance and its corporate financial performance is therefore identified (Margolis & Walsh, 2003). Viewed from an agency theory perspective this contribution of CSR performance to corporate financial performance might indicate that the firm’s resources are being used to advance the interests of shareholders.

The relationship between CSR performance and corporate financial performance was studied by, at least, the following two studies in 1997. Waddock and Graves (1997) study the causal relationship in an US context and find, in support of what they termed the ‘good management theory’, that better corporate social performance may lead to improved corporate financial performance. The most plausible explanation for this relationship is argued to be that acting proactively to concerns that reflect a variety of stakeholder interests make these stakeholder relations more prominent in the link to financial performance (Waddock & Graves, 1997). The other study is a study from Preston and O’Bannon (1997) that introduces multiple hypotheses about the relationship between the two studied concepts. These hypotheses includes, among others, the social impact hypothesis, the trade-off hypothesis and the positive or negative synergies hypothesis (Preston & O’Bannon, 1997). The first one expects that favorable social performance will lead to favorable financial performance, since external reputation among corporate stakeholders is positively related to financial performance. The second one states that social accomplishments involve financial costs that put the firm at a relative disadvantage compared to firms that are less socially active. This hypothesis therefore assumes a negative association. The last hypothesis that is introduced here argues that social and financial performance are synergetic, either positive or negative. This means that the two concepts interact, either positive or negative, through time. The authors find merely positive associations between social and financial performance in their dataset of US firms. This implies no support for the trade-off or negative synergy hypotheses introduced above (Preston & O’Bannon, 1997).

21 Additionally, a study by Berman, Wicks, Kotha, and Jones (1999) studied the relationship between stakeholder management and firm financial performance. The authors compare the view that managerial concern for stakeholders is solely determined by the perceived ability to improve financial performance as a consequence and the view that firms have a (moral) commitment to treating stakeholders in a positive way, which may then be seen as part of their strategy and impacting corporate financial performance (Berman, Wicks, Kotha, & Jones, 1999). The authors find that two dimensions of stakeholder management, employees and product safety/quality, have strong effects on financial performance. These dimensions are perspectives from the MSCI database that is explained in upcoming sections. Additionally, the perspectives from the MSCI database can be seen in Appendix 1. This suggests that firms may be better of focusing merely on these perspectives (Berman et al., 1999). The authors find no significant influence of stakeholder management on financial performance through the corporate strategy. For this study, an aggregated measure of CSR performance is used instead of a measure per perspective. This is discussed in the methodology section.

A more recent study from Makni et al. (2009) studied the causality between corporate social performance and financial performance in a Canadian context. The authors use the hypotheses introduced by Preston and O’Bannon (1997) introduced above to support the conclusions they draw upon their research. They find no significant relationship between the aggregated CSR score and corporate financial performance, except for market returns (Makni, Francoeur, & Bellavance, 2009). They, however, find a robust negative relationship between the environmental dimension of CSR, which are explained in the methodological section of this study, and return on assets, return on equity and market returns. This is consistent with the trade-off hypothesis and, in part, with the negative synergy hypothesis from Preston and O’Bannon (1997), which argues that firms that are socially responsible experience lower profits, which then leads to limited socially responsible investments (Makni et al., 2009).

Finally, a study by Perrini, Russo, Tencati, and Vurro (2011) deconstruct the relationship between CSR performance and corporate financial performance. Adopting a stakeholder-based view of the engagement of firms in CSR practices, the authors provide guidance to better understand

22 the mechanisms by which CSR practices translate into corporate financial performance (Perrini et al., 2011). Perrini et al. (2011) develop a framework that, according to the authors, can be used by companies and the investment community to better understand how CSR practices can positively affect corporate performance. CSR efforts lead to drivers of performance, which subsequently leads to performance outcomes with regard to revenue-related outcomes and cost-related outcomes (Perrini et al., 2011). These performance outcomes include respectively growth opportunities and competitive positioning on the one hand and operational efficiency and reduced costs of capital on the other.

2.4.3. Prior literature on influence of corporate financial performance on CEO compensation Additionally, prior literature studied the impact of corporate financial performance on CEO compensation to a large extent. One of the most cited articles concerned with this topic is the article from Jensen and Murphy (1990) about performance pay and top-management incentives. The authors study the many mechanisms through which compensation policy can provide value-increasing incentives, including performance-based bonuses and revisions of salary, stock options, and dismissal decisions based on performance (Jensen & Murphy, 1990). The resulting incentives per category of compensation varies significantly. A change in shareholder wealth of $1,000 leads to an average increase in the salary and bonus of this year and next year of about two cents; wealth consequences associated with revisions of salary, outstanding stock options and dismissals decisions based on performance of about 75 cents; and an average increase of $2.50 with regard to stock owned by the CEO. The pay-performance sensitivity is thus estimated at $3.25 per $1,000, or 0.325% (Jensen & Murphy, 1990). The authors add that this result depends on the size of the studied firm.

A study from Hall and Liebman (1997) follows up on the study performed by Jensen and Murphy (1990) by using a broader measure of executive compensation with more focus on holdings of stock and stock options by CEOs. This shifted focus results in a “very strong link between the fortunes of CEOs and the fortunes of the managed companies” (Hall & Liebman, 1997). Hall and

23 Liebman (1997) find that roughly the complete pay-to-performance sensitivity is driven by these CEO stock and stock options holdings. The wealth of CEOs changes by millions of dollars when firms perform moderately above average in comparison with firms performing moderately below average (Hall & Liebman, 1997). The authors further show that the level and the sensitivity of CEO compensation to firm performance have risen significantly during their research period of 15 years. The reasons for this are, as explained above, the increased popularity of compensation in the form of stock holdings and options and additionally the fact that the US stock market has increased substantially (Hall & Liebman, 1997).

Finally, another study by Hall (1998) analyzes the pay-to-performance incentives that are created by stock options. Therefore, Hall (1998) studies to what extent CEO compensation, the equity component especially, is affected by the financial performance of the firm. The author finds that stock options have a considerably higher pay-to-performance sensitivity than rewarded stock. Additionally, the value of stock option grants is positive and statistically significantly related to past performance (Hall, 1998). According to Hall (1998), the relationship between stock price performance and option grants is even five to eight times stronger than the relationship between stock price performance and salary and bonuses, which is referred to as short-term compensation in this study.

2.4.4. Prior literature on studied relationship

This study studies the impact of being socially responsible, thus having a high CSR performance, on the composition of CEO compensation. The relationship that is studied in this paper is to my knowledge one of the first. A study performed by Coombs and Gilley (2005) did, however, study the effect of lagged values of stakeholder management on CEO compensation, both as a main effect and as an interaction effect with the financial performance of the firm. The authors propose that stakeholder management will have a positive effect on CEO compensation levels. Using five types of stakeholder management variables, namely community relations, diversity, employee relations, environmental impact, and product safety/quality, the authors conclude

24 that boards consider financial and stakeholder-related performance when determining CEO compensation (Coombs & Gilley, 2005). As far as base salary is concerned, Coombs and Gilley (2005) found that CEOs appear to be given disincentives to engage in stakeholder management issues. For cash bonuses, stock options and total compensation, the authors found almost no significant effects of stakeholder management compensation (Coombs & Gilley, 2005). Coombs and Gilley (2005) also test whether stakeholder management and financial performance have a combined effect on CEO compensation. The authors test the interaction of financial performance with the five types of stakeholder management variables and include them in the model, but a consistent pattern was difficult to find (Coombs & Gilley, 2005).

Additionally, a study performed by Benson and Davidson (2010) studied the relationship between one year lagged stakeholder management, firm value, and CEO compensation. The primary purpose of the study was to explore the correlation between stakeholder management and firm value. However, by proposing that firms will compensate its CEOs based on firm’s goals the authors investigate whether stakeholder management and/or shareholder value maximization either individually or jointly influence the CEOs compensation (Benson & Davidson, 2010). The authors find no statistical evidence that changes in stakeholder management directly influence CEO compensation. They do, however, find that stakeholder management is positively related to firm value and that firm value is positively related to CEO compensation (Benson & Davidson, 2010). Benson and Davidson (2010) conclude that the findings of their study are not consistent with either stakeholder theory or strict shareholder value maximization theory exclusively and that the ultimate goal of firms is to maximize firm value. CEOs compensation is structured to comply with this goal, but the maximization of firm value also requires sufficient stakeholder management (Benson & Davidson, 2010).

Thirdly, a study performed by Cai et al. (2011) studied the impact of being socially responsible on CEO compensation. However, their research study did not study the exact same relationship that is studied in this paper. Cai et al. (2011) studied the impact of CSR performance on CEO compensation, but only to the extent of the effect of this variable on total compensation and cash (bonus) compensation. Additionally, Cai et al. (2011) use the lagged values as a proxy for

25 CSR performance. They investigated the difference in CEO pay between socially responsible and socially irresponsible firms. Cai et al. (2011) firstly conclude that CEOs in socially responsible firms receive significantly less compensation than CEOs in otherwise similar but socially irresponsible firms. Secondly, the authors conclude that higher levels of CSR lead to lower levels of CEO compensation in the following year (Cai et al., 2011). This found relationship is consistent with the viewpoint based on the stakeholder theory that was discussed above (Cai et al., 2011). Finally, Callan and Thomas (2011, 2014) studied the relationship between CSR performance and executive compensation, while including corporate financial performance in their studies from 2011 and 2014. The paper from 2011 uses regression analysis to identify the pay-for-performance relationship, which represents the marginal effect of changes in firm pay-for-performance on executive pay. Identifying and measuring the determinants of and relationship among CSR performance, corporate financial performance and executive compensation is the primary objective of this study. The authors develop a multi-equation model that allows for the endogeneity of the studied variables (Callan & Thomas, 2011). The authors find that the variables are indeed endogenous and that the variables can influence one another through feedback or indirect effects and that independent variables can affect the studied variables. The authors also find that CSR performance is positively influencing financial performance.

In their paper from 2014, Callan and Thomas develop a multi-equation model for the three key variables that they assumed endogenous in their earlier study: CEO compensation, CSR, and corporate financial performance. They use this model to test the influence of each key variable on the other two and to examine whether the measure of compensation, being either short-term, long-term or total compensation, is of influence (Callan & Thomas, 2014). This is where this study of Callan and Thomas is becoming most relevant to this study, since their study involves the divergence between different types of CEO compensation as well. However, by including data from approximately the last two decades instead of merely three years, this study tries to present a more extensive analysis. The study from Callan and Thomas (2014) concludes that CEOs seem to be paid according to corporate financial performance. Additionally, CSR is found to influence short- and long-term compensation, both in absolute terms and

26 proportionately. Corporations might consider long-term compensation to motivate CEOs more effectively to achieve CSR objectives (Callan & Thomas, 2014).

2.4.5. Prior literature on other (economic) determinants of CEO compensation

Besides prior literature on the influence of corporate financial performance and CSR

performance on CEO compensation, accounting literature is concerned with other (economic) determinants that determine CEO compensation. In order to provide a complete view of the determinants of CEO compensation, this prior literature is included in this literature review as well.

First of all, a study by Core et al. (1998) studied whether corporate governance variables explain variations in CEO compensation. This paper is one of the mostly cited papers concerning the topic of CEO compensation. The authors divide corporate governance variables into two

categories: board-of-director characteristics and ownership structure. Core et al. (1998) control for demand for managerial talent, firm financial performance, and firm risk in order to

investigate the association of board and ownership structure with the level of CEO compensation. The authors find that both categories have a substantive cross-sectional

association with the amount of CEO compensation, while controlling for standard determinants for the level of CEO compensation, as discussed above (Core, Holthausen, & Larcker, 1999). The authors furthermore test if the amount of CEO compensation is associated with subsequent firm performance. They perform this additional test to test whether the board-of-director and ownership variables are reflecting the relative effectiveness of governance structures in controlling agency problems, as opposed to that these corporate governance variables are additional proxies for the demand for a high-quality CEO. Since a negative association with subsequent firm financial performance exists, the corporate governance variables are suggested to reflect the effectiveness for controlling agency problems (Core et al., 1999).

27 Additionally, five years prior to the study performed by Core et al. (1999), Boyd (1994) focused on the influence of board of director characteristics on CEO compensation. Prior research mainly included this determinant as a part of a wider analysis, but this does not clarify the unique effect of the board in setting executive compensation (Boyd, 1994). Boyd (1994) studied whether CEO compensation is affected by board control, which includes CEO duality, the ratio of insider directors, the percentage of stock owned by the directors, and the compensation of the

directors, while he controlled for firms size and profitability. The author finds that board control has a significant negative relationship with CEO compensation and that firm size and firm profitability are positively, but not significantly, related to CEO compensation (Boyd, 1994). Finally, with regard to corporate governance determinants, a study by Armstrong, Gow and Larcker (2013) focused on the influence of corporate governance on the corporate policy with regard to executive compensation. The authors focus especially on the impact of shareholder voting outcomes on equity-based compensation plans on firms’ executive compensation policies (Armstrong, Gow, & Larcker, 2013). The authors find virtually no evidence of decreased CEO incentive compensation as a result of low shareholder support for, or even rejection of, proposed equity compensation plans. This indicates that corporate governance variables, as discussed by Core et al. (1998), are of more importance in determining CEO compensation. However, the literature on the determinants of CEO compensation dates back even further. Agarwal (1981) developed a model in which he viewed executive compensation as a function of three basic factors: job complexity, employer’s ability to pay and executive human capital. The relationship between job complexity and CEO compensation is logically expected, since the complexity of the job measures the nature and magnitude of responsibilities (Agarwal, 1981). Secondly, the employer’s ability to pay is taken as a major determinant of executive compensation, since companies’ ability to pay higher wages may enlarge the number and quality of applicants, minimize employee turnover, and stabilize wage costs (Agarwal, 1981). Thirdly, according to Agarwal (1981), the amount of human capital is expected to influence productivity and therefore earnings. The author finds that all three determinants are significantly influential on CEO compensation, but the employer’s ability to pay is found to be of most importance.

28 Therefore, the first and third determinant are, based on the prior literature on the relationship between CSR performance and CEO compensation discussed above and the study of Agarwal (1981), perceived to be of less relevance to this study.

Additionally, a study by Finkelstein and Hambrick (1989) present a model of the determinants of CEO compensation that includes several additional factors compared to Agarwal (1981). They divide between market factors, such as firm size, corporate performance, corporate complexity and human capital, and political forces, such as CEO power and board vigilance (Finkelstein & Hambrick, 1989). They conclude that firm size is positively related to total compensation and base salary and that firm profitability is positively related to total compensation. Additionally, complexity is not related to one of the CEO compensation measures, which is a different conclusion than Agarwal (1981) found. With regard to the political factors, CEO power had a mixed association with CEO compensation (Finkelstein & Hambrick, 1989). CEO tenure is not significantly related to one of the three measures of CEO compensation, CEO holdings was only significantly related to CEO base salary and CEO’s family holdings was negatively associated with all of the three CEO compensation measures. Furthermore, vigilant boards are not significantly linked to one of the three executive compensation measures, while for bonuses the interaction between vigilance and ROE was negative (Finkelstein & Hambrick, 1989). This surprising finding suggests that the pay-for-performance association is lessened due to vigilant boards.

The studies presented above that examine other (economic) determinants of executive compensation focus on several variables that have an effect on the variation in executive compensation, but these articles do not focus on the metrics on which executive compensation is based. A study by Ittner et al. (1997) does so by focusing on the relative weights placed on financial and non-financial performance measures, which may include CSR performance, in CEO bonus contracts. Non-financial measures may be introduced to increase the efficiency of contracting, and thus decrease the agency problem, since they increase the information about managerial effort desired by the owner (Ittner, Larcker, & Rajan, 1997). On the other hand, according to Ittner et al. (1997) non-financial measures may be incorporated by powerful CEOs to increase compensation that is justifiable by financial performance. The authors find that the

29 strategy of the firm, specifically a more prospector strategy, may increase the relative weight on non-financial performance in annual bonus contracts (Ittner et al., 1997). This supports the argument made by McGuire et al. (2003) that firms attempt to link executive compensation to objectives of the firm in order to align management incentives and organizational goals. Additionally relevant to this study is the finding that the relative weight placed on non-financial performance is dependent on the noise in the measurement of financial performance. As the noise in financial measures increases, firms place more weight on non-financial measures (Ittner et al., 1997). Relating this to this study means that in theory more weight will be placed on CSR performance if financial performance measurement will be more noisy, which will be the case if corporate financial performance provides relatively less information about the manager’s actions.

A more recent study by Graham et al. (2012) added to prior literature that the variation in executive compensation is also significantly determined by unobservable firm and managerial characteristics. Unobservable firm characteristics may include the culture of a firm, while unobservable managerial characteristics may include the innate ability or personality of the manager (Graham, Li, & Qiu, 2012). The authors conclude that these fixed firm and managerial effects explain the majority of the variation in executive compensation can be explained by these time-invariant effects. Including these firm and managerial fixed effects can furthermore lead to alterations in the magnitudes of the coefficients estimated for other independent variables. (Graham et al., 2012). An example of this is firm size, for which the effect on executive compensation is significantly smaller when fixed effects are included in the analysis. However, the authors find no change in the signs of the coefficients. Therefore, including these fixed effects does not lead to a completely different empirical model. Interpreting this, the argument that Graham et al. (2012) present is that firm size as such is not explaining the variance in executive compensation in total, but that managerial fixed effects may have a contribution to, or in other words explain part of, the identified association.

30 Table 1 below presents a chronological overview of the papers used per prior literature area. Based on the extensive literature review of these papers, the research methodology is developed and hypotheses are formulated in the upcoming section.

Table 1: Chronical presentation of scientific articles used in studying relationship in prior literature

Author(s) Year Title

Prior literature on inverse relationship

Stanwick & Stanwick 2001 Titled: “CEO compensation: Does it pay to be green?”. McGuire et al. 2003 Titled: “CEO incentives and Corporate Social Performance”.

Deckop et al. 2006 Titled: “The effect of CEO pay structure on Corporate Social Performance”. Mahoney & Thorn 2006 Titled: “An examination of the structure of executive compensation and

Corporate Social Responsibility: A Canadian investigation”. Berrone &

Gomez-Mejia

2009 Titled: “Environmental performance and executive compensation: an integrated and agency-institutional perspective”.

Merriman & Sen 2012 Titled: “Incenting managers toward the triple bottom line: An agency and social norm perspective”.

Prior literature on influence of CSR performance on financial performance

Waddock & Graves 1997 Titled: “The corporate social performance-financial performance link”. Preston & O’Bannon 1997 Titled: “The corporate social-financial performance relationship”.

Berman et al. 1999 Titled “Does stakeholder orientation matter? The relationship between stakeholder management models and firm financial performance”.

Margolis & Walsh 2003 Titled: “Misery loves companies: Rethinking social initiatives by business”. Makni et al. 2009 Titled: “Causality between corporate social performance and financial

performance: Evidence from Canadian firms”.

Perrini et al. 2012 Titled: “Deconstructing the Relationship Between Corporate Social and Financial Performance”.

Prior literature on influence of corporate financial performance on CEO compensation

Jensen & Murphy 1990 Titled: “Performance pay and top-management incentives”. Hall & Liebman 1997 Titled: “Are CEOs really paid like bureaucrats?”.

Hall 1998 Titled: “The pay to performance incentives of executive stock options”. Studied relationship in prior literature

Coombs et al. 2005 Titled “Stakeholder management as a predictor of CEO compensation: Main effects and interactions with financial performance”.

Benson et al. 2010 Titled: “The relation between stakeholder management, firm value and CEO compensation: A test of enlightened value maximization”.

Cai et al. 2011 Titled: “Vice or virtue? The impact of corporate social responsibility on executive compensation”.

Callan and Thomas 2011 Titled: “Executive Compensation, Corporate Social Responsibility, and Corporate Financial Performance: A Multi-Equation Framework”.

Callan and Thomas 2014 Titled: “Relating CEO Compensation to Social Performance and Financial Performance: Does the Measure of Compensation Matter?”.

31 Agarwal 1981 Titled: “Determinants of Executive Compensation”.

Finkelstein & Hambrick 1989 Titled: “Chief Executive Compensation: A Study of the Intersection of Markets and Political Processes”.

Boyd 1994 Titled: “Board Control and CEO Compensation”.

Ittner et al. 1997 Titled: “The choice of performance measures in annual bonus contracts”.

Core et al. 1999 Titled: “Corporate governance, chief executive officer compensation, and firm performance”.

Graham et al. 2012 Titled: “Managerial attributes and executive compensation”.

Armstrong et al. 2013 Titled: “The Efficacy of Shareholder Voting: Evidence from Equity Compensation Plans”.

32 3 Development of hypotheses and research methodology

Based on the theoretical background and prior literature on the relationships between the studied variables that was discussed in the previous sections, this section provides the development of the hypotheses and the research method that is used to test these hypotheses for this study.

3.1. Development of hypotheses

In this section the hypotheses with regard to the influence of CSR performance on short- and long-term and total compensation are developed as this are the key relationships studied in this paper. Based on prior literature, the other (economic) determinants that are perceived to be of most relevance to this study are used as control variables and therefore no hypotheses are formulated with regard to these variables. The methodology for testing these hypotheses is further explained thereafter.

The hypotheses are mainly developed using the stakeholder theory rooted in the agency theory. Firstly, the stakeholder theory rooted in the agency theory assumes that the main objective of executive compensation is to align the incentives of the CEO with those of the principles (Jensen & Murphy, 1990). If on the other hand is assumed that CSR performance leads to performance on the long term, CEOs that enhance CSR performance are not expected to receive short-term compensation for their achievement with regard to CSR performance. Therefore, short-term compensation is expected to not or negatively be related to CSR performance. Additionally, based on the general idea from the agency theory, short-term compensation does not incentivize CEOs to have a long-term focus, since this would not lead to private benefits (McGuire et al., 2003; Murphy, 1999). Compensating CEOs in this manner will therefore not benefit socially responsible firms.

A different argument, specifically based on the stakeholder theory viewpoint, is that CEOs of socially responsible firms will possibly receive less short-term compensation to mitigate potential conflicts of interests among managers and other stakeholders (Cai et al., 2011).

33 Furthermore, CEOs that engage CSR practices are perceived to have more self-control, modesty, moderation, unselfishness, and humility compared to other CEOs (Cai et al., 2011). Cai et al. (2011) go on and state that CEOs that are perceived to be socially responsible should refuse exorbitant compensation, since it would be out of character. Finally, Cai et al. (2011) acknowledge that ethics would suggest that a more modest pay is more suitable for CEOS with high social and ethical standards.

Additionally, Gerhart and Milkovich (1990) state that to achieve specific organizational goals, such as being more socially responsible, compensation schemes should be more focused on variable and achievement-dependent pay instead of fixed pay. Based on the stakeholder-agency theory that was discussed above, short-term compensation of CEOs is therefore suggested to be not or negatively associated with multiple stakeholder management dimensions (Coombs & Gilley, 2005). This is due to the fact that short-term compensation contains a fixed pay component in the form of base salary. On the other hand, you can argue that CSR performance is a type of performance for which a CEO should be compensated and that this will be taken into consideration when the annual cash bonuses are determined. This can imply a positive association between the two concepts. However, Benson and Davidson (2010) found no significant evidence that CEOs are compensated for managing relations with their stakeholders. This paper anticipates in accordance with the mixed results from prior literature by expecting no direction in the hypothesis with regard to the influence of CSR performance on short-term compensation.

With regard to the effect of CSR performance on long-term compensation, the general agency theory suggests a positive association. If a firm is wholly owned, meaning hundred percent of the equity is owned by one person, this owner will make decisions that maximize his utility and therefore the utility of the firm (Jensen & Meckling, 1976). Concerning the effect of being socially responsible on CEO compensation, this would mean that in order for CEOs to behave in a socially responsible way that focuses on the long-term, these CEOs have to be compensated accordingly.

34 According to the stakeholder theory, companies need to respond to growing pressures from stakeholders to behave in more sustainable ways (Font, Walmsley, Cogotti, McCombes, & Häusler, 2012). Sustainability initiatives may require substantial up-front investments that affect short-term financial performance negatively, but which will increase performance in the longer run (Merriman & Sen, 2012). It therefore may be in the interest of shareholders and stakeholders to use compensation for the CEO to encourage actions that promote CSR practices and are focused on longer term objectives (Mahoney & Thorn, 2006). Additionally, Mahoney and Thorn (2006) state that firms that are socially responsible are willing to forgo short-term profits in order to invest in social and environmental objectives with no immediate payoff. Based on the stakeholder-agency theory that is set forth by Coombs and Gilley (2005), CEOs are compensated for maximizing value for the stakeholders because, in doing so, these CEOs are enhancing the overall effectiveness of the firm. Incentives should therefore be structured for enhancing long-term firm performance to maximize stakeholder value (Coombs & Gilley, 2005). This longer-term focus suggests a positive relationship between the variables.

Since CEO compensation includes roughly two types of compensation (Murphy, 2012), namely short-term and long-term compensation, cumulative this would result in an expected positive relationship regarding total CEO compensation. Furthermore, this had been hypothesized and found by Callan and Thomas (2014) in their study using seven indicators for CSR performance. On the other hand, they found that the total CEO compensation is significantly negatively affected by the lagged value of CSR performance (Callan & Thomas, 2014). This is in line with the results from Benson and Davidson (2010). According to the authors, based on the stakeholder theory, managers are directed to pursue the welfare of the stakeholders and should be compensated with this objective in mind (Benson & Davidson, 2010). Additionally, this is in line with the results from Cai et al. (2011). Therefore, this paper anticipates in accordance with the cumulative hypotheses for short- and long-term compensation and the prior literature on the effect of the non-lagged value of CSR performance on total CEO compensation by expecting a positive relationship.