The effect of Integrated Reporting on the compensation structure of CEOs

RSAC and MSc thesis

The effect of Integrated Reporting on the compensation structure of CEOs

Name: Elise Rosenkötter Student number: 10267018 Thesis supervisor: Dr. Q. Yang Date: 26-06-2017

Word count: 18,331

MSc Accountancy & Control, specialization: [Accountancy] Faculty of Economics and Business, University of Amsterdam

Statement of Originality

This document is written by student Elise Rosenkötter 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

The purpose of this thesis is to investigate the effect of implementing Integrated reporting on the structure of CEO compensation. Since the Integrated Reporting framework was only recently developed there is limited research available on this topic, especially empirically based research. This thesis provides empirical evidence and a better understanding of the effect of implementing Integrated Reporting on a firm’s compensation. The effect is analyzed using a dataset containing 5,734 observations of which 423 have implemented Integrated Reporting. The dataset includes 29 different countries globally and uses a timeline of 2011-2016.

The results indicate that firms that implemented Integrated Reporting use different types of compensation than firms that have not implemented Integrated Reporting. The regression analysis provides evidence that there is a change in the structure and that effect on incentive based, long-term incentive based and total compensation is negative, this is contrary to the positive effect that was expected based on the literature.

Other interesting results found, indicate that implementing Integrated Reporting has a positive effect on firm performance and firm performance has a positive effect on CEO compensation. This shows that there is not only a direct link, but also an indirect link between implementing Integrated Reporting and CEO compensation. Furthermore, firm size can also describe part of the effect of implementing Integrated Reporting on CEO compensation. Key words:

Table of contents

2.1.2. The importance of IR reporting ... 11

2.1.3. The development of IR ... 12

2.1.4. The value of implementing Integrated Reporting to the company ... 14

2.2. Compensation ... 16

2.2.1. The importance of compensation structure ... 16

2.2.2. The link between IR and Remuneration ... 18

3. Research methodology ... 25

3.1. Sample and data collection ... 26

3.2. Sample selection and sample period ... 26

3.3. Description of the variables ... 27

3.3.1. Dependent variable ... 28

3.3.2. Independent variable ... 28

3.3.3. Additional variables ... 29

3.4. Research models and research question ... 30

3.4.1. Research models ... 30

4. Descriptive statistics and sample characteristics ... 32

4.1. Descriptive statistics ... 32 4.2. Correlation matrix ... 33 4.3. Two-sample t-test ... 34 5. Regression analysis ... 35 5.1. Model 1 ... 35 5.2. Model 2 ... 36 5.3. Model 3 ... 39 5.4. Robustness check ... 41

5.4.1. Robustness check operationalization dependent variable ... 42

5.4.2. Robustness check outliers ... 42

6. Discussion ... 43

6.1. Analysis of the results ... 43

6.1.1. Analysis model 1 ... 43

6.1.3. Analysis model 3 ... 48

6.2. Limitations and future research ... 49

7. Conclusion ... 50

Reference List ... 53

Appendixes ... 61

Appendix 1: Overview of the five-stage model (PwC, 2015) ... 61

Appendix 2: Two-sample t-test output ... 61

Appendix 3: Regression output robustness check 1 ... 63

Appendix 4: Regression output robustness check 2 ... 66

1. Introduction

1.1. Objective

There has been an increase in the amount of information demanded by stakeholders (Deloitte, 2015). Currently not only financial information is important, but the amount of non-financial information is becoming even more important (Eccles et al., 2011; NBA, 2013; Reimsbach et al., 2017). The physical assets and financial assets no longer represent how the company creates value (IIRC, 2011). The amount of value created by physical and financial assets versus non-financial factors has changed. Where in 1975, 83% of the value created by companies was explained by the physical and financial assets, this has changed to only 19% in 2009 (IIRC, 2011). The additional value can be explained by so-called “other factors” (IIRC, 2011). There is a large gap between the market value of the firm and the book value reported by the firm (EY, 2013b). In response to this problem firms have been trying to fill this gap and reduce the information asymmetry between firms and stakeholders. More information is provided to stakeholders by publishing reports on for instance Corporate Governance, Remuneration and Corporate Social Responsibility (Reimsbach, 2017). This has led to an increase in the number of reports and also to an increase in the number of pages stakeholders have to review when making investment decisions (Deloitte, 2015; EY, 201; de Villiers et al., 2014).

In order to reduce the information gap and provide stakeholders with a full picture of the value of the firm, corporate reporting has to evolve. A relatively new way of reporting that provides a possible solution to reduce the information gap between book value and market value is called Integrated Reporting (IR) (IIRC, 2013). The International Integrated Reporting Council (IIRC) created the IR framework to provide firms with guidance to create a report that gives an insight into how value is created by the firm. Current financial

statements are mainly backward looking, since they rely on historical information (ACCA, 2017; Ball et al., 2012; Ball, 2013; EY, 2014; Khoroshilov and Narayanan 2008). IR also provides information on the medium and long run (IIRC, 2013).

Since Integrated reporting (IR) is a relatively new way of reporting information there has not been a lot of academic research published about this topic. Some literature discusses whether IR will change the reporting climate (Adams, 2015; Flower, 2015), however few are empirically based. The existing literature indicates that the structure of CEO compensation will change when implementing IR, however there have been few studies that provide any empirical evidence on whether firms actually change the structure. This study is to my knowledge, the first to provide empirical evidence on the effect of IR on CEO compensation. There is a gap in literature on the effect of IR, there are studies that summarize for instance,

the proposed value that IR can have for the firm. The existing literature does not elaborate on whether something has actually changed within firms that have implemented IR.

This chapter further elaborates on the background of this study, the research question, methodological approach and the research outline.

1.2. Background

As mentioned previously, in recent decennia the amount of information requested from companies by stakeholders has increased (Deloitte, 2015). During the global financial crisis of 2007-2008 some companies, which had been performing well according to their financial statements in the years before the crisis, faced bankruptcy nonetheless. The degree of transparency required by investors has increased after the financial crisis (IIRC, 2011). Stakeholders increasingly demand non-financial information in an effort to further substantiate their estimates of total company value (Adams and Simnett, 2011). In recent years the reporting standards have been amended in an effort to increase the reliability of financial statements (Rezaee, 2005). Since non-financial information is difficult to measure objectively, not all of the firm’s value can be captured by the financial statements. There is a growing market interest in disclosure of information on Environmental, Social and

Governance (ESG), so firms are providing this information on a voluntary basis to their stakeholders (Eccles et al., 2011). An IR includes financial and non-financial information in one report, it focuses on the interconnectedness between these two types of information and describes how this creates value for the firm (IIRC, 2013).

CEO compensation has been a topic of interest in academic papers (Jensen and Murphy, 1990b; Healy and Palepu, 2001; Tosi et al., 1997; Vemala et al., 2014). The amounts and the structure of compensation have been highly criticized (Core et al., 1999; Hall and Liebman, 1998; Jensen and Murphy, 1990; Mehran, 1995). There is a request from stakeholders that demand firms to provide more transparency on how CEO compensation has been determined (Ferri and Maber, 2013). Companies have to indicate how compensation is linked to the performance of the company.

The main focus of this study is on the effect implementing IR has on the structure of CEO compensation. As previously indicated, both reporting on non-financial information and CEO compensation have received high amounts of attention. Therefore, the effect of IR on CEO compensation is studied. After the financial crisis, information disclosure for both non-financial information and compensation has increased, this has already resulted in some changes, but implementing IR can improve disclosure even more. When firms implement IR, the link between performance information and value creation has to be explained (IIRC and

KPMG, 2016). The increase in knowledge can be used by firms to change the compensation structure, to a structure based on value creation. Firms can hereby provide a better

explanation to shareholders on how the CEOs compensation is determined.

1.3. Research question

The purpose of this paper is to add knowledge to the existing literature that describes the relationship between IR and CEO compensation. Currently there is limited empirical knowledge on what the effect of IR is. Since employees are an important component in

implementing IR, it is important to understand the relationship between IR and compensation. This research is an empirical study, which aims to determine if the structure of the CEO compensation has changed for firms that have implemented IR.

In order to meet the research purpose the following main research question is formulated:

“What is the effect of Integrated Reporting on the structure of the CEO compensation?”

1.4. Methodological approach

This study is composed based on a literature study. The literature study used various computerized databases, search engines, including Google Scholar and the UvA Catalogue Plus. Furthermore, the search strategy and selection criteria have resulted in a selection of 96 academic articles.

The literature review gives a theoretical background using basic economic theories and describes how IR can fill the gap between information requested and information presented by the firms. Based on the aforementioned review of the extant literature, the hypotheses are formulated. The research models that are used test for three different

structures of CEO compensation: incentive compensation, long-term incentive compensation and total compensation.

The first model gives an insight into the effect of implementing IR on the structure of CEO compensation. Three simple OLS regressions using robust standard errors are

performed to obtain evidence on whether there is a different effect on the three types of compensation.

The second model obtains evidence on whether firm performance positively mediates the effect of IR on CEO compensation. To conclude this, a three-step approach is used (Baron and Kenny, 1986). The first step indicates whether the predictor has an effect on the outcome, which is performed in the first research model. The second step indicates whether

the predictor has an effect on the mediator, which is tested by performing an OLS multiple regression analysis using robust standard errors. The third step adds firm performance measure as a third variable in the model including the predictor and the outcome.

In the third research model, evidence is obtained on whether firm size is a positive moderator on the effect of IR on CEO compensation. Evidence on this effect is obtained by adding the moderator and an interaction variable between the moderator and the predictor to the model. Three OLS multiple regressions using robust standard errors are performed to indicate the effect of the moderator on the different compensation structures.

1.5. Results

The research is based on a dataset containing 5,531observations of which 388 have an IR. The dataset includes 29 different countries globally and uses a timeline of 2011-2016. The main results of this study indicate that firms that have implemented IR have a different structure of the CEO compensation than firms that have not implemented IR. The results are, however, not in line with the expectations and suggest that firms that have implemented IR have lower incentive based, lower long-term incentive based and lower total compensation than firms that have not implemented IR. This study also provides evidence on the influence of firm performance and firm size on the relationship between implementing IR and CEO compensation. The results indicate that both variables have a significant effect on CEO compensation, however when correcting for outliers the results are less significant.

1.6. Research outline

To answer the research question, this report is structured as follows: chapter two presents the results of the literature study, the development of Integrated Reporting and the link between compensation structure and IR. In the third chapter the research design is explained by

providing an overview of the retrieval of the data, the variables and the research models used. The fourth chapter will include the data descriptive statistics and an analysis of the findings. Chapter 5 discusses the results. Finally, chapter 6 describes the conclusions and limitations of this study and gives insight in the recommendations for further research.

2. Literature Review

In this section the extant literature will be discussed. The first paragraph discusses Integrated Reporting (IR); an introduction will be given, the importance of reporting, the development of the framework and the benefits will be discussed. In the second paragraph remuneration

will be discussed and how this relates to IR. In the third paragraph the relationship between firm performance and IR will be reviewed.

2.1. Integrated Reporting 2.1.1. Introduction

After the global financial credit crisis of 2008, markets were characterized by a lack of trust in financial reporting. The aforementioned has prompted a change in the amount of

information demanded by the stakeholders (Deloitte, 2015). According to Deloitte (2015) Stakeholders want more transparency and information on how value is created by a firm. Not just information on the financials is important, however, also information on the

non-financials is becoming important (Eccles et al., 2011; NBA, 2013; Reimsbach et al., 2017). With this new demand for information, information on long-term goals has to be disclosed as well (PwC, 2015). According to PwC (2015) this development requires companies to explore different ways of reporting the requested information to their stakeholders. Firms started publishing reports on the different parts of their organization. Firms started publishing not only a mandatory report for the financial information; the financial statements, but also voluntarily information on Corporate Governance, Sustainability, et cetera. However, the amount of new information is overloaded, which makes it difficult for stakeholders to obtain the information they needed (IIRC, 2011). The amount of reports and also the amount of pages stakeholders have to take into account when deciding to invest in a certain company have increased (Deloitte, 2015; EY, 2014; de Villiers et al., 2014). The amount of different reports has made the information more quantitative rather than qualitative to the stakeholders (IIRC, 2011). Because of the above listed problems in existing literature, a group of experts established IR, a relatively new way of reporting. With IR all the separate reports that are published by the companies are formed into one concise report that contains information on all aspects that allow a company to create value.

IR includes information on both non-financial and financial factors in one report (Eccles and Saltzman, 2011; IIRC, 2011). Information about non-financial factors has become more important to stakeholders (García-Sánchez et al., 2013). IR is not just about adding

information from the annual and sustainability together in one report, when firms implement IR, a concise report that explains the link between the financial and non-financial

performance measures has to be published (Maroun, 2017). The goal of IR is to establish integrated thinking within a firm’s business practice. Integrated thinking is described as taking into account the interconnectedness between the capitals the firm uses or affects. Financial reports are based on historical information, so more backward looking and less

timely. The IR also focuses on the medium and the long-term timeframe and thereby having a more forward looking approach (ACCA, 2017; Ball et al., 2012; Ball, 2013; EY, 2013; Khoroshilov and Narayanan 2008).

2.1.2. The importance of IR reporting

This paragraph elaborates on the importance of IR, first a review will be shown of why reporting exists. The review discusses some fundamental theories regarding IR: agency theory, legitimacy theory and stakeholder theory. The theories provide an explanation of why companies disclose information voluntarily. Since implementing an IR is not mandatory in most countries it can be seen as a form of voluntary disclosure.

There is a gap between the amount of information stakeholders have about a company compared to the amount of information management has. This is due to the separation of ownership and control (Fama and Jensen, 1983). The stakeholders own the company but management will make the day-to-day decisions, so management is in control of how the company operates (Healy and Palepu, 2001). The information asymmetry between

management and stakeholders is known as the agency problem (Jensen and Meckling, 1976). This information asymmetry can cause management to act in their own interest

instead of supporting stakeholders in reaching their goals. Agency theory addresses the problems and provides solutions on how to align the incentives of management and the stakeholders (Donaldson and Davis, 1991). According to Walsh and Seward (1990) agency theory is an important framework to help understand the conflict between the principal and the agents and how this conflict can be reduced. Annual reporting is one way to reduce the information asymmetry, by providing a clear view on how the firm is performing. Enabling stakeholders to make better investment decisions. Thus, financial reporting increases the transparency of information (Barth and Schipper, 2008). Regulations for financial reporting will ensure that the information is reliable and correctly presented.

Firms can provide information to stakeholders in several ways. Reports can be categorized in different ways, one way of categorizing is that firms can offer reports about financial reports about non-financial information. The reason that firms provide non-financial information to their stakeholders can be explained by the legitimacy theory. Legitimacy theory is based on the fact that companies can be successful if they act in a socially acceptable way (Guthrie and Parker, 1989; O’Donovan, 2002). The actions are based on social norms, values and beliefs. With legitimacy theory the amount of voluntarily disclosure on, for instance, environmental and social information can be explained (Suchman, 1995).

The stakeholder theory is another theory that provides insight into why firms

voluntarily disclose information on corporate social responsibility (CSR). A stakeholder can be an individual or a group of people that can influence the achievement of the company’s goals or is affected by the company’s objectives (Freeman, 1999). There are two main variants of the stakeholder theory: an ethical and a managerial approach (Mahadeo et al., 2011). Mahadeo et al. (2011) argue that the ethical perspective means that companies have to account for their actions. The managerial approach is however seen as the more important; it focuses on the need to control stakeholders who have a greater and more critical impact on the company (Mahadeo et al., 2011).

With IR the information provided by companies give an insight on how the company creates value (IIRC, 2013). This information will provide stakeholders with a better view of the company, so the amount of information asymmetry will be reduced. The reason why firms would implement IR, which is not mandatory in most countries, could be explained by the legitimacy theory and the stakeholder theory.

2.1.3. The development of IR

In 2010 the International Integrated Reporting Council (IIRC) was created to guide the development of IR. The council was initiated by the Global Reporting Initiative (GRI) and the Prince’s Accounting for Sustainability Project. The council consists of regulators, investors, companies, standard setters, the accounting profession and NGOs from all over the world (IIRC, 2013). According to the IIRC an IR is defined as a way for a company to communicate with their stakeholders. An IR provides stakeholders information about how value is created by the company over the short, medium and long-term (IIRC, 2011).

IR has first been implemented as a pilot in 2011 and at the end of October 2011 the program contained approximately 40 global leading companies (GRI, 2012). The group of companies is comprised of different sectors and different sizes, this gives a clear overview of the effect based on different firm characteristics. This pilot program focused on the first implementation phase of the IR. The pilot took three years and contained three different phases (GRI, 2012). The first phase was called ‘Dry run’, in this phase the companies participating could express their view on the current knowledge of IR, what they saw as opportunities and threats that could occur during the implementation of the IR. The second phase “Pilot Cycle 1”, the IR is implemented with the first group of participants. Phase three “Pilot Cycle 2”, the last phase contained the second group of participants implementing the IR using the experiences of the first round (GRI, 2012). In 2013 the first IR Framework was published (IIRC, 2013). The pilot group contained in 2013 already 100 companies in 25

different countries. IR is not obligatory and can be implemented voluntarily. However, the IIRC’s long-term goal is that IR becomes the global reporting norm (IIRC, 2013). Black Sun Plc. in cooperation with the IIRC (2014) interviewed participants of the pilot program and concluded that companies that have implemented the IR were positive about the results. Companies that implemented IR for more than one year experienced more benefits. Benefits reported include a better understanding of how the company creates value and how this improves decision making, improvement of management information, a better relationship with their stakeholders and an improvement of the relationship between the departments (Black Sun PLC and IIRC, 2014).

The Association of Chartered Certified Accountants (ACCA) (2017) reviewed 41 reports, which have implemented IR, for the period until March 2016. The review was conducted to identify benefits and challenges in the implementation of IR. The review summarizes experiences by firms that have implemented IR to give recommendations and guidance to firms that want to implement IR.

The most important benefits of the report that were identified by the ACCA are: - More integrated thinking;

- A better view of performance;

- A better relationship with stakeholders; - A better reputation;

- More efficient reporting; - Improved gross margins.

Integrated thinking is achieved through a better communication between the departments with the non-financial and financial information (ACCA, 2017). According to the ACCA (2017) the better picture of performance and improved gross margins are realized because there is a better insight into what drives the firm’s performance. More transparent information gives the firm a better reputation. More efficiency is created for the preparers of the report, but also for the users. According to the ACCA (2017) this is due to the fact that preparers have more interconnected departments and for the users it is easier to find the information needed without contacting the company. However, there were also some areas that still need improvement and which were found difficult to implement by the participants, for example, value creation in relation to the business model, showing an integrated view of factors that have an effect on value creation, conciseness, determining materiality, report good an bad news and consistency and comparability (ACCA, 2017).

Before IR was developed, there was no report that integrated for instance sustainability, remuneration or corporate governance, all information is published in separate reports

Reimsbach et al., 2017). Instead of publishing separate reports, by implementing IR all relevant information is published in one report (IIRC, 2011). IR will make all the separate reports interdependent (NBA, 2013). There is a growing demand from stakeholders to provide more non-financial information (Eccles et al., 2011). According to Eccles et al. (2011) one of the reasons why stakeholders demand more disclosure of non-financial information, is that the market value of the firm that can be assigned to tangible assets has decreased from 80% in 1975 to less than 20% in 2009.

According to Adams & Simnett (2011) IR started to develop because of three changes; (1) evolution in regulation and standards, (2) the introduction of combined

frameworks and new ways to measure environmental, (3) social and governance factors are more reliable. The framework is based on several guiding principles: strategic focus and orientation, connectivity of information, stakeholder relationships, materiality, conciseness, reliability and completeness and consistency and comparability (IIRC, 2013). These

principles give an overview of what the IR framework governs, companies can use these principles when implementing IR. With IR the IIRC wants to increase accountability and stewardship for intellectual, human, financial, natural, social and relationship and

manufactured capitals Fundamental to the framework, all capitals will eventually create value in the long run (Churet et al., 2013). The capitals function as a way to quantify the non-financial measures and they represent any store of value (IIRC, 2013).

Although IR is being used more and more it is not obligatory in most countries. South Africa is currently the only country that has a mandatory requirement for companies to publish an IR and this is since March 2010 (Rensburg and Botha, 2014). Studies looking at the results of the implementation conclude conflicting results. A study by Atkins and Maroun (2012) investigates the reactions of different stakeholders towards the implementation of the IR framework in South Africa by using a qualitative research approach. Atkins and Maroun (2012) indicate that the information on environmental social and governance (ESG) has increased and the non-financial and financial have become more integrated. However there is still a long way before IR is accepted as the main way of reporting.

2.1.4. The value of implementing Integrated Reporting to the company

In order to conclude whether implementing IR is useful, the proposed value to the firm is discussed. The Big 4 accounting firms have all published their consideration on IR, how they can assist in implementing IR and how they think about the benefits for the company. PwC (2015) mentions direct and indirect benefits from implementing IR. The direct benefits can be in the category of cost reduction or revenue growth. The indirect benefits can be in the risk

management or brand and intangibles category (PwC, 2015). EY (2013) argues that the most important benefit of implementing IR is the communication of value to the stakeholders. Other benefits they mention are: greater focus on the long-term business sustainability, more integrated thinking and an improvement to the reporting. KPMG (2016) in collaboration with the IIRC argue that companies should implement an IR, because it will lead to an improved link between performance indicators and value creation, a clearer view of the implementation of the strategy by management and a better insight into how the firm is performing. Deloitte (2015) also emphasizes the benefit of using more integrated thinking. Deloitte (2015) focuses on the benefits mentioned by Black Sun Plc. and the IIRC (2014) in their survey of the participants joining the pilot program. Black Sun Plc. argues that there are multiple benefits to implement IR: (1) positive benefits regarding investors, (2) improved data quality, (3) common understanding of the value creation process, (4) enhanced quality of

communication, (5) looking at reporting through a new lens which also emphasizes the importance of non-financial information. The IIRC (2011) argues that implementing IR will prevent publishing an overload of information, which will make it easier for stakeholders to find the relevant information.

Eccles and Krzus (2010) argue that IR has the potential to change the way companies operate and investors think. According to Eccles and Krzus (2010) IR will make a

commitment to both CSR and a sustainable society. The value of implementing IR for the company, according to Eccles and Krzus (2010), is related to the following benefits:

- Greater clarity about relationships and commitments: if the company gets a better view of the relationship between financial and non-financial performance it can indicate where the opportunities and risks are and use this to improve its capability to implement sustainable strategies

- Better decisions: Developing the link between financial and non-financial outcomes requires a high level of collaboration between functions and business units. Since every function gets a better idea of their role in the company this will lead to better decisions.

- Deeper engagement with all stakeholders: If companies obtain more internal

collaboration they will also seek more external collaboration to better understand the expectations from its stakeholders.

- Lower reputational risk: Implementing IR provides a better view of how value is created, this will give the company a better insight on how their strategic choices impact society. Deeper engagement with the stakeholders will lead to a company’s strategy being more in line with the society’s needs.

Mervelskeper and Streit (2015) also argue that publishing information on ESG using an IR does not lead to different results then when publishing information on ESG in a separate report. Flower (2015) argues, using four objectives in the proposal of 2011 that did not support the original objectives of IR, that IR will not have a big impact on corporate reporting. The four objectives Flower mentions are: (1) Firms are not required to use IR as its primary report, (2) The IR framework does not cover sustainability, it is only mentioned once in the framework, (3) The IR framework focuses only on shareholders instead of all

stakeholders, (4) There are few obligations on the firm when implementing IR, for example, firms are not obliged to use the IIRC’s capitals and firms can choose to not include

information if it harms their competitive advantage.

In response to the arguments proposed by Flower, Adams (2015) argues that IR has already an impact on corporate reporting. Flower’s argument of IR not being used as a primary report is not valid according to Adams (2015). Adams (2015) argues that the long-term view of the IIRC might be that IR will be the primary report, however from the consultation drafts this is not entirely clear yet. Also according to Adams (2015) the IR framework does cover sustainability, however it has never been the main purpose of IR. Since sustainability is not a core mission of IR it is not mentioned in the Discussion paper of the IIRC in 2011. Furthermore, Adams (2015) supports the third and fourth argument

presented by Flower.

2.2. Compensation

In the following paragraph the function of remuneration and the link to IR will be explained.

2.2.1. The importance of compensation structure

Remuneration can be used to align the goals of CEOs and stakeholders. CEOs will act in their self-interest if they are not properly motivated (Jensen and Meckling, 1976; Walsh and

Seward, 1990). The information asymmetry between the principal and the agent is called the agency problem. The agency problem can be reduced in two ways; monitoring the executives or motivating them in a certain way (Healy and Palepu, 2001). However, monitoring is a very expensive way to reduce information asymmetry. Tosi et al. (1997) therefore argue, using a laboratory design research model, that equity holders find that the alignment of incentives of CEOs and stakeholders is more efficient than monitoring, when reducing information

asymmetry.

Compensation consists of different parts: base salary, cash compensation, long-term incentives and stock options (Murphy, 1999). In order to align the incentives of the CEO and

the stakeholders, the structure of the compensation plays an important role (Jensen and Murphy, 1990). To make sure that the CEOs are motivated to obtain the goals selected by the stakeholders, instead of serving their own interests, the compensation can be structured in different ways. Compensation generally consists of a base salary and an incentive part (Bushman et al., 1996). Base salary shields executives from being affected by events outside their control (Gray and Cannella, 1997). Mehran (1995) argues that the incentive part of the compensation can motivate agents to take more risk, potentially creating more value for stakeholders. Mehran (1995) also argues that firm performance is positively related to

percentage of the total compensation being rewarded as incentive based compensation. When motivating employees the structure of the compensation is more important than the amount of compensation (Mehran, 1995). Murphy (1999) argues that there are two types of aspects to link pay to performance; explicit and implicit aspects. The explicit aspects are: bonuses, restricted stock and the option awards. The implicit aspects are: target bonuses, options and restricted stock grant sizes. According to the agency theory a performance measure will only be included in the contract if it provides additional information in assessing the unobservable effort of the agent (Holmström, 1979).

Rewarding CEOs with equity-based compensation is seen as the most powerful tool to increase the alignment of the goals of the principal and the agent (Jensen and Murphy, 1990; Firth et al., 1999; Matolcsy and Wright, 2011). When CEOs own shares they will

immediately profit from increasing share prices. It will also create a more long-term vision, because the more the stock price increases, the more profit the CEO will make when they sell the shares. There are different types of stock options, for instance the options have to be held for a certain time before being sold or they cannot be sold at all (Ittner et al., 2003). So equity-based compensation can be classified as a long-term incentive. The potential negative effect of equity-based compensation is that managers can become risk averse in order to protect the value of their equity. This could eventually result in equity-based compensation being more expensive than cash compensation (Lambert et al., 1991).

Cash compensation creates short-term incentives; CEOs will try to make the target to obtain their cash bonus. However cash compensation does not create incentives to maximize the company’s value (Jensen and Murphy, 1990). The threat with cash bonuses is that CEOs will start working towards obtaining the short-term target. However this might not be serving the best interest in the long run. Healy (1985) argues that bonus contracts given CEOs

incentives to manipulate accruals and accounting procedures to create the maximum cash bonus. Healy (1985) concludes that the manipulation is mostly present at the upper or lower bound of the bonus to increase the net income. Holthausen et al. (1995) also argue that CEOs

that are at the upper bound of their bonus contracts manipulate earnings downwards to increase the present value of the bonuses. So cash compensation could lead to manipulation of earnings by the agent instead of maximizing the value of the company.

2.2.2. The link between IR and Remuneration

2.2.2.1. The relationship between IR and incentive based compensation

As mentioned before, incentive based compensation can be used to align the incentives of employees with the incentives of stakeholders. The payout of incentive-based compensation is based on a number of different firm performance measures (Bushman et al., 1996). One of the elements mentioned in the IR framework, which has been published in December 2013, is that companies need to show how the performance indicators and remuneration are linked to value creation in the short, medium and long-term. The purpose of IR is to cover the

complete value of a company, from both financial and non-financial perspectives. IR emphasizes not only short-term, but also medium and long-term value creation of a firm. Firms also need to include an explanation on how the performance indicators are linked to the IIRC’s six capitals formulated in the framework (IIRC, 2013). IR requires companies to provide information on the link between incentives, compensation and short and long-term value creation. This entails that implementing IR will provide an increase in the information on incentive-based compensation.

PwC (2015) published a report containing information on how to implement IR. In this report a five-stage model is used that includes aligning ones internal processes to ones strategy. In stage 3 (See Appendix 1 for the other stages) PwC discusses that the culture of the firm has to change to a culture where its purpose is creating value for the stakeholders. According to PwC (2015) to achieve the change in culture, employees need to be motivated to ensure that they put effort towards supporting this strategy. Organizations can change the remuneration structure to align the new strategic incentives of the company with the

incentives of the employees (Black Sun PLC and IIRC, 2014; PwC, 2015). The new

information the IR provides will lead to a decrease in information asymmetry which in turn will reduce the agency costs. Implementing IR will provide an opportunity to use the obtained knowledge about how CEO compensation is linked to value creation, to align the remuneration structure to the new strategic incentives.

Black Sun PLC and the IIRC (2014) also argue that organizations, which are in the early process of adopting IR, have to focus on changing their management systems to include non-financial information for non-financial performance indicators. This change in the

management system includes linking these non-financial performance indicators to executive compensation (Black Sun PLC and IIRC, 2014). When implementing the IR framework the relation between remuneration and the KPI’s has to be explained by the firms (ACCA, 2017; Black Sun PLC and IIRC, 2014; Deloitte, 2014; EY, 2014; IIRC, 2011; KPMG, 2012; PwC, 2015). When implementing IR firms need to obtain more information on how value is created in the short, medium and long run and link this to the KPI’s. This encompasses an increase in knowledge on the link between compensation and the firm’s performance indicators. The increase in information provides the firm with the opportunity to use more incentive based compensation and the increase in opportunity will lead to firms using more incentive based compensation.

According to PwC (2016), the relationship between compensation and value creation is one of the ways in which a firm’s commitment to integrated thinking can be measured. The long-term vision of the IIRC’s reporting framework (2013) is that firms implement integrated thinking within their business practices. The IIRC argues that using an IR facilitates

integrated thinking and they want IR to become the reporting norm (IIRC, 2013). The South Africa Institute of Chartered Accountants (SAICA) published an exploratory survey in 2015, which describes the key drivers to integrated thinking according to the respondents.

According to the respondents, one of the drivers to obtain integrated thinking is the

remuneration strategy. To obtain integrated thinking firms have to actively considerate the interconnection between the firm’s capitals. By using the knowledge obtained by

implementing an IR, on the link between remuneration and the performance indicators, the remuneration strategy can be changed to include more incentive based compensation. This will increase the link between compensation and both the financial and non-financial

performance indicators. So, by implementing IR, thereby changing the remuneration strategy, firms can achieve more integrated thinking (IIRC, 2011; SAICA, 2015).

Park et al. (2000) argue that increased disclosure will make shareholders more aware of how CEOs are compensated. This will increase pressure from shareholders to provide information on how compensation is linked to performance. To satisfy the investors boards will have to provide evidence that the CEO compensation is linked to firm performance. According to Park et al. (2000) this will lead to a shift from salary to incentive based compensation. Implementing IR will increase the disclosure of information to the

stakeholders, therefore the stakeholders will increase the demand for evidence on the link between CEO compensation and firm performance. Therefore, implementing IR will increase incentive based compensation.

2.2.2.2. The relationship between IR and long-term incentive based compensation

Incentive based compensation consists of a short-term and a long-term incentive plan. This section will focus on the effect of IR on long-term incentive based compensation. Some examples of long-term incentive based compensation are: shares, options and restricted stock awards (Ittner et al., 2003).

Khoroshilov and Narayanan (2008) researched the structure of incentive based compensation. They argue that when the market can fully observe the managers’ actions that compensation will be based on short-term incentives and not on long-term incentives. The firm’s profit reflects the employees’ ability and effort. The employees’ ability is related to the reputation of the employee, it reflects how the performance of the employee is evaluated by the market. Effort is needed to formulate and implement strategies (Khoroshilov and

Narayanan, 2008). According to Khoroshilov and Narayanan (2008) the managers’ ability and effort cannot be fully observed by shareholders, because the risk that the manager will manipulate earnings to increase the perception of their ability is high. Therefore, according to Khoroshilov and Narayanan (2008) if there is no full transparency, compensation will be based on long-term incentives to make sure that manager’s act in the shareholders best interest. If there is complete information transparency and the managers’ ability can be observed, compensation will be more based on short-term incentives (Khorshilov and Narayanan, 2008). Implementing IR increases the knowledge on how value is created, this increase in knowledge can be used in performance measures. When implementing IR there is an increase in information on the manager’s ability and effort. Firms that implement IR have to indicate how compensation is linked to value creation. The increase in transparency of the current actions will decrease the need to motivate CEOs for the long-term, since there is a lower risk of CEOs acting in their own interest. Analysing the paper of Khorshilov and Narayanan (2008) shows that implementing IR will lead to more short-term incentive based compensation. This theory is however based on transparency of information and does not take into account the types of information that is disclosed. With IR there is more

transparency, but the strategy of the firm also change to being based more on the medium and long-term.

FCLT Global (Focusing Capital on the Long Term), which was founded by McKinsey & Company and others, is an organization that focuses on motivating businesses to adopt long-term behaviours in their business and investment decision-making. Current reporting is more backward looking, since it is based on historical data (Ball et al., 2012; Khoroshilov and Narayanan 2008). To illustrate how FCLT Global thinks companies can focus more on the long-term, 10 elements were published that companies should include when they formulate a

long-term strategy. One of these elements is firms should make sure that the executives will focus more on the long-term. In order to align the new strategy with the motives of the executives their compensation will have to include more long-term incentive based compensation (Bebchuk and Fried, 2010; Healy, 1985). In response to the publication of FCLT Global, the IIRC (2017) indicates how the 10 elements published by FCLT Global are linked to the elements in the IR framework. According to the IIRC (2017) implementing IR will ensure a firms focus on long-term value creation. According to the IIRC (2017), linking the executive and director compensation to the long-term value creation and the strategic goals is achieved by three standards included in the IR framework. The standard that is relevant in this context shows that firms have to explain how the executive compensation is linked to medium- and long-term value creation. Firms that have studied the relationship between compensation and medium- and long-term value creation will have more knowledge on this relationship. The increase in knowledge will provide the firm with the opportunity to use this knowledge to focus more on long-term value creation. Motivating employees to focus more on long-term value creation, thereby aligning their incentives with the incentives of the firm, is obtained by focusing more of the CEO compensation on long-term incentives (Bebchuk and Fried, 2010; Healy, 1985).

Falck and Heblich (2007) argue, based on a literature study, that when firms want to focus more on CSR that compensation should be based on long-term incentives. CSR is a long-term goal, it is an investment in the company’s future (Falck and Heblich, 2007). According to Falck and Heblich (2007), if firms implement the CSR strategy well, it can enhance the firm’s reputation. When a firm has a separation of ownership and control, the firm should use the incentive structure to motivate employees to focus on long-term goals. If CEOs receive term incentive based compensation, employees will focus on the short-run to maximize their compensation. This will not profit the long-term goal of managing the stakeholders’ needs of focusing more on CSR effectively (Falck and Heblich, 2007).

Implementing IR will incorporate non-financial information on, for example, sustainability in one report (PwC, 2015). Therefore, when implementing IR, the CEO’s compensation has to be more focused on long-term goals. According to Falck and Heblich (2007), increasing long-term incentive based compensation will motivate the CEO to focus more on long-term goals.

2.2.2.3. The relationship between IR and total compensation

IR will integrate all different reports on non-financial and financial information in one report. Implementing IR is however not mandatory at the moment, so having an IR can be

seen as a form of voluntary disclosure. Sustainability information can be seen as a component of IR (PwC, 2015). Some studies have concluded that there is a relationship between CSR and compensation (Cai et al., 2011; Callen and Thomas, 2010; Mahoney and Thorne, 2005; Rekker et al., 2014; Riahi‐Belkaoui, 1992).

Analysing the papers that have been published regarding the relationship between sustainability reporting, CSR and compensation generates conclusive results. Callan and Thomas (2010) argue, using data on CSR from firms in the S&P500, that CSR is one of the significant determinants of CEO compensation. The regression equations used to obtain this result include multiple measures of firm performance and other firm characteristics, for example: gender, age, research and development expenditures and net sales (Callen and Thomas, 2010). Cai et al. (2011) argue, using an empirical study based on US companies, that higher levels of CSR are related to lower total compensation. Firms that behave socially responsible have lower compensation than firms which do not behave socially responsible, to mitigate potential conflicts with stakeholders (Cai et al., 2011). Secondly, Cai et al. (2011) argue that due to the controversy of extensive compensation, it is desirable for CEOs with high social and ethical standards to have a modest compensation. The regression analysis of Cai et al. (2011) includes firm and governance characteristics. Rekker et al. (2014) also argue that high CSR performance will lead to a lower compensation.

However, there are some differences between sustainability reporting and IR (ACCA, IIRC and NBA, 2013). The key differences between these two forms of reporting are the audience and the way the information is published. Where the sustainability report focuses on a large group of stakeholders, IR focuses on stakeholder that provide financial capital. A sustainability report focuses on providing information on environment, society and the economy. IR focuses more on the effect that the capitals have on the value creation.

Therefore the key difference is that IR is more focused on connectivity between the capitals and an IR provides more specific information. The differences between sustainability

reporting and IR indicate that; the implementation of IR could have a different effect on total CEO compensation than implementing a CSR report.

Park et al. (2001) argue, using an empirical based study based on Canadian firms, that an increase in disclosure will increase total CEO compensation. Park et al. (2001) study the effect of a mandatory increase in disclosure of executive compensation. While IR is not mandatory, the implementation will also result in an increase in the disclosure of CEO compensation. Park et al. (2001) argue that total CEO compensation will increase due to the fact that increased disclosure will increase the price competition for talented CEOs (Park et

al., 2001). Therefore, implementing IR will also lead to an increase in total CEO compensation.

The previous paragraphs outline the relationship between IR and CEO compensation for three levels of compensation. This leads to the following hypothesis:

“H1: Firms that have implemented an IR use different types of compensation”

H1a: Firms that have implemented an IR use more incentive based compensation in their compensation structure;

H1b: Firms that have implemented an IR will focus more of their compensation on long-term incentives;

H1c: Firms that have implemented an IR will have higher total compensation.

2.2.2.4. Implementing IR, CEO compensation and firm performance

The relationship between CEO compensation and firm performance has been extensively discussed in previous literature (Core et al., 1999; Hall and Liebman, 1998; Jensen and Murphy, 1990b; Mehran, 1995; Tosi et al., 2000; Zhou, 2000). Jensen and Murphy (1990b) argue that there is a significant relation between pay and performance, for every $1000 increase in wealth the CEO wealth increases by $3.25. Hall and Liebman (1998) argue, using a dataset of CEO in the largest public traded US companies, that the relationship between pay and performance can almost fully be described by the stock value and stock option value. Zhou (2000) then concludes, using data from Canadian companies, that there is a positive relationship between firm performance and firm size.

Previous literature is inconclusive about the effect of implementing IR on firm performance. Some research indicates there is a positive effect and other studies don’t find a relationship at all. Some Implementing IR will improve the relationship between overall performance and executive compensation (Deloitte, 2015; EY, 2014b). According to Black Sun Plc and the IIRC (2014) implementing IR will give firms a new perspective on their performance. This new information can help firms integrate performance on non-financials and use the new information to make better decisions (EY, 2014b). Churet and Eccles (2014) conclude that there is no significant relationship between firm performance and having an IR. The study of Churet an Eccles (2014) is based on data for the years 2011-2012, so before the IR framework was published. Churet and Eccles (2014) suggest that an explanation for the results is that there was not a lagged time variable included in the model, it takes time before ESG turns has an effect on financial performance. The study of McKinsey & Company

(2017) argues that focusing on long-term incentives will increase firm performance. The research focuses on the period of 2001 until 2014 and they conclude that firms with a long-term view on average have 47% higher revenue than their peers (McKinsey & Company, 2017). Implementing IR will make firms focus more on the medium and long-term (IIRC, 2013). So when firms implement IR, the focus will be shifted towards a more long-term perspective and this will lead to a higher firm performance.

This leads to the following hypotheses:

“H2: Firm performance positively mediates the effect of an IR on CEO compensation”

H2a: Firm performance positively mediates the effect of an IR on incentive based compensation;

H2b: Firm performance positively mediates the effect of an IR on long-term incentive based compensation;

H2c: Firm performance positively mediates the effect of an IR on total compensation.

2.2.2.5. Implementing IR, CEO compensation and firm size

Firm size has been proven to be an important indicator of the amount of CEO

compensation (Baker et al., 1988; Core et al., 1999; Firth et al, 1999; Murphy, 1985; Tosi et al., 2000; Zhou, 2000). Tosi et al. (2000) argue firm size explains more of the variance of CEO compensation than firm performance. Firm size explains more than 40%, while firm performance only explains 5% of the variance of CEO compensation (Tosi et al. (2000). Baker et al. (1988) argue, based on a literature review, that a 10% increase in firm size leads to a 3% increase in CEO compensation. According to Baker et al. (1988) this can be

explained by the fact that larger firms have better qualified and better paid CEOs.

Buzby (1975) and Gamerschlag et al. (2011) both state that providing disclosure to stakeholders is related to firm size. Buzby (1975) argues there are a few reasons why firm size would be positively related to disclosure: gathering and dispersing the information is very costly, small firms might not have the resources; large firms use more external financing so feel the need to disclose more information to obtain investor confidence; and small firms in comparison with large firms may fear for their competitive advantage when disclosing more information. Gamerschlag et al. (2011) conclude that large firms provide more

voluntary CSR disclosure. Large firms disclose more information to legitimize their actions (Gamerschlag et al., 2011). Marvelskeper and Streit (2015) argue that implementing an IR

requires a large amount of financial and human capital. Since implementing IR is voluntary, firms size will be a factor that influences the probability of firms implementing IR.

Buitendag et al. (2017) argue that bigger firms will publish an IR of higher quality than smaller firms. The study of Buitendag et al. (2017) focuses on the EY Excellence in Integrated Reporting Survey that is performed annually. The survey of EY includes the top 100 firms that are listed on the Johannesburg Stock Exchange. EY ranks the IR according to their quality and categorizes the reports as ‘excellent’, ‘good’, ‘average’ or ‘progress to be made’. Buitendag et al. (2017) use the data obtained from the survey to investigate whether the quality of the IR can be determined by entity’s characteristics. One of the entity’s characteristics that Buitendag et al. (2017) use in their research is firm size; they conclude that larger firms are more likely to produce higher quality IR than small firms. Large firms have more pressure from stakeholders to disclose information and size will have an effect on CSR disclosure, which can be seen as a part of IR (Buitendag et al., 2017). Larger firms will have higher quality IR, this lead to larger firms having more knowledge on how the firm creates value. Also since the IR framework is voluntary, the higher quality IR of large firms indicates that large firms might have implemented the IR more extensively than small firms. A bigger improvement in the knowledge of how value is created by the firm and a more extensive implementation will lead to a bigger change in the compensation structure. This indicates that firm size has an effect on the relationship between IR and CEO compensation.

This leads to the following hypothesis:

“H3: Size can positively moderate the effect of an IR on CEO compensation”

H3a: Size positively moderates the effect of an IR on incentive based compensation; H3b: Size positively moderates the effect of an IR on long-term incentive based compensation;

H3c: Size positively moderates the effect of an IR on total compensation.

3. Research methodology

This chapter presents a description of the data collection and the methodology. In the first section the sample and data collection are discussed. In the second section the variables, which are used in the regression, are described. The dependent variables, the independent variables and the additional variables are explained. The third section presents the research model used in the regressions and the research question that is answered by this study.

3.1. Sample and data collection

The data was collected from three different databases. The Global Reporting Initiative (GRI), which initiated the IIRC council together with the Prince’s Accounting for Sustainability Project, provided a list of companies that have implemented an IR. This dataset contains information on companies for the years 1999 until 2017 and specifies per year whether firms have implemented IR or other GRI sustainability guidance.

The data for the CEO compensation is obtained from Wharton Research Data Services (WRDS). The dataset Capital IQ-People Intelligence contains information on 4.5 million professionals and over 2.5 million people. Compensation for public and private companies is available for firms all over the world. Annual data for CEO compensation is retrieved from this database for the period 2011-2016.

The other variables included in this research are firm performance and firm size. The dataset for these variables is also obtained from WRDS. From the Institutional Brokers’ Estimate System (IBES) database, owned by Thomson Reuters, the annual return on assets (ROA), return on equity (ROE) and Sales (revenue) have been retrieved. The database IBES contains actual data for approximately 69,000 companies across a global level. ROA and ROE are used as an indicator for firm performance and the Sales is used as a measure for the size of a company. Annual data is again retrieved for the period 2011-2016.

3.2. Sample selection and sample period

For this study the years 2011-2016 are used. This study does not utilize data prior to 2011 as 2011 was the first year that the IIRC started the pilot cycle of the implementation of IR (GRI, 2012). The pilot program consists of three phases; the Dry run, Pilot Cycle 1 and Pilot Cycle 2 which will take three years in total (GRI, 2012). Most current data for the other variables (CEO compensation, firm performance and size) is used. The database contains data until the year 2015 and there are some observations for the fiscal year 2016 available. IR

implementation is used as the independent variable the dataset includes firms that have implemented IR and for firms that have not implemented IR. .

After merging the datasets of Capital IQ- people intelligence and IBES the number of observations was 49,093. The dataset of GRI, containing information on firms that have implemented IR, has to be merged to the datasets containing information on compensation and firm characteristics. Since there is no identical variable in both datasets that can be used to merge them a dummy variable for IR is created by hand. This creates a unique dataset for this research. Merging the databases Capital IQ- People Intelligence and IBES creates observations that could not be merged. After removing the missing observations the

remaining dataset contains 6,437 observations. While checking the data there were some duplicates left in the dataset. These duplicates are removed since they are copies of other observations. Finally, the data was checked for values that were economically impossible, for instance, compensation of 0 or a negative sales value. This gives a final sample of 5,531, which contains 5,143 firms without an IR and 388 with an IR. Table 1 provides an overview of the final sample.

Table 1: Sample

Number of observations

Initial Sample 49,093

Removing the missing values 42,656

Removing duplicates 703

Removing economically incorrect values 203

Final Sample 5,734

The final sample contains information on firms from 29 different countries. However, 83.2% of the total observations are from companies in the United States of America (USA). The observations are reported in the home currency of the countries. In order to obtain

meaningful results the variables used have to be reported in the same currency. Therefore, the observation for compensation and sales are transformed into US dollars (USD), since a large part of the data is from firms that are situated in the USA.

The publishing date of the report is used to obtain the currency rate. Currency rates were obtained from OANDA Corporation, a leading company that provides foreign exchange currency data and is used by a wide range of organizations and investors. The currency rates are obtained using the USD/home currency rate. Hereafter, the observations the

compensation and sales variables are multiplied by the currency rates to obtain the USD amount.

3.3. Description of the variables

This section introduces the variables that are used in the research models. First the dependent variable for CEO compensation are discussed, secondly the independent variable IR and at last the mediator variable firm performance and the moderator variable firm size.

3.3.1. Dependent variable

In order to answer the research question three types of compensation are used in the

regressions: total compensation, incentive based compensation and long-term incentive based compensation.

The total compensation is measured by:

Total compensation = Salary + Bonus + Other Annual + Restricted Stock Grants + LTIP (long-term incentive plan) Payouts + All Other + Value of Option Grants

This measure is included as a variable in the dataset obtained from Capital IQ- People

Intelligence. In the variable description list other annual is explained as income that could not be classified as salary or bonus. Examples of this are: perquisites, tax reimbursements and other personal benefits. All other consists of all other compensation paid to the CEO, for instance life insurance premiums and discounted share purchases.

The second measure of compensation, the incentive based compensation, is measured by all parts of the compensation that are based on the performance of the CEO. Therefore, the model excludes salary, other annual and all other, since these parts are not based on

incentives. This leads to the following equation:

Incentive based compensation = Bonus + Restricted Stock Grants + LTIP (long-term incentive plan) Payouts + Value of Option Grants

The third measure of compensation, long-term incentive based compensation, is calculated by adding Restricted Stock Grants, LTIP and Value of Option Grants. These are all measures that are based on the long term. Bonus is a short-term incentive since this rewarded each year. The long-term incentive based compensation is calculated by:

Long-term incentive based compensation = Restricted Stock Grants + LTIP (long-term incentive plan) Payouts + Value of Option Grants

In line with Core et al. (1999) in this model the natural logarithm of the CEO compensation is used. The data of the compensation is largely spread and is therefore not normally distributed. Using the natural logarithm of the compensation will transform the distribution of the data into a normal distribution (Cai et al., 2011; Duffhues and Kabir, 2008; Park et al., 2000; Zhou, 2000).

3.3.2. Independent variable

In this research IR is measured with a dummy variable. It is 1 if the company has an IR and 0 if the company does not have an IR. Using a dummy variable will indicate whether the implementation of an IR has had an effect on the structure of the CEO compensation. As

mentioned before the data for which firms have implemented an IR is obtained from GRI. The dummy variable is thereafter merged by hand to the compensation dataset.

3.3.3. Additional variables

In this study the following additional variables will be used to test whether the effect of the implementation of the IR on CEO compensation is due to other factors. The additional variables used are firm performance and firm size. Firm performance is a mediator in this research and firm size is a moderator.

3.3.3.1. Firm performance

Firm performance is in this research a positive mediator for the effect that IR has on CEO compensation. A mediator is a variable that explains the effect that the independent variable has on the dependent variable (Baron and Kenny, 1986). Baron and Kenny (1986) describe the mediator as a way in which the predictor influences the outcome variable. From the literature review it can be concluded that firms tend to have higher firm performance when implementing IR and firms with higher performance have higher compensation. So there is a direct effect of implementing IR and having higher compensation. However, there is also an indirect effect since part of the effect is due to the mediator firm performance.

Firm performance can be measured in a lot of different ways. For this study the return on assets (ROA) and return on equity (ROE) are used. These performance measures are widely used in previous research (Attaway, 2000; Callen and Thomas, 2011; Core et al., 1999; Duffhues and Kabir, 2008; Tosi et al., 2000; Zhou, 2000). The return on assets is calculated by dividing the net income by the average total assets. The return on equity is calculated by dividing the net income by the equity.

3.3.3.2. Firm size

Firm size is used as a positive moderator in this research. A moderator variable is a variable either quantitative or qualitative that has an effect on the strength or the direction of the

relation between the independent and the dependent variable (Baron and Kenny, 1986). The literature review describes the effect that firm size has on the amount of compensation being paid to the CEO and the possibility of implementing IR.

In the research model firm size is measured using the sales value, also called revenue (Murphy, 1999). To adapt the sales value to the normal distribution the research models will use the natural logarithm of the sales value, which is common in preexisting literature (Core et al., 1999; Duffhues and Kabir, 2008; Mehran, 1995; Murphy, 1985; Park et al. 2000; Tzioumis, 2008).

3.4. Research models and research question

In this section the research models and the research question are discussed. In the first

paragraph the research models that are used to obtain evidence to conclude the hypotheses are introduced. In the second paragraph the research question that will be answered in this study is introduced.

3.4.1. Research models

In order to answer the research question three models will be used to answer the three hypotheses. To answer the first hypothesis the effect of IR on CEO compensation will be tested using the three types of compensation. The second model will provide evidence on whether firm performance is a mediator in this research. The third model will provide

evidence on whether firm size is a moderator. The second and third models will be tested for two performance measures; ROA and ROE. The regressions are performed using the OLS regression method with robust standard errors.

The first research model is used to find evidence on whether implementing IR has an effect on the structure of CEO compensation. This model is a simple linear regression using the independent and the dependent variable.

Ln (CEO compensation) = α + β1 * IR

1) Ln (Incentive Based Compensation) = α + β1 * IR

2) Ln (Long-Term Incentive Based Compensation = α + β1 * IR 3) Ln (Total Compensation) = α + β1 * IR

To test whether firm performance positively mediates the effect of IR on CEO compensation a three-step approach is used (Baron and Kenny, 1986). The first step detects whether IR is related to CEO compensation. This step has been performed in research model 1. For the second step the regression provides evidence on whether having an IR has an effect on firm performance. So research model 2a provides evidence on whether the predictor is related to the mediator. In the third step a multiple regression model is used, so both the predictor and the mediator are included in the research model while using CEO compensation as the dependent variable. Research model 2b is used to obtain these results. To conclude whether firm performance is a positive mediator in this research the variables in the first and the second models need to be significant. If the effect of the predictor on the dependent variable is significant in the third model this implicates that there is partial mediation. If the effect of the predictor on the dependent variable is not significant in the third model this indicates that there is full mediation.

Research model 2:

Model 2a: Firm performance = α + β1 * IR

Model 2b: Ln (CEO compensation) = α + β1 * IR + β2 * firm performance

1) Ln (Incentive Based Compensation) = α + β1 * IR + β2 * firm performance 2) Ln (Long-Term Incentive Based Compensation = α + β1 * IR + β2 * firm

performance

3) Ln (Total Compensation) = α + β1 * IR + β2 * firm performance

To obtain evidence for the third hypothesis, whether size positively moderates the effect of IR on CEO compensation, an interaction term is added to the model (Baron and Kenny, 1986). So, the third research model adds the moderator variable and an interaction term between predictor and the moderator. In order to conclude that size is a positive moderator the interaction variable has to be significant.