Are CSR reports of controversial industries value relevant?

(1)

1

Are CSR reports of controversial industries

value relevant?

Master thesis

Julia Cozijnsen

6073824

February - June, 2014

University of Amsterdam

Faculty of Economics & Business

Accountancy & Control

Supervisor: B.G.D. O’Dwyer

(2)

2

3 3.5 Control variables 31 3.6 Research model 33 4. Results 35 4.1 Descriptive statistics 35 4.2 Correlation matrix 37 4.3 Regression Analysis 39 4.4 Testing of hypotheses 42 5. Discussion 43 6. Conclusion 46 Appendix I: References 48 Appendix II 52 Appendix III 53

(4)

4

Abstract

Currently, the European Commission is proposing mandatory social and environmental reporting for large firms to improve transparency of social and environmental issues. This proposal assumes value relevance of CSR reports. Prior literature also seems to suggest this. However, legitimacy theory and stakeholder theory seem to suggest that firms disclose social and environmental information merely for strategic purposes. This thesis examines how these findings relate to CSR reports of controversial industries. To examine this, two hypotheses are posed of which the first is: ‘CSR reports of controversial industries are not value relevant’ and the second is: ‘CSR reports of controversial industries that are externally assured do not have more value relevance than CSR reports of controversial industries that have not been

externally assured.’ An European sample of large listed firms is examined using a regression model based on the Ohlson model (1995). This model places the market value of equity as a function of the book value of equity, income before extraordinary items and a dummy variable for CSR reports and external assurance. In addition, the model controls for size and institutional differences between common law and code law countries. Results show that CSR reports of controversial industries are somewhat negatively value relevant, however external assurance on CSR reports does not provide additional value relevance. Overall, it can be concluded that the market somewhat negatively values CSR reports of controversial

industries. Implications of these findings justify the proposal of the European Commission, however they also suggest that there is a need for more strict rules regarding CSR reporting.

(5)

5

Samenvatting

Recentelijk heeft de Europese Commissie een voorstel gedaan tot het verplichten van de vastlegging van bedrijfsactiviteiten met betrekking tot maatschappelijk verantwoord ondernemen (MVO) voor grote bedrijven. Dit voorstel zou de transparantie van MVO

activiteiten moeten vergroten. Hierbij veronderstelt het voorstel dat zulk soort rapporten value relevant zijn, dat wil zeggen het veronderstelt dat de markt deze rapporten gebruikt bij

investeringsbeslissingen. De bestaande literatuur lijkt dit ook te suggereren. Legitimacy theorie en stakeholder theorie, beiden afgeleid van de overkoepelende political economy theorie, lijken echter te suggereren dat bedrijven voornamelijk rapporteren over MVO vanuit een strategisch oogpunt en dat er weinig werkelijk verantwoordelijkheidsgevoel achter zit. Deze scriptie onderzoekt hoe deze bevindingen relateren tot MVO rapporten van

controversiële industrieën. Om dit te onderzoeken zijn er twee hypothesen opgesteld, waarvan de eerste stelt dat MVO rapporten van controversiële industrieën niet value relevant zijn. De tweede hypothese stelt dat MVO rapporten van controversiële industrieën met externe assurance niet meer value relevant zijn dan MVO rapporten zonder externe assurance. Het onderzoek voert een regressieanalyse uit op een steekproef van beursgenoteerde Europese bedrijven in controversiële industrieën. Het regressiemodel is gebaseerd op het Ohlson model en beschrijft de marktwaarde als een functie van de boekwaarde, inkomen voor bijzondere items en overige relevante informatie. Deze overige relevante informatie wordt ingevuld door een dummy variabele voor MVO rapporten. Daarnaast wordt er nog een dummy variabele voor externe assurance aan het model toegevoegd. Tot slot controleert het model voor bedrijfsgrootte en institutionele factoren. Bevindingen laten zien dat MVO rapporten van bedrijven uit controversiële industrieën een enigszins negatieve value relevance hebben. Hiertoe wordt de eerste hypothese verworpen. De tweede hypothese kan niet worden

verworpen. In het geheel kan er geconcludeerd worden dat MVO rapporten van controversiële industrieën een redelijke mate van value relevance hebben. Deze bevinding rechtvaardigt het voorstel van de Europese Commissie om MVO rapportage te verplichten voor grote

bedrijven. Echter, het feit dat de value relevance slechts van redelijke mate is, duidt mogelijk ook aan dat er behoefte is aan striktere wetgeving met betrekking tot MVO rapportage.

(6)

6

1. Introduction

1.1 Background

Recently, the European Commission is considering to mandate social and environmental reporting for large firms. This proposed directive aims to improve transparency relating to social and environmental issues of large firms and requires large firms to report on relevant and material social and environmental information in their annual reports (European

Commission, 2013). This proposal assumes that reporting on CSR is value relevant. Generally, existing literature also suggests that the market values CSR reports and CSR performance. Moreover, the relation between firms’ market value and CSR reports and performance is mostly positive (Berthelot, Coulmont & Serret, 2012; Bird, Momenté & Reggiani, 2011; Cormier, Magnan & Morard, 1993; Kwon, 2013; Milne & Chan, 1999; Moneva & Cuellar, 2009; Wang, Qiu & Kong, 2011).

However, it is not clear whether the market also values CSR reports of firms

operating in controversial industries and whether this valuation is positive or negative. These industries are characterized by moral debates and are not seen as the most socially responsible businesses (Cai, Jo & Na, 2012). Our understanding of whether and how such controversial industries gain legitimacy through CSR engagement is limited (Cai et al., 2012; Jo & Na, 2012). This is because there is no direct research on this topic and literature on CSR

engagement by these industries is casuistic (Barraclough & Morrow, 2008; Du & Vieira Jr., 2012) or focuses on only one industry sector (Frynas, 2005; Palazzo & Richter, 2005; Yani-de-Soriano, Javed & Yousafzai, 2012).

CSR engagement by controversial industries appears to be a contested topic though. Opponents question whether it is possible for controversial industries to engage in CSR, as they argue that in essence their business is not ethical and therefore any attempt at being socially responsible will be seen as distrustful (Yani-de-Soriano et al., 2012; Palazzo & Richter, 2005; Hong & Kacperczyk, 2009). Proponents, on the other hand, argue that it is completely legitimate for these industries to engage in CSR to enhance their reputation and image, and improve their business (Jo & Na, 2012). As of now, whether controversial

industries can engage in CSR remains an open question (Cai et al., 2012). This thesis adds to this debate and the limited literature by examining whether CSR reports of controversial industries are valued by the market, by focusing on investors’ behavior as reflected in stock prices.

(7)

7

1.2 Research question

In order to examine this, the following research question is formulated: “Are CSR reports of controversial industries value relevant?”

1.3 Research method & results

Using a sample of listed European firms in the period 2005-2011, the incremental effect of a CSR report and external assurance of such a report on the market value of equity of a firm is measured. This is done by using the Ohlson model, which models the market value of equity as a function of the book value of equity, income before extraordinary items and other relevant, non-accounting information (Ohlson, 1995). Results show that adding a dummy variable for CSR reports to the regression model provides some weak incremental value relevance. However, adding a dummy variable for external assurance does not provide incremental value relevance. Furthermore, CSR reports of controversial industry firms have a negative effect on the market value of equity.

1.4 Contribution

The scientific relevance of this thesis relates to its contribution to the limited knowledge on the value relevance of CSR reports of controversial industries. In addition, this thesis can add to the debate on whether it is possible for controversial industries to engage in CSR by providing evidence on whether investors perceive CSR reports of controversial industries trustworthy and decision useful.

The societal relevance of this thesis relates to the proposal of the European

Commission to mandate CSR reporting for large firms, as this thesis will provide insights into the effectiveness of the proposal with regard to large listed firms operating in controversial industries. Research shows that the lack of strict mandatory guidelines for CSR reporting impacts the decision-making of users by making comparisons between firms difficult. Furthermore, it is argued that CSR reporting is not much of a systematic activity (Gray, Kouhy & Lavers, 1995). Evidence on the value relevance of CSR reports can have implications for stricter regulation with regard to CSR guidelines, which can improve transparency and comparability.

1.5 Structure of the thesis

The remainder of this thesis is structured as follows. Section 2 presents a literature review and the resulting development of hypotheses. Section 3 presents the research method of the thesis.

(8)

8

The findings of the paper will be reviewed in section 4. Section 5 presents the discussion. Finally, section 6 provides a conclusion.

(9)

9

2. Literature review and hypotheses development

2.1 What is CSR

This paragraph will explain what CSR is. To this end, the definition of CSR is explained, which is followed by a review of some of the most important CSR guidelines. Then, the proposal of the European Commission to mandate social and environmental reporting for large firms is discussed in some more depth. Finally, opposing views regarding mandatory or voluntary CSR reporting are presented.

2.1.1 Definition

There are several definitions of Corporate Social Responsibility (CSR) which all seem to come down to a responsibility of firms beyond making profit to take account for their impacts on society and the environment (Barraclough & Morrow, 2008). Reporting on CSR refers to the disclosure of a firm’s engagement in CSR and enables companies to present themselves to and to communicate with their stakeholders (Unerman, Bebbington & O’Dwyer, 2007).

CSR is defined by the European Commission as ‘the responsibility of enterprises for their impacts on society’. This holds that enterprises ‘should have in place a process to integrate social, environmental, ethical human rights and consumer concerns into their business operations and core strategy in close collaboration with their stakeholders’ in order to fully meet their social responsibilities (European Commission, 2011). The World Business Council for Sustainable Development (WBCSD) takes more of a business perspective with regard to CSR and perceives CSR as a business and risk-mitigating tool (Unerman et al., 2007). The WBCSD defines CSR as ‘the continuing commitment by business to behave ethically and contribute to economic development while improving the quality of life of the workforce and their families as well as of the local community and society at large’ (Moir, 2001). Cai, Jo and Pan (2012) suggest on the basis of several papers that CSR ‘generally refers to serving people, communities, society, and environment in ways that go above and beyond what is legally required of a firm.’ Furthermore, it is argued that CSR is

multidimensional, as it at least covers issues related to human rights, employee well-being, the environment and anti-corruption and bribery (European Commission, 2011).

Firms are increasingly reporting on their CSR engagement (Berthelot, Coulmont & Serret, 2012; Jo & Na, 2012). The European Commission (2011) states that the number of European organizations that publishes CSR reports in line with the Global Reporting Initiative (GRI) has increased from 270 in 2006 to more than 850 in 2011 according to numbers of the GRI. This increased voluntarily engagement to report on CSR also holds for sin and other

(10)

10

controversial industries, such as the tobacco industry and the oil industry (Cai et al., 2012; Yoon, Gürhan-Canli & Schwarz, 2006).

2.1.2 CSR guidelines

A number of different organizations support and promote the engagement in sustainability reporting and establish reporting guidelines, such as the Global Reporting Initiative (GRI), AccountAbility (AA) and the World Business Council for Sustainable Development

(WBCSD) (Unerman et al., 2007). These guidelines will be elaborated upon separately below. The Global Reporting Initiative (GRI) is an independent institution that aims to make sustainability reporting a common practice for organizations. To this end, GRI has established a Sustainability Reporting Framework, which increases transparency and accountability of organizations and ultimately will result in increased public trust in organizations. This framework provides organizations with guidance on methods of measuring and reporting sustainability issues in the form of guidelines (“What is GRI?”). These guidelines are

generated through a global multi-stakeholder process (“Reporting Framework Overview”) and have their origin in several reporting principles, such as transparency, inclusiveness,

auditability, completeness, materiality, relevance, comparability, and timeliness. Minimum requirements for the content of a sustainability report include vision and strategy,

organizational profile, governance structure and management systems, GRI content index and performance indicators (Unerman et al., 2007). The most recent guidelines are the G4

Guidelines and these are said to have increased user-friendliness and accessibility over the previous guidelines (“G4”). Currently, the Sustainability Reporting framework of the GRI is the most widely used framework for sustainability reporting (Nikolaeva & Bicho, 2011).

Another set of guidelines for reporting on CSR is AccountAbility. AccountAbility developed the AA1000 series of standards to increase accountability, sustainability and performance on different aspects, such as measurement of key performance indicators, quality management, external stakeholder engagement, risk management, governance and relations with regulators. The standards are principles-based and are, like the GRI guidelines, generated through a multi-stakeholder consultation process to enable organizations to be more

accountable, responsible and sustainable (“The AA1000 Standards”). The core principles upon which the standards are based are inclusivity, materiality and responsiveness. Inclusivity stands for the incorporation of the ideas and concerns of stakeholders in the decision making. Materiality means that only issues that are of importance should be included into decision making. Furthermore, responsiveness stands for the need of organizations to be transparent

(11)

11

about their decision making and actions (“About us”). It is argued that these AA1000 series of standards improve an organization’s long-term viability (Unerman et al., 2007). The AA1000 standards differ from the GRI guidelines in that the GRI guidelines are more focused on the content of sustainability reporting, whereas the AA1000 standards are more process standards and thus do not give much guidance on what should be reported but merely state which processes ensure transparent and reliable reporting (Unerman et al., 2007).

A third standard on sustainability reporting is developed by the World Business Council for Sustainable Development (WBCSD). The WBCSD provides guidance for organization on several aspects, such as sustainable livelihoods, energy and climate and accountability and reporting. To encourage sustainability reporting by organizations, the WBCSD published a document that provides guidance to organizations for reporting on sustainability reporting and that emphasizes why sustainability reporting is of value.

Specifically, this document assists organizations by stating how stakeholder groups should be addressed in a sustainability report and by developing a framework for sustainability

reporting. Furthermore, the document gives some specific guidance on the defining of reporting objectives, the planning, constructing and distributing of the report and the collecting and analyzing of feedback. Overall, the WBCSD takes more of a business case perspective with regard to sustainability reporting as compared to the GRI guidelines and the AA1000 standards (Unerman et al., 2007).

2.1.3 European Commission’s proposal

As of now, CSR reporting is voluntary in most of the countries in the European Union. However, the European Commission is recently considering a proposal for a directive to mandate social and environmental reporting for large firms. The proposal aims to improve transparency relating to social and environmental issues of large firms and requires firms of more than 500 employees to report on relevant and material social and environmental information in their annual reports. This holds that firms are required to report on policies, risk and results with regard to social and environmental issues, such as employee well-being, human rights, anti-corruption and the diversity of the board of directors. Firms with less than 500 employees are not required to stick to this new regulation as for these firms the

administrative costs would likely outweigh the benefits. Firms retain a lot of flexibility as they can decide themselves which social and environmental issues are material and should be disclosed. In addition, the manner in which firms want to disclose social and environmental

(12)

12

issues is also left at their own discretion and both national and international guidelines are approved (European Commission, 2013).

This proposal of the European Commission is related to the in October 2011 published report on the Commission’s Strategy for Corporate Social Responsibility for 2011 to 2014. This report argues that CSR is not only in the best interests of society, but also in the best interests of organizations themselves. Furthermore, it is stated in this report that although CSR efforts are increasing across European organizations, there is still a lot to win. In order to achieve this, the European Commission established several factors that aim to increase the impact of CSR policies. Examples of these factors are the need for a balanced

multi-stakeholder approach, the need to provide rewards for socially responsible businesses and the need to focus more on transparency relating to social and environmental issues. Another point on the new strategy agenda is the improvement of company disclosure of social and

environmental information. Disclosure should come from increased engagement with stakeholders and should enable the identification of material risks related to sustainability issues. In addition, it should increase accountability and result in increased public trust. Furthermore, the European Commission argues that full disclosure of non-financial information contributes to better capital allocation by investors. Other points relate for example to the improvement of public trust in businesses, enhancement of the market reward for CSR and better alignment of European and global approaches to CSR. These agenda points will be reviewed halfway 2014 in addition to periodical monitoring activities by the European Commission with member states, organizations and stakeholders, in order to assess its effectiveness (European Commission, 2011).

2.1.4 Voluntary versus mandatory CSR reporting

As described in the previous paragraph, reporting on CSR is mostly voluntary in the European Union to date, but recently the European Commission is considering to mandate social and environmental reporting for large firms. Whether CSR reporting should be mandatory or voluntary appears to be a contested topic. Opinions are mixed and literature on the effects of mandatory CSR reporting is limited.

Proponents of mandatory reporting argue that without compulsory guidelines, CSR reporting can be misused and might not serve stakeholder needs (Adams, 2004). In addition, it is suggested that mandatory reporting will satisfy a certain level of uniformity across reports and will ensure a minimum requirement for those reports which voluntary reporting cannot give (Berthelot, Cormier & Magnan, 2003). Furthermore, it is argued that regulation will

(13)

13

improve disclosure and will result in more transparency, more accountability and eventually in improved sustainability performance (Unerman et al., 2007). Also, standards on CSR reporting can increase the relevance of CSR information for investors and other stakeholders. It is argued that current CSR reports are not comparable and presumable have been positively biased (Frost, 2007; Deegan & Rankin, 1997). These shortcomings of voluntary

environmental information are said to provide demand for mandatory CSR regulation (Berthelot et al., 2003).

Opponents of mandatory reporting however suggest that voluntary reporting might be a stronger and more credible signal for investors than mandatory reporting can provide. This is the result of the fact that publishing a CSR report is costly for a firm in terms of effort and financial resources. When CSR reporting would be mandatory, these costs are only made as the result of compliance and might lose their signal to the market. This might reduce the credibility of reports (Berthelot et al., 2012; Murray, Sinclair, Power & Gray, 2006). Furthermore, opponents argue that even when CSR reporting is mandatory, there is still a possibility of selective and subjective reporting (Unerman et al., 2007).

Few studies have examined the effects of mandatory CSR reporting as a result of the often missing compulsory reporting standards. Bebbington and Thy (1999) study the situation in Denmark, where large firms are required to report on a comply and explain basis regarding their CSR policies (Adelphi, 2011). They conclude that approximately half of the firms that comply, experiences financial benefits in excess of the costs of compliance and thus conclude in favor of mandatory CSR reporting. Another study enables to distinguish the value

relevance of environmental information between the situation before and after mandatory reporting for Spanish listed firms (Moneva & Cuellar, 2009). Moneva and Cuellar find that mandatory reporting has more value relevance than voluntary reporting as with voluntary reporting, firms selectively disclose only positive environmental information. Moneva and Cuellar conclude that environmental reporting standards ensure value relevance of

information and this can potentially mitigate the shortcomings of voluntary reporting. In addition, Hassel, Nilsson and Nyquist (2005) also find more value relevance as the result of an increase in CSR regulation. Specifically, they find a stronger relation between market value and environmental performance after an enforcement of CSR regulation.

Perspectives on whether CSR reporting should be mandatory or voluntary thus appear to differ widely. Furthermore, studies on the effects of mandatory environmental reporting are limited but in general seem to conclude that mandatory reporting is beneficial, as it increases the value relevance of information and provides financial benefits.

(14)

14

2.2 What are sin and other controversial industries

This thesis will examine the value relevance of CSR reports of sin and other controversial industries, which can be defined as industries ‘which are typically characterized by social taboos, moral debates, and political pressures.’ Furthermore, these industries ‘include sinful industries, such as tobacco, gambling, alcohol, and adult entertainment as well as industries involved with emerging environmental, social, or ethical issues, i.e. weapons, nuclear, oil, cement and biotech’ (Cai, Jo & Pan, 2012, p.468). Du and Vieira Jr. (2012) argue that

industries can be controversial because of the goods or services they provide or because of the way in which they achieve their organizational objectives. This thesis will focus on the first category.

2.3 Why do firms engage in CSR

It is important to know what motivates firms to engage in CSR and CSR reporting, as this might have implications for the value relevance of their CSR engagement and reports. Value relevance shows the extent to which the market, i.e. investors, values certain information (Scott, 2012). Three theories will be discussed that can explain firms’ voluntary engagement in CSR. These are political economy theory, legitimacy theory and stakeholder theory and are interrelated and might together be helpful in explaining the motivation of firms to engage in CSR. Especially, both stakeholder theory and legitimacy theory are derived from political economy theory and all three are systems-oriented theories, which take the view of a firm as part of a broader social system that impacts the organization and on which the firm has impact. Corporate disclosure is then perceived as a way to influence the external perceptions of society regarding the organization (Deegan, 2002; Gray, Owen & Adams, 1996).

2.3.1 Political economy theory

Political economy theory is derived from the concept of political economy, which is defined by Gray et al. (1996) as ‘the social, political and economic framework within which human life takes place’ (p.47). Political economy theory posits that within society there are power conflicts between various groups. As a result, society, politics and economics cannot be viewed apart from each other and economic issues can only be examined in relation to the political, social and institutional framework of the issue. It is therefore important to involve the political economy of the organization when studying organizations and organizational behavior. For example, the political economy can explain the information disclosure behavior

(15)

15

of a firm. Political economy theory perceives accounting reports as political, social and economic instruments which firms can use as legitimizing instruments to meet their own best interests. Specifically, firm’s engagement in and reporting on CSR is perceived as a tool for managing firms’ position in the social system.

Political economy theory falls apart into a classical approach and a bourgeois approach. The classical approach centers at conflicts within society and explains corporate disclosure as a tool for powerful individuals to achieve a certain position at the disadvantage of less powerful individuals. The bourgeois approach does not focus on structural conflicts in society, but instead takes the view of an unanimous society that forms the activities of

organizations. Legitimacy theory is argued to be derived from this approach (Unerman et al., 2007).

2.3.1.1 Legitimacy theory

As discussed above, legitimacy theory is derived from political economy theory and is a systems-based theory which posits that the firm is part of a broader social system upon which it impacts and by which it is impacted. This social system has specific expectations about the behavior and activities of the firm and these are implicitly laid down in the social contract. The social contract is a theoretical construct with intangible clauses which depend on a social license to operate, resulting from being perceived as legitimate by the social system.

Legitimacy is defined by Lindblom (1994, p.2) as ‘a condition or status which exists when an entity’s value system is congruent with the value system of the larger social system of which the entity is a part’. When these value systems differ or when a difference in value systems is perceived, there is a threat of a legitimacy gap. This can be the result of a change in expectations of society with regard to what acceptable and desirable firm behavior is, as was the case with tobacco companies (Palazzo & Richter, 2005) or it can be the result of some event that has occurred which has impacted the legitimacy of the organization or industry in which it operates. Moreover, it can occur because of previously unknown information

becoming known. Legitimacy thus depends on perceptions of society regarding organizational behavior and is a relative concept as the value system of a society changes over time and across cultures.

Legitimacy theory is based on the idea that there is no obvious right for organizations to resources nor to even exist. The existence of organizations is only granted to the extent that society perceives these organizations as legitimate. Legitimacy is necessary for the survival of firms. However, this can be influenced by the firm by using corporate disclosure to influence

(16)

16

the perceptions of society (Deegan, 2002). Particularly, legitimacy theory predicts an increase in social and environmental reporting by a firm or industry when there exists a legitimacy threat for that particular firm or industry. Voluntary CSR reports are thus seen as a strategy to gain, maintain or repair legitimacy and are driven by survival concerns as opposed to a feeling that it is important to provide accountability. Overall, legitimacy theory seems to suggest that corporate disclosure is the result of firms’ desire to legitimize several aspects of their firm, which can be especially valuable for controversial industries. However, this rationale for CSR reporting implies that CSR reporting does not follow from a sincere motivation and concern to report on CSR to be accountable to those who have the right to this information, but merely from a strategic motivation (Deegan, 2002). This notion has implications for this thesis, as for controversial industries legitimizing activities might be even more important. The question then arises whether the market will see through this strategic motivation and if so, whether the market values this strategic motivation negatively or positively.

2.3.1.2 Stakeholder theory

As discussed previously, stakeholder theory is also a systems-oriented theory and conceptualizes the organization as part of a broader social system. However, the main

difference between legitimacy theory and stakeholder theory is that the latter argues that there are different groups of stakeholders with unequal power which have different perceptions of organizations. Furthermore, stakeholder theory assumes that there are various social contracts with several stakeholder groups. Legitimacy theory, however, takes the view of society as a whole that impacts the firm and to which perceptions the firm should adhere in order to be seen as legitimate (Deegan, 2002).

Freeman (1984) defines a stakeholder as “any group or individual who can affect or is affected by the achievement of the firm’s objectives” (p.25) and argues that it is important for firms to address the needs of their stakeholders in order to achieve firm objectives. According to stakeholder theory, information is disclosed to signal to stakeholders that the firm is

addressing their concerns and conforms to their expectations. However, stakeholder theory argues that firms will respond to the demands of stakeholders that are necessary for the survival of the firm only and will neglect the demands of powerless stakeholders. Stakeholder pressures are often conflicting though and firms address these pressures by assessing the relative importance of them with respect to power, urgency and legitimacy (Mitchell, Agle & Wood, 1997). The survival of organizations is dependent upon the extent that organizations are effective in the managing of these conflicting stakeholder demands. To gain or maintain

(17)

17

the support of these particular powerful stakeholders, firms can use strategic disclosures, such as for example CSR reports. Stakeholder theory thus also assumes that the disclosure of (social and environmental) information might be the result of a certain strategy and might not be the result of a sincere felt responsibility, similar to the view of legitimacy theory which argues that argues a use of disclosure for legitimizing strategies (Deegan, 2002).

2.3.2 Conclusion

Political economy theory and derived from it, legitimacy theory and stakeholder theory, all seem to conclude that (social and environmental) information disclosure is often used by firms as a mere legitimizing device without much sincere concern for the underlying social and environmental issues. This notion might have implications for the value relevance of CSR reports, as the market potentially sees through this strategic use of CSR reports. Moreover, implications might be even larger for controversial industries, as these industries are not the most socially responsible and can be argued to be in bigger need of legitimacy.

2.4 How do investors use information

As this thesis aims to examine whether investors use CSR reports of controversial industries for their investment decisions, it is important to understand how investors use information for such decisions. To examine this, the decision usefulness theory and the efficient securities markets theory will be discussed in this paragraph. In addition, the definition of value relevance is discussed. On the basis of these theories, it is argued that if CSR reports of controversial industries are value relevant, investors will use these for their investment decisions and as a result this should be observable in market valuation of the firm.

2.4.1 Decision usefulness theory

The objective of financial statements is to provide information to investors to help them make informed investment decisions. The usefulness of financial statements for such decisions depends upon the relevance and reliability of the information in the financial statements. It is argued that ‘information is relevant if it is capable of making a difference in the decisions made by users’ (p.94) and that information is reliable when it is ‘a faithful representation of what it purports to represent’(p.96). Information can then be defined as ‘evidence that has the potential to affect an individual’s decision’ (Scott, 2012, p.76). In line with this, the decision usefulness theory holds that information is decision useful if it changes investors’ prior beliefs.

(18)

18

2.4.2 Efficient securities markets theory

The efficient securities markets theory explains the result of the investment behavior of all rational investors together in a market. The theory describes an efficient securities market in its semi-strong form as ‘one where the prices of securities traded on that market at all times fully reflect all information that is publicly known about those securities’ (Scott, 2012, p.110). The market becomes efficient when enough investors act in the following way. At first, investors use information to revise the prior probabilities they have and are thus continuously updating their subjective estimates of firm’s future performance. However, information is not free in non-ideal conditions, so investors must make a cost-benefit tradeoff to decide how much information to obtain. In addition, investors must quickly act upon this new

information, as the benefit of new information will otherwise be faded out by the activities of other investors (Scott, 2012).

2.4.3 What is value relevance

As mentioned in the previous paragraph, rational investors continuously revise their prior probabilities as new information leads to changes in beliefs. This continuous updating leads to buy and sell decisions and ultimately to security price changes. As a result, stock prices at all times reflect all publicly available information according to the efficient securities markets theory. However, rational investors only demand and use decision useful information for their investment decisions. According to decision usefulness theory, information is decision useful when it changes investors’ prior beliefs. As security prices at all times reflect all publicly available information, this means that a change in volume or price is only expected as a result of new, unexpected information. Only new and unexpected information will cause a change in beliefs and will result in buy and sell decisions. Thus, a volume or price change is a measure of the extent to which prior beliefs of investors change and is thus a measure of the

usefulness, i.e. the value relevance, of information. This volume or price change is larger, the greater the differences in prior beliefs are (Scott, 2012).

2.5 Literature review

This paragraph will review relevant literature related to the value relevance of CSR reports of controversial industries. It is important to assess what is already known in prior literature regarding the CSR engagement of controversial industries to find out whether CSR by controversial industries is valued by the market. In this paragraph, literature on CSR

(19)

19

engagement of controversial industries is discussed first in paragraph 2.5.1. This is followed by a review of studies on the value relevance of CSR, which is decomposed into studies on the value relevance of CSR reporting and studies on the value relevance of CSR performance activities in paragraph 2.5.2.1 and paragraph 2.5.2.2 respectively. Finally, in paragraph 2.5.3 literature on external assurance of CSR reports is presented. On the basis of all this literature, a conclusion is presented in paragraph 2.5.4 which will lead to the hypotheses development.

2.5.1 CSR in controversial industries

There are ongoing debates on whether or not controversial industries can engage in CSR (Cai et al., 2012; Jo & Na, 2012). Opponents argue that CSR engagement by controversial

industries is a paradox and merely is a window dressing strategy (Cai et al., 2012; Palazzo & Richter, 2005). As a result, the quality and sincerity of CSR engagement by controversial industries is highly doubted (Barraclough & Morrow, 2008; Frynas, 2005; Miller & Michelson, 2013; Palazzo & Richter, 2005). Proponents, on the other hand, argue that all firms and industries should be able to engage in CSR and CSR reporting to enhance firm reputation and become more socially responsible. Moreover, it is claimed that firms have the right to pursue any preferred strategy (Cai et al., 2012).

Barraclough and Morrow (2008) provide evidence of the perspective of opponents as they show in a case study of British American Tobacco Malaysia (BATM) that health advocates heavily critique BATM’s effort relating to CSR. This is the result of one of many contradictions between the firm’s business and its CSR engagement, of which one example is the firm’s engagement in youth preventing projects and its refusal to report on customer health and safety issues related to their products in their reports at the same time. In line with this, Palazzo and Richter (2005) also claim that CSR engagement by the tobacco industry seems impossible. Similar to Barraclough and Morrow, they show that NGOs heavily critique CSR engagement of the tobacco industry and claim that the tobacco industry merely uses CSR to hide the actual business as usual activities. In addition to the tobacco industry, Frynas (2005) poses that the effectiveness of the increasing engagement in CSR activities by oil companies is also questioned. A gap seems to exist between such aims of oil companies and their activities and the effect on the society. Frynas mentions that this is because the use of CSR to achieve business as usual objectives reduces the benefit of CSR for the society. In addition, Yani-de-Soriano, Javed and Yousafzai (2012) and Miller and Michelson (2013) show that although CSR activities by the gambling industry are increasing, they lack credibility and are therefore not effective.

(20)

20

In line with these previous discussed studies that conclude on controversial industries having a difficult time engaging in CSR, Hong and Kacperczyk (2009) find that stock prices are affected by social norms. Specifically, they find that publicly traded companies in the alcohol, tobacco and gaming industry have higher risk and expected return as a result of being held less by norm-constrained institutions, having lower coverage from analysts and being subject to greater litigation risk, in relation to comparable stocks not engaged in these

industries. Confirming the findings of Hong and Kacperczyk, El Ghoul, Guedhami, Kwok and Mishra (2011) find that firms that engage in CSR have a significant lower cost of equity capital. This effect appears to be stronger more recently, which can be interpreted as evidence for increased social responsible awareness of investors. An additional analysis shows that firms in the tobacco and the nuclear power industry have a significant higher cost of equity capital, indicating a higher perceived industry risk. This is effect is not found for other controversial industries however, such as alcohol, gambling and military.

Different findings come from the survey of Halpern and Snider (2011), which

suggests that the weapon industry can be seen as socially responsible. This survey shows that defense firm managers are significantly less concerned with economic issues and are on average significantly more concerned with combined non-economic issues, compared to ‘regular’ managers. Halpern and Snider interpret this as defense firm managers potentially being more concerned with social and environmental issues than the average ‘regular’ manager. Furthermore, studies that look at no specific industry sector also conclude on positive effects of CSR engagement by controversial industries. Cai, Jo and Pan (2012) find a positive effect of CSR engagement by controversial industries on firm value. This is

interpreted as supporting their value enhancement hypothesis, as opposed to the window dressing hypothesis, which holds that CSR results in increased transparency and firm value for controversial industries. In line with these findings, Jo and Na (2012) find that

controversial industries that engage in CSR have lower firm risk. Furthermore, this effect of CSR engagement on risk is more significant for controversial industries than for

non-controversial industries. This finding is interpreted as supporting their risk-reduction hypothesis, as opposed to the window-dressing hypothesis, which poses that CSR reduces firm risk potentially by providing ‘insurance-like protections, improving risk management, providing market appeal to customers, improving information transparency, or by making access to financial markets easier’ (p.452).

(21)

21

2.5.1.1 Conclusion

Overall, it can be concluded that the findings of studies that examine CSR engagement by controversial industries are mixed. Several studies that focus on one specific controversial industry show that these industries are not successful in their CSR engagement (Barraclough & Morrow, 2008; Frynas, 2005; Miller & Michelson, 2013; Palazzo & Richter, 2005; Yani-de-Soriano et al., 2012). Furthermore, it is found that stock prices are affected by social norms (Hong & Kacperczyk, 2009) and notwithstanding CSR engagement, the cost of equity capital of several controversial industries is higher compared to non-controversial industries (El Ghoul et al., 2011). However, Halpern and Snider (2011) suggest that the weapon industry can be socially responsible and studies that focus one no specific industry show that

controversial industries that engage in CSR have higher firm value and lower firm risk (Cai, Jo & Pan, 2012; Jo & Na, 2012).

2.5.2 The value relevance of CSR

This paragraph will discuss prior literature on the value relevance of CSR engagement. Reviewing this literature is important to find out whether, how and when the market values CSR. The literature will be subdivided into CSR reporting and CSR performance and activities, as both provide evidence on the value relevance of CSR and enable investors to form an opinion on the performance of firms (Dhaliwal, Li, Tsang & Yang, 2011).

2.5.2.1 CSR reporting

Berthelot, Coulmont and Serret (2012) find that investors value sustainability reports of Canadian listed companies as showed by a significant higher market value of firms that publish such a report compared to firms that do not. Cormier and Magnan (2007) also conclude that environmental reporting is value relevant. They argue that environmental disclosure provides investors insights in the firms’ operating and financial risks and in addition, enhances the credibility of its financial statements, resulting in a positive effect on the market to book premium regardless of the sign of the environmental information. However, their findings shows that the effect of environmental reporting on firms’ stock market valuation depends on reporting context and the aspect of environmental issues that is reported. In addition, Dhaliwal, Li, Tsang and Yang (2011) also document a positive effect of CSR engagement. Specifically, they find that firms that disclose CSR reports reduce their cost of equity capital and that having a high cost of equity capital in the previous year increases the likelihood of disclosing a CSR report in the current year. In addition, Dhaliwal et al. find a

(22)

22

significant lower cost of equity capital for firms that have superior CSR performance compared to their industry peers. In line with these findings, Holm and Rikhardsson (2008) find that investors consistently invest a larger amount in a firm that discloses environmental information compared to a firm that does not. Positive environmental information is

concluded to be decision useful for investors and it is concluded that environmental

information affects investment decision making. Milne and Chan (1999) however, find that narrative corporate social disclosures are only moderately useful for investment decisions as they do not show a direct effect on a company’s risk and return. Furthermore, Milne and Chan find no significant differences in mean amounts invested in firms with non-financial narrative disclosure and firms with non-financial narrative corporate social disclosure. These results indicate that narrative corporate social disclosures do not significantly influence the

investment behavior of investors. It is argued that social and environmental disclosures will be more decision useful when quantitatively presented. Confirming this finding, Moneva and Cuellar (2009) find that only financial environmental information has value relevance. Specifically, Moneva and Cuellar examine the value relevance financial and non-financial environmental disclosures. Their results show that financial environmental disclosures related to environmental costs have negative value relevance and that these disclosure have higher value relevance for more polluting industries. They also find an overall increase in value relevance after enactment of a compulsory environmental reporting regime. Overall, Moneva and Cuellar suggest that the market significantly values financial environmental disclosures of Spanish listed firms, but that this does not hold for non-financial information. Furthermore, Murray, Sinclair, Power and Gray (2006) do not find a direct relation between social and environmental disclosure and stock returns. Specifically, tests reliant on cross-sectional data suggest no relation between market performance and social and environmental disclosure. Longitudinal data however, does show an association between market performance and social and environmental disclosure, indicating that firms with high (low) market performance also have high (low) social and environmental disclosure rates. However, this association is not significant on a year basis and in addition, is not stable and attenuates over time.

In general, literature on the value relevance of CSR reports suggests that the market more or less values CSR reports and this valuation is mostly positive (Berthelot et al., 2012; Cormier & Magnan, 2007; Dhaliwal et al., 2007; Holm & Rikhardsson, 2008; Milne and Chan, 1999; Moneva & Cuellar, 2009; Murray et al., 2006).

(23)

23

2.5.2.3 CSR performance and activities

Hassel, Nilsson and Nyquist (2005) find that environmental performance ratings have

negative value relevance, suggesting that investors perceive environmental costs to be a mere cost to the firm without benefits. Hassel et al. argue that this can be the result of investors’ perception of environmental performance as window dressing strategy or because of the short-term orientation of the market. In addition, Hassel et al. find stronger negative value relevance as the result of a strengthening in environmental disclosure regulation. Gupta and Goldar (2005) also find that environmental performance ratings are value relevant. They find that the market respectively penalizes and rewards firms if they have environmentally

unfriendly or friendly performances. In addition, Konar and Cohen (2001) also find that environmental performance has positive value relevance. Moreover, using voluntary donation expenditures as a proxy for environmental performance, Kwon (2013) finds that

environmental performance is value relevant. Measuring the market reaction to CSR donations by assessing future abnormal returns, it is shown that there is a positive and

significant relation between donation expenditures and future abnormal returns. Furthermore, Al-Tuijwari, Christensen and Hughes (2004) find a significant positive relation between environmental performance and economic performance, indicating that the market values and rewards good environmental performance. In addition, they find that firms that are under more public exposure and scrutiny have higher environmental performance than comparable firms. Decomposition of the sample into different industries shows no significant differences in environmental performance and environmental disclosure. In addition, Hughes (2000) shows that the market penalizes firms for polluting. Furthermore, the value relevance of nonfinancial pollution information increases as a result of increased environmental legislation. More evidence for the value relevance of environmental performance comes from Thomas (2001), who finds that there exists a relation between abnormal stock returns and the

engagement of firms into an environmental protocol. Specifically, results show that for more polluting firms, the engagement in an environmental protocol leads to less negative abnormal returns. Another study suggesting the value relevance of environmental performance is that of Gunthorpe (1997), who finds that the market penalizes firms for unethical behavior.

A lagged value relevance effect of bad environmental performance is found by Lorraine, Collison and Power (2004. Their results show that the market does not respond to good environmental performance. Bird, Momenté and Reggiani (2011) find evidence for differing value relevance across different countries and regions. Specifically, they find a

(24)

24

stronger positive relation between CSR ratings and market returns for European companies compared to companies in the United States, Australia and Japan. It is concluded that whether the market values CSR activities depends upon region and country. Cormier, Magnan and Morard (1993) also find evidence for a more moderate value relevance and find that firms with more positive pollution records are traded at a somewhat premium in the market. In line with this finding of a more moderate value relevance of CSR reports, Wang, Qiu and Kong (2011) find that value relevance of CSR performance is contingent upon an event and only up to a certain threshold. They find only a positive effect of CSR performance on investors’ trading behavior after a specific environmental event and this only holds for CSR scores above a particular threshold. Furthermore, the results show that there is a diminishing marginal benefit of CSR on investors’ behavior, which Wang et al. conclude to indicate that investors perceive high CSR performance to be negative for the financial performance of firms. Wang et al. argue that the finding of a contingency effect of a CSR event on investors’ perceptions of CSR performance can explain the contradicting findings of prior literature regarding the relation between CSR and financial performance.

Literature on the value relevance of CSR performance and activities overall indicates that the market values CSR performance and activities and this valuation is mostly positive (Al-Tuwaijri et al., 2004; Bird et al., 2011; Gunthorpe, 1997; Gupta & Goldar, 2005; Hughes, 2000; Konar & Cohen, 2001; Kwon, 2013; Lorraine et al., 2004; Thomas, 2001; Wang et al., 2011).

2.5.3 External assurance of CSR reports

The number of firms that have their sustainability reports externally assured is rising

(O’Dwyer & Owen, 2005; Perego & Kolk, 2012). External assurance of sustainability reports is argued to largely resemble the financial audit of the annual report (Park & Brorson, 2005). Furthermore, it is suggested that while the reporting on CSR matters increases transparency, the external verification of this reporting enhances credibility (Deloitte, 2010). Because of the quite recent development of the external assurance practice, guidelines remain limited

although they are increasing. Several bodies that occupy themselves with providing guidance on CSR reporting, such as the Global Reporting Initiative (GRI) and AccountAbility (AA) as mentioned in paragraph 2.1.2, are also involved in the guiding of the assurance process.

(25)

25

Different benefits are imputed to be the result of independent third party assurance. First and foremost, it is argued that third-party assurance of CSR reports can increase the credibility of the reports to stakeholders (Park & Brorson, 2005). By having a CSR report voluntarily assured, the firm sends a signal of trust to the market which can mitigate the credibility gap that exists in society. This signal is strengthens as the result of the assurance process being mostly voluntary, but costly for firms. Furthermore, the voluntary engagement in the assurance process leads to the assumption that the benefits of assurance probably outweigh the costs for firms (Simnett, Vanstraelen & Chua, 2009). Besides increased credibility, several other benefits of third-party assurance are identified by companies and assurance providers, such as assistance in the development of efficient internal reporting systems (Park & Brorson, 2005), reduced agency costs and greater confidence of users regarding the accuracy and validity of the information in the report (Simnett, Vanstraelen & Chua, 2009). Moreover, it is claimed that external assurance might be helpful in giving management insights in how they manage their social and environmental risks (O’Dwyer & Owen, 2005).

Overall, external assurance seems to result in improved CSR engagement by

increasing the credibility and accountability of the report (Dando & Swift, 2003; Manetti & Becatti, 2009). However, there are also critics that argue that the benefits of assurance practices are limited as it often involves too much management override. As a result, the external assurance process is not performed independently and transparency and credibility to stakeholders is reduced (O’Dwyer & Owen, 2005). Furthermore, the lack of clear standards guiding the assurance process is argued to result in limited accountability and comparability to stakeholders. In addition, there appear to be large inconsistencies in the actual

implementation of the assurance process and stakeholders appear often to be not consulted in the process (O’Dwyer & Owen, 2005; Perego & Kolk, 2012). Literature is thus mixed on the actual benefits and value relevance of the assurance process for stakeholders.

With regard to controversial industries, assurance might be especially important as these industries are not seen as the most socially responsible businesses and therefore it is likely that their CSR reports have credibility and trust problems (Miller & Michelson, 2013). In line with this, Simnett et al. (2009) find that firms that are in need of more credibility regarding their sustainability reports and want to enhance their corporate reputation have higher incidences of assurance. Especially, they find that the mining, production, utilities and financial industry, which are argued to have a larger social footprint and are more closely

(26)

26

monitored as a result of the nature of their business compared to other industries, have higher incidences of assurance regarding their sustainability reports.

Literature on the assurance process is mixed in terms of its benefits and value relevance for stakeholders. Several benefits are imputed to third party assurance, such as increased credibility and transparency (Park & Brorson, 2005). However, critics address numerous shortcomings of the assurance process which largely impair its credibility and transparency (O’Dwyer & Owen, 2005; Perego & Kolk, 2012). Regarding external assurance in

environmentally sensitive industries, it appears that there is greater demand for such external verification (Simnett et al., 2009). However, whether external assurance also is more value relevant in such industries remains unknown.

2.6 Hypotheses development

To provide an answer on the research question of this thesis: “Are CSR reports of controversial industries value relevant?” two hypotheses are stated.

2.6.1 Hypothesis 1

The findings of the literature reviews of CSR engagement by controversial industries and the value relevance of CSR reporting and performance appear to be incompatibility. The literature on CSR engagement by controversial industries is mixed, as most studies that focus on one specific industry show unsuccessful CSR engagement, while other literature that focuses on more than one industry shows beneficial effects of CSR. The literature on the value relevance of CSR reports and performance, however, shows overall a positive valuation of CSR

engagement. Due to this incompatibility and the lack of direct research on the value relevance of CSR engagement of controversial industries, it is not possible to propose a direction for the first hypothesis of this thesis and as a result, this hypothesis will be stated in a null-form:

H1: “CSR reports of controversial industries are not value relevant.”

2.6.2 Hypothesis 2

Literature on the assurance process provides mixed findings related to its benefits and value relevance for stakeholders. On the one hand, some studies suggest that independent third party assurance provides several benefits for stakeholders and management. On the other hand,

(27)

27

other literature seems to suggest that the assurance process is often not carried out properly and lacks clear standards, resulting in reduced credibility and transparency for stakeholders. Furthermore, literature suggests that there is a greater demand for external verification in environmentally sensitive industries. However, it is not clear whether external assurance also is more value relevant in these industries. Due to these mixed findings, the second hypothesis will also be stated in a null-form:

H2: “CSR reports of controversial industries that are externally assured do not have more value relevance than CSR reports of controversial industries that have not been externally assured.”

3. Research methodology

3.1 Research method

To answer the research question “Are CSR reports of controversial industries value relevant?” this thesis will use archival data. This data is then used to measure the market value of equity as a function of the book value of equity, income before extraordinary items and other

relevant, non-accounting information based on the Ohlson model (1995) following the studies of Trueman, Wong & Zhang (2000), Hassel et al. (2005) and Moneva and Cuellar (2009). The Ohlson model is useful in providing evidence on how accounting numbers and

non-accounting information are valued by the market and thus serves as an useful instrument to measure value relevance. Particularly, the model enables to link accounting numbers with valuation (Hassel et al., 2005; Moneva & Cuellar, 2009; Ohlson, 1995). Furthermore, the model can be easily used for international studies that cover data from more than one country as the model requires accounting numbers that are internationally accepted. Also, the model is argued to have reasonable explanatory power and can therefore be useful for policy

recommendations. Moreover, a benefit of the model is that it allows for several other

variables to be added to the ‘standard’ accounting numbers to find out whether they also can help explain market value of equity. Finally, the model lends itself for examining the

investment decisions of investors without the time consuming process of directly asking investors (Lo & Lys, 2000). These aspects of the Ohlson model make it highly useful for this thesis as it covers international data and is linked to a policy recommendation of the European Commission’s proposal to mandate social and environmental reporting for large listed firms in Europe.

(28)

28

The Ohlson model is based on several assumptions, of which the first is that the market value of equity is the present value of all future dividends. The second assumption is that a clean surplus relation holds, which means that all changes in equity other than

transactions with shareholders flow through the income statement. This holds that the book value changes by the amount of net income minus dividend payouts. The third and final assumption is that abnormal earnings, defined as the excess of the risk free rate of return times the current book value of equity over current earnings, follow a linear progress (Lo & Lys, 2000). These assumptions result in the following model:

𝑀𝑉_𝑡= 𝐵𝑉_𝑡+ 𝛼₁𝓍_𝑡𝛼+ 𝛼₂𝜈_𝑡 (1)

Where 𝑃_𝑡 stands for the market value of equity at time t, 𝐵𝑉_𝑡 stands for the book value of equity at time t, 𝓍𝑡𝛼 refers to abnormal earnings at time t and 𝜈𝑡 is other relevant,

non-accounting information at time t. Furthermore, 𝛼₁ = 𝜔

𝑅𝑓−𝜔> 0 and 𝛼2 =

𝑅𝑓

(𝑅𝑓−𝜔)(𝑅𝑓−𝑦)> 0,

where ω denotes the persistence of abnormal earnings, 𝑦 denotes the persistence of other relevant, non-accounting information and 𝑅𝑓 is the risk free rate of return (Ohlson, 1995).

Following Hassel et al. (2005), substituting the definition of abnormal earnings in equation (1) results in the following equation:

𝑀𝑉𝑡= 𝐵𝑉𝑡+ 𝛼1𝑁𝐼𝑡− 𝛼1𝑅𝑓 ∗ 𝐵𝑉𝑡−1+ 𝛼2𝜈𝑡 (2)

Where 𝑁𝐼_𝑡 denotes net income at time t. When using the clean surplus relation, the equation changes as follows:

𝑀𝑉_𝑡= 𝐵𝑉_𝑡−1+ 𝑁𝐼_𝑡− 𝐷_𝑡+ 𝛼₁𝑁𝐼_𝑡− 𝛼₁𝑅𝑓 ∗ 𝐵𝑉_𝑡−1+ 𝛼₂𝜈_𝑡 (3) Where 𝐷𝑡 refers to dividends at time t. Equation (3) can then be rewritten into:

𝑀𝑉_𝑡+ 𝐷_𝑡 = (1 + 𝛼₁)𝑁𝐼𝑡+ (1 − 𝛼1∗ 𝑅𝑓)𝐵𝑉𝑡−1+ 𝛼2𝜈𝑡 (4)

Finally, equation (4) equals:

𝑀𝑉_𝑡+ 𝐷_𝑡 = 𝛽₀+ 𝛽₁𝐵𝑉_𝑡−1+ 𝛽₂𝑁𝐼_𝑡+ 𝛽₃𝜈_𝑡+ 𝜀 (5)

Based on this final equation (5), Moneva and Cuellar (2009) and Trueman et al. (2000), the final empirical model that will be used in this thesis is:

𝑀𝑉𝐸𝑗𝑡 𝑇𝐴_𝑗𝑡−1= β0 + β1* 𝐵𝑉𝐸𝑗𝑡−1 𝑇𝐴_𝑗𝑡−1 + β2* 𝐼𝑛𝑐𝑜𝑚𝑒𝑗𝑡 𝑇𝐴_𝑗𝑡−1 + β3*LAW + β4*CSRjt + β5*ASSUjt + ε

In this empirical model, 𝑀𝑉𝐸𝑗𝑡 means the market value of equity of firm j at the end of year t,

(29)

29

income before extraordinary items at the end of year t for firm j. These variables are deflated by total assets of the previous year (𝑇𝐴_𝑗𝑡−1) to control for size, which will be further

explained in paragraph 3.5 Another control variable is 𝐿𝐴𝑊, which is an indicator variable which takes the value of one when the firm is a common law country and zero otherwise. These control variables are further explained in paragraph 3.5. The variable which is of most importance for this thesis is 𝐶𝑆𝑅, which is a dummy variable which equals one if firm j has published a CSR report in the year t and zero otherwise. Lastly, 𝐴𝑆𝑆𝑈 is an indicator variable that takes the value of one when a CSR report of firm j has been externally assured in year t and zero otherwise. These two variables will be further explained in paragraph 3.4.

The accounting variables needed for the empirical model are taken from WRDS - Compustat Global and are combined with Datastream where necessary. Market value of equity is taken directly from Datastream. Book value of equity is computed by subtracting total liabilities from total assets (Moneva & Cuellar, 2009). The calculation of the other variables will be discussed in the following paragraphs.

Value relevance of CSR reports will be inferred from changes in the market value of equity as the result of firms publishing a CSR report and/or having this report externally assured, as market value of equity is a function of CSR reports. Furthermore, value relevance of CSR reports will be inferred by comparing the relative change in the 𝑅2_{of the empirical}

model when adding the dummy variable for CSR reports to the model, which is called a relative association study (Holthausen & Watts, 2001).

3.2 Sample

The sample will consist of firms that are operating in the following sin and other controversial industries: tobacco, alcohol, gambling, oil, nuclear, cement, biotech and weapon industry, following the study of Cai, Jo & Pan (2012). The adult entertainment is sometimes also mentioned as controversial industry, but will be excluded from the sample in this thesis, following Hong and Kacperczyk (2005) and Jo and Na (2012). These studies show that companies in this industry are mostly not publicly traded companies and therefore, excluding them from the sample will not significantly change results.

The firms are identified using Standard Industrial Classification (SIC) codes and North American Industry Classification System (NAICS) codes, following Cai et al. (2012) and Hong and Kacperczyk (2005) from the Compustat Global database. The specific data that is requested from Compustat is the following: fiscal year-end, company name, total assets, total liabilities, income before extraordinary items, common shares outstanding and current ISO

(30)

30

country code headquarters. As the Compustat Global database does not allow market value of equity to be requested for firms, Datastream is also consulted for this variable.

The sample consists of listed European firms, as this thesis is partly motivated by the proposal of the European Commission to mandate social and environmental reporting for large firms. Due to practicality issues, it is chosen to restrict the sample to listed firms from France, Germany, The Netherlands, Switzerland and the United Kingdom (UK), following Adams, Hill & Roberts (1998). The restriction of the sample is firstly because of data availability reasons, as most large listed firms that operate in controversial industries are situated in these countries as can be seen from the Compustat database. In addition, as Adams et al. (1998) indicate, this restriction of the sample rules out several confounding factors between different countries in Europe that might also impact the value relevance of CSR reports.

The sample covers the time period 2005-2011, which is chosen because the European Commission mandated listed firms in the European Union to report in line with International Financial Reporting Standards (IFRS) as of January, 1 2005 (“IAS regulation”). This ensures that the sample chosen consists of firms with comparable financial reporting.

This results in an initial sample of 250 firms identified from Compustat. From this initial sample, 78 firms are excluded as the result of having a year-end other than on December (N=172). Furthermore, twelve firms with missing data are removed from the sample (N=160). When the additional data is requested from Datastream, another fifteen firms are removed from the sample as the result of missing data in Datastream (N=145). Finally, the sample is reduced even more during the research stage as SPSS detected several outliers with casewise diagnostics and two missing values. Following De Vocht (2004), the outliers that have very large standardized residuals, i.e. larger than five, are removed from the sample data. This results in a final sample of 143 firms and 804 firm-year observations. Panel A and B present the sample and industry composition.

Table 1 Sample and industry composition

Panel A: Sample derivation

Description N

Initial sample from Compustat 250

Removal of firms with year-end other than December -78

Removal of firms with missing data in Compustat -12

Removal of firms with missing data in Datastream -15

Sample before removal of outliers 145

Are CSR reports of controversial industries value relevant?