Corporate Social Responsibility and financial performance : in the European context

(1)

Corporate Social Responsibility

and financial performance

In the European context

19-6-2015 Final version University of Amsterdam Romée Kruiswijk 10254331 Word count: 16400

Supervisor: Dhr. prof. dr. V.S. Maas

MSc Accountancy & Control, variant Accountancy Amsterdam Business School

(2)

2

Statement of originality

This document is written by Student Romée Kruiswijk who declares to take full responsibility for the contents of this document.

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

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

(3)

3

Abstract

Purpose – The paper provides empirical evidence on the relationship between CSR and financial performance in the European context. Whereas the majority of the existing literature is focused on the United States this paper provides insights on the impact of CSR on the financial

performance of companies from 25 European countries.

Methodology / approach – By conducting empirical archival research the relation between CSR and financial performance is studied. The CSR scores were obtained from the Reputation

Institute which designed a framework to rank the social performance of companies. This resulted in 25 companies for which data was available. For these companies the financial data was then recorded.

Findings – The measurement method of CSR can have significant impact on the results. First no relation between CSR and financial performance was found, but when the measurement method for CSR was changed, and Asset4 ESG data was used, there were some significant results. These results indicated a negative relation between CSR and ROS.

Limitations – A first limitation of the research is the size of the sample, which consists of only 25 companies. Another limitation can be the construction of the sample since there needed to be a complete set of data for all firms in the sample over the period 2012-2014, which could result in a survivor bias.

Key words: Corporate Social Responsibility, financial performance, social performance, return

(4)

4

Table 9: Results of the regression for the first equation (a one-year time lag, 2013-2012). Table 10: Results of the regression for the second equation (a one-year time lag, 2014-2013). Table 11: Results of the regression for the third equation (a two-year time lag, 2014-2012). Table 12: Descriptive statistics CSR data from Asset4 database.

Table 13: Results of equation 1 with Asset4 ESG data as proxy for CSR. Table 14: Results of equation 2 with Asset4 ESG data as proxy for CSR. Table 15: Results of equation 3 with Asset4 ESG data as proxy for CSR. Table 16: Results for equation one, changed control variables.

Table 17: Results for equation two, changed control variables. Table 18: Results for equation three, changed control variables.

(6)

6

1. Introduction

The interest in Corporate Social Responsibility (CSR) experienced exponential growth over the last few years (Tsoutsoura, 2004). A good indicator of the increased enthusiasm of companies to involve in CSR are the rapidly growing companies such as Ben & Jerry’s and The Body shop, which make extensive use of CSR in their marketing strategies (McWilliams & Siegel, 2000). However the recent increase in CSR engagement is not only due to enthusiasm: scandals such as the BP Oil disaster that caused damage to the Mexican coast, the collapse of the clothing factory in Bangladesh in 2013, or the horse meat affaire in Europe have also increased the need for corporate accountability. This accountability goes beyond the financial aspect and includes also moral, social and legal aspects (Van Beurden & Gossling, 2008).

Research carried out by KPMG (2013) has documented the increase in CSR reporting and shows that in Europe, America and Asia more than 70% of the companies report on their CSR

activities. A remarkable change in comparison to the report of 2011 is that there has been an increase of 22% in CSR reporting by companies in emerging economies such as India and Chile. Due to the increased interest in CSR many articles about this topic have been published, but the recently published paper by Servaes and Tamayo (2013) concludes that even though much has been written about this topic it is still not possible to draw a clear conclusion about how

involvement in CSR impacts a company’s financial performance since the literature is divided.

1.2 Research question

The CSR description of Van Beurden and Gossling (2008, p. 407) is: “CSR comprises the notion

that organizations have to meet the expectations of society’’, which implies that companies are

mainly undertaking CSR actions and report about them because they need to meet the

expectations of society. This raises the question whether companies solely make use of CSR to satisfy societal expectations or whether they are driven by other motives such as the expectation of a positive effect on the company’s financial performance. Therefore, this thesis will try to formulate an answer on the question:

‘What is the impact of Corporate Social Responsibility on the financial performance of European companies?’

(7)

7

1.3 Relevance of the study

Several studies are looking into the impact CSR has on the financial performance of the firm. Even though this debate started years ago and various studies have been published since then there are still research opportunities in this field.

Scientific contribution

A scientific contribution that can be made in this field is to investigate companies that operate in a specific country or continent. This could be interesting since most of the currently published articles use a sample that consists either of solely American companies or a mixed sample with companies from all over the world. Therefore this study uses a sample that consists of various large companies across the European Union. The reason behind this specific sample choice is that there appears to be a difference in the way CSR is used in companies across the United States, which is researched frequently already, and European countries about which not many studies have published results.According to Maignan and Ralston (2002), who have researched how and why companies across France, the U.K., the Netherlands and the U.S. make use of CSR, there is a different dedication to CSR in these four countries. The main finding of their paper is (2002, p. 511): “European companies often presented CSR as an activity enhancing the

firm's success and survival […] while most U.S. firms introduced CSR practices as an expression of their own organizational culture’’.This difference between using CSR only because it is expected from society and on the other hand paying attention to CSR driven by internal motivations can have an impact on the intensity, and therefore the costs spend on CSR, by European and US companies. When a company dedicates more resources to their CSR activities there might be a higher risk that it will not pay off compared to when less is invested in CSR. To conclude: when companies would spent just enough to satisfy the expectations of society it might be more likely that CSR has a positive (or at least no negative) effect on financial performance in comparison to a company that has high CSR costs because from a company perspective CSR might be valued even more than by society.

Societal contribution

Also from a societal point of view an investigation of the impact of CSR on financial

performance of European listed companies can be interesting. Two specific groups of people who will potentially value the outcome of the study are investors and the top and middle

(8)

8 management of the company.

The thesis will contribute to the knowledge of the management of the investigated companies, and will probably also be helpful for companies in general. Since one of the main goals of

companies is to make sustainable profit, knowledge about whether the relationship between CSR and financial performance is positive, negative or neutral can be helpful in making decisions about the level of CSR reporting and activities.

On the other hand potential investors could be interested in the effect of CSR on firm

performance. Before an investor is able to make an investment decision he will have to make choices about the company in which they should invest. If this research shows that companies which undertake CSR actions also have enhanced performance this would have impact on their decision of which company to invest in.

A second group for which the results of this thesis could be interesting are top and middle management. These management layers are the ones who make decisions about the intensity of CSR. When they know the effect of CSR on their financial performance they can make better informed decisions about the trade-offs between these two.

The remainder of this thesis is organized as follows. In the next paragraph some background information regarding the concept of CSR will be provided. Section 3 will outline the existing literature by reviewing the conclusions from articles that have used different measures for CSR and financial performance. In section 4 attention will be paid to the methodology. This section is then followed by a discussion of the results in paragraph 5 and a sensitivity analysis in paragraph 6. Finally in paragraph 7 the conclusion can be found.

2. Background information & theory

The next paragraph provides background information about the concept of CSR. After reviewing various definitions of CSR the changes in the concept over the years will be discussed. Besides a short description of the most important regulations and guidelines regarding CSR is included. The last part of the background section pays attention to the underlying theories that explain the involvement in CSR.

(9)

9

2.1 Definition of CSR

Since one of the key subjects of this research is CSR it will be helpful to provide a clear

description of what this concept entails in order to understand the main purpose of the research. Because CSR is a complex concept that exists for several decades there is a substantive amount of literature that describes the meaning of it (Maignan and Ralston, 2002). A possible

explanation for this substantive amount of literature is that the concept of CSR has changed over the past couple of decades, as will be elaborated upon in later paragraphs, and as a consequence the definition of the concept also changed slightly. Despite various publications on this subject there is still no universal definition of the concept. This problem does not only exist in the academic world, the uncertainty about how CSR should be defined also exists in the corporate world (Dahlsrud, 2008).

The paper written by Sprinkle and Maines (2010) provides two different definitions of CSR. One of these is the widely used definition of the European Commission, which defines CSR as (2010, p. 445):

‘‘a concept whereby companies integrate social and environmental concerns in their business operations and in their interaction with their stakeholders on a voluntary basis.’’

The other definition is one that is used commonly in management literature, which is originally developed by Davis (2010, p. 445):

‘‘the firm’s considerations of, and response to, issues beyond the narrow economic, technical, and legal requirements of the firm to accomplish social [and environmental] benefits along with the traditional economic gains which the firm seeks.’’.

What these two definitions have in common is that they both refer to the social aspect of CSR. But besides the social aspect there are other dimensions used in CSR definitions such as the economic, environmental, stakeholder and voluntariness dimension. These five dimensions are used by Dahlsrud (2008) to classify 37 different descriptions of CSR that exist in literature. The review by Dahlsrud points out that in 97% of the definitions there are at least three different dimensions found. The dimension that is used the least is the environmental dimension, which is probably the result of developments in the definition of CSR: in early years the environmental aspect of CSR was not considered that often.

Besides the use of various dimensions most definitions also include the group of people involved. Whereas most authors, Van Beurden and Gossling (2008) and Tsoutsoura (2004)

(10)

10 amongst others, consider a very large public including all stakeholders of the company, their families and society at large, there are also definitions that exclude a certain group of people such as investors (Sprinkle and Maines, 2010).

The overall conclusion by Dahlsrud (2008) is that even though it seems to be a problem that there is no universal definition of the concept of CSR at first glance, it eventually turns out that all the definitions are congruent and therefore the lack of a single accepted definition is not considered to be a real problem. Since the definition produced by the European Commission includes all five dimensions it can be considered as a complete definition and will therefore be used as the definition of CSR in this thesis.

2.2 Developments relating to CSR

Van Beurden and Gossling (2008) recognize that the debate about CSR is going on for many years already. McWilliams and Siegel (2000) point out that the intensity of the debate changed since CSR has gained a lot more attention in de past three decades. Not only the fact that the debate has been active in the past decades but also the intensity of the debate is important to analyze because the level of attention to CSR can have consequences for the findings in previous studies that researched the relationship between CSR and financial performance.

Two papers that describe the evolution of CSR in great detail are the papers from Cochran (2007) and Carroll (1999). Whereas Cochran starts with describing the underlying academic reasoning behind CSR, Carroll just investigates the literature from the last 50 years.

According to Cochran (2007) there was a Professor Dodd that asked the question to whom companies were responsible which he answered by: ‘society as a whole’, rather than only

shareholders since the company serves society at the first place rather than existing only to make profit. Cochran (2007) continues by stating that this forms the basis of CSR.

From the 1950’s more attention by the business world is paid to what would later on be called CSR. Carroll (1999) explains that in this decade the term Social Responsibility is used instead of Corporate Social Responsibility. One of the important books in which this term is used is Social

Responsibilities of the Businessman, written by Howard Bowen, who is called ‘the father of

CSR’ by Carroll (1999). The first time period, from 1950 until 1960, is characterized by corporate philanthropy which means that companies started to donate money to for example

(11)

11 schools or universities in order to contribute to a better society without any direct benefits for the company itself. This expression of social responsibility changed into a more strategic way of making donations in the years that followed. With the movement to strategic philanthropy the goal to do something good for society was combined with the economic goals of the company. Just as the 1950’s the 1960’s were also characterized by the establishment of activist groups and non-governmental organizations (NGO’s) (Cochran, 2007). Both were involved in the debate about what good business behavior entails. By attracting media attention the activist groups and NGO’s were able to convince companies of their duty to be responsible for their actions and be accountable for a larger group than only their shareholders since negative attention in the media could have negative consequences for these companies with respect to their financial

performance. This aggressive strategy is described by Deegan and Blomquist (2005), who also explain that in the course of the years the strategy from NGO’s and activist groups changed from confrontation to cooperation and collaboration.

From the 1970’s on companies in the U.S. started to undertake action and Corporate Social Responsibility changed into Corporate Social Responsiveness whereby CSR was no longer just a debate about social responsible behavior and ethics (Cochran, 2007). Whereas in the 1970’s people tried to come to a single definition for CSR, in the 1980’s the focus was not so much on defining the concept but more on how to measure it (Carroll, 1999). From the 1980’s on the measurement of CSR has gained attention and until today no universal measurement technique is accepted.

2.3 Costs and benefits of CSR

A wide variety of benefits that arise from CSR are provided by the existing literature. One of these benefits is that the company signals to the market that it provides high quality products since only companies that are willing to invest in product quality will involve in CSR. For companies that are solely profit driven the investment in CSR is too expensive (Servaes and Tamayo, 2013). As a result of this higher product quality the firm is able to ask a higher price for its products and therefore can improve its margins, whereby the company’s market share could also be increased according to Sprinkle and Maines (2010). A further potential consequence of this higher product quality is a lower risk that the company has to recall product lines that turn out to be defective (Tsoutsoura, 2004).

(12)

12 and Maines (2010) and Tsoutsoura (2004) describe the contracting benefits that can result from CSR. These contracting benefits are said to be helpful in recruiting employees and even to retain employees. In this case the company can save money on training and development costs,

recruitment costs and turnover costs.

Besides the above mentioned benefits of increased customer and employee goodwill, there are benefits related to reduced risk. For example when a company voluntary reduces its emissions, decreases its carbon footprint or reduces its waste production it may be able to avoid certain litigation risks (Sprinkle and Maines, 2010). Tsoutsoura (2004) indicates that there is also a lower risk that a company has to pay any penalties in the form of fines for their level of

pollution. Furthermore, she recognizes the decreased likelihood of negative social effects which can cost a company millions of dollars.

Reduced risks often imply reduced costs as well. One way in which a company can reduce its operating costs is by calculating a more optimal route for the delivery process or by using less packaging materials. At the same time these actions reduce the environmental impact the

company has (Sprinkle and Maines, 2010; Tsoutsoura, 2004). Costs savings can also be made on advertising costs. According to Sprinkle and Maines (2010) it is not uncommon for media, such as television, radio or newspapers, to publish articles about companies that engage in CSR activities. In this way companies enjoy the benefits of ‘free’ publicity. Lastly Sprinkle and

Maines (2010) and Tsoutsoura (2004) point out the benefits of attracting capital more easily from investors or financial institutions as a result of an enhanced brand image and reputation and receiving more favorable credit terms.

With this large variety of benefits that result from CSR the question rises which costs the company has to incur. According to McGuire, Sundgren and Schneeweis (1988) there are three perspectives regarding the costs that are related to CSR. A pessimistic view which reasons that a company that incurs costs of CSR activities will be at an economic disadvantage in comparison to companies that do not engage in CSR. A more optimistic view that the costs related to CSR are not substantive and will therefore be out weighted by the benefits. And a third view that argues that the costs are significant.

Even though previous literature shows that customer goodwill can increase as a result of CSR there is also a risk that consumers decide to buy their products elsewhere as a result of CSR. This can for example be the case when companies ask their customers for donations to charity, which

(13)

13 can evoke the feeling that the company enjoys the benefits while the customer is paying for it. Another example of a cost to consider when deciding on engaging in CSR is the opportunity cost which is related to donations of cash or products. The above examples of the different costs a company can incur confirm the statement of Sprinkle and Maines (2010) that costs are not just the cash outflows, but also the reduced cash inflows.

2.4 Company motivations to use CSR

The motivation to study the influence of CSR on financial performance of European companies stems from the research conducted by Maignan and Ralston that showed that companies in European countries had different motives to make use of CSR than their American competitors. It appeared that European companies are forced by the public to engage in CSR, whereas the majority of the American companies are often involved in such activities because it is part of their corporate culture (2002). In this paragraph the more general motivations behind the engagement in CSR will be evaluated.

As the definition of the European Commission stresses CSR is a choice for companies, it is voluntary for them to involve in these kind of activities. Therefore the question rises what the literature describes the motivations behind undertaking CSR activities are: are they related to company performance? Is there an internal motivation or is the commitment to CSR more related to pressure from outside parties?

The overview that is presented by Maignan and Ralston (2002) reveals that there are in general three categories of motivations that drive the commitment to CSR: it serves as an instrument that helps achieving certain performance objectives related to sales or profitability, or because

companies are trying to comply with the norms that stakeholders have set, or because of the internal motivation to have a positive impact on the environment or community. Other authors distinguish only economical or ethical (Windsor, 2006) or moral and practical forms of CSR (Van Beurden and Gossling, 2008). Windsor (2006) furthermore explains that the economic and ethical forms of CSR are mutually exclusive and these two concepts cannot overlap, which means that a company is either involved in CSR in its own interest or because it wants to contribute in solving social problems in a unselfish and altruistic way.

(14)

14 internally driven and not so much focused on helping the community or environment but more on the self-interest of the company and thus can be classified as economic or practical form of CSR. According to Maignan and Ralston (2002) this approach can be considered as a utilitarian approach, under this approach only the interest of a certain group in society is considered. In this case the main interests that are considered are the interests of the investors, this is called

utilitarian ethics by Windsor (2006). One possible way in which CSR can contribute to the company’s performance is outlined by McWilliams and Siegel (2002). They explain that a firm can use CSR in order to create a reputation of reliability and honesty. For certain customers these characteristics are important and they will buy their products at a company they see as honest and reliable. In some cases the products of a company that is involved in CSR are believed to be of higher quality. Although the concepts of economic and ethical CSR cannot overlap according to Windsor (2006), it is possible that the results that are produced by CSR activities which are initiated by ethical motives also have positive economic consequences for the company as explained by McWilliams and Siegel (2002).

The second motivation, companies are trying to comply with the norms that stakeholders have set, is also called the negative duty approach (Maignan and Ralston, 2002). This duty is clearly described by Tsoutsoura (2004): according to her there has been a shift of power to the

corporations and since they have this power it is simply their duty to play a role in resolving social problems. She continues by stating that it is impossible for a firm to ignore events such as political unrest, poverty of citizens and exhaustion of natural resources, because all these

problems can also have an impact on the company. This explanation by Tsoutsoura implies that even though involvement in CSR is not mandated by any laws or regulations, it is not completely voluntary either. Either companies are forced to reduce or resolve the problems because they need to minimize the impact on the company or their stakeholders require to do so. When the latter is the case the use of CSR is mainly a legitimacy instrument. By labelling this motivation as ‘window-dressing’ Sprinkle and Maines (2010) highlight the negative aspect of this duty. Thirdly a company can involve in CSR because of the internal motivation to have a positive impact on the environment or community, called the positive duty approach. In this case the use of CSR is part of the corporate identity (Maignan and Ralston, 2002). When a company is involved in CSR because of such altruistic intentions this is also labelled as ‘being a good corporate citizen’ by Sprinkle and Mains (2010).

(15)

15 Finally Maignan and Ralston (2002) state that all three approaches, the positive- and negative duty as well as the utilitarian ethics, can be seen as a form of impression management since they are used to influence the perceptions that various stakeholders have of the company.

2.5 CSR guidelines and regulations

In this paragraph an outline will be provided of the most important frameworks relating to CSR. In 2004 Tsoutsoura reported that it is not very common for companies to have their CSR reports externally verified which raises the concern that the quality of such a report can be questionable since management is able to exaggerate the company’s CSR performance.

As stated before the involvement in CSR is on a voluntary basis which means that there are no strict rules or regulations that a company has to adhere to, but due to the increased attention from society for CSR companies started not only to engage more in CSR activities but also disclosed and reported more about these practices and therefore in response to the increase in CSR reporting various guidelines were developed (Chen and Bouvain, 2009). Brown, de Jong and Levy (2008) state that the GRI framework is the best-known amongst the various frameworks and guidelines that exist. Chen and Bouvain (2009) add to this that the UN Global Compact also belongs to the most important ones whereas the AccountAbility’s AA1000 standard, Social Accountability International’s SA8000 standard and The ISO 14000 environmental management standard can be considered as the emerging standards.

The GRI framework

The in 1999 introduced Global Reporting Initiative, which was established by a collaboration between the Coalition for Environmentally Responsible Economies (CERES) and the United Nations Environment Programme (UNEP), is considered to be one of the most important CSR frameworks (Chen and Bouvain, 2009). According to recent research by Bonsón and Bednárová (2014) of all companies that publish a CSR report forty percent uses these guidelines, especially in Europe this is the preferred framework by companies. The framework has two main goals of which the first goal is to bring the various reporting systems that were used at the time

together. The second goal is related to becoming a platform for societal dialogue about the content of CSR performance (Brown et al., 2009).

(16)

16 Under the GRI framework a distinction is being made between the social and environmental issues that need to be discussed in a company report besides the financial matters. In the category of social issues topics such as human rights, society, labor practices and product responsibility are addressed. The environmental issues that need to be described are issues such as materials used and the percentage of recycled materials, water usage and the consumption of energy (Chen and Bouvain, 2009).

UN Global Compact

In the same year that the GRI framework was introduced the United Nations also published their framework, the United Nations Global Compact, which was intended to encourage companies from all over the world to develop, adopt and report on their CSR policies (Chen and Bouvain, 2009). Initially this framework consisted of nine principles, but a tenth principle was added in 2004. Topics that are addressed by the GRI framework correspond to those addressed by the UN Global Compact. The main difference is that the latter has divided the ten principles in four categories: human rights, labor, environment and corruption (Williams, 2007). The principles relating to human rights prescribe that businesses should protect and respect internationally recognized human rights and they should not being accused of human rights abuse (Chen and Bouvain, 2009). As outlined by Williams (2007) the labor section asks companies to refrain from any form of forced and child labor. Furthermore it should not discriminate on the basis of

employment or education and it should recognize the right of association and collective bargaining. Regarding the environmental category it demands companies to promote the development of environmental friendly technologies and encourages initiatives relating to enhanced responsibility for the environment. Lastly Chen and Bouvain (2009) describe that to comply with the corruption guidelines of the framework companies should avoid involvement in all forms of corruption.

Other frameworks and standards

Three other frameworks for CSR reporting were mentioned at the beginning of this paragraph. The Account Ability AA1000 which has been established in order to restore and improve public confidence in reports about social and economic issues and to enhance the performance of companies in these fields (Göbbels and Jonker, 2003). This framework focuses on the social and

(17)

17 ethical issues a company has to deal with (Marimon, del Mar Alonso-Almeida, del Pilar

Rodríguez and Aimer Cortez Alejandro, 2012).

The focus area of the in 1997 introduced SA8000 standard is more on human rights and

conditions in the workplace (Göbbels and Jonker, 2003). According to Marimon et al. (2012) the main goal of this standard is to provide guidelines on international human rights norms and national labor laws (2012).

Lastly the ISO14000 was mentioned by Chen and Bouvain (2009). This standard focuses especially on environmental reporting (Marimon et al., 2012). The framework aids in designing plans, policies and actions regarding environmental issues.

2.6 Theories that explain the use of CSR

In this paragraph two main theories that explain the engagement in CSR by companies will be discussed, these are stakeholder theory and legitimacy theory. Even though theories cannot completely explain certain behavior, since they are only abstractions of reality, they can still provide valuable insights (Deegan & Unerman, 2006). Both stakeholder theory and legitimacy theory are derived from the same theory which is called political economy theory. The idea behind this theory is that the concepts of society, economy and politics cannot be separated which means that it is not possible to study one of the aspects without taking the other two aspects into consideration. The political economy theory is divided into two different categories: the classical stream and the bourgeois stream. The former focuses on the structural conflicts and inequity that exist within society, whereas the latter takes the interactions amongst various groups in society into account. It is this latter stream of political economy theory from which both the stakeholder theory and legitimacy theory are derived.

2.6.1. Legitimacy theory

Together with institutional theory the legitimacy theory and stakeholder theory can be classified as systems-oriented theories, which are described by Gray, Owen and Adams (1996, p. 45) as: “... a systems-oriented view of the organization and society ... permits us to focus on the role of

information and disclosure in the relationship(s) between organizations, the State, individuals and groups”. The theory assumes that the connection between the organization and society is

(18)

18 influenced by society (Deegan & Unerman, 2006). It is especially the influence of society on the organization that is stressed in most definitions of legitimacy theory. Wilmshurt and Frost (2002) recognize the reaction of corporate management on societal expectations as the driver of

legitimacy theory. Deegan and Unerman (2006) add that the bounds and norms of society are guidelines for how a company should operate. According to Guthrie and Parker (1989) it is of critical importance that an organization acts upon these desires and expectations of society in order to achieve its objectives and being able to survive.

A distinction must be made between the terms legitimacy and legitimation (Lindblom, 1994), whereby the former can be considered as a condition or status and is therefore more fixed than the latter, which is the process that ensures that the companies actions are seen by society as legitimate. This process of legitimation is continuous since Deegan and Unerman (2006) report that the bounds and norms of society are changing from time to time which implies that

companies need to continuously adapt to changes in societal perceptions.

Lindblom (1994) further describes that there are four different strategies to reach legitimacy. The first strategy is to divert attention to other areas in which the company performs well. Secondly an organization can be seen as legitimate by educating the public and convincing them that some expectations are unrealistic or third by informing the public that expectations and actions are in line with each other. Lastly companies can change the perceptions the public has of the company and its activities while not actually changing the company’s actions.

The interaction between the organization and society can been seen in the form of a contract. As highlighted before by Guthrie and Parker (1989) the survival and achievement of objectives of the company depends on if it complies with the expectations of the community, if it does not comply it will not reach its goals just like when a contract is breached.

The link between CSR and legitimacy theory should be evident by now: from the point of view of legitimacy theory companies should comply to the social contract with society in order to guarantee their continued existence. In order to meet the requirements of this contract a company can decide to engage in CSR. According to Maignan and Ralston (2002) there is evidence that proves that CSR can be an effective instrument to increase the legitimacy of a company.

(19)

19

2.6.2. Stakeholder theory

Even though the main idea behind legitimacy theory and stakeholder theory is the same, that organizations are influenced by society and vice versa, there are also several important

differences between the two theories (Deegan and Unerman, 2006). One of the key differences is that under stakeholder theory it is believed that the firm has to adhere to many social contracts as opposed to what is believed in legitimacy theory, that the firm has only one contract with society as a whole. The groups that have a contract with the organization can be classified as

stakeholders which are defined by Freeman as (2006, p. 286): “any group or individual who can

affect or is affected by the achievement of the firm's objectives”. Examples of these groups

include employees, creditors, governments, owners, media, suppliers and customers (Roberts, 1992). A distinction can be made between primary and secondary stakeholders. The first category, primary shareholders, are the ones that the company needs in order to

survive. Secondary stakeholders can have an impact on the organization, but not in a way that primary stakeholders can, since secondary stakeholders do not engage in transactions with the organization. Groups such as media and governments are classified as secondary stakeholders (Deegan and Unerman, 2006).

In one of the two branches of stakeholder theory, the managerial (positive) branch, primary stakeholders are considered to be the most important ones which management should always consider in making decisions. They are considered to be of greater importance since they are critical for the survival of the company. The other branch, which is referred to as the ethical branch, argues that all groups of stakeholders should be treated equally, regardless of the power they exert. Besides this branch focuses on how an organization should behave and therefore it is sometimes called the normative branch.

The explanation for the involvement in CSR from the perspective of stakeholder theory is that the company has to comply with the expectations of certain groups in society. These

stakeholders all may have different expectations, which do not only have to relate to their own welfare but can also include the effects on others such as expectations regarding not making use of child labor (Maignan and Ralston, 2002). Based on the relative importance of these

stakeholder groups the company can decide on the intensity and variety of the CSR activities it undertakes.

(20)

20

3. Literature review

The next section will outline the existing literature on the relation between CSR and financial performance. In total seventeen articles from the 1980’s until now will be analyzed. This literature is divided into three different sections based on the finding of a negative, neutral or positive relation between the two variables.

3.1 Negative relationship

One of the possible explanations for a negative relation between involvement in CSR and the financial performance of a company is that companies that do involve in CSR are facing a competitive disadvantage compared to companies that do not engage in such activities since socially responsible companies incur costs that have a direct impact on profitability (Tsoutsoura, 2004). McGuire et al. (1988) also document this explanation and add that the economic

disadvantage can be due to foregoing certain strategic alternatives as a consequence of being a socially responsible company. An example of such strategic alternatives is not producing lines of weapons or pesticides. A study that has found such a negative relationship is that of Boyle, Higgins and Rhee (1997). These authors investigate how investors react on social responsible initiatives by companies. In contradiction to what Boyle et al. a priori expected to found they documented a negative relation because involvement in CSR was found to have a negative impact on future cash flows.

A remarkable finding that is obtained by Griffin and Mahon (1997) is that the majority of studies that show a negative relation between CSR and financial performance comes from the 1980’s. Griffin and Mahon (1997) have analyzed 51 studies which they categorized into three different time periods, 1970’s, 1980’s and 1990’s. In the 1980’s more than 60% of the analyzed studies reported a negative relation whereas this was only 6% in the 1970’s and 37% in the 1990’s, which can be considered as a significant difference. An explanation for this large discrepancy between these time areas is the intensity of the CSR debate. McWilliams and Siegel (2000) report that CSR gained importance especially over the last three decades. This in combination with the assertion by Arlow and Gannon (1982) that the effects of CSR are not immediately visible but only appear after a couple of years may explain the negative effect in these studies. In other words the negative effect on financial performance that is documented in the 1980’s can be due to the increase in CSR activities which involves extra costs while the benefits of these activities are not directly visible.

(21)

21 Even though it seems that a negative relationship between CSR and financial performance is mostly found in earlier decades there is also a more recent study, carried out by Brammer, Brooks and Pavelin (2006), which documents a negative effect. What this study has in common with the study by Boyle et al. (1997) is that Brammer et al. (2006) also make use of the

investors’ perspective in their research. By using stock returns to investigate how financial performance is affected by CSR activities. Just like Boyle et al. (1997) Brammer et al. (2006) conclude that companies with higher CSR scores show lower returns and the other way around. As clarification for the results that Brammer et al. (2006) obtained they provide the same

explanation that was given by Tsoutsoura (2004) and McGuire et al. (1988): the costs of CSR do have a negative effect on the returns a company is able to gain and therefore the performance of these companies is lower compared to other companies which do not score well on social aspects.

3.2 No relationship found

In comparison to the negative relation that was described before, approximately the same amount of articles have found no relationship between CSR activities and financial performance. A general explanation of why there would be no relationship between the two variables is given by Tsoutsoura (2004) who argues that there are so many intervening variables between CSR and financial performance that one should not expect this relation to exist. Another reason why several studies have not found a relationship between CSR and financial performance can be explained by the motivation which is called the negative duty approach. Companies that engage in CSR because of the pressure from stakeholders will put just enough effort in these activities to satisfy stakeholders and thereby preventing themselves from any negative consequences on financial performance.

In total there are four articles found that come to the conclusion that there exists no relation between CSR and financial performance. Whereas three of them are published in the last decades, only one of these studies is carried out already in 1985. This study, carried out by Aupperle, Carroll and Hatfield (1985), criticizes the existing body of literature at that time because all these studies either made use of incomplete and simplistic methodologies or suffered from ideological bias. The reason the authors have this critique is mainly because they recognize that there are no accurate measures for CSR which leads to the following approach: Aupperle et

(22)

22 al. (1985) made use of the construct of Carroll, a definition of CSR by four different components and these components were then evaluated by the CEO’s of the company. Even though this method to measure CSR can be considered very original, as no other studies have been found that make use of this method, it is also very subjective. When CEO’s have to rate how their own company scores on various aspects of CSR there is a significant risk of bias. Therefore the finding that no relationship exists between CSR and financial performance is questionable. The three other studies that have reported that there is no relationship between the two variables are those of Aras, Aybars and Kutlu (2010), Brine, Brown and Hackett (2007) and Fiori, Donato and Izzo (2007). Besides the fact that these studies all have been published in the past decade they have also in common that they are investigating the influence of CSR on financial performance in a specific country. Aras et al. (2010) examine the Turkish context, Brine et al. focus on Australian companies and Fiori et al. make use of Italian companies in their study. In the study that is carried out by Arras et al. (2010) accounting based measures were used to measure financial performance and for measuring CSR the authors made use of content analysis. By the end of the study Arras et al. (2010) were not able to find a relationship between CSR and financial performance they did find a correlation between CSR and firm size. Regarding the explanation for their first finding they state (2010, p. 210): “It is interesting to speculate on the

reason why no significant relationship is found but difficult to determine a reason”. Even though

Aras et al. (2010) do not provide a clear reason for the fact that they did not find a relation between CSR and financial performance they do exclude the possibility that the finding is due to the fact that Turkey is a developing country since previous research has shown that there are no differences between developed and developing countries in this respect. This statement is confirmed by the study of Brine et al. (2007) who have made use of companies in Australia which can be considered a developed economy. Besides the fact that this study also makes use of accounting based measures, ROA, ROS, ROE just like Aras et al. (2010), they do also make use of content analysis. Just as many other authors Brine et al. (2007) mention the difficulty of measuring the concept of CSR. In the final paragraph of the paper Brine et al. (2007) come to the conclusion that due to the difficulty of measuring CSR their study should be seen as a first step in researching the relationship between CSR and financial performance in Australia. Just like in the study of Aras et al. (2010) no clear reason is provided for the findings, only several

(23)

23 Lastly there is the research by Fiori et al. (2007), who have studied Italian companies, that

documented that there is no relationship between CSR and financial performance. The main difference between this study and the other three studies is that the authors made use of market based measures to draw conclusions on the impact of CSR on financial performance. Measuring financial performance by using the company’s stock price is an approach that is not much used in other studies, but the way Fiori et al. (2007) measure CSR is by content analysis, just like the three papers that have just been discussed. Different from the articles written by Aras et al. (2010) and Brine et al. (2007) who did not provide clear reasons for their finding that no relationship exists between CSR and financial performance Fiori et al. (2007) present a set of three possible reasons for this finding. First they argue that CSR is still a relatively new concept in Italy and therefore most investors do not take the information provided about CSR into consideration. The second argument relates to the measurement of CSR which is considered to be complicated since it is difficult to determine the quality of CSR disclosure due to the absence of generally accepted principles. Thirdly Fiori et al. (2007) mention the difference in the

orientation: shareholders tend to be more short-term oriented whereas the impact of CSR is in the medium or long term.

A general conclusion that can be drawn based upon the literature that claims that no relationship exists between CSR and financial performance is that the measurement of CSR seems to be a problem. Furthermore, when market based measures are used for financial performance, the attitude of shareholders seems to be of importance. If the shareholders do not take the CSR into consideration this might explain that CSR has no impact on financial performance.

3.3 A positive relation

Compared to the previous two paragraphs which analyzed the literature that concluded that there is either no relation between CSR and financial performance or a negative relation this paragraph discusses various articles in which a positive relation is documented. In total seventeen articles about the relationship between CSR and financial performance were found of which ten have reported a positive relation.

A remarkable fact about these articles is that the majority is published in the years from 2000 until 2014. Only two articles are from the 1990’s and none of them was published in the 1980’s. A possible explanation for the fact that most of these articles that document a positive relation

(24)

24 are still recently published is that the importance of CSR has increased over the last years

(Tsoutsoura). The two exemptions here are the papers of Herremans et al. (1993) and Waddock and Graves (1997). Besides the time period of publishing another similarity is that both studies investigate the American context. The former have especially focused on several large

manufacturing companies in the United States. Not only have Herremans et al. (1993) found that companies with higher CSR scores have higher profitability but these companies also turn out to be less risky. Since this study makes use of market based measures for financial performance they are also able to make a statement about the perceptions of investors about CSR. These stakeholders are not only aware of the difference in CSR performance but they also seem to act upon this information according to this statement (Herremans et al., 1993, p. 600): “evidenced by

positive abnormal returns accruing to the stocks of companies with superior reputations during the period studied”. Lastly Herremans et al. (1993) mention that in the years of their study there

was a transition point in thinking: even though it was still assumed that CSR negatively impacted financial performance because it was purely seen as an extra cost of doing business, some

arguments for a positive impact were suggested for the first time. Several years later, Waddock and Graves (1997) also found a positive relation between CSR and financial performance. While most other papers have only studied the impact that CSR has on financial performance, Waddock and Graves (1997) decided to investigate also whether financial performance has impact on CSR. Making use of a self-constructed CSR index and accounting based measures they have

documented that this is indeed the case for their sample that consists of the S&P 500 firms. The authors argue that there might be a logical explanation for this finding: when firms are in financial distress it is less likely that they will invest in CSR. On the other hand firms that already perform well in the financial aspects have more opportunities to invest in CSR. The theory behind this explanation is called slack resources theory. A study that is highly comparable to the study of Waddock and Graves (1997) is that of Tsoutsoura (2004). Not only did this study make use of the same sample (S&P 500), it also uses ROA, ROE and ROS as measures of financial performance. The main difference between these two is that Tsoutsoura (2004) has studied a different time period, approximately ten years later. In accordance with the findings of Waddock and Graves (1997), Tsoutsoura (2004) also documents that CSR has a positive impact on financial performance. The argument that is based on slack resource theory that was

(25)

25 investments in CSR can have strategic impact on the long term. It may lead to the attraction of better employees who are, due to their skills and expertise, able to help the company perform better on the financial aspects.

Just like Herremans et al. (1993) the study of Ngwakwe (2009) also focuses on manufacturing companies but has a sample of Nigerian firms instead of American. What distinguishes this study from most other studies is the method Ngwakwe (2009) uses: the sample is divided into two groups, one group with ‘socially irresponsible’ firms and the other consisting of ‘socially responsible’ firms. Three elements of CSR are used to test whether a relation exists with financial performance, measured by return on total assets (ROTA) in this case. Besides the finding that this relation is positive the author also documents that companies that are socially responsible encounter less fines and penalties. It is questionable to which extent the findings of this study can be generalized since Nigeria can be classified as a developing country and besides the sample used by Ngwakwe (2009) is rather small since it consists of sixty firms.

A recent study by Servaes and Tamayo (2013) focuses in particular on the effect of customer awareness of CSR on the financial performance of firms. The proxy that is used to measure the customer awareness is the amount of advertising. Only for firms with higher customer awareness the relation between CSR and financial performance is positive, when the awareness is low the effects of CSR on financial performance are insignificant or negative (Servaes & Tamayo, 2013). This awareness can also has disadvantages: when the public is aware but there are concerns about CSR performance this can lead to penalties. This can be seen as a confirmation of the motivation stated earlier in this study that customers are more willing to buy from companies that perform well on social dimensions.

Van Beurden and Gossling (2008) contribute to the literature by reporting about the factors that are most important in explaining the relation between CSR and financial performance. These are size, risk, industry and R&D. What distinguishes this study from most other studies is that it draws conclusions about the relation between CSR and financial performance based on a meta-analysis of previous literature. The finding that four factors mainly influence the relation is important since it is based upon the existing literature. Van Beurden and Gossling (2008) have studied 34 articles of which 68% have found a positive relation between CSR and financial performance.

(26)

26 conducting a meta-analysis instead off by doing empirical research is from Orlitzky, Schmidt and Rynes (2003). The mixed results of previous research have motivated them to investigate this relation again by adopting a more rigorous method. Furthermore Orlitzky et al. (2003) decided not only to investigate the impact of CSR on financial performance but also whether financial performance has impact on the engagement in CSR, which was also investigated by Waddock and Graves (1997). This bi-directional effect is confirmed by their study. Another interesting finding of this study is that is it not only the case that social responsibility pays off, but the authors separately concluded that environmental responsibility also pays offs even though this is to a lesser extent than social responsibility.

Montabon, Sroufe and Narasimhan (2007) investigate the impact of CSR, which they describe as environmental management practices (EMP’s), by analyzing the environmental and business performance of 45 international companies. In this study the focus is more on the environmental practices of the company than on the social practices which follows from the elements that are used to carry out the contents analysis. These are inter alia proactive waste reduction, recycling and remanufacturing. Even though the focus is somewhat different than from most other articles that investigate the relation between CSR and financial performance, Montabon et al. (2007) also document a positive relation. This might be due to the fact that better environmental policies lead to fewer fines and penalties as suggested by Ngwakwe (2009) before.

Based upon stakeholder theory the hypotheses of Barnett and Salomon (2006) is that there is a financial loss of investing in social responsible investment funds, since these portfolios are poorly diversified compared to other funds, but that this loss is offset as social screening

increases. This makes their approach different from previous studies that investigated the relation between CSR and financial performance.

Ziari and Peters (2002) do not only document a positive relation between CSR and financial performance but argue that acting in a socially responsible way is critical for the performance of a company. The main threat that Ziari and Peters (2002) identify which possibly prevents

companies from not adhering to the best ethical and social practices is that there are no standards that are mandated for companies. Currently some organizations are therefore ‘making half gestures’ (Ziari and Peters, 2002) which only focus on the self-interests of the company on the short run. Pressure groups can play an important role here according to Ziari and Peters (2002). In conclusion this overview of previous literature clarifies that even though the articles make use

(27)

27 of different measures for financial performance and CSR, have different samples and different methodological approaches they all documented that CSR has a positive impact on financial performance. An observation that came to the surface is that over time literature has more often found a positive relation between these two variables. This trend can be explained by the increasing importance of CSR in the last decades.

3.4 Hypothesis development

The previous paragraphs have outlined the rationale behind the involvement in CSR. In order to formulate a hypothesis about the sign and the existence of a relationship between CSR and financial performance the main arguments will now shortly be reviewed. The arguments are categorized into three categories: arguments derived from existing literature, arguments based on existing theories and arguments relating to the outcomes of previous empirical research.

One of the papers from the existing literature is that of McWilliams and Siegel (2000) which points at the increased intensity of the debate on and attention to CSR. More attention to social performance may imply increased impact on financial performance. Besides the fact that more attention is paid to social performance another reason for companies to engage in CSR activities is to avoid negative media attention. Avoiding such negative attention can save companies a large amount of money to improve their reputation. Based on this argument a positive relation between CSR and financial performance can be expected.

Sprinkle and Maines (2010) state the argument of increased customer goodwill which can be the result of signaling higher product quality. The consequence of higher product quality is that the company is able to ask higher prices and thereby increase their margins. In the end the market share can be increased as well. These effects have a positive result on financial performance, which again indicates a positive relationship.

Not only customer goodwill can be increased by involving in CSR but also employee goodwill. It is argued (Tsoutsoura, 2004) that a company which performs well on social dimensions is able to attract better employees. Besides employees are retained more easily which reduces training and recruitment costs. A final cost saving argument relates to reduced costs on advertisement, penalties and fines, defective product lines and packaging materials. Again there is a positive impact on financial performance.

(28)

28 society and vice versa. According to Guthrie and Parker (1989) a company has to act upon the desires and expectations of society in order to survive. From this statement it can be concluded that when a company does not perform well on the social dimensions it will have negative consequences for its financial performance.

After analyzing the results from previous empirical research about the relationship between CSR and financial performance from the 1980’s until today a major conclusion that can be drawn is that the majority of the contemporary studies has documented that CSR has a positive impact on financial performance. From the seventeen articles that were analyzed in the literature review 17% documented a negative relationship, 23% reported that no relationship was found, but the majority of research that has been conducted (60%) found a positive relationship between CSR and financial performance. These results are comparable to the percentages of positive, neutral and negative relations that were found in the meta-analysis by Van Beurden and Gossling (2008). Based on the above mentioned argument the following hypothesis will be tested:

H1: CSR has a positive effect on the financial performance of European companies.

4. Methodology

This chapter discusses the various measurement methods that are used in the existing literature and which one will be used in this study. Furthermore the data collection process and the resulting sample are discussed. Besides the empirical model that is used to perform the regression is provided.

4.1 Measures of CSR

Herremans, McInnes and Akathaporn (1993) argue that the reason why there is still no consensus about the relationship between CSR and financial performance is that there are still no clear guidelines about how to measure CSR. The reason for this, according to Van Beurden and Gossling (2008), is that it is a difficult concept to measure since it is not a variable and therefore cannot directly be measured. The authors continue by stating that various studies therefore make use of the concept of Corporate Social Performance which enables them to put CSR into practice since this concept can be transferred into variables. This concept consists of three categories that relate to the extent of CSR disclosure by the company, the activities that are carried out and

(29)

29 corporate reputational ratings as for example KLD or Fortune (Van Beurden and Gossling, 2008).

Cochran and Wood (1984) describe that there are in general two different ways in which CSR performance can be measured. These are the reputation index and content analysis. The first method, using a reputation index, involves the rating of firms by professional or knowledgeable observers on one or more dimensions of CSR. Cochran and Wood (1984, p. 43) provide three arguments in favor of using this method: “First, it tends to be internally consistent because one

evaluator is applying the same (albeit usually subjective) criteria to each firm. Second, it makes no pretence of applying a rigorous objective measure to a dimension that may be innately subjective. Third, it may summarize the perceptions of a key constituency of various firms.”. A

disadvantage on the other hand is that the method is subjective since the companies are rated according to the observations of people, not by objective information. The other method to measure CSR, content analysis, is a more objective method compared to making use of a reputation index because the company’s CSR score depends on particular variables that have been chosen. Another advantage of this method is that it is more mechanical and therefore it is possible to include larger samples. But also this method has drawbacks. The most important disadvantage is that, even though less than the other method, subjectivity is still involved since the analyzed reports only provide information about what the companies say there are doing to be socially responsible and not what there are actually doing.

Various indices exist that have collected data about the social responsible performance of

companies. Examples of these indices are the Kinder, Lydenberg, and Domini (KLD), the Domini 400 Social Index and the Dow Jones Sustainability index. For this study these indices are not suitable since they contain data about US firms mostly. The CSR data that is used in this study is produced by the Reputation Institute, hereafter referred to as R.I., which will be described in the paragraph about the data collection and sample description.

4.2 Measures of financial performance

As already described in the section that analyzed the previous literature there are various ways in which financial performance can be measured. In general all these measures can be subdivided into two broad categories: accounting-based measures and market-based measures, of which the latter is sometimes called investor returns (Cochran and Wood, 1984). Accounting-based

(30)

30 measures of financial performance which are frequently used in previous research are

profitability measures such as return on assets (ROA), return on equity (ROE) and return on investment (ROI). Examples of market-bases measures are the share price, market return and the ratio of market value to book value (Van Beurden and Gossling, 2008).

Both methods have some advantages and disadvantages which will now be reviewed. The main disadvantage of accounting-based measures is that they only provide information about historical performance and therefore it is questionable what they can tell about future performance

(Tsoutsoura, 2004). Furthermore these measures are more susceptible to accounting

manipulation by members of the management who have incentives related to their compensation and besides there is a higher risk of bias due to differences in accounting procedures (McGuire et al., 1988). On the other hand McGuire et al. (1988, p. 868) give an argument that supports the use of accounting-based measures: “If perceptions of social responsibility are firm-specific

(unsystematic), accounting measures of return should be more sensitive to them than stock-market measures, which reflect systematic stock-market trends.”. Bruner (2002) points out that the

credibility of accounting-based measures is high since the financial statements have been audited by an external auditor.

Compared to accounting-based measures market-based measures have the advantage that the risk of bias resulting from differences in accounting procedures is lower. Besides there are less opportunities for management to manipulate market-based measures of financial performance (McGuire et al., 1988). Orlitzky et al. (2003) explain why marked-based measures can be good predictors of the relation between CSR and financial performance. They state that investors determine the stock price, and hence the market value of the company, by their expectations of future stock returns, which are influenced by the social performance of the company.

Even though the literature agrees upon the fact that financial performance is more

straightforward to measure than CSR, various authors have different viewpoints about whether accounting-based measures or market-based measures are a better predictor for the financial performance. In the studies of Orlitzky et al. (2003) and McGuire et al. (1988) both marked-based measures and accounting-marked-based measures are used, but the majority of analyzed literature uses accounting-based measures as ROA and ROE. Based on the credibility argument of Bruner (2002) the measures that will be used in this study as proxies for financial performance are ROA, ROE and ROS.

(31)

31

4.3 Sample and data collection

As mentioned before the focus is on US firms in the current literature. Therefore indices such as the Down Jones Sustainability index and the KLD, which mainly contain data about S&P500 firms, are not suitable for this research. The data about the CSR performance of the companies in the sample that is used in this study comes from the reports of the RI. This institute, which was founded in 1997, engages in researching reputations and advising their clients on reputational issues. The model that is currently used to measure reputation, which is called RepTrak, is based on the academic models of one of the founders of the institute, Charles Fombrun (RepTrak, n.d.).The seven dimensions that are used to measure reputation are performance,

products/services, innovation, workplace, governance, citizenship and leadership as shown on figure 1. (Reputation Institute, 2013). Every year the RI publishes reports about the hundred companies with the best reputation. This is done globally, on a per country basis and in the regard of CSR. The reports about the latter category will be used here. For the CSR reputation ratings three out of the seven dimensions are used. The first is citizenship which relates to being a good corporate citizen by supporting good causes and taking care of the environment. The second dimension is governance which prescribes that the company gives transparence of its business, is run in a responsible way and behaves ethically. The third dimension is workplace which entails the treatment of its employees (RI, 2012).

There are reports available for 2012, 2013 and 2014. Because these reports include the top hundred companies in the world, first the non-European companies must be filtered out. This was first done by researching where the firms are listed and

in the case that a firm was not listed the location of the headquarter was used to determine the country of the company. Due to this process the sample was reduced from 100 to 38 companies. Since some the data of these 38 companies was not available for all three years the sample was further reduced to 28 companies.

The second step in the data collection process is to search for the financial performance data of the remaining 28 companies. This data was collected via the AMADEUS

(32)

32 of 20 million European companies over the last ten years. After obtaining the ticker symbols for the companies in the sample the AMADEUS database provided the financial data for 2011-2013 for 21 companies. Since the data for 2014 was not yet available this data needed to be manually collected from the annual reports of the companies, these were downloaded from their websites. The same procedure was applied for missing values in the AMADEUS data. For the remaining companies the annual reports of all four years needed to be collected in order to obtain the data. For most variables the calculation was straightforward since the underlying figures, such as net income or total assets, were stated in the financial statements. An exemption was the calculation of long term debt which was needed to calculate the control variable risk. Since this variable was often not directly traceable from the financial statements the amount was calculated by using the same elements as were used in the amount for the previous year that was available from

AMADEUS. If this was not possible either, because previous year data was not available, the following elements were included:

Long term debt = loans and borrowings + tax liabilities + employee benefits.

The reason behind including these specific elements is that these were the ones that were usually included in the long term debt amount provided by AMADEUS. Including the same elements contributes to the comparability and reliability of the figures.

For three companies the annual reports for all or the last year were not provided. These three companies were excluded from the sample. This finally results in a sample of 25 companies with the following graphic dispersion:

Geographic dispersion of the sample

Country Number of companies Percentage Denmark 2 8% France 5 20% Germany 7 28% Italy 1 4% The Netherlands 3 12% Spain 1 4% Sweden 3 12% Switzerland 2 8% United Kingdom 1 4% Total 25 100%

Corporate Social Responsibility and financial performance : in the European context