Cultural systems and determinants of Corporate Social Responsibility

Cultural systems and determinants of Corporate Social Responsibility

1

Author:

Reinier Kamermans Student Number: 1875213 Phone Number: +31 (0)6 496 189 81

Prinsesseweg 11 9717 BA Groningen reinier.kamermans@gmail.com

Abstract: This study investigates the determinants of the corporate social responsibility. Determinants that are measured are time, economic firm performance and past controversies. The influence the cultural system of a country has on these determinants is investigated by looking at Hofstede’s (1997) individualism. The relation between the determinants and CSR is derived from institutional theory, legitimacy theory, slack resources theory and good management theory. I find an autonomous growth of CSR over time and that economic firm performance is significantly correlated with the CSR score of a company. I also find a positive significant relation between past controversies and CSR score. Individualism has a positive moderating effect on the relations between time and CSR and economic firm performance and CSR. These findings confirm the link between the different theories and CSR. The results vary significantly if the US is left out of the sample. The data used originates from the ASSET4 ESG dataset from Thomson Reuters, BoardEx and DataStream with a sample containing 1395 unique listed firms from 41 countries from the years 2002 till 2012.

Supervisor:

dr. R.B.H. Hooghiemstra

Second assessor:

Paul Klaassen

January 13

, 2014 Word count: 13.443

University of Groningen Faculty of Economics and Business

MSc. Accountancy

Contents

1. ! Introduction 4 !

1.1. ! Research goal 4 !

1.2. ! Research questions 5 !

1.2.1. ! Research main question 5 !

1.2.2. ! Research sub questions 5 !

1.3. ! Research plan 5 !

1.4. ! Scientific relevance 6 !

1.5. ! Structure of thesis 7 !

2. ! Theoretical Framework 8 !

2.1. ! Corporate social responsibility 8 !

2.2. ! Institutional theory 8 !

2.3. ! Legitimacy theory 10 !

2.4. ! Slack resource theory 11 !

2.5. ! Good Management Theory 12 !

2.6. ! Hypothesis 12 !

2.6.1. ! CSR and time 12 !

2.6.2. ! CSR and economic firm performance 13 !

2.6.3. ! CSR and past social & environmental controversies 14 !

2.6.4. ! CSR and country level individualism 14 !

3. ! Research methods 17 !

3.1. ! Sample and data source 17 !

3.2. ! Dependent variables 17 !

3.2.1. ! Corporate social responsibility 17 !

3.3. ! Independent variables 18 !

3.3.1. ! Time 18 !

3.3.2. ! Economic firm performance 18 !

3.3.3. ! Social & Environmental Controversies 18 !

3.3.4. ! Individualism 18 !

3.4. ! Control variables 19 !

3.4.1. ! Free Cash Flow 19 !

3.4.2. ! Firm size 19 !

3.4.3. ! Sector dummy 19 !

3.4.4. ! Board characteristics 19 !

4. ! ^{Results 21} !

4.1. ! Sample 21 !

4.2. ! Multicollinearity 22 !

4.3. ! Hypotheses testing 22 !

4.3.1. ! Hypothesis 1: CSR and time 25 !

4.3.2. ! Hypothesis 2: CSR and economic firm performance 25 !

4.3.3. ! Hypothesis 3: CSR and past social & environmental controversies 25 !

4.3.4. ! Hypothesis 4 till 6: CSR and country level individualism 26 !

4.4. ! Robustness 27 !

4.4.1. ! Split environmental and social score 27 !

4.4.2. ! Without US 27 !

5. ! Conclusion and discussion 28 !

6. ! References 30 !

1. Introduction

In recent years corporate social responsibility (CSR) has become a popular subject in both academic literature (e.g. McWilliams & Siegel, 2001; Margolis & Walsh, 2003; Orlitzky, Schmidt & Rynes, 2003) and business. Non-governmental organizations (NGOs) are publishing reports grading companies on their sustainability goals. One of these recent reports was from Oxfam and was not very satisfying for most companies. None of the monitored companies scored an overall good score. This bad score can, in a worst case scenario develop into consumer actions, such as boycotts from consumers (The Guardian 26

February 2013)

. Scoring badly on CSR has developed to become a potential threat to companies as we can see in some recent examples; Starbucks in the UK had a completely legal tax avoidance technique which it had to abandon once a public disturbance arose. (The Week 24

June 2013)

. The tax evasion of Starbucks gave it a lot of negative media attention. This damaged the image of the company and resulted in a reduction in costumers because of a boycott from British people. In a recent Report from Cone Communications / ECHO 90% of consumers said they would boycott a product if they learned of the producing company’s irresponsible or deceptive business practices. The same survey showed that 55% of consumers boycotted a company’s products or services upon having learned it behaved irresponsibly in the last 12 months

. Apart from the potential threats and or real adverse effects companies can also gain a competitive advantage from CSR through for example a good corporate image. A few examples of companies that profit from CSR are the Body Shop or DSM.

1.1. Research goal

There are a number of theories that help explain the way companies act on CSR. First of all is legitimacy theory; this theory has been used to explain CSR activities (Deegan 2002).

According to this theory CSR activities are used to give the organization a permission to exist in society. To earn this right a company has to please all of its stakeholders and thus be socially responsible. Another theory is instrumental stakeholders theory (Clarkson 1995;

Cornell and Shapiro 1987; Donaldson and Preston 1995; Freeman 1984; Mitchell et al. 1997).

Companies expect their stakeholders to be satisfied if they undertake certain CSR activities.

Via this way we could explain why companies act the way they do. This theory suggests that companies don’t randomly undertake CSR activities but do this in order to achieve a desired result. Freeman and Evan (1990) use legitimacy theory to explain the practice of balancing and addressing claims of the stakeholders. In this way managers can more efficiently adapt their organization to external demands and thus strive for the maximum amount of corporate financial performance. Another theory regarding CSR activities is stakeholder-agency theory (Hill and Jones 1992; Jones 1995) this theory can give a reason for why companies try to mislead their stakeholders by making information extra complicated and less transparent and thus making it impossible to see if the company is fulfilling its CSR obligation to society. Via this way managers may try to get approval from society while it doesn’t really conduct CSR activities. All of the previously mentioned theories suggest that (the facade of) CSR activities are essential for companies to stay relevant to society and survive. There are also theories that suggest that CSR activities don’t exist to survive but are developed after corporate financial prosperity. In these theories CSR is considered non-critical for survival and is only created when all critical things are taken care of. Mcguire et al. (1988) find that CSR activities depend on excess funds and since most and thus a higher amount of CSR activities can be expected after profitable years. One could say that to start off with CSR activities, companies need an

1

http://www.theguardian.com/sustainable-business/oxfam-multinational-companies-failing-csr

2

http://www.theweek.co.uk/companies/53781/starbucks-pay-tax-first-time-five-years

3

http://www.conecomm.com/stuff/contentmgr/files/0/fdf8ac4a95f78de426c2cb117656b846/files/20

initial investment. The lack of resources can repress the development of CSR according to slack resource theory (Ullmann 1985; Waddock and Graves 1997). On the other hand Waddock and Graves (1997) also suggest an alternative possibility called the ‘good management theory’. This theory suggests that by participating in CSR activities companies satisfy their stakeholders and thus gain an advantage over their competitors. This in turn will lead to extra profit. Apart from these different theories on what forces affect CSR activities, in what temporal order, one can argue that the determinants of CSR have different effects in different countries. Country specific factors can influence the relations, this can be explained by institutional theory (Huntington 1969) which says companies are subjected to their surrounding and are a result of it. Ioannou & Serafeim (2012) suggest there are four different areas in which country differences influence CSR activities of companies. These areas are political system, education and labor system, financial system and cultural system. They find several statistically significant relations regarding these systems on companies CSP. As one can see a lot of theories can be used to explain why a company takes part in CSR activities.

Given the nature and reach of this research not all theories can be used. A selection of the theories mentioned above will be reviewed and taken along in deriving the hypotheses.

1.2. Research questions 1.2.1. Research main question

Given the above theories and previous research the following research question surfaces:

- ‘How can the different determinants of corporate social responsibility between companies be explained across country borders’

1.2.2. Research sub questions

From the main question and the above theories mentioned the following sub questions are formed:

- ‘Does CSR autonomously develop over time and how can this be explained.’

- ‘In what order do corporate financial performance and CSR influence each other and how can this be explained’

- ‘What influence do controversies have on the CSR of a company and how can this be explained.’

- ‘Are determinants of CSR influenced by cultural system and how can this be explained’

1.3. Research plan

To empirically examine the research question a set of variables, originating from the databases “DataStream’, ‘ASSETS 4 ESG’ and ‘Hofstede’s individualism index’, will be used. The first variable is ‘time’, over time the public has become more demanding of companies regarding CSR and NGOs have become better at tracking companies weather they actually fulfill their obligations to society. Therefore I would expect an autonomous growth of CSR activities over time.

The second set of variables is previous corporate financial performance and next year corporate financial performance. I would expect this variable to interact with CSR activities given that four different theories suggest an interaction. Slack resource theory (Ullmann 1985;

Waddock and Graves 1997) argues that CSR activities would follow corporate financial

performance since an amount of excess resources is needed to participate in any other

activities than core business activities. In slack resource theory excess funds theory (McGuire

et al. 1988) is sort of included since it is very similar to slack resource theory only it is less

broad since only excess funds are considered of importance. Another theory, good

management theory (Waddock and Graves 1997), suggests that CSR activities are deployed

by management since it is the right thing to do. This would mean that CSR activities and

corporate financial performance would occur at the same time. Since management immediately seizes the opportunity to deploy CSR activities as soon as the opportunity arises CSR and corporate financial performance occur together. Legitimacy theory (Deegan 2002) says that companies need to be given the right by society to exist. If the company loses this right from society the company can no longer obtain critical resources from society and goes bankrupt. Thus, legitimacy theory would suggest that corporate financial performance follows CSR activities. If a company has better CSR more resources are granted to it by society and thus it can perform better financially. These four theories have a different suggestion of the temporal relation between financial performance and CSR activities. There has been much discussion on the direction of the relationship and thus is it worthwhile do research this topic again. By comparing the previous year and future year financial results from a company to its CSR score during the current year, a correlation should surface with the strongest relation and thus giving more clarity on the direction of influence.

The third variable will be controversies. Some companies are better watched than others and thus be more sensitive to actions from consumers such as boycotts. A reason for this can be past controversies. These companies have to be on their toes since small adverse events can be magnified by the intense monitoring from the public. These better watched companies are the more eager to prevent negative publicity. The better the company is known by consumers the better the quality of its CSR activities will be (Du & Vieira 2012). A higher number of past controversies would therefore be expected to be related to a better CSR.

Legitimacy theory can explain that the company is striving for it’s right to exist in society (Deegan 2002). This right to exist has been damaged by the recent controversies and thus needs to be repaired.

The fourth variable will be cultural system as a country specific factor as a moderator.

All the variables mentioned above might interact different across countries due to of cultural, political or educational system differences as found by Ioannou and Serafeim (2012). In their study they research the effect of country specific factors on the corporate social performance of companies. Their finding is backed by institutional theory (Huntington, 1969) which claims that companies are embedded in a nexus of formal and informal rules (North, 1990). This means that in different countries, different formal and informal rules make companies act in different ways.

1.4. Scientific relevance

Over time the attention given to failure to comply with CSR has risen as is shown by the research of Lee & Carroll (2011). With NGOs closer monitoring and politicians having more attention for CSR an autonomous rise in CSR activities over time could be expected. If this link is found an extra part of the explanation regarding the question; ‘what determines the level of CSR’ is explained. When more of the determinants of CSR are explained we can get a better understanding of all influencers of CSR. When more influencers are known theories can be tested more rigidly since the importance of different forces becomes clear.

The link between corporate financial performance and CSR has been found (Ullmann

1985; Waddock and Graves 1997; Orlitzky, Schmidt & Rynes, 2003) but in these same papers

discussion remains about the direction of the relationship. Slack resource theory (Ullmann

1985; Waddock and Graves 1997), Excess funds (McGuire et al. 1988), Good management

theory (Waddock and Graves 1997) and Legitimacy theory (Deegan 2002) all suggest

different temporal order of the relationship between corporate financial performance and CSR

activities. In the past different directions have been researched and found but the results are

not consistent. In this paper both past and future economic performance will be used and

therefore hopefully a more conclusive result will surface on the direction of the relationship

between corporate financial performance and CSR. These findings are relevant because this

will elaborate on which theory best describes the process of CSR activities coming in to place.

A better understanding of the way managers make decisions can be obtained and possibly an indication on how crucial CSR is to companies.

Controversies in specific sectors have been researched to establish a relation with increased CSR (Du & Vieira 2012). A research without taking sectors into account, regarding controversies and it’s link to increased CSR is new. This variable is closely linked to legitimacy theory since the companies that recently have been in controversies have lost at least some of their right to exist. By conducting a lot of CSR activities the company can try to regain its right to exist and regain access to crucial resources. Islam & Deegan (2010) find in their study that two companies that have recently been subjected to negative press because of bad CSR now have better CSR practices on the criticized topics. Legitimacy theory is still an under developed theory according to the paper of Deegan (2002). If this link is found an extra argument can be made to confirm the existence of the legitimacy theory. If found statistically significant this will also contribute to the understanding of the determinants of CSR.

Most prior studies are single country studies (McWilliams & Siegel, 2001; Margolis

& Walsh, 2003; Orlitzky, Schmidt & Rynes, 2003). A number of studies regarding the effects off cross-country variables on CSR have been conducted before, most of them found significant differences among countries but remain unclear about what variable determines the difference (Maignan and Ralston 2002; Chapple and Moon 2005). Aguilera et al. (2007 p836) argue that “because business organizations are embedded in different national systems, they will experience divergent degrees of internal and external pressures to engage in social responsibility initiatives”. These different national systems have been split up into cultural, political or educational system differences by Ioannou and Serafeim (2012). They find relations between all three of these systems and CSR. The cultural system may be the reason to explain why managers react different to firm economic performance and controversies in relation to CSR. The expectation of the public will depend on cultural system and this can be linked to both legitimacy and institutional theory. Consumers may expect companies to share their profit with more shareholders depending on cultural system. Also, managers may react different under different cultural system as the different unwritten rules oblige them to take part in CSR activities following changing times, firm economic performance or controversies which would advocate institutional theory. Although Ioannou and Serafeim (2012) have also used cultural system, this study does bring something new to the table because culture is being researched as a moderator. If the moderator link is found with any of the previously mentioned variables then a lot of inconclusive results and opposing findings in previous studies may be explained. A further understanding of both institutional and legitimacy theory may follow. Both towards the theory as a whole and the theories regarding cross-national issues.

1.5. Structure of thesis

This paper will continue with the theoretical framework where the main theories used in this

research will be explained. The main theories used are legitimacy theory, institutional theory

and slack resource theory. Based on these theories hypothesis will be formed. Next is the

research method section where all the variables will be made measurable and the used dataset,

ASSET4 by Thomson Reuters, will be explained. In the result section the statistical results

regarding the interaction of the variables will be presented and analyzed. The conclusion and

discussion section will present the limitations of the findings and suggest further research.

2. Theoretical Framework

In this chapter the most important theories that will be used are discussed. Hypothesis will be formed based upon in the second part of this chapter. The reasoning in this study is based on three different theories; Legitimacy theory (Suchman, 1995), Institutional theory (Huntington, 1969) and Slack resource theory (Ullmann, 1985). First the concept of social responsibility will be explained since a unanimous definition still does not exist. Second the different theories will be briefly explained in relation to CSR. Hypothesis will be formed later on in this chapter.

2.1. Corporate social responsibility

In order to find determinants of CSR we first need to identify what CSR is and what definition will be used. One of the first important papers on CSR was the essay from Friedman (1970). Friedman claimed that companies cannot have responsibilities but only people can. Therefore companies should not spend money on CSR but the shareholders and employees should with the proceeds from their stock return and wages if they want to. Later on, the tone gradually shifted and CSR became more than just a hobby of managers and actually became of strategic relevance.

Then the question arises when does something qualify as CSR? Campbell (2007) explains that social responsibility depends on point of view. Should one mainly look at the wellbeing of the workers such as, wage, safety job security or should one look mainly at the environmental impact of the company? Should only objective criteria be selected or are there some subjective criteria also? It is also notable that the public opinion on what CSR consists of changes over time. Campbell (2007) points out: in the industrial revolution a reduction from a 14 hour workday for factory workers to a 10 hour workday was considered as CSR where today 10 hours would be outrageous. The development of the conceptualization of CSR over time has been described by Brown: (2008, p3.) “CSR arose, at least initially, not as a model example of organizational proactivity, but rather as reactions to crises. CSR was triggered by the cratering of public opinion—the declining perception of major corporations, among a whole range of other economically, socially, culturally and politically dominant institutions.” A number of researchers agreed to consider CSR as the following to prevent differences between different researches: “CSR is actions on the part of the firm that appear to advance, or acquiesce in the promotion of some social good, beyond the immediate interests of the firm and its shareholders and beyond that which is required by law.” (Wolfe and Aupperle, 1991; Waddock and Graves, 1997; Hillman and Keim, 2001; Waldman, Siegel and Javidan, 2006).

As a result of this in this paper a combination of social and environmental responsible factors will be used to measure the CSR. Corporate social responsibility is influenced by different factors. Some factors will be researched in this paper and therefore need theoretical backing. Below the theories that will be used to explain the way the variables influence corporate social responsibility will be presented.

2.2. Institutional theory

Institutional theory is a theory on how social beliefs, values, constraints, relations and

expectations lead to a complex of formal and informal rules in which companies are

embedded (North 1990; Crossland & Hambrick 2011). The previously mentioned contributors

to the formal and informal rules are called institutions within institutional theory. These

institutions have been defined by Huntington as: “stable, valued, recurring patterns of

behavior,” defined by their “adaptability, complexity, autonomy, and coherence” (Huntington

1969: p12). The behavior of a company is dependent on the institutions in place. Companies

will build their strategy on the institutions around them and if this requires adoption or

implementation of CSR they will deploy these (Jackson & Apostolakou, 2010).

According to institutional theory organizations will behave different if the institutions are different and will adapt to the external expectations (DiMaggio & Powell 1983). As the surrounding of a company has a certain structure regarding for example governance it is most likely that a company will feel pressure to also change its governance structure and comply with the institutions which are the result of the environment (Deegan 2002).

Some scholars claim that CSR is the result of the institutions around companies. The national business system and cultural system influences the roles of different stakeholders and the influence they have on companies (Matten & Moon, 2008; Jackson & Apostolakou, 2010). The differences between stakeholders in different countries means that they have to be addressed in different ways. The CSR actions that might have positive effect in a certain set of countries might have a totally different outcome in another set of countries. Assuming that companies strive for profit, the effectiveness of CSR and thus the decision of companies to practice it is greatly dependent on the institutional structures in a country.

In the past studies have shown that institutions differ a lot across countries. Whitley (1999) has divided these differences in different national business systems which are based on

“distinctive patterns of economic organization that vary in their degree and mode of authoritative coordination of economic activities, and in the organization of, and interconnections between, owners, managers, experts, and other employees” (Whitley, 1999, p.33). Ioannou & Serafeim (2012) recognize Whitley’s framework as a way to account for the key role of the cultural system and as relevant for empirically researching the influence of culture on CSR. In Whitley’s framework “business systems” are grouped based on a few key institutions within a national business system. By this a comparison between different countries can be made which is necessary do discover nationwide factors.

Matten & Moon (2008) give an example of the difference between CSR in Europe and the U.S.. They say that the U.S. has developed a system where wealthy businessman and corporations give back to society through big charitable donations whereas in Europe a system of public organizations has been put in place to ensure the giving back of wealthy people and businesses. From this example can be seen that most likely the CSR in these two regions is very different. Whereas in the U.S. companies will be expected to participate in a lot of voluntary CSR activities, companies in Europe can get by by mainly just obeying the law. The European people expect less of the CSR of companies since other institutions have been formed to ensure a socially responsible environment.

Crossland & Hambrick (2011) studied these cultural differences on a more micro level and find that a fundamental element of cultural systems is whether they are based on autonomous or consensus-based action. This can be translated to how individualistic

a society is. A very individualistic country will allow for displaying success and one’s contribution to society. In a very collectivistic country one would merely comply with the display of others and keep ones success to itself. If we look at these differences from a CSR perspective this would mean that countries with high levels of individualism would have more CSR as managers want to display the image of the company and be out there with “their”

success. In countries with a higher level of collectivism a big display of CSR and linking one’s name to it could be seen as self-promotion which is selfish and unacceptable in countries with a high level of collectivism. The research of Ioannou & Serafeim (2012) found

4

“Individualism can be defined as a preference for a loosely-knit social framework in which individuals are expected to take care of themselves and their immediate families only. Its opposite, Collectivism, represents a preference for a tightly-knit framework in society in which individuals can expect their relatives or members of a particular in-group to look after them in exchange for

unquestioning loyalty. A society's position on this dimension is reflected in whether people’s self-

image is defined in terms of “I” or “we.”” (Hofstede, 1997)

that countries with a high level of individualism scored higher on corporate social performance thus supporting above reasoning.

2.3. Legitimacy theory

Legitimacy theory is a systems-oriented theory which means that a key aspect is that an entity is influenced by society it operates in but also influences the society it operates in (Gray et al.

1996; Deegan 2002). The legitimacy theory is defined by Suchman (1995, p574) as follows:

“Legitimacy is a generalized perception or assumption that the actions of an entity are desirable, proper, or appropriate within some socially constructed system of norms, values, beliefs, and definitions.” According to legitimacy theory companies do not have any inherent right to resources or even to exist. Organizations only exists because society grants them legitimacy (Mathews 1993; Deegan 2002).

In today’s society companies will have to comply to certain expectations of stakeholders and agree to social contracts, to be socially acceptable (Maignan and Ferrell 2004; Peloza and Shang 2011). Companies have to oblige to these social in order to maintain their right to certain resources the company needs to survive. If society feels that the company is not acting in an acceptable, legitimate manner they will revoke its social contract. As a result of the revoked social contract, consumers may reduce the use or completely abandon the company’s products. Also labor could no longer be available to the company because people are no longer willing to work for this “wrong” company. The financial capital the company requires can become harder to acquire so a premium has to be paid. A lobby movement to convince the government to enforce extra taxes, fines or new laws can arise.

These are all serious threats to the existence of any company that do not conform with the expectations of society (Deegan, 2002).

The social contract exist of two kinds of terms. The legislated terms and the non- legislated terms. Little discussion exists about the legislative terms since they usually clearly state what is acceptable and what is not. The non-legislative terms on the other hand are based on perceptions of the manager and thus might vary a lot between firms (Gray et al. 1996). In different social environments different actions are required from companies to meet the expectation of stakeholders. Fombrum (2005) showed that the requirements of CSR differ per country through different criteria for CSR, CSR reports, prizes and awards and CSR guidelines and regulations.

Legitimacy can be considered a resource by management and in turn be treated as one.

If management thinks this resource is vital to organizational survival than legitimacy theory says they will try to secure a continuous supply of it (Pfeffer and Salancik 1978). From this the activities in relation to CSR can be explained.

Obtaining and maintaining legitimacy does not necessarily lead to good financial performance but the loss of legitimacy could lead to organizational failure because the resources needed by the company are no longer granted by its environment. CSR can be seen as the way a company acts to fulfill the obligations it has to society.

There are several reasons why a company may be seen as not being legitimate. One

reason can be that the expectations of society have changed while the company didn’t change

with it. An example of this is fair-trade chocolate. 10 years ago very few food producers used

fair trade cacao. Over the years most suppliers changed to fair-trade and the ones that did not

were publicly burdened in the newspapers. As Lindblom (1994) explains: legitimacy is a

dynamic concept that changes over time. Another reason why a company might no longer be

perceived as legitimate can be because of a certain event that damaged the reputation or

legitimacy of a company or its industry (Patten 1992). The events one can think of are for

example adverse social or environmental impacts. Examples of this are a huge oil spill like

BP had in the Caribbean or Apple’s factory workers condition in China which became public.

Depending on a manager’s beliefs on non-legislated terms his actions to the adverse event may differ. As a reaction to adverse events managers that believe that the non-legislated terms of the social contract have been breached will undertake action. One of the options they have is to change their bad practices and clearly communicate this to the public (Pfeffer and Salancik 1978; Lindblom 1994; Deegan 2002). The study of Islam & Deegan (2010) gives an example of how past bad practices lead to improved CSR. The clothing companies Nike and Hennes & Mauritz, that have recently been subjected to negative press because of adverse social and environmental events. Their research shows that these companies now have better CSR practices on the criticized topics and therefore an improvement was made as a result of pas controversies.

2.4. Slack resource theory

Slack resources are according to theory defined as spare or uncommitted resources, a cushion of resources beyond the minimum necessary to maintain the organizational coalition (Cyert &

March, 1963), or excess resources beyond those needed to produce a given level of output (Nohria & Gulati, 1996). In previous studies a variety of things has been considered slack resources (e.g. unused machine capacity, extra raw materials or labor or excess work in progress inventory) but the most common resource in the slack resources theory is excess cash (Sharfman et al 1988). Seifert et al. (2004) advocate that in order to search for a financial performance-social performance link cash flow is much more relevant than accounting returns since cash flows allow for spending while accounting returns may have problems regarding liquidity once they start giving to charity.

The slack resources theory suggests that the availability of slack resources create the ability to take part in CSR by spending some of these resources on something CSR related. If there are no slack resources there is nothing to spend on CSR. These resources can both be monetary and non-monetary. Examples of non-monetary goods that can be used for CSR include finished goods, use of facilities, services, managerial expertise or offer employees time for volunteer work. The research of Brammer & Millington (2002) found evidence that the availability of inventory and labor were positively related to corporate community involvement.

In the U.S. the law allows companies to give either directly to charities or through corporate sponsored charitable foundations who in turn donate to charity with the same fiscal consequences. This gives the companies a big incentive to transfer money to their foundation while they perform well and thus have slack resources. This means that companies rather give a large amount of money all at one’s to charity rather than the same amount over several years (Thayer, 2003). Since the practices of their foundation is considered CSR, giving to these foundations is considered CSR. Because large sums are transferred to these foundations in financially positive years slack resource theory surfaces here (Seifert et al 2004).

Some other evidence on slack resource theory comes from the study of Buchholtz et al. (1999). They find a relation between the perceived organizational slack of a company by its CEO and the donations given to charity among medium-sized firms in two industries.

Other studies have used publicly available financial data instead of managers perception to measure organizational slack. Navarro (1988) found that the payout of dividends were positively related to charitable giving. Even though one would expect dividends to limit charitable giving since they have to be paid from slack resources they can apparently co-exist.

This means that the slack resources that are available for extra dividends can also be partly used for CSR activities.

Regarding the temporal sequence the slack resource theory is very clear. CSR follows

corporate financial performance (Seifert et al 2004). This is backed up by several studies that

found that financial performance was a better predictor of future CSR activities than that CSR

was a predictor of future financial performance (McGuire et al. 1988; Preston & O’Bannon, 1997; Waddock & Graves, 1997). The authors gave as an explanation that the firms had more slack resources which is in line with slack resource theory.

In slack resource theory financial performance is not the initiator. Financial performance is considered to be a reason slack resources can exist but does not necessarily lead to slack resources (Sharfman et al., 1988; Singh, 1986). As a result of this Buchholtz et al. (1999) criticize the use of accounting returns as indicator for slack resources. They claim that even though a firm has high accounting returns it’s free cash flow can be low and thus have little slack resources available. They therefore advocate to take cash flow as the measure for slack resources available rather than accounting returns.

2.5. Good Management Theory

According to ‘good management theory’ managers don’t use CSR activities as an instrument but they use CSR activities to do the right thing. According to good management theory improved CSR results in better relations with important stakeholders (Freeman 1984). A few examples of this are good employee relations which will lead to more productive workers, less strikes and easier hiring of good new employees, a good community relation which might lead to tax breaks or more cooperating (local) governments. In the long run good management can also lead to better schooled workers and thus a more skilled workforce (Waddock and Graves 1997).

The image of being socially and environmentally responsible can also lead to a competitive advantage as consumers are becoming more aware of durable products as pointed out by the introduction of this paper points out and previous research (Prahalad and Hamel, 1994). Previous research from McGuire et al. (1988) finds this link between good management and subsequent financial performance.

2.6. Hypothesis

In this section the hypotheses regarding the interaction of the independent variables and CSR will be formed based on theories presented above, prior literature and reasoning.

2.6.1. CSR and time

The institutions in society are based on a set of social beliefs, values, constraints, relations and expectations according to institutional theory. The concept of institutions is dynamic, which means that they can change over time (North 1990; Crossland & Hambrick 2011). This means that CSR can change also a lot over time since it is determined by the institutions. Over the years the amount of CSR activities has risen a lot. Legitimacy theory states that the social contracts companies have with society change over time. CSR is a result of the social contracts companies have with society. As the social contracts change a change in CSR will likely follow. Therefore both institutional theory and legitimacy theory expect a change in CSR over time.

The findings of Campbell (2007) indicate that the expectations of society on CSR has increased over time. As an example he gives the work hour reduction from the industrial age from 14 to 10 hours a day. Back then this was seen as corporate social responsibility but nowadays this would still be an unacceptable workday. Even though this example may be a little extreme the point that CSR changes over time is clear.

In another research Lee & Carroll (2011) also find that CSR attention changes a lot over time. Their sample consisted of articles form U.S. newspapers over a 25-year period.

They measured CSR as whether a company was mentioned in relation to one of the four CSR dimensions in a positive or negative matter. They distinguish between different kinds of CSR;

Economic responsibility, Legal responsibility, Ethical responsibility and Philanthropic

responsibility. Even though the different kinds of CSR all spiked at different periods in the period 1980 till 2004 overall there was a growth.

From these two prior studies the direction becomes clear: Over time society starts to expect more from the same companies and thus in order to stay legitimate companies have to increase their CSR activities. This leads to the first directional hypothesis.

H1: There is a positive relation between time and the level of CSP 2.6.2. CSR and economic firm performance

Numerous theories have found a relation between CSR and economic firm performance (Ullmann 1985; Waddock and Graves 1997; Orlitzky, Schmidt & Rynes, 2003). The part that is still unclear is the direction of the relation. There are several theories that can be used here.

One of the theories that can explain the direction of the relation is Slack resource theory. The resources that are available can according to slack resource theory be used for other than core business activities. CSR activities come after the purely core business activities and thus a relation between the availability of funds and CSR activities should be likely. Logically a link between firm financial performance and availability of funds also exists. As support for this theory Buchholtz et al. (1999) found that companies decide to spend resources on CSR when they think they have enough resources to spare. This was researched through a questionnaire for the CEO of mid-size companies about the amount of slack resources in the organization. A significant correlation was found between the perception of the CEO on the amount of slack resources and the amount of money spent on corporate philanthropy. This would suggest that CSR follows firm financial performance. In another study Seifert et al. (2004) find that corporations transfer large amounts of money to their charity entities in years of good economic firm performance. Since the transfer happens when the money is available it is expected that CSR follows firm economic performance based on their findings.

Another theory that can be used to explain the direction of the relation between economic performance and CSR is good management theory. This theory assumes that managers undertake CSR because it is the right thing to do and as a result society will reward the company for being responsible. This theory would suggest that undertaking CSR activities will result in better economic performance.

Waddock & Graves (1997) find that firms that have done well financially often spend some of this money on CSR. As a contrast to this they also find some evidence that promotes the opposite, This would suggest that Good management theory is also in place.

Orlitzky et al. (2003) find in their meta-analysis about the effects of management’s discretion and values on the link between CFP and corporate philanthropy that resource level is indeed important for the level of corporate philanthropy when isolated. As a possible explanation they say that the large amount of slack is the result of past good performance which gives the management more freedom from the supervisory board. This freedom is then used to spend resources on corporate philanthropy.

Because there are two relevant theories and the prior research is inconclusive about the direction of the relation but certain about a relation two directional hypothesis is formed.

H2a: There is a positive relation between previous year economic firm performance and the level of CSP

H2b: There is a positive relation between next year economic firm performance and

the level of CSP

2.6.3. CSR and past social & environmental controversies

When social or environmental controversies happen the legitimacy of a company is threatened. According to the legitimacy theory managers have to try to regain this legitimacy or the existence of the company is threatened. In order to regain legitimacy a company has to show that it is fulfilling the social contract it has with society (Maignan and Ferrell 2004;

Peloza and Shang 2011). Managers can try to regain the acceptance of society by undertaking CSR activities. Therefore a positive relation between past controversies and CSR is expected from theory.

Prior research from Deegan (2002) shows that companies should have legitimacy from society in order to have access to crucial resources. If these crucial resources are revoked because the company is no longer perceived as legitimate the survival of the company is threatened. Therefore an action from managers is expected when legitimacy is reduced.

In the research of Patten (1992) events that can possibly damage the reputation and social contract is described. Such an event can be a social or environmental controversy. In order to regain legitimacy the company can take part in a CSR activities. In the research from Du & Vieira (2012) an example is given of two oil companies that have been involved in environmental controversies. After these controversies surfaced the CSR had risen substantially because the managers tried to regain the legitimacy of the company.

In the discussion of the paper from Patten (1992) Is stated that it appears that at least for environmental disclosures, threats to a firm's legitimacy do entice the firm to include more social responsibility information in its annual report. Therefore an increase in CSP can also be expected. The suggestion that environmental disclosures are used as a legitimizing tool is also found in the paper of Cho & Patten (2007). They used the KLD database and found the relation to exist amongst all corporations.

Islam and Deegan (2010) also find evidence in their case study on clothing and sports manufacturers that media attention to controversies leads to improvement of the specific subject. A rise in CSP as a result of controversies would therefore also be likely.

From the theory and previous research one can say that social or environmental controversies most likely have a positive relation with CSR. The following directional hypothesis arises.

H3: There is a positive relation between past social and environmental controversies and the level of CSP

2.6.4. CSR and country level individualism

To research the effect of a country’s cultural system a specific nationwide variable will be researched. Institutional theory says that companies are influenced by their surroundings and thus the cultural system leads to different behaviors of companies. Crossland & Hambrick (2011) find that the fundamental element of a country cultural system is whether it is based on a consensus based actions or autonomous actions. This can be translated to how individualistic a country is (Ioannou & Serafeim 2012).

Individualism might have an influence on the development of CSR in a country. The more individualistic a country is the more CSR is expected since the managers of the company want to display their success. Matten & Moon (2008) gave as an example of this the difference between the U.S. and Europe. In the U.S. successful businessman give back to society by charitable donations. The same can be expected for companies led by successful managers in the form of CSR.

With individualism as a moderating factor on the development of CSR over time, one

would expect a positive relation. CSR has received more attention over time (Lee & Carroll,

2011) and an individualistic manager looking to display his success would seize this

opportunity to get more publicity for his achievements. A stronger link between economic performance and CSR is also expected when individualism is high because self-promotion is more important with high individualism (Ionnou & Sarafeim 2012). The initial incentive for management to undertake a CSR activity regardless of timing will be increased because with high individualism managers also have an incentive to display their success. Legitimacy theory states that managers have to undertake action to prevent loss of legitimacy after adverse social or environmental events (Patten, 1992; Cho & Patten, 2007). When individualism is high managers will see the success and failure of a firm more as their own success and failure (Ionnou & Sarafeim 2012). This means that when individualism is high, the effect from social and environmental controversies on CSR is strengthened because managers have more incentive to change the image of the company. Another incentive for the managers to undertake action is to prevent regulatory action because this will limit their freedom and thus their personal power which is very important in individualistic countries (Hofstede, 1997). The division of Crossland & Hambrick (2011) in consensus based action and autonomous based action can also affect the way managers act when a social or environmental controversy arises. Since managers want to protect the image of “their” success a bigger rise in CSR is expected when the level of individualism is higher in a country (Ioannou & Serafeim 2012). As Hofstede (1997) states Individualism can be defined as a preference for a loosely-knit social framework in which individuals are expected to take care of themselves and their immediate families only. Therefore individual freedom is very important. As a result of this the laws of the country grant high individual freedom. Since there are little rules in indivualisitc countries, the higher individualism is the more one would expect rules to be formed if controversies do happen since a lot of freedom exists unless things go wrong. To prevent this managers will probably decide to undertake extra CSR activities so obligatory rules are not implemented by the government (Deegan 2002).

In their research Ioannou & Serafeim find a significant relation between Individualism and CSR. In the combination with time a positive moderating relation can therefore be expected. When a company is successful and has a lot of slack resources one would expect that individualism would make the managers use these resources for displaying their success and giving back (Ioannou & Serafeim 2012). When we look at the good management theory instead of the slack resource theory a positive moderating effect of individualism can also be expected. Campbell (2007) has found that CSR had become more important over time. This means that manager that are doing “the right thing”, according to good management theory, will have a competitive advantage because others want to be associated with this company in order to increase their own stats. The legitimacy theory combined with the status-seeking buyers leads to a positive moderating effect of individualism on the effect of economic performance on CSR.

To divide countries I follow Whitley (1999) who divides business systems based on how much their institutions differ. Ioannou & Serafeim (2012) also use this method introduced by Whitley and recognize it as a key role for cultural system. By taking individualism as indicator for the key difference between different cultural systems we continue to give a direction. The following directional moderating hypotheses are formed.

H4: Individualism strengthens the relation between time and the level of CSP

H5a: Individualism strengthens the relation between previous year economic firm

performance and the level of CSP

H5b: Individualism strengthens the relation between next year economic firm performance and the level of CSP

H6: Individualism strengthens the relation between social & environmental controversies and the level of CSP

Corporate Social Performance Time

Previous and Future Economic Firm

Performance

Social and environmental

controversies

Country level individualism H1

H4

H5

H6 H2

H3

3. Research methods 3.1. Sample and data source

Waddock and Graves (1997) already mentioned the measurement problem. In different industries different measures are often used. The consistency between different researchers can also cause measurement differences. As a reaction to this they decided to use the Kinder, Lydenberg, Domini (KLD) database to prevent the consistency problem between researchers.

The methodic of using professional researchers will be repeated here. The limitation of the KLD database is that only data from U.S. firms is included. Therefore this sample consists of a combination of different databases. The environmental, and social metrics come from Thomson Reuters ASSET4 ESG database. This database is specialized in providing objective, auditable, relevant and systematic environmental and social information and investment analysis tools for professional investors. An estimated €2.5 trillion assets are managed using the ASSET 4 data by combining the database with traditional investment analysis. This selection of data is comparable with the way the data sample of Ioannou & Serafeim (2012) was selected. The nation-level variables come from the paper of Hofstede (2001) where the individualism score from Hofstede (1997, 2001) per country is stated. The sample exists mostly of companies from the U.S., the U.K. and Japan but also a substantial amount of companies from Continental Europe, Australia, Hong Kong and Singapore. The sample consists of 5274 entries.

3.2. Dependent variables

3.2.1. Corporate social responsibility

The measurement for CSR is very difficult to choose. Often only a single part of CSR is choosen such as corporate philanthropy. This provides a very limited result because CSR is much wider than any single aspect (Lydenberg et al., 1986; Wolfe and Aupperle, 1991). The need for a multidimensional measure of CSR was also highlighted by Waddock and Graves (1997, p304) as they requested the “need for a multidimensional measure applied across a wide range of industries and larger samples of companies”. A wide range of different measures have been used in the past but in this paper the methodology used by Ioannou &

Serafeim (2012) will be used. The dataset used, global ESG dataset from Thomson Reuters ASSET4, is collected by specially trained research analysts. 900 point per firm are collected from publicly available data. After the data is collected the data is transformed into consistent units to enable quantitative analysis. Thomson Reuters says that every data point has multi- step verification and process control to ensure high quality of data. From the 900 data point 250 key performance indicators (KPIs) are formed. Out of these 250 KPIs 18 categories are formed within 4 pillars. The pillars are: environmental performance score, social performance score, corporate governance score and economic performance score. A z-score is formed for each of the pillars, this creates a benchmark of its performance compared to the rest of the firms.

The first pillars used will be the Environmental score from ASSET 4 Data index which is described as follows: The environmental pillar measures a company's impact on living and non-living natural systems, including the air, land and water, as well as complete ecosystems.

It reflects how well a company uses best management practices to avoid environmental risks and capitalize on environmental opportunities in order to generate long-term shareholder value. The second pillar used is Social score which is described as follows: The social pillar measures a company's capacity to generate trust and loyalty with its workforce, customers and society, through its use of best management practices. It is a reflection of the company's reputation and the health of its license to operate, which are key factors in determining its ability to generate long term shareholder value. This way of measuring is the same as Ioannou

& Serafeim (2012)

CSR will be measured as a combination of two pillars in het ASSET 4 Data index the same as Ioannou & Serafeim (2012) did in their research. Their decision of taking equal weight is based on the convention established by Waddock and Graves (1997), Hillman and Keim (2001) and Waldman, Siegel and Javidan (2006) among others who used the KLD dataset, in constructing a composite CSP index by assigning equal importance (and thus equal weights) to each of the two pillars.

3.3. Independent variables 3.3.1. Time

Time is measured as a number starting at 0 for the first year of entry +1 one for each next year continuing till the newest entry. This is a linear scale which is used to show the influence of the advancing time on CSR. The first year of entry will be t = 0. From then on the count will start till the last year of entry.

3.3.2. Economic firm performance

The method that was used by Waddock & Graves (1997) with ROA ROE and ROS will be used here. Only return on assets will be used with the methodology of Waddock & Graves (1997). Both the relations between past and future economic performance is tested the return on assets and free cash flows scaled to total assets from t-1 and t+1 will be used compared to fixed set of CSR in time.

3.3.3. Social & Environmental Controversies

Social & Environmental controversies are recorded as one of the data points in the ASSET 4 ESG dataset. For a total of 16 different types of controversies will be tested. The number of media covered controversies will be counted and used as data entry. The different controversies counted will consist of the following: Emission Reduction/Biodiversity Controversies, Emission Reduction/Spills and Pollution Controversies, Product Innovation/Product Impact Controversies, Resource Reduction/Environmental Resource Impact Controversies, Community/Critical Countries - Indigenous People Controversies, Community/Public Health Controversies, Community/Bribery, Corruption and Fraud Controversies, Diversity and Opportunity/Diversity Controversies, Employment Quality/Wages or Working Condition Controversies, Human Rights/Freedom of Association Controversies, Human Rights/Child Labour Controversies, Human Rights/Human Rights Controversies, Health & Safety/Health & Safety Controversies, Product Responsibility/Social Exclusion Controversies, Product Responsibility/Customer Controversies. The controversies independent variable will be assembled in two different ways. As the natural log of the number of controversies, Ln (#controversies). And as a Dummy where either the value 0 is given when no controversies took place or the number 1 if one or more controversies have happened. These methodic are used because the difference between one controversy compared to no controversies is much bigger than the difference between 1 or more controversies.

3.3.4. Individualism

The cultural system will be characterized by a well-established measure of cultural difference identified by Hofstede (1997, 2001): Individualism. Individualism is identified as follows:

"The degree to which individuals are integrated into groups". Individualism has been recorded

since 1973. The individualism score for the country will be linked to the company. The

country that is assigned to the company depends on the location of its headquarters.

3.4. Control variables 3.4.1. Free Cash Flow

In this study free cash flow scaled to total assets will be used as a control variable. FCF is measured as cash flow from operating activities minus common and preferred dividends scaled by total assets (Lang et al. 1991; Core and Guay 1999).

3.4.2. Firm size

The size of a firm can be measured in many different ways but I will take total assets just like Seifert et al. (2004). The information for this will be extracted from DataStream. Size as a control variable makes sense because of the findings of de Villiers et al. (2011) found. They found that larger firms are more likely to identify environmental issues as a separate management priority and manage it effectively (Al-Tuwaijri et al., 2004; Clarkson, Li, Richardson, & Vasvari, 2008; McKendall et al., 1999). This may lead to a relation between size and CSR score. The natural log of total assets will be used in the quantitative analysis just like Villiers et al. (2011) have.

3.4.3. Sector dummy

Some industries are environmentally more sensitive than others which might lead to more disclosure (Cho & Patten, 2007; Halme & Huse, 1997; Patten, 2002). Therefore a sector dummy will be taken along as a control variable. The methodology of Fama and French (1997) and Villiers et al. (2011) is used by taking SIC codes to identify the industries. The results will be controlled for sectors. For the division into portfolios the methodology of Fama and French (1997) with 17 sectors will be used.

3.4.4. Board characteristics

de Villiers et al.(2011) find that Board independence and board size is of influence on CSR.

Board independence is thought to increase CSR because independent board members also

take their own reputation into consideration when making decisions on CSR. The independent

board members have to think about their future employment (de Villiers et al. 2011) . Board

size is thought to increase CSR because large boards have better access to financial resources

which are needed to deploy CSR activities (de Villiers et al. 2011) . Therefore these factors

shall be used as control variables.

Table 1: Overview of variables and how they are measured

Name Meaning Measure

CSRi Corporate social

responsibility Score in CSR ASSET4 CSP and CEP

combined

TIMEi Calendar year Time in years Year from ASSET 4

ROAPi ROA previous year Return on assets in the year prior to

measurement ASSET4 ROA

ROANi ROA next year Return on assets in the year after

measurement ASSET4 ROA t

DCONi Controversies Dummy Dummy based on controversies in previous years

ASSET 4 Controversies dummy

HOFi Hofstede individualism score

Score of individualism in country where the HQ is located

Hofstede individualism index

FCF/TA FCF devided by TA Free cash flow in the current year

divided by total assets DataStream

SIZEi Company size Total Assets in year of measurement Total Assets on balance sheet

BOARDSZi Board size Number of directors in the supervisory

board Boardex board size

BOARDINi Board independence How independent the board is of the managers

Boardex board independence

SECi Sector dummy Dummy based on the SIC code SIC code form DataStream

CSR

= β0 + β

TIMEi + β

FCFP/TAi + β

SIZEi + β

BOARDSZi + β

BOARDINi + β

SECi + εi,t (H1) CSR

= β0 + β

ROAPi + β

FCFP/TAi + β

SIZEi + β

BOARDSZi + β

BOARDINi + β

SECi + εi,t (H2a) ROAN

= β0 + β

CSRi + β

FCFP/TAi + β

SIZEi + β

BOARDSZi + β

BOARDINi + β

SECi + εi,t (H2b) CSR

= β0 + β

DCONi + β

FCFP/TAi + β

SIZEi + β

BOARDSZi + β

BOARDINi + β

SECi + εi,t (H3)

CSR

= β0 + β

HOFi TIMEi + β

TIMEi + β

HOFi + β

FCFP/TAi + β

SIZEi + β

BOARDSZi + β

BOARDINi + β

SECi + εi,t (H4)

CSR

= β0 + β

HOFi ROAPi + β

ROAPi + β

HOFi + β

FCFP/TAi + β

SIZEi + β

BOARDSZi + β

BOARDINi + β

SECi + εi,t (H5a)

ROAN

= β0 + β

HOFi CSRi + β

CSRi + β

HOFi + β

FCFP/TAi + β

SIZEi + β

BOARDSZi + β

BOARDINi + β

SECi + εi,t (H5b)

CSR

= β0 + β

HOFi DCONi + β

DCONi + β

HOFi + β

FCFP/TAi + β

SIZEi + β

BOARDSZi + β

BOARDINi +

β

SECi + εi,t (H6)

Where βi shows the different coefficients and εi is the error term.

4. Results 4.1. Sample

The sample for this study is a combination of DataStream, BoardEx and ASSET4 ESG. The data entries are made from fiscal years between 2002 and 2012. I was left with a sample of containing 5274 firm year observations for which I have data on all variables build up from 1395 unique firms from 41 countries. In this sample USA has respectively 293 unique firms and 1299 firm year observations. The UK has 1045 firm year entries and 206 unique firms.

Also Continental Europe, japan and Australia are well represented by respectively; 1309, 326 and 382 firm year observations and 306, 174 and 128 unique firms.

Table 2 provides the descriptive statistics of the dependent and independent variables.

The dependent variable, CSR, which is a rating between 0 till 100 has a mean of 56.46 in the sample. The median is a bit higher with 59.31. The highest CSR score recorded was 97.78 and the lowest score was 6.55. The standard deviation amounted to 28.43. In comparison the sample of Ioannou & Serafeim (2012) had a mean of 49 and a standard deviation of 29.

From the independent explanatory variables Time reaches from 2003 to 2011 with a mean of 2007.56 and, median of 2008 and a standard deviation of 2.07. From this one can see that there a more entries from the later years. The variable ROAP is not identical to variable ROAN since they contain a different set of firm year observations. If you have a set of firm year entries from the same firm, ROAP removes the first year entry and ROAN removes the last in order to be computed. ROAP has a mean of 7.79, median of 6.95 and a standard deviation of 8.24. ROAN has a mean of 7.70, median of 6.87 and standard deviation of 7.74.

The minimum and maximum are for both the same, -23.25 respectively 37.21. This is due to the fact that the ROA entries were winsorised before transforming them to ROAP and ROAN.

The Hofstede individualism index, which ranges from 1 till 100, has a mean of 74.33 in this sample. The median is 89.00 and the standard deviation is 21.63. The lowest score in this sample is 14.00 and the highest 91.00. There was no need for winsorizing since the highest and lowest values were within the limits of tree times the SD.

The control variables consist of FCF, in USD, in the same year as the CSR value.

After winsorizing, the mean of the FCF is 4,585,000 and the median is 363,000 with a

standard deviation of 49,569. The highest entry recorded was 1,062,311,000 and the lowest -

310,206,000. Free cash flow is scaled to total assets when used as a control variable. Size,

measured as total assets in USD is also used as a control variable. The mean of the sample is

57,970,000 with a median of 5,063,000 and a standard deviation of 579,858. The lowest entry

was 5,815,000 and the highest entry was 9,255,470,000. Board size with a mean of 10.07 and

a standard deviation of 3.16 was also used as a control variable. The median consists of 10.00,

the minimum is 1.00 and the maximum was 20.42 due to winsorizing. The last control

variable is board independence which is set on a scale from 0 to 100. The mean in this sample

is 60.17 with e standard deviation of 26.22. The median is 62.50 and the minimum and

maximum are set on the very edge of the scale.