The impact of CSR motives and firm size on CSR performance, and the influence of firm size on stakeholderdriven CSR motives

(1)

The impact of CSR motives and firm size on CSR

performance, and the influence of firm size on

stakeholder-driven CSR motives.

Marco van Dijk

(6052193)

Amsterdam, June 29, 2015

Thesis seminar Business Administration First supervisor: Mw. Dr. K.J.P. Quintelier Second supervisor: Mw. Dr. F.M. Bridoux Academic year: 2014 - 2015

(2)

Statement of Originality

This document is written by Marco van Dijk 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)

4.2 Measurements ... 24 4.2.1 CSR performance ... 24 4.2.2 CSR motives ... 25 4.2.3 Firm size ... 25 4.2.4 Control variables ... 26 4.3 Procedure ... 26 4.4 Analysis ... 27 5. Results ... 28 5.1 Sample ... 28 5.2 Assumptions of normality ... 28 5.3 Correlations ... 30 5.4 Results ... 33 6. Discussion ... 38 6.1 Summary ... 38 6.2 Discussion ... 38

6.3 Critique & Limitations ... 40

6.5 Implications for stakeholders ... 41

6.4 Implications for future research. ... 41

6.6 Conclusion ... 42

Bibliography ... 43

Appendix A - Firms ... 48

(4)

Abstract

Although research identifies different motives for firms to engage in CSR activities, little research has been done on the influence of CSR motives on a firm’s CSR performance. Drawing on attribution theory, we propose that the motivation of a firm, specified goal setting and the attitude of the firm towards CSR activities lead to better CSR performance. More specifically we hypothesize that a firm that displays (1) more CSR motives, (2) more value-driven motives, (3) more performance-value-driven motives or (4) more stakeholder-value-driven motives in year x has a higher CSR performance in year x+1. Also we hypothesize a larger firm to have higher CSR performance and more stakeholder-driven motives than a smaller firm. We test these hypotheses with multiple linear regressions. A sample of 111 firms was used. All firms are active in the “Food & Drinks” or “Apparel & Textile” industry and originate or are foremost active in the Unites States of America. No evidence was found that more CSR motives of any kind, positively effect CSR performance. The hypothesis suggesting a positive effect of firm size on CSR performance was supported. However, no evidence was found that firm size has a positive effect on the number of stakeholder-driven motives displayed by the firm, but regarding this last hypothesis a trend in the data suggesting this relationship was recognized.

1. Introduction

In the last decades a lot of research has been done on the topic of CSR. However one aspect of CSR received less attention in earlier research. Although there is literature that states the importance of CSR motives to engage in CSR activities, almost no existing literature

researches the relationship between these motives and the actual CSR performance of a firm (Maignan & Ralston, 2002). A confronting example of such a possible relationship is the case of British Petroleum (i.e., BP). BP is one of the market leaders in the gas and oil industry, and according to their financial statement of 2009 it is their aim to become the largest firm in the industry in the near future. Besides this performance-driven motive to engage in CSR

activities, they state in their CSR report of that same year that also based on stakeholder demands CSR activities are taking place. These activities could be identified as stakeholder-driven (BP, 2009). Unfortunately, and despite their CSR motives, in 2010 one of the largest

(5)

disasters in the oil industry ever recorded took place. This disaster, which was named “The Deepwater Horizon Accident”, resulted in the death of 11 employees and a spill of more than 600 million litters of oil into the ocean. The total costs were estimated to be around 20 billion dollars. Obviously the communicated CSR motives to engage in CSR activities did not lead to any desired CSR results (Brantley, Harlow & Harlow, 2012; Valentine, 2015).

Besides this example were we can see the motives of a firm to engage in CSR

activities, CSR motives are also discussed in the literature. Kramer & Porter (2006) argue that there is a shift in customer demands regarding CSR activities. Instead of a firm just producing and selling products, stakeholders nowadays expect firms to be good to its own employees, the environment and the society (Aguinis & Glavas, 2012; Lindgreen & Swaen, 2010). This “doing good” is explained by Holme & Watts (1999) as CSR, and is defined as the continuing commitment of a firm to behave ethically and contribute to economic development by

improving the quality of life for everyone involved.

While research done by Maignan & Ralston (2002) categorizes these firm motives into three groups, namely: (1) value-driven, (2) performance-driven and (3) stakeholder-driven, we do not know if these motives have a relationship with the actual CSR performance, and if this is the case, if a firm that displays more motives performs better on its CSR activities. Knowing if this relationship is present could be very valuable information for investors, potential buyers or other people that are affected by the firm. For example, if this relationship is true, investors could investigate the firm’s motives beforehand and predict its CSR

performance. Based on this knowledge the investors than could decide to invest or not. Knowing if this relationship is present could really help stakeholders around the firm. We therefore think examining this gap in the literature is necessary. Therefore our research question is: “ Does a firm that displays more motives to engage in CSR activities has a better CSR performance?

Firstly the relationship between the total number of motives a firm has to engage in CSR activities will be researched. Later on, this question will be answered in more detail regarding the categorization of different types of motives as stated by Maignan & Ralston (2002).

Lastly, research done by Udayasankar (2008) and Adams & Hardwick (1998) suggests that firm size is an influential factor on the willingness (i.e., motives) of firms to engage in CSR activities. Therefore we will also examine the influence of firm size on the firm’s stakeholder-driven motives, as explained by Maignan & Ralston (2002). This is done to improve the generalizability of our findings and the potential value for stakeholders.

(6)

In the reminder of this article we will firstly discuss relevant literature regarding CSR and CSR performance. A theoretical framework that introduces the hypotheses based upon existing theories follows this literature review. These hypotheses will be tested, which is described in the method section. Then based on our findings the results are presented. Lastly we present a discussion, implications for future research, implications for stakeholders and a conclusion.

2. Literature review

In order to provide a clear image about the relationship between a firm’s CSR motives, size and its CSR performance, this section will discuss the essential findings of previous research in this field. It briefly discusses the different interpretations of CSR, CSR performance and the causes of CSR performance.

2.1 CSR as a Concept

While the term ‘CSR’ originates from the United States of America, it nowadays is a well-known topic all over the world (Carroll, 2008). Unfortunately and despite years of research, there is still no widely accepted definition of the CSR concept, but the different insights do overlap. Therefore different insights on the topic of CSR are discussed in this section.

CSR is often understood as a firm its obligation to be accountable to all of its

stakeholders, in all its operations and activities, as described by (Doane, 2005). According to Doane (2005), CSR is about balancing the needs of stakeholders and the need to make profit in the decision making process. This means that firms are responsible for the effects that their actions have on people inside and outside the firm. Aguinis (2011, p.855) adopts in his research the same perspective, but adds his believes about the Triple bottom line theory to the CSR concept. This is to say he defines CSR as all actions of a firm that take into account stakeholder expectations and the triple bottom line of economic, social and environmental performance.

An accepted model related to this multilevel adoption of CSR as described by Aguinis (2011) is the four-part model for CSR as presented by Carroll (1979). He states that a firm has its responsibilities on an economic, legal, ethical and philantropical level. The order of these levels is crucial and the model is shaped to importance, alike a pyramid. Carroll (2008) suggests that true CSR should meet the expectations and desires of the society on all these

(7)

levels. Other researchers use more simplified definitions of CSR. Frederick (1986) is an example; he argues a firm always has the obligation to work for social betterment. This concerns issues beyond the economic, legal and traditional requirements of the firm and always benefits the society (Davis, 1973).

This CSR obligation for a firm arises from a situation where business ethics and the laws the firm has to abide to often result into a grey area where it is unclear what the responsibilities are of the firm (Crane & Matten, 2010). Sometimes the law, for example, does describe the minimum requirements to prevent from environmental damage, but does not describe what the community ethically desires from the firm. The actions taken by the firm to manage these differences can be identified as CSR activities.

The consideration that firms are indeed responsible for their actions, and influence the environment around them, leads to a situation where a firm almost should be forced to adopt CSR activities. Crane & Matten (2010) state that firms cause social problems, such as pollution and have the responsibility to help solving these problems. Also firm size, with globalization in mind, is suggested as a reason for accountability. CSR problems related to environmental or social problems are not restricted to country borders. Therefore firms could potentially be better in taking actions than countries by themselves (Reich, 2008). By law this responsibility is hard to command, but ethically speaking firms should accept the

responsibilities for the greater good. Rugman (2000) disagrees with these suggestions. He states that even while firms have gained power due globalization, firms are still weak, for example because of their dependence on national governments.

CSR has also been linked in the literature to corporate social responsiveness. This more strategic insight is about how firms are responding to CSR concerns (Frederick, 1994). Carroll (1979) suggests four strategies to respond to these CSR concerns. These range from reactive or denial of any form of responsibility by the firm, to a proactive attitude where the firm does more than is expected (Aguinis & Glavas, 2006; Becker-Olsen, Cudmore & Hill, 2006).

But firms do not necessarily have to engage in CSR activities on their own.

Collaborations between pressure groups or Civil Society Organisations (i.e., CSO’s) are often hard to accomplish, but a possibility as well (Ackers & Payne, 1998; Baron, 2009; Fassin 2009). As a result the charitable feeling of CSR activities decreases and becomes more of a partnership between the firm and its stakeholders (Crane & Matten, 2010).

Summarizing these insights on the topic of CSR we could synthesize that CSR is about creating value for the firm while taking the influences the firm has on the society,

(8)

environment and economy in mind (Aguinis, 2011). But also the attitude the firm adopts and the way it reacts to CSR engagement are part of the CSR concept.

Now the basic aspects regarding CSR, CSR activities and CSR engagement are discussed we will move on to the next section where the outcomes of CSR, the CSR performance is discussed in detail.

2.2 CSR Performance

Stating that firms are actively engaging in CSR activities is not something new. But asking the question what the results are of these CSR activities is what is most important. The CSR outcomes, namely, accomplishments, future plans and current activities are often presented on large firms websites or in their CSR reports. As an example we could take a look at the earlier mentioned BP CSR report of 2009. Here BP reports the improvements accomplished on employee safety and the reduced levels of pollution in Africa (BP, 2009)

Unfortunately these CSR performances realised by firms are very hard to measure. Asking the simple questions: “what should we measure?” or “What is performance?” are questions often stated in existing literature. As starting point we take Griffin & Mahon (1997). In their overview of literature about CSR performance, also called Corporate Social

Performance (CSP), they argue that this performance should be measured in different dimensions. Carroll’s (1979) statements are in congruence with this standpoint. In his research he states multiple measurements and sources should be used to describe CSR performance. He debates that performance consists out of three different indicators namely, (1) social issues, (2) philosophies and (3) social responsiveness. Wood (1991) has the same line of thought. She identifies CSP as a combination of the social impacts the firm has, the programs in place to prevent or handle CSR related issues and the existing policies inside the firm to guide decision-making with the CSR goals in mind. Lastly, Clarkson (1995)

advocates the necessity to distinguish stakeholder issues from society issues in social

performance. According to Clarkson (1995), managers have to deal with stakeholder groups, but not with the society as a whole. This is the result of individual stakeholders forming a team or pressure group in order to persuade the firm to listen to them. The message of such a pressure group is therefore not the same as the message of the society. Consequently the perceived CSR performance could differ between the stakeholders (groups) and society.

(9)

Based on the concept of CSR performance, it is clearly necessary to measure CSR performance on different dimensions. A great example, and widely accepted tool that measures CSR performance is the KLD Index, as developed by the financial analysis firm Kinder, Lydenberg, Domini & Co. Inc.. This model measures CSR performance on different dimensions. In the Method section the KLD index is explained in more detail.

Now we discussed what is important in measuring CSR and how CSR performance should be measured, the next paragraph describes the causes of CSR performance.

2.3 Causes of CSR performance

In the section above we discussed how CSR performance should be measured, but it is also interesting to know what the influential factors or causes are of CSR performance. If we do know for example what causes great CSR performance, we could focus on these causes and improve the CSR performance. Therefore different thoughts on the causes of CSR

performance are discussed in this passage.

While researching the elements driving CSR performance we found literature to be very scarce. Most research we found on successful CSR implementation merges financial performance and sustainable (i.e., environmental & social) performance together into one variable (Doh, Miller, Perez-Batres & Pisanim, 2012; Sangle, 2010). Identifying the causes of the firm’s CSR results could therefore be partly distorted by this combination. However, different important factors for CSR performance are found in the literature. Sangle (2010) describes the importance of funding for CSR activities. Investment, often in the form of money, is a critical factor enabling CSR success. Second, at the management level,

commitment is needed. This means quality management of the CSR objectives and support from top-level management are necessary to achieve desired CSR performance. Lastly, Sangle (2010) argues the need for integration with other functional strategies of the firm. Here, synergy between the various goals, namely, financial performance and the firm’s social performance is stated as crucial, meaning that the CSR goals should be part of the overall goals of the firm.

Clarkson (1995) argues that in managing the CSR objectives and in merging the different goals of a firm into one cohesive strategy there is a need for a clear structure. Processes and programs are a great help when it comes to managing, analysing and planning CSR activities. Policy development and implementation of CSR activities also benefit from

(10)

such a clear structure. Therefore the processes and programs in place, managing CSR activities, could be recognized as causes of CSR performance.

Since Clarkson (1995) argues that integration of CSR strategies, financial strategies and structures of to the firm are identified as causes of CSR performance, we could question if the implementation of these strategies always leads to desired results. An answer to this question is found in the research done by Borial (2007) concerning ISO 14001

implementation. ISO 14001 is a management tool to translate the CSR strategy of a firm into a job structure. He argues that even when programs like ISO 14001 are integrated in the firm’s strategy and structures there is still a chance for lack of CSR performance. The cause identified in Borial’s (2007) research is two sided and about the motivation inside the firm. Firstly he argues that when the motivation for a firm to adopt a program as ISO 140001 is just to be certificated as a “green firm” CSR performance is usually low. Secondly he suggests that the motivation of low-level employees and managers in their daily work activities has a relationship with the CSR outcomes (i.e., CSR performance) of the firm. He finds that the employees are mostly performing CSR activities because top-management tells them to do so, but often they do not know the reason why. Involvement and motivation in these activities is as a result said to be low. Since Borial (2007) states that this leads to undesired CSR

performance we could think of lack of true motivational factors by the top-level management and a lack of motivational factors with the employees to engage in CSR activities. Increased communication for example, ten could lead to increased motivation and better CSR

performance. These motives firms or employees have, could also be important when it comes to the influential factors on CSR performance as discussed by Clarkson (1995) and Sangle (2010) as described before. We could assume that factors as investment, support from top level management and the right structures and programs would benefit (i.e., implemented faster and better) when the (true) motivation of the firm and employees to engage in CSR activities is higher. Therefore, according to the above, motivation could have an influence on the firm its CSR performance, on multiple levels.

The thought of importance of a firm its motivation to engage in CSR activities is even more strengthened when we adopt a more strategic point of view. Besides the synergy that is needed between the financial and CSR goals (i.e., strategy) of the company also the way in which new goals or strategic focal points are adopted is identified as influential (Romanelli & Tushman, 1994; Sangle, 2010). Research done by Romanelli & Tushman (1994) argues that incremental changes in strategy are unlikely to yield desired performance improvements. Also these incremental changes are stated to reduce the motivation of employees inside the firm.

(11)

Revolutionary changes are according to Romanelli & Tushman (1994) more likely to yield performance improvements, because motivation of employees is retained through the power of the radical changes. Also the motivation, to implement such a new, radical changed strategy needs to be high at the top management level. Therefore, according to this theory, the way a firm adopts new strategic directions tells something about the top-level

management’s motivation, and influences the motivation of employees inside the firm. Summarizing we could say that the motivation a firm has to adopt CSR practices could be a crucial factor in the success a firm has on its CSR activities (i.e., CSR

performance). Therefore the theoretical framework, which follows next, will discuss the motivators of the firm to engage in CSR activities in more detail.

3. Theoretical framework

In the literature review the basic principles of CSR, CSR performance and causes of CSR performance are reviewed. In this theoretical framework, which is based upon the basic principles discussed in the literature review, the motives a firm has to engage in CSR activities are discussed in more detail. Based upon the theories discussed hypotheses are introduced. Also the possible influence of firm size on CSR performance and stakeholder-driven motives is explained.

3.1 Firm motives to engage in CSR

In what follows, we will talk about CSR motives to denote the motivation a firm has to adopt CSR practices. Most important for these motives were the findings of Borial (2007)

concerning ISO 14001 and the theory of Romanelli & Tushman (1994) advocating radical CSR to lead to better CSR performance and retained motivation of employees. This last finding, the employees their motivation and its relationship with performance can be clarified by an example using ourselves, as researchers. The more a researcher is motivated to dig deep into the existing literature and learn every aspect of the subject he is researching, the greater his knowledge becomes on the research topic. When he then uses this knowledge to underpin his research results with the theories he learned, he is able to draw quality conclusions out of the data. In this example, his motivation to study the subject therefore is positively related to

(12)

the quality of his performance. In other words, a higher level of motivation leads to better performance of the researcher.

This is also found by Cavender, Petty & McGee (1984). In their research it is argued that employees with higher levels of motivation perform better on their jobs. The better performing on the job could be explained as a higher level of performance. Taking into account that these theories are based on employees, managers, and in our example a single researcher, we could take this reasoning one step further. We could make the transition from a single person to the firm as a whole, reasoning that a firm basically is a group of employees working together on a common goal. Therefore we could think as a similar relationship between a firm its motivation and performance. This relationship is already discussed by theories of Romanelli & Tushman (1994), but is also described by Deshpande & Webster (1989) and Schein (2010). They state that the business culture of a firm predicts the firm its future goals (i.e., strategy, common goals). Here a business culture is explained as the shared values, beliefs and the social context of the employees, indicating a similarity between a single person and a firm as a whole. Therefore, expanding these theories again to the firm level, we could assume that the higher the motivation of a firm is to engage in a certain activity, the better its performance on this activity becomes. Applying this reasoning to our research, we could say that a firm with more motives to engage in CSR activities will have a higher level of CSR performance. This idea is enforced by the findings of Latham, Locke, Saari & Shaw (1981). In there research, which focuses on setting and accomplishing goals, they find a positive relationship between the specific definition of certain goals and the quality of the accomplishments. This indicates again that when a firm is more specific with their motives to engage in CSR activities this leads to a higher level of CSR performance.

3.2 Firm motives and their stability over time

CSR motives can only influence CSR performance if they are sufficiently stable over time. Without stability in a firm its motives, we cannot derive any form of true relationship between these motives and CSR performance. If we take for an example a firm that would change its motives to engage in certain activities every week, our analysis regarding CSR motives as a cause of CSR performance over time, would not be trustworthy. Fortunately reasoning can be found in the literature why a firm’s motives are stable over time.

(13)

Firstly, and according to Romanelli & Tushman (1994), radical changes in strategy are best for CSR implementation and CSR performance. Radical changes can be explained as fundamental changes in a firm its strategy, and therefore changes in a firm its core activities. Besides the motivation needed to engage in such a change of strategy, explained as a firm long-term goals, also commitment needs to be high to pursue this new strategy (Porter, 1979). In other words, the firm needs to have enough motivation to follow the current, or newly adopted strategy; otherwise the firm’s management would change direction by changing the strategy. As long as a firm pursues its current strategy, we could argue that a firm has roughly the same motives for a longer period of time. Therefore the motives of a firm can be identified as stable over time.

Bigley, Felps & Jones (2007) argue the stability of firm motives as well. In their article they argue that firms differ in their CSR motives as a result of different business cultures inside the firms. A business culture is explained earlier as the shared values and beliefs of employees inside the firm and are also argued to be stable over time (Deshpande & Webster,1989; Schein, 2010). Schein (2010) also argues that a business culture influences the firm its strategy. Therefore a firm its business culture enlarges the effect of stability generated by the strategy, as discussed above.

Assuming that a firm its CSR motives are stable over time, that more specified motives lead to higher levels of performance and that the amount of motives a firm a has influences the CSR performance of the firm, we could hypothesize that firms that display more motives to engage in CSR activities will have an higher level of CSR performance. In order to identify the relationship between the CSR motives (i.e., causes) and CSR

performance (i.e., results), we use the CSR motives in year x and the firm its CSR performance in year x+1. Hypothesis 1 summarizing our expectancies is stated next.

Hypothesis 1: Firms that display more motives to engage in CSR activities in year x, will have a higher CSR performance in year x+1

(14)

3.3 Identifying different CSR motives

Now we hypothesized that the motives of a firm have their impact on the firm’s CSR performance, we will discuss the different categories of motives coming forward out of the literature, in more detail.

In the literature different reasons for firms to engage in CSR activities are explained. Doane (2005) identifies four drivers that would impel a firm to adopt a CSR program: (1) managing risk and reputation, (2) Protecting human capital assets, (3) avoiding or deferring regulations and (4) as a response to consumer demands. All, except for the last driver focus on the benefits CSR has for the firm itself, this last driver focuses solely on the stakeholders involved. Greening & Turban (2000) add the goals of attracting new employees and the desire to make a positive contribution to society to these drivers (Kramer & Porter, 2006). These drivers could be categorized using the performance-driven and stakeholder-driven motives as identified by Maignan & Ralston (2002). Their research addresses three motives for a firm to engage in CSR. These are (1) value-driven motives, (2) performance-driven motives and (3) stakeholder-driven motives. The value-driven motives are an addition to the model of Doane (2005) and Greening & Turban (2000) and are only slightly described by Kramer & Porter (2006). These imply a situation where managers, shareholders or owners of the firm, have the strong believe and willingness to do something good. Their motive to engage in CSR is based on their personal believes, norms and values and is, maybe the most of all, about doing something good for the community, environment or society.

These three motives can be examined even further when taking the distinct tactics regarding CSR engagement into account. These range from reactive, or a form of denial, to a pro-active attitude towards accountability (Aguinis & Glavas, 2006; Becker-Olson et al., 2006; Carroll, 1979). Especially Aguinis & Glavas (2006) define CSR more deliberately. Their separation focuses on a reactive and a proactive way of engaging in CSR activities and taking responsibility. Firstly the proactive approach is identified as the willingness of the firm to undertake actions based on the firm its own initiatives. Here the intentions and actions emerge from inside the firm. The second strategy is more reactive. Here the firm adopts a more cautious attitude and waits with engagement in CSR activities until factors from outside the firm persuade or force them to take action.

These tactics to engage in CSR activities do have similarities with the motives

explained by Maignan & Ralston (2002). When looking at the value-driven and performance-driven motives we could say these are in line with the proactive strategy as defined by

(15)

are for a large part identified as proactive when a firm engages in CSR to benefit their stakeholders, even when law or regulations do not force them to do so.

Becker-Olsen et al. (2006) also found that firms with a more proactive attitude towards CSR activities receive better (i.e., more positive) response from their stakeholder than firms with a reactive attitude. Also Baron (2009) argues that the perceived sincerity and being genuine of the firm, lead to more positive response from stakeholders. Since the KLD index for CSR performance is measured by the perceived performance of the stakeholders, we could expect differences in the performance measures between these CSR motives.

Because of the possible differences in performance between the CSR motives, the next section will describe the three CSR motives as described by Maignan & Ralston (2002) in more detail. After linking these motives to existing theories found in the literature hypotheses are presented.

3.4 Value-driven motives to engage in CSR activities

In this section the value-driven motives of a firm are discussed in more detail. Regarding the article written by Maignan & Ralston (2002) the core of this CSR motive is the intention the firm has, and the strong belief that those actions need to take place for the greater good, even without benefits for the firm itself. These value-driven motives are linked to the description of philanthropy by Kramer & Porter (2002; 2006). They describe philanthropy as creating the largest amount of social benefit for the community, without any economic benefits for the firm itself. Also no stakeholder demands are necessary for the firm to take action and therefore these value-driven motives can be identified as proactive tactics to engage in CSR activities. This means that these proactive, value-driven motives of the firm take the concept of CSR to a higher level. This is because, as described in the literature review, CSR often tries to fill the gap between ethical beliefs and the law in a business environment. The mentioned grey area is the place where the firm is not required by law to take action, but wants to take action as a result of the ethical beliefs. In this light, we could assume these value-driven motives to be the most sincere motives of all, as defined by Maignan & Ralston (2002) and Porter & Kramer (2002).

As we have seen before, motivation, specified goal setting, the proactive attitude of the firm towards CSR and the perceived sincerity by stakeholders are positively related to CSR performance (Baron, 2009; Becker-Olson et al., 2006; Borial, 2007; latham et al., 1981;

(16)

Romanelli & Tusman, 1994). Therefore for these value-driven motives this positive relation should also be the case. To express value-driven motives towards the people around the firm, motivation for these motives needs to be high, especially since the firm does not directly benefit from these taken actions. According to the positive relationship between motivation and CSR performance more-value-driven motives would impel an increase in CSR

performance.

Secondly the value-driven motives are not demanded by anyone from outside the firm and can therefore be seen as proactive. Also they are not intended to benefit the firm. As a result the public will implement these actions as genuine and sincere (Baron, 2009). This leads to a higher rating by the public on the CSR performance KLD index (Kramer & Porter, 2002). Considering these factors we should expect CSR performance of a firm to become higher when a firm displays more proactive value-driven motives. This thought is even more strengthened when the theory as discussed by Latham et al. (1981) is taken into account. Here more specified goal setting, and in this situation displaying more detailed CSR motives, should lead to better performance. Using this theory in the CSR context this would impel that a firm with more value-driven motives (i.e., goals) regarding CSR would perform better on CSR activities. Based on these findings in the literature we hypothesize that a firm that

displays more value-driven motives in year x will perform better on CSR Performance in year x+1.

Hypothesis 2a: Firms that display more value-driven motives to engage in CSR activities in year x, will have a higher CSR performance in year x+1

3.5 Performance-driven motives to engage in CSR activities

This section focuses on the performance-driven motives as defined by Maignan & Ralston (2002). Despite the term for this motive as dictated by Maignan & Ralston (2002), this motive does not imply every form of performance. Especially the business performance regarding financial aspects and growth are meant by this description, meaning that a firm displaying performance-driven motives is performing CSR to increase its profits. The existence of this motive is not surprising if we think of how CSR became a more prominent aspect of a firm its strategy in recent years. The traditional shareholder approach that firms adopted in the

(17)

decades before the CSR approach claims that the goals of the firm should solely be based on shareholder interests (Kramer & Porter, 2006). Friedman (1970) also argues that firms do not have responsibilities that go any further than the basic rules of society, and that their sole interest should be profit making.

Despite firms being more CSR orientated nowadays, the firm still has a huge desire to grow and generate profit, for the simple reason they need profit to survive. Therefore it is not surprising that financial goals of a firm are motives for firms to engage in CSR activities.

The question then is, if these performance-driven motives lead to desired CSR results. We think this is the case, for the simple reason that even when their motives to engage in CSR activities are based on this more traditional, profit orientated goals, we still could expect a firm engaging in CSR to have better CSR performance than a firm not engaging in CSR. Resource allocations as investments in money or employees mean that a firm is actively engaging in CSR activities and therefore we could expect CSR performance to be higher than when no activity is present. Therefore, even when the firm only engages in CSR as a result of performance-driven motives we could still expect CSR performance to be higher than when a firm does not engage in CSR, or has no motives to engage in CSR.

Secondly, the performance-driven motives are not demanded by anyone from outside the firm, and can therefore be seen as proactive. Becker-Olson et al. (2006) found that firms with a more proactive attitude towards CSR activities receive better (i.e., more positive) response from their stakeholders than firms with reactive attitude. Since stakeholders rate the firm’s CSR performance we can therefore expect a firm that displays more performance-driven motives to have a higher CSR performance.

Also, as we have seen before, motivation and specified goal setting, are positively related to CSR performance (Becker-Olson et al., 2006; Borial, 2007; Latham et al., 1981; Romanelli & Tushman, 1994). For the performance-driven motives this could also be the case. A firm that displays more performance-driven motives to engage in CSR activities could be said to be more motivated to take actions. This higher level of motivation should therefore lead to better CSR performance. Lastly more specified goal setting, where displaying more different CSR motives to engage in CSR is similar to, is argued by Latham et al. (1981) to lead to better performance results. Therefore we hypothesize that a firm that displays more performance-driven CSR motives in year x, will have better CSR performance in year x+1

(18)

Hypothesis 2b: Firms that display more performance-driven motives to engage in CSR activities in year x, will have a higher CSR performance in year x+1

3.6 Stakeholder-driven motives to engage in CSR activities

Another prominent factor as an influence on a firm its CSR motives are the stakeholders the firm has (Doane, 2005; Kramer & Porter, 2002; Maignan & Ralston, 2002). In their research stakeholders are explained as people or groups around and inside the firm that are affected by the activities of the firm. According to Crane & Matten (2010) this affection includes both benefits and harm. Evan & Freeman (1993) suggest that if a firm wants to identify its stakeholders two simple principles can be applied. The first principle “The principle of corporate effects” says that firms are accountable and responsible for the effects their actions have on others. The second principle “Principle of corporate rights” emerges from the rights people around the firm have. This principle demands that the firm has the obligation to not violate the rights of others. When a firm uses these two principles, their stakeholder can be identified.

Once a firm identified its stakeholders Agle, Mitchell & Wood (1997) argue managing these stakeholders the right way is crucial. They suggest that the power, legitimacy and urgency of the stakeholder need to be taken into account. Using these layers a firm is able to rank these stakeholders by their importance. These theories as discussed, add up to the stakeholder theory as described in existing literature. This theory is probably one of the most approved theories in the CSR field and has emerged from management and ethical studies (Crane & Matten, 2010). This more stakeholder oriented view, has merely replaced the

traditional shareholder approach that firms adopted in the previous decades (Kramer & Porter, 2006). This more traditional view claims, as earlier discussed in this research, that firms should be managed in the interest of the shareholders alone. In the stakeholder theory this shareholder interest is not abandoned completely, but advocates there has to be room for the interests of stakeholders simultaneously (Crane & Matten, 2010; Kramer & Porter, 2002).

Porter & Kramer (2002) also describe stakeholders to expect more from firms on the CSR level than ever before. Crane & Matten (2010) and others argue why this change is happening. First they find firms to be bigger than ever before, sometimes even larger than countries. It is said these firms are in some occasions even more powerful than local

(19)

governments. As a result, when stakeholders want to change a situation that affects them, it is likely they will turn to a firm instead of the government. Also, with the on going globalization in mind, firms become active in more different countries and as a result have more different stakeholders. Lastly, Fassin (2009) states in his article, by watching large pressure groups and non-governmental organizations that globalization, increase in the world’s economy and the increased power of multinational firms enlarge stakeholder pressures on firms. We suggest that this increase in pressure could partly be explained by the easiness with which

stakeholders can communicate with firms nowadays. Social media, or the Internet itself have made it very easy to connect with other stakeholders, form pressure group, or get in contact with firms. As a result more people are enabled to speak up for themselves.

Accepting the rise of stakeholder demands for CSR activities, and the increased pressure on the firm, we should identify what the exact demands are as well. Since firms are active at different locations and in different markets we could expect the firm its stakeholders to differ from time to time, in different projects and between firms it self. Crane & Matten (2010) underwrite this, and Fassin (2009) adds the inconsistencies between stakeholder groups to this line of thought. He advocates that different stakeholder groups can demand contradicting actions from the firm. This results in a situation where an action taken by the firm is good for a certain stakeholder group but negatively impacts another stakeholder group. The consequence of these contradicting demands and differences between stakeholder groups is that a firm has to take lots of stakeholder demands into account (Bigley et al., 2007).

Besides a situation where stakeholders demand actions of the firm, this has not necessarily to be the case. For example a firm that would like to have a stronger relationship with its employees could also have these motives from it self. Because taking action based on stakeholder demands when this not mandatory by law and intrinsic motivation of the firms regarding stakeholders, we could identify stakeholder-driven motives as proactive. Becker-Olson et al. (2006) found that firms with a more proactive attitude towards CSR activities receive better (i.e., more positive) response from their stakeholders than firms with reactive attitudes. Since stakeholders rate the firm’s CSR performance we can therefore expect a firm that displays more performance-driven motives to obtain a higher level of CSR performance.

Also if we take the range of different stakeholder demands into account, we could argue that a firm listening to these demands, and excepting these demands as legit, would actively engage in CSR activities, in order to satisfy its stakeholders. This engagement results in a certain level of CSR performance. We could suggest that when a firm engages in CSR activities based on pressures by stakeholder groups, its motivation to be high. According to

(20)

Borial (2007) higher levels of motivation lead to better performance. As a result, when a firm is under more pressure from different stakeholder groups, its motivation should become higher and its CSR performance should increase as well.

This extra motivation is also shown in the number of stakeholder-driven motives a firm displays. More motives displayed should be a result of higher pressures and according to Latham et al. (1981) these more detailed motives to engage, lead to better performance at the same time.

Lastly, since the firm engages in CSR activities based upon the stakeholder pressures, the stakeholders will be delighted with the firm its engagement in CSR activities in the directions they desire. According to Baron (2009) this results in higher CSR rating by the stakeholders and a higher rated CSR performance as well. Therefore we hypothesize that a firm witch displays more stakeholder-driven motive to engage in CSR activities in year x, will also have a higher CSR performance in year x+1.

Hypothesis 2c: Firms that display more stakeholder-driven motives to engage in CSR activities in year x, will have a higher CSR performance in year x+1

3.7 Firm size and firm performance

As mentioned in the introduction of this article, firm size is another factor that possibly explains CSR performance. One reason is the strong relationship between firm size and the strategy used within the firm. Smaller firms are more likely to focus on fast growth or profits and less on legal, philanthropically and ethical activities. Large firms are argued to be more open to these activities (Adams & Hardwick, 1998; McElroy & Siegfried, 1985; Orlitzky, 2001; Udayasankar. 2008). Identifying philanthropic and ethical activities as CSR orientated we could expect larger firms to perform better on CSR activities.

A second aspect of the willingness to engage in CSR activities is the access to resources a firm has. Brammer & Millington (2006) suggest that it is easier for large firms than small firms to get access to resources. Also large firms are faster in obtaining these resources. This is also argued by Doh et al. (2012). In their research a relationship between the availability of resources and the willingness to follow a CSR course in the firm is found.

(21)

Meaning that the less resources a firm has available, the less likely it is to focus on CSR activities. Based on these findings regarding resource access we could expect large firms to engage in CSR more often than small firms and have better CSR performance.

Further, large firms are argued to have the advantage of economies of scope. Teece (1980) explains this theory as a situation where a firm is able to produce output “y1” & “y2” together for fewer costs than producing them separately from each other. The larger a firm is, the more likely it is it has this advantage. Krugman (1980) identifies a theory closely related to economies of scope, namely, economies of scale. Economies of scale is present when a firm is able to differentiate in products or processes, without extra costs. This is also more likely for larger firms than small firms. According to these theories larger firms are able to perform better on financial performance, since they are able to produce or undertake more with the same amount of resources or money (Krugman, 1980; Teece. 1980). These

economies of scale and scope could also be influential in CSR activities. Since a larger firm is able to differentiate and combine different CSR activities, by using the same processes, and accomplish more with the same resources, we expect larger firms to perform better on CSR activities. This idea is strengthened by the articles written by Cochran & Wood (1984) Hall & Weiss (1967) and McWilliams & Siegel (2000). These researchers state a positive

relationship present between a firm’s financial performance and corporate social performance. Meaning that a firm with higher financial performance also performs better on CSR activities. According to the economies of scale and scope theories as argued by Krugman (1980) and Teece (1980) we could assume this positive relationship to be stronger for larger than for small firms. Based on these findings we hypothesize that a positive relationship is present between the size a firm has and its CSR performance.

Hypothesis 3a: Larger firms will have a higher CSR performance than smaller firms.

3.8 Firm size and stakeholder-driven motives

Besides the above hypothesized positive relationship between firm size and CSR performance and the theories building op to this hypothesis, we could also assume that larger firms will most likely be more differentiated than smaller firms. More specifically, a large firm will have

(22)

a more diversified portfolio and be active in more markets (Krugman, 1980). Orlitzky (2001) underwrites this in his article and also states that this leads to more visibility for the large firm. The extra market activity leads to more stakeholder involvement, which also enlarges the attention the media has for the firm. This media attention enlarges the visibility of the firm again (Cowen, Ferreri & Parker, 1987; Udayasankar, 2008). Other literature suggests firms often adopt CSR as risk management, to ensure and maintain a good reputation. Since reputation management is related to the visibility of the firm this magnifies the thought that larger firms are more willing to listen to their stakeholder demands (Ambec & lanoie, 2008; Doane, 2005; Visser, 2010). Therefore we could expect a larger firm to be more willing to take stakeholder pressures into account, and actively engage in CSR activities. This could lead as stated in hypothesis 3a to a higher level of CSR performance for larger firms, but also to more specific motives regarding the firm its CSR activities. Based on the discussed theories we therefore hypothesize that larger firms will have more stakeholder-driven motives

regarding CSR than smaller firms. All the hypotheses are summarized in figure 1 below.

Hypothesis 3b: Larger firms in year x will have more stakeholder-driven CSR motives in year x than smaller firms.

(23)

4. Methodology and research design

After stating the relevant literature and introducing the hypotheses for our research in the earlier sections, this section will address the set up of our research. First the design of our study and the sample are discussed. Further the measurements, procedures and analysis are discussed in more detail.

4.1 Design & Sample

The data to test our hypotheses is collected by a cross-sectional study based upon the

information extracted from the corporate social performance index named KLD, the financial index Compustat and an analysis of the annual reports of the firms targeted. The KLD Index will provide the research with the firms score on their CSR performances, while the

Compustat index will provide the data regarding firm size and financials. Lastly the annual reports will provide the CSR motives of the firms.

The extracting and coding of the data is done by five students all working on their bachelor or master thesis, while a single researcher does the data analysis. By making a coding scheme on forehand the five researchers prevented the data from being distorted by being reviewed by different researchers. This was needed to enable the required consistency throughout the process. A total of 150 firms was analysed in this study but because of

incomplete datasets the total firms used for our analysis was 111. This number of firms is still sufficient to draw conclusions with the statistical tests used in this research. The names of the firms analysed in this study are available in appendix A. The firms analysed are active in the “food & drinks” and “apparel & textiles” industries and are all and foremost active in the United States of America. These industries were targeted because collaborative researchers are especially interested in these industries. Also the firms inside these industries tend to differ from each other in their CSR practices, therefore they are suitable for our analysis. The specific firms were targeted because of their presence in COMPUSTAT and the KLD index. Also their firm size was a crucial factor regarding hypotheses 3a and 3b.

Lastly the firms do not know they are part of this study. Because time has to pass before a clear relationship between cause (i.e., CSR motives) and effect (i.e., CSR

performance) can be researched, our analysis is based on the data from previous years. Since the CSR adoption rapidly increased in the last years we did not want the data to be to old. Therefore we choose 2011 as our year to research.

(24)

4.2 Measurements

In our analysis the three variables as earlier introduced; CSR performance, CSR motives and Firm size will be used. CSR performance is categorized as a dependent variable, while the other two, CSR motives and Firm size are independent variables. All variables used in our research are discussed next.

4.2.1 CSR performance

Our dependent variable CSR performance is measured by the KLD index. This index is developed by an independent group of researchers from outside the firms, working for the financial analysis firm Kinder, Lydenberg, Domini & Co Inc.. All firms are rated based on information from inside and outside the firm. Sources of information are business press, general media, trade magazines (i.e., external), proxy statements and annual or quarterly reports (i.e., internal) (Graves & Waddock, 1997).

The eight dimensions used to rate the firms are: (1) community relations, (2) employee relations, (3) environment, (4) product, (5) treatment of women and minorities, (6) military constrains, (7) nuclear power and (8) South Africa involvement. The first five dimensions are rated on a 4-point scale ranging from major strength to major weakness. The other three dimensions are rated on a dichotomous scale; major weakness or major strength (Griffin & Mahon, 1997). As a result of the difference in scales used in the KLD index some dimensions need to be reverse-coded to enable a computed variable with all dimensions included. This variable is referred to as CSR performance and is used as a dependent variable in hypothesis 1, hypotheses 2a,b,c and hypothesis 3a.

A great advantage of this KLD index is its objectiveness as a result of its

independence and the large variety where CSR performance is measured on. Unfortunately the way of ranking, the 4-point scale and dichotomous scale are not very specified, and therefore could have been better. But since this database uses the largest amount of

dimensions and is the largest database available we will use the KLD in our research. In our analysis all the dimensions available in the index are used, except the controversial options for nuclear power, military and Tabaco, where for most of the firms no data was available. Also we assume that these dimensions are not key CSR performance measurements for firms in the “food & Drinks” and “Apparel & Textiles” industry.

(25)

4.2.2 CSR motives

The independent variable CSR motives relates to the motives a firm has to engage in CSR activities. In our analysis we will use this variable in hypothesis 1 as an independent variable. This variable will also be used as a control variable in the other hypotheses of our research. The firm’s annual reports of the year 2011 are used to extract the total amount of motives displayed regarding CSR activities. This total number of motives to engage in CSR, results in the variable: Total CSR Motives. The total motives then are categorized in three groups according to Maignan & Ralston (2002), namely: (1) value-driven motives, (2) performance-driven motives and (3) stakeholder-performance-driven motives. This categorization enables our research to use these categories as new variables, which are needed for hypotheses 2a,b,c and hypothesis 3b. The coding scheme used to identify the different CSR motives of the firms is available in appendix B.

4.2.3 Firm size

Earlier research suggests firm size to refer to the scale of operations within an organization (Evens, 1987). More precise descriptions of firm size are also found. Described methods for measuring firm size are the total assets, annual revenue, number of employees or the total amount of yearly orders placed by customers (Samiee & Walters, 1990; Wolff & Pett, 2000; Czinkota & Johnston, 1983). Since other research suggests that CSR performance is related to the financial performance of a firm, we think measuring firm size by a firm’s total assets, the place were a firms profits are stored, could deliver interesting results (McWilliams & Siegel, 2000; Cochran & Wood, 1984). Measuring Firm size by a firm’s total assets is also in line with the findings of Sangle (2010) as earlier discussed. Sangle (2010) identified the amount of investment as a cause of CSR performance. We identify a firm’s total assets as a place where these investments are coming from. Doh et al. (2012) also stated that the availability of resources increases the firm’s willingness to engage in CSR. The total assets of a firm could also be recognized as resources available. Also the total assets a firm has enable economies of scale and scope as mentioned before. As a result we identify the total assets of a firm as a legitimate measurement for firm size in relation to our hypotheses. In this research the

variable firm size is used as an independent variable in hypotheses 3a and 3b. Firm size is also used as a control variable in hypothesis 1 and hypothesis 2a,b,c

(26)

4.2.4 Control variables

Regarding the multiple regressions used in our research, control variables are needed to ensure the validity of our findings. A total of four control variables is recognized in this research. The first two, as stated before are the independent variables Total CSR motives and Firm size. These variables are used as control variables when they are not used as independent variables. Two more control variables are discussed next.

The Industry a firm is active in is used as a control variable. This variable is coded with a “0” or “1” dependent on their activities in the “Food & Drinks” or “Apparel & Textile” industry.

Lastly the variable Number of employees is used as a control variable. Pett & Wolff (2000) and Moine (1995) used the number of employees to measure firm size, and came up with satisfying results. However in our research, regarding the relationship Total assets has with our hypotheses and findings in the literature, we will not use the number of employees as a measurement for firm size, but as a control variable. Pett & Wolff (2000) and Moine (1995) used a categorization of small and large firms regarding their Numbers of employees variable. Here small firms are stated to have less then five hundred employees. They also divide the firms in four different groups according to their number of employees. Since this research uses the KLD and Compustat indexes, which consist of much larger firms, we cannot use the classification “less or more than five hundred employees”. Therefore we decide to use the number of employees as a continuous variable. By doing this we stay as close to the data as possible and do not have to make any assumptions to categorize the firms employees into certain employee groups. As a result our control variable Number of employees is as objective as possible.

4.3 Procedure

While the research will take approximately twelve weeks to complete, the aim is to extract all the data needed in little over two weeks. This is needed to ensure a reliable analysis of the results. As stated before the researchers made a coding scheme on forehand to overcome possible bias by firms being reviewed by different researchers. Then the (1) value-driven motives, (2) performance-driven motives and (3) stakeholder-driven motives are counted separately and threated as separated variables. These CSR motive variables are also

(27)

size is extracted from the Compustat index. Further the corresponding KLD scores are added to the database also.

Lastly the different variables are put together in one dataset and checked for inconsistencies. After ensuring that the dataset is correct, the analysis is pursued.

4.4 Analysis

Before testing our hypotheses, we will conduct several tests to preclude any form of non-normality in the data. Therefore our 8 variables, as described in the measurement section will be tested with descriptive statistics. The skewness, kurtosis and Kolmogorov-Smirnov tests are part of this descriptive analysis. If normality cannot be assumed the variables are transformed using transformation techniques to improve the normality of the variables. If a transformation is needed for a variable, this is mentioned in the result section. Further, after ensuring the variables are normally distributed the interrelatedness (correlation) between the variables is analysed.

Based on Baron & Kenny’s (1986) model, multiple regression analyses are used. All

hypotheses are tested using these multiple regression analyses. For every hypothesis first the control variables are added to model x, where after the control variables and the independent variable are added to model x+1. By examining the significance and the differences in variance explained by these models, we are able to draw conclusions solely based on the influence of the independent variable. Lastly regarding hypotheses testing, multicollinearity tests were included. These tests analyse if the independent variable in the multiple regression is highly correlated with the control variables. Obviously we do not want this to be the case because this would negatively influence the trustworthiness of our hypotheses tests. Lastly the betas of the independent variables are analysed in order to reject or support a hypothesis.

(28)

5. Results

In this section the results as found after analysing the data are presented. First the sample and sample size, assumptions of normality and the correlations between the variables are

described. Further, the hypotheses as stated in the theoretical framework are tested and the outcomes are discussed.

5.1 Sample

During our data collection, conducted by five researchers, a total of 150 firms was analysed in this study. Because of incompleteness or untrustworthiness of some of the data 39 of these firms were discarded. As a result our analysis is conducted based on a total of 111 firms. 51.4% of the firms were active in the “Food & Drinks” industry and 48.6% of the firms were active in the “Apparel & Textiles” industry. After using several variable transformation techniques needed to ensure the normality of the data, which is discussed in more detail in the next section, only a few missing entries remained. These are excluded whenever this is

important for testing the hypotheses. On average the KLD score for the firms is 0.077 (SD=0.093). The average amount of motives displayed is 11.95 (SD=11.97). Also the average amount of value-driven motives 4.83 (SD=6.63), performance-driven motives 3.41 (SD=4.64) and stakeholder-driven motives 3.71 (SD=3.45) were identified. Lastly, the average total assets of a firm were 11299 million dollars (SD=29114,02).

5.2 Assumptions of normality

Different normality tests were used to measure if the variables were normally distributed. As described in the method section these are descriptives, boxplots and Kolmogorov-Smirnov tests.

Regarding the first variable Total CSR motives the skewness (2.145) and kurtosis (5.234) of the sample were both positive, indicating that the scores were clustered to the left and its distribution to be peaked in the centre, meaning many scores were situated in the tails. The Kolmogorov-Smirnov test was significant meaning normality of the variable could not be assumed. In order improve the normality of the variable Total CSR motives we made use of a variable transformation technique. The variable was transformed using the log10(x+1)

(29)

transformation technique. As a result the skewness became slightly negative (-.258), and the kurtosis had a positive value of (0.191). Also the Kolmogorov-Smirnov test was no longer significant, indicating that the variable was normally distributed. As a result we could assume the variable Total CSR motives to be normally distributed.

The second variable Value-driven motives had a positive skewness (2.720) and kurtosis (9.416), indicating that the scores were clustered to the left and the distribution to peaked in the centre, meaning that many scores were situated in he tails of the distribution. Using the same transformation technique as described above, the skewness changed to a smaller positive value (0.262), and a negative kurtosis of (-0.730). The Kolmogorov-Smirnov still was significant, but regarding the small sample size and the scree plot showing a straight-line indicating normality, normality of this variable is assumed.

The variable Performance-driven motives showed a positive skewness (1.826) and kurtosis (3.176), indicating the scores to be clustered to the left and a pointy distribution. Also the Kolmogorov-Smirnov test was significant meaning non-normality of the variable. Using the transformation technique log10(x+1) as described before, the variable was transformed. This resulted in a negative skewness (-0.942) and positive Kurtosis (3.405). The Kolmogorov-Smirnov was significant, but regarding the decrease in skewness and kurtosis, the small sample size and the scree plot indicating a straight line, we assume normality of this variable.

The last variable regarding a firm’s CSR motives, Stakeholder-driven motives showed a positive skewness (2.537) and kurtosis (10.733). These scores indicated scores clustered to the left. Also regarding the kurtosis, the normal distribution was pointy and heavy tailed. The Kolmogorov-Smirnov test was significant indicating non-normality of the variable. By using the log10(x+1) transformation technique again, the new values of the skewness (0.001) and kurtosis (0.026) were lower. The Kolmogorov-Smirnov test was significant but regarding the scree plot indicating a straight line, the small sample size and the decrease in skewness and kurtosis, normality is assumed.

The variable regarding CSR performance had a positive skewness (1.515) and kurtosis (1.798). These scores indicate the data to be clustered to the left and slightly heavy tailed. The Kolmogorov-Smirnov test indicated this variable not to be normally distributed. Using

log10(x+1) as a transformation technique a new variable was computed. The skewness (1.352) and kurtosis (1.144) of this variable were positive and lower. The Kolmogorov-Smirnov test was significant but regarding the decrease in skewness and kurtosis, and the scree plot indicating a straight line normality is assumed.

(30)

The skewness (5.088) and kurtosis (30.689) value for Firm size were both positive. Also the kurtosis value was positive. These scores indicate more scores on the left side of the distribution, and many scores situated in the tails. The Kolmogorov-Smirnov test indicated this variable to be not normally distributed. Using log10(x) as a transformation technique a new variable was computed. The skewness of this variable was positive (0.455). The kurtosis value was negative (-1.188) The Kolmogorov-Smirnov test proved not to be significant indicating the variable to be normally distributed. Also the plots showed a straight line. As a result we assume the variable Firm size to be normally distributed.

The control variable Industry regarding the industry a firm is active in showed a positive skewness (0.055) and negative Kurtosis (-2.034). Regarding the plots the variable Industry is assumed normally distributed.

The control variable Number of employees had a positive skewness (4.379) and kurtosis value (23.641) meaning a left centred distribution with many scores in the tails. The Kolmogorov-Smirnov test was significant indicating non-normality in the variable. By computing a new variable using the log10(x) transformation technique, we tried to improve the normality of the variable. The skewness became slightly negative (-.0241) and the kurtosis positive (0.475). Lastly the Kolmogorov-Smirnov test was no longer significant, indicating normality of the variable. Therefore the control variable Number of employees is assumed normally distributed.

5.3 Correlations

After finding our variables to be normally distributed, the variables are tested for their correlation. By doing this we try to identify early relationships between the different variables. First the correlations regarding the hypotheses as stated in the theoretical

framework are discussed. Second other correlations found are described in more detail. An overview of all the correlations between the variables is available in Table 1.

First the correlation between the total number of motives a firm displays and its CSR performance is tested. No significant positive correlation was found between the variable total motives and CSR performance r(94)=.139, ns).

Second the correlations between the categorized motives a firm displays and its CSR performance was tested. The positive correlation between Value-driven motives and CSR performance did not prove to be significant r(94)=.165, ns). Also Performance-driven

The impact of CSR motives and firm size on CSR performance, and the influence of firm size on stakeholderdriven CSR motives