Master Thesis July 2014
“Motivation of Corporate Social Responsibility from a
Corporate Financial Performance Perspective”
Ronald Klijsen Student nr. 10545352
Institution: Amsterdam Business School (UvA) Business Studies – Strategy track
Date of submission: July 2014 Supervisor: Lars Moratis
2 ACKNOWLEDGMENT
This thesis is written as part of the Master program Business Studies with specialization Strategy at the Amsterdam Business School (UvA). The subject of this study was chosen as a result of my profound interest in strategic behaviour of businesses and my believe that sustainable decisions making is an important part of modern business.
There are a few people I would like to thank. First, my supervisor Lars Moratis for his helpful assistance during the design and writing of my thesis. I want to thank my family for their unconditional support during my studies. And there is one person that I want to acknowledge in particular. Fleur, my love, thank you for your advice, support and love during the writing of this thesis.
Ronald Klijsen Amsterdam July 2014
3
Content
1. Introduction to the topic 5
1.1 Corporate Social Responsibility 5
1.2 CSR and business performance 6
1.3 CSR and motivation 7
1.4 Problem statement & Academic relevance 7
1.5 Reader’s Guide 9
2. Literature Review 10
2.1 Impact of CSR 10
2.2 Communication of CSR 12
2.2.1 Credibility & Application 13
2.3 Corporate Social Motivation 13
2.3.1 Self-serving 14
2.3.2 Other-serving 14
2.3.3 Empirical Research CSM 15
3. Research design 17
4. Methodology Part 1: CSR motivation 18
4.1 Data 18 4.2 Methodology 19 4.2.1 Content Analysis 19 4.2.2 Ranking 21 4.2.3 Motivation classification 22 5. Results on CSR Motivation 24
6. Methodology Part 2: CSM & CFP 25
6.1 Methodology 25 6.2 Data 26 7. Results on CSM & CFP 27 7.1 Return on Assets 27 7.2 Net income 28 7.3 Efficiency 30
8. Discussion, limitations and recommendations 32
8.1 Discussion 32 8.2 Limitations 34 8.3 Recommendations 35 9. Conclusion 36 10. References 37 11. Appendix 43
4
Abstract
Corporate Social Responsibility (CSR) has become increasingly important for businesses. Many scholars investigated the concept of CSR and different topics within CSR. The reason for companies to act social responsible is called Corporate Social Motivation (CSM). This CSM can vary from financial oriented motives to altruism and moral obligation. It seems that CSR motivation is one of the topics that lacks a profound background, because it got only the attention of a few researchers. Nevertheless, since motivation for CSR is considered to be an underlying driver of CSR, this creates the potential for new insights. Corporate Financial Performance (CFP) is one of the most studied subjects since the start of business research. Various causes for CFP are indicated in academic research over time. The fact that some companies seem to have a financial motivation for CSR shows the relevance to investigate whether CSM is a cause of CFP. This study therefore focuses on the motivation behind CSR and the consequences of different types of motivation on CFP. The research question for this study is as follows: “Does a different CSR motivation lead to different corporate financial performance?”. A review of existing literature leads to two types of motivation, namely self-serving versus other-serving. Additionally, the literature review on this subject let come to mind that most of the research lacks of empirical studies, it requests knowledge from experience. In this study a unique method is developed in order distinguish these motives empirically in a real market setting. The Dutch market is analysed, and different motives are distinguished empirically. Quantitative methods are used to test whether a different CSR motive leads to different financial performance. This study finds that a different CSR motivation does not lead to different financial performance.
5
1.
Introduction to the topic
This first chapter introduces the topics of this study. First the concept of CSR is discussed, followed by the relation between CSR, motivation and business performance. Subsequently the research question and structure of the research are explained.
1.1
Corporate Social ResponsibilityCorporate Social Responsibility (CSR) refers to the notion that corporations need to be aware and take care of their environment. The degree to which companies succeed in this is called Corporate Social Performance (CSP). The concept has become a major topic in both management and business literature (Bontis & Serenko, 2009; Carroll & Shabana, 2010; McWilliams, Siegel, & Wright, 2006). This study builds on the definition of Elkington (1998), where CSR follows the triple bottom line (TBL) of economic prosperity, environmental quality and social justice. These TBL dimensions are commonly called the three Ps: People, Planet, Profits (3Ps). The SER (Social Economic Council of the Netherlands) states that the Dutch market approaches CSR in a similar way. In addition they argue: “CSR is the conscious direction of business activities towards creating value in three dimensions in the longer term: not only in terms of financial-economic variables, such as profitability and shared value, but also in an ecological and social sense” (SER, 2001, pp. 17–18). In this way, they follow the TBL dimensions of Elkington (1998).
In order to acquire an understanding of the concept of CSR, a short summary of the development of CSR over time is provided. The use of the concept of corporate social responsibility started around sixty years ago by the work of Bowen (1953). His work elaborates on earlier work, for example from Adolf Berle and Merrick Dodd which questioned the concept of rightful stakeholders since 1930 (Macintosh, 1999). Bowen’s 1953 article argues that the social consciousness of managers goes beyond the profitability of the firm, an important expansion on the traditional view of managerial responsibilities. Another important development is the introduction of the stakeholders theory (Freeman, 1984). This theory introduces the ‘concept of interdependency’ and implies a complex, multi-directional relationship between various stakeholders of an organization (Donaldson & Preston, 1995; Freeman, 2010).
Currently, CSR is considered to be a more integral part of firm’s management rather than a normative concept (Carroll, 1999). The discussion addressed by Porter & Kramer (2011) is an example of this. They argue for a more balanced perception of CSR by focussing more on firm profit rather than solely on public benefits. Very influential in that discussion is the work of Friedman (2007), who argues against pro-CSR arguments by stating that businessmen are first and foremost responsible for increasing the profit of the shareholders.
In line with this integral approach is the increasing focus in academic literature on the strategic implications of CSR (Aaronson, 2009; Hart & Ahuja, 1996; Husted & De Jesus Salazar, 2006; Porter & Kramer, 2006; Vaara & Durand, 2012); on the one hand research indicates a relation between CSR
6 and business performance (Husted & De Jesus Salazar, 2006; Orlitzky, Schmidt, & Rynes, 2003), while on other hand researchers doubt the added value of CSR in business performances (Jensen, 2002; Windsor, 2001; Wright & Ferris, 1997). This discrepancy on strategic implications underlines the importance of a thorough understanding of the circumstances under which CSR can be part of a firm’s success.
1.2
CSR and business performanceFrom a theoretic perspective, the concept of Corporate Financial Performance (CFP) is the time test of any strategy. Therefore it is at the centre of strategic management (Schendel & Hofer, 1979). From an empirical perspective business performance is used to examine a variety of strategy contents and processes (Ginsberg & Venkatraman, 1985). For example, Nash (1983) indicates the evident managerial importance of business performance, as it offers prescriptions for performance improvement. Although the effects of CSR on business performance have received a lot of attention in academic literature, it seems to be found ambiguous and the theme is being approached inconsistently.
Even though CFP is of high importance, in academic CSR literature there are incongruent results. Recent studies often refer to the research from Orlitzky, Schmidt, & Rynes (2003) that states: “Most theorizing on the relationship between corporate social/environmental performance and corporate financial performance assumes that the current evidence is too fractured or too variable to draw any generalizable conclusions.” (p. 403). In their meta-analyses the researchers suggest that CSR is likely to pay off. This almost neutral outcome seems to lie in the middle of the two extreme views of on the one hand (strong) positive relation (Al-Tuwaijri, Christensen, & Hughes Ii, 2004; Porter & Kramer, 2011; Porter & Van der Linde, 1995; Russo & Fouts, 1997; Van Beurden & Gössling, 2008) and on the other hand (strong) negative relation (Benito & González-Benito, 2006; Khanna & Damon, 1999; Wagner, Van Phu, Azomahou, & Wehrmeyer, 2002).
Hence, it is interesting that the research on business performance, for example Al-Tuwarijri et al. (2004) uses a variety of variables to measure business performance, as most research on CSR and CFP does (Griffin & Mahon, 1997). Venkatraman & Ramanujam (1986) restate that performance measurement is a recurrent theme for as well academic scholars as practicing managers. Additionally the researchers argue that although the importance of the performance concept is widely recognized among theoretical, empirical and managerial perspective, the treatment is diverse among research settings. The scope of applied performance measurements ranges from solely financial performance to operational performance and efficiency performance (Hofer, 1983). This suggests that proper selection of the different indicators to measure business performance is essential. This study therefore explicitly explains which indicators of financial performance are used in this study and what their specific implications are. Due to proper selection of indicators this study extends current academic understanding of the relation between CSR and CFP.
7
1.3
CSR and motivationResearch indicates that effects from Corporate Social Performance (CSP) are influenced by Corporate Social Motivation (CSM), the motivation behind CSR. For example Aguilera, Rupp, Williams, & Ganapathi (2007) argue that “An important new line of inquiry within this field is no longer if CSR works, but rather what catalyses organizations to engage in increasingly robust CSR initiatives and consequently impart social change” (p. 3). Other researchers agree on the influence of CSM on CSP (Brønn & Vidaver-Cohen, 2009; Galaskiewicz & Colman, 2006; Graafland & Mazereeuw-Van der Duijn Schouten, 2012; Graafland & van de Ven, 2006). For example Aguinis & Glavas (2012) argue that CSM is an important predictor of CSP. Additional to this, some researchers investigated the influence of CSM on Corporate Financial Performance (CFP): the CSM-CFP relation (Becker-Olsen & Hill, 2005; Ellen, Webb, & Mohr, 2006; Husted & De Jesus Salazar, 2006; Lougee & Wallace, 2008). Hence, this last approach is innovative since the majority of research only investigates influences of CSM on CSP or separately the influence of CSP on CFP.
As an example of CSM research Becker-Olsen & Hill (2005, p. 48) differentiate between “firm-self serving (e.g., to increase profits, sales or boost a specific brand)” and “public serving (e.g., help needy citizens, assist with community development or raise awareness for a specific cause)”. They demonstrate that some socially motivated programs may negatively impact consumer behaviour responses to promotions. Another example is the research of Graafland & Mazereeuw-Van der Duijn Schouten (2012) that differentiates between extrinsic motivation (financial motivation) and intrinsic motivation (non-financial motivation). In their research intrinsic motivation is further divided into CSR as moral duty and CSR as an expression of altruism. By digging deeper into the different categories of CSP they argue that the type of motive that is effective differs per CSR category.
Former researchers, who studied CSM in a real market setting, have based their findings on various methods. These methods vary from interviews (Babiak & Trendafilova, 2011) to questionnaires (Brønn & Vidaver-Cohen, 2009; Graafland & Mazereeuw-Van der Duijn Schouten, 2012) and analyses of CSR communication (Deegan & Rankin, 1996; Hedberg & von Malmborg, 2003).
This study reacts on the quest of Brønn & Vidaver-Cohen (2009) and Becker-Olsen & Hill, (2005) who both underline the importance of further empirical research on CSM. The former indicates a potential added value in the research to the CSM-CSP relation, while the latter stresses the importance of empirical study in the research to the CSM-CFP relation.
1.4
Problem statement & Academic relevanceAcademic literature proposes further analysis of CSR, CSM and their relation with CFP. The first quest, to expand knowledge of CSR, can be found in the extensive recent literature review of Aguinis & Glavas (2012). They emphasize the importance to expand the knowledge about CSM, since it is one of the underlying mechanisms of CSR. This is in line with other studies that indicate the knowledge
8 gap regarding CSR and, in particular, it’s foundations (Aguilera et al., 2007; Margolis & Walsh, 2003; Wood, 2010). Williamson, Lynch-Wood & Ramsay (2006) specifically address the lack of empirical research in this regard. The second quest, further exploration of the CSM field, can be found in the research of Babiak & Trendafilova (2011). They highlight both the opportunity and the importance of further investigation on CSM as driver of CSP. This is in line with the notification of Williamson et al. (2006) that argues the need for empirical research. Siegel & Vitaliano (2007) studied the drivers of CSR empirically and agree with McWilliams et al. (2006). Siegel & Vitaliano (2007) further indicate the limitations of the use of cross-sectional data and advise to use different moments in time instead. This study empirically investigates the combination of CSM and CSP and by doing this it recognizes development over time. The third and last quest comes from researchers like Orlitzky et al. (2003), who investigated the academic field by meta-analyses, indicate the lack of consistent findings concerning CFP.
Literature lacks a thorough understanding of CSM, CSP and CFP, and moreover the inter-relation of these three. Figure 1 (appendix I) shows a schematic representation of this interinter-relation. Although not extensively, the CSM-CSP relation is investigated by different authors (Aguinis & Glavas, 2012; Brønn & Vidaver-Cohen, 2009; Galaskiewicz & Colman, 2006; Graafland & Mazereeuw-Van der Duijn Schouten, 2012; Graafland & van de Ven, 2006) and is represented by ‘A’. As indicated the CSP-CFP relation also requires further investigation, since there are contradicting results. This CSP-CFP relation is represented by ‘B’ in Figure 1. It seems to be the case that academic literature for the larger part ignores the direct influence of CSM on CFP. When following that line of reasoning, the influence of CSM on CFP needs to occur via CSP, as can be seen in Figure 1. This implies that CSP acts as a connector of CSM and CFP, and consequently: CSM can only influence CFP by change in CSP.
As indicated the research of Becker-Olsen & Hill (2005), Husted & De Jesus Salazar (2006), (Ellen et al., 2006) and Lougee & Wallace (2008) can be classified as innovative since they highlight the CSM-CFP relation. These studies do not just elaborate on the concept of CSP when investigating drivers like CSM, but they also make a direct link from CSM to CFP. Taking these findings into account, it appears that the majority of former literature lacks recognition of the relationship between CSM and CFP. The schematic
representation of
Figure1 seems to
lack coverage of the bigger picture of CSR, CSM and CFP. A triangular schematic approach of the relations is given by Figure 2 (appendix II). The CSM-CFP relationship is added in here and represented by ‘C’.Based on the findings in this chapter, this study has the following research question:
“Does a different CSR motivation lead to different corporate financial performance?”
This study investigates the direct relation between CSM and CFP. First of all deep understanding of CSM is necessary. Subsequently, methods needs to be explored in order to distinguish CSM
9 empirically. A real market setting, the Dutch market, is investigated to 1) distinguish CSM and 2) test whether different CSM leads to different CFP.
In order to answer the research question, the following sub questions are: - Which different motivations for CSR can be distinguished in theory? - How can different CSM be distinguished empirically?
1.5
Reader’s GuideThis study contains nine chapters. In the first chapter the concept of CSR is discussed, followed by the relation between CSR, motivation and business performance. Subsequently the research questions and structure of the research are explained. The second chapter focuses on the theories related to the research question. Here is discussed which different motivations for CSR can be distinguished in theory (sub question one) and this leads to three hypotheses. Chapter 3 explains the research approach of this study. Chapter 4 contains the first part of the research methodology. Here is explained how different CSM can be distinguished empirically (sub question two). The answers of the first part can be found in chapter 5. Chapter 6 investigates whether a different CSR motivation lead to different financial performance. Results of this second part can be found in chapter 7. Chapter 8 discusses all findings of this study and indicates the limitations of this study. A general conclusion is given in chapter 9.
10
2.
Literature Review
Different researchers address the fields relevant for this study; Corporate Social Responsibility (CSR), Corporate Social Performance (CSP), Corporate Social Motivation (CSM) and Corporate Financial Performance (CFP). Most of the work includes a combination of multiple fields. This literature review discusses the state of the art of academic literature concerning these topics. By exploring these specific fields of literature, this study aims to find an answer to the first sub question of this thesis; What type of motives can be found?
2.1
Impact of CSRNowadays literature is highly concerned with the incongruence of former literature, that analysed the effects of CSR on business performance (Orlitzky et al., 2003). In order to investigate effects on business performance it is important to understand the different arguments of former literature and to discuss these differences. As indicated by McWilliams et al. (2006), varying definitions of the concept business performance may be the reason of the different outcomes. Therefore this section highlights the differences of various definitions and approaches that are used in academic literature.
The research of Orlitzky et al. (2003) covers an extensive review of former literature on the CSP-CFP relation. They point towards an increasing recognition of a positive relation. This is confirmed by the work of Giménez Leal, Casadesús Fa & Valls Pasola (2003) who found strategic benefits from CSP. Also Du, Bhattacharya, & Sen (2010) argue in favour of this positive relation, by finding evidence of increased financial performance due to CSP. Reason for a positive impact is for instance given by Sen, Bhattacharya, & Korschun (2006), who found positive company related associations among consumers (e.g. purchase intentions). This is in line with the research of Becker-Olsen & Hill (2005) that highlights the increase of positive consumer perception when companies’ CSR motivation fits their communication. This way the latter shows CSM is a key driver of financial rewards. Other authors that agree on the positive relation are Joyner & Payne (2002), Margolis & Walsh (2003) and Orlitzky, Siegel, & Waldman (2011).
Nevertheless, for many years academic researchers have been less convinced about the positive impact of CSR. Friedman (1962) is considered the first author to propose this view. Other researchers found different, positive findings, but he does not agree with this (Husted & De Jesus Salazar, 2006). To understand his reasoning Husted & De Jesus Salazar (2006) suggested the thoughts of Friedman as follows: “Although social performance may be compatible with profits under certain conditions, it is not consistent with maximizing profits and thus with creating value for stockholders” (Husted & De Jesus Salazar, 2006, p. 75). Also Hillman & Keim (2001) argue that social issue participation is negatively associated with shareholder value after studying S&P500 firms and their market value. They suggest that easiness of copying by competitors is the reason. Hence, this research ends with suggesting further research to analyse the motivation behind social issue participation.
11 Griffin & Mahon (1997) emphasize the methodological inconsistencies of former literature when investigating the CSP-CFP relation. One solution they use to overcome these inconsistencies is the selection of five most commonly applied accounting measures in the CSR and CFR literature. Their results indicate that “the a priori use of measures may actually predetermine the CSP-CFP relationship outcome” (p. 6). The article stresses the importance of a proper selection of CSM, CSP and CFP indicators.
Alternative solutions to study the impact of CSR are suggested. For example the recent study of Barnett & Salomon (2012), that recognizes the inconsistency of academic literature in defining this relation. Different from most research they dig deeper into the problem, by searching for a U-shaped relation between CSR performance and financial performance. In line with the arguments of Griffin & Mahon (2012) in the former paragraph they used various, explicitly described indicators of CFP. The main indicator they use for financial performance is the Return on Asssets (RoA), in line with the method of Waddock & Graves (1997). This continuous ratio variable is defined as net income of a firm divided by its total assets. Next to RoA Barnett & Salomon (2012) use net income as indicator for financial performances. This continuous variable is defined as the earnings (after interests, taxes, depreciation and amortization) of a firm.
Another innovative research on the impact of CSR is done by Lys, Naughton & Wang (2013). The latter elaborate further on the finding that CSR has a positive relation with financial performance. After they provide evidence on the fact that CSR is not charity and neither improves future financial performance, they rather state that companies undertake CSR expenditures in the current period when they anticipate stronger future financial performance. This way they meddle in a discussion of causality and suggest that: “the causality of the positive association between CSR expenditures and future firm performance is different from what is claimed in the vast majority of the literature, and that corporate accountability disclosures are another channel by which firms convey financial prospects to outsiders” (Lys et al., 2013, pt. abstract). So in their view, CSR expenditure can be seen as a signal for expected stronger future financial performance.
It is also argued that the impact of CSR also covers deeper mechanism of the organization. For example Porter & Kramer (2011) argue that CSR not only leads to higher revenue, due to positive attitude of consumers, but that CSR also leads to lower production costs, due to a more sustainable production. Following that line of reasoning CSR would lead to more efficient organizations.
Other researchers include the motivation behind CSR in order to determine the impact of CSR. For example Husted & De Jesus Salazar (2006) find a positive relationship between CFP and strategic CSR investments. The third section of this chapter elaborates further on CSR motivation and its impact.
12
2.2
Communication of CSRFormer CSR literature has shown the difficulty to measure CSR, CSP and CSM (Wood, 2010). For example, most of the scholars that investigated CSP were not able to measure the amount of pollution, or the impact on society, manually. Neither do researchers in the field of CSM have a comprehensive look into the mind of managers. To overcome these kind of problems, large part of the CSR literature includes CSR communication into its analyses (Wood, 2010). Similarly for this study, it is important to get insight into the different views on communication about CSR. It gives insight into how companies communicate their CSR and why they communicate about CSR. Deeper insights into communication allows deriving empirical information from companies’ communication. Closely related to CSR communication is CSR credibility, since no credible communication leads to misinterpretation of the real facts and thoughts. It is essential to get a deep understanding of the credibility of CSR and CSR communication in order to derive any conclusion from it.
Businesses’ attention to CSR communication increased over the years. The reason for this increase is given by Tate, Ellram, & Kirchoff (2010) argue: “Firms are increasingly under pressure from stakeholders to incorporate the triple-bottom line of social, environmental and economic responsibility considerations into operations and supply chain management strategies” (p. 19). Research of KPMG (2011) towards CSR indicates that current sustainability reporting is still not uniform. Hence, besides the increased and dominant usage of standalone reports some companies combine social and environmental information with financial information in their annual report or publishes the CSR information on their websites. Different from other reporting types, a standalone report solely discus CSR issues. Kolk (2004) reviews sustainability reporting and agrees on the existence of different channels to communicate about CSR.
Due to the fact that stakeholders become more demanding towards companies, companies’ use of sustainability reports has increased in the past 50 years (Tate et al., 2010). What started as employee reporting moved via social reporting, environmental reporting and triple bottom line reporting to the current state of sustainability reporting (Buhr, 2007). As further argued by Buhr (2007) by the turn of the millennium the focus of sustainability reporting moved towards linking social, economic and environmental aspects of corporate performance. This is known as triple bottom line (TBL) reporting (Buhr, 2007). Although the terms social reporting, environmental reporting, corporate social responsibility (CSR) reporting, triple bottom line reporting and sustainability reporting are not identical, they are strongly related. Hence, in order to facilitate the readability of this study, the term sustainability reporting is used as umbrella term.
By analysing the sustainability reports of companies, Tate et al. (2010) found “unique insights regarding corporate communications that other methodologies would not found” (p. 19.), which indicates how powerful analyses of sustainability reports can be. Kolk (2008) indicates that sustainability reporting can be done by both the annual report and a separate sustainability report.
13 Based on these arguments it can be concluded that CSR communication is useful and widely accepted in academic research.
2.2.1 Credibility & Application
Former literature discusses the contents and benefits of CSR communication, and the credibility of CSR communication. In order to be able to draw conclusions from sustainability reporting understanding the different views of their credibility is necessary.
Forehand & Grier (2003) provide insight into the credibility of CSR communication by investigating when honesty is the best way for companies to communicate with their consumers. They state that ‘‘the negative effect of consumer skepticism regarding a firm’s motives can be inhibited by public acknowledgment of the strategic benefits to the firm” (Forehand & Grier, 2003, pt. Abstract). This is confirmed by a recent study of Kim & Lee (2012) that confirms the importance of being honest about CSR motives. Du et al. (2010) argue that companies should use credible communication channels to inform stakeholders about their CSP. The latter agreed that sustainability reports are an effective manner to do so. Other scholars are less positive about the credibility of sustainability reporting. Adams (2004) for example found in his case study a portrayal gap between corporate performance and reporting.
Some authors imply influence of CSM on CSR communication. Du et al. (2010), for example indicate that financial benefits from CSR engagement requires effective CSR communication. Other researchers argue that some companies used CSR solely for marketing purposes (Bronn & Vrioni, 2001; Gallego-Álvarez, Prado-Lorenzo, Rodríguez-Domínguez, & García-Sánchez, 2010).
The way sustainability reports are used in academic literature varies as well. Most studies analyze sustainability reports in order to find proxies, for example for environmental disclosure. One method to analyze reports is to focus on the quality of the reports (Dragomir, 2010). Another method is to focus on quantity of the reports, which is a more objective volumetric measure (Verrecchia, 1990). Villiers & van Staden (2011) argue that “the volume of disclosure is regarded as measure for the importance that managers place on the information.” (Villiers & van Staden, 2011, p. 512). Although a proxy for CSR motivation is not clearly developed yet, former methods can be used to develop a such a proxy for CSR motivation.
2.3
Corporate Social MotivationAcademic scholars have accepted the impact of Corporate Social Motivation (CSM). As Aguilera et al. (2007) argue, the field of CSR is searching for underlying drivers of CSR, instead of solely the impact of CSR as such. The recent developed shared value view on CSR refers to motivation as being one of the main success factors for successful CSR (Porter & Kramer, 2011). The majority of the studies focus on CSM is concerned with the different types of CSM. It divides these different CSM types based on companies’ reasoning behind CSR. Consequently this division is related to companies’
14 objectives to create CSP. This section further elaborates on the different types of motivation, how they are being referred to in business literature and how this study uses them.
It can be seen that most of the researchers divide CSM into two opposite types. Forsyth, Berger, & Mitchell (1981) introduced the terms self-serving versus other-serving when studying human behaviour. Both terms are used in this study, as umbrella term for the two motives; self-serving motivated (SS-CSM), where the main aim is financial related, versus other-serving motivated (OS-CSM), where the main aim is non-financial related. The next paragraphs explains the use of the terms self-serving and other-serving in this research and how there is referred to in academic literature.
2.3.1 Self-serving
There are different ways in which academic literature refer to behaviour that can be expressed by the term self-serving. Graafland & van de Ven (2006) investigated the Dutch market and refer to it as positive strategic view on CSR which “implies that companies believe there exists a win-win relationship between CSR and financial success of the company” (Graafland & van de Ven, 2006, p. 2) and that “there are sound reasons to believe that in many cases ethics pays, which provides an important incentive to adopt a proactive stance with respect to CSR.” (Graafland & van de Ven, 2006, p. 2). Like others they refer to it as profit(able) motive, financial motive, economic motive or consequential reasons. Other studies also refer to it as extrinsic (Graafland & Mazereeuw-Van der Duijn Schouten, 2012) or shareholders motive (Tudway & Pascal, 2006).
The reasoning behind this motivation type can be found in the study of Miles & Covin (2000), who argue that reputational advantages (that enhance marketing and financial performances) can be created by environmental stewardship. This is in line with Sen et al. (2006) who argue: “CSR activity has the potential to increase not only CSR associations, attitudes, and identification but also the intent of stakeholders to commit personal resources (e.g., money, labour etc.) to the benefit of the company” (p. 164). This quotation explains why SS-CSM companies participate in CSR. But also current and potential employees can be positively influenced by a good CSR reputation (Greening & Turban, 2000). The latter further argue that positive influence employees can lead to better financial performance. As Graafland & van de Ven (2006) also argue, in line with Hart & Ahuja (1996) and Moore (2001), there is evidence found that CSR, without explicitly considering reputation, pays off for companies. Furthermore it is argued that CSR can help businesses avoid regulation (Lougee & Wallace, 2008). The benefits as mentioned above are part of the created value, the win-win, that Porter & Kramer (2011) refer to.
2.3.2 Other-serving
Business literature refers in different ways to behaviour that can be expressed by the term other-serving. Graafland & van de Ven (2006) refer to it as a positive moral view on CSR. They argue that “Many companies have a business culture that upholds certain business principles according to which
15 CSR is perceived as a moral duty of the firm.” (Graafland & van de Ven, 2006, p. 2). Other terms for OS-CSM in business research are moral obligation, non-financial motive (Graafland & Mazereeuw-Van der Duijn Schouten, 2012), social motivation (Becker-Olsen & Hill, 2005), social performance (Husted & De Jesus Salazar, 2006) and deontological motive (Etzioni, 2010). Graafland & Mazereeuw-Van der Duijn Schouten (2012) refer to is as intrinsic motivation when combining the principles of altruism and ethics. They explain the differences between self- versus other-serving by stating that “Non-financial motives often reflect intrinsic motives that perceive CSR as an end in itself, independent from (financial) benefits.” (Graafland & Mazereeuw-Van der Duijn Schouten, 2012, p. 380). The latter make a separation of OS-CSM into moral duty and altruism. The moral duty to be socially responsible can be derived from ethical principles of moral philosophy or from religious principles (Graafland & Mazereeuw-Van der Duijn Schouten, 2012). As Etzioni (2010) argues this means that one is not acting because it is enjoyable, but rather because one feels obliged to do something. Ribar & Wilhelm (2002) divide altruism into pure altruism, which is valuation of CSR because of positive societal consequences, and impure altruism, which is private enjoyment from performing CSR. Both moral duty as the different forms of altruism can be covered by the concept of OS-CSM.
2.3.3 Empirical Research CSM
There are several studies concerning the motivation of CSR behaviour. Brønn & Vidaver-Cohen (2009) found ambiguous results when studying Norwegian companies on strategic versus moral drivers for CSR. They argue that on average both SS- as well as OS-CSM are relevant, with self-serving being slightly dominant. The study explicitly highlights the need for quantitative research in the field of CSR motivation, since most of the research regarding the topic is qualitative. Research by Becker-Olsen & Hill (2005) shows how consumers’ perception of companies’ CSR motive seems to be very influential on CSR effectiveness. They emphasize the importance of fit and state the following about it: “Interestingly, when the firm is viewed as motivated by firm centred interests there is not a reduction in perceived corporate credibility.” (p. 50). They argue that the influence of motivation is less compared with the influence of fit. This is consistent with Forehand and Grier’s (2003) notion that scepticism is not driven simply by a firm being profit motivated but rather by a discrepancy between stated objectives and firm actions. Slightly different, the research of Ellen et al. (2006) focuses on the motivation of customers. They find that there is no influence of being self-serving or other-serving motivated when judging a companies’ CSP.
There are studies that emphasis that the impact of SS-CSM is bigger than the impact of OS-CSM. For example Lougee & Wallace (2008), who find that companies use CSR mainly as a form of risk management. An important assumption in their research enables them to disentangle the influence of different motivation types. They assume that other-serving companies are more likely to invest in both increasing CSR strengths and decreasing CSR concerns. Self-serving companies solely focus on
16 maximizing profit from their CSR investment. Lougee & Wallace (2008) finally argue that SS-CSM is dominant in the contribution to CSP. Graafland & Mazereeuw-Van der Duijn Schouten (2012) divide CSR into the categories; labor, environmental and social aspects. They find that “[…] for the social aspects of CSR, executives are significantly more driven by intrinsic motives than by extrinsic motive. But also for environmental aspects of CSR, intrinsic motives provide stronger stimulus than extrinsic motive” (Graafland & Mazereeuw-Van der Duijn Schouten, 2012, p. 377)
Other researchers found stronger impact of OS-CSM, compared to SS-CSM. For example the study of Galaskiewicz & Colman (2006), who investigates manager’s motives for corporate philanthropy. They argue the other-serving motive to be more important that other personal interests or interests of the company. The reason is that respondents genuinely believed in the duty of businesses to improve local communities in order to create a better world. Miles & Covin (2000) argue that CSR can improve a firm’s reputation and this way improve CFP. They found a weak correlation between SS-CSM and CSP, however they found the CSR implementation to be strongly influenced by OS-CSM.
As both former paragraphs indicate there are different views on which CSM type has the most impact. Nevertheless it is important to mention that firms often have both types of motivation to a certain degree. Research of Ellen et al. (2002) indicate the existence of mixed-motives that consumers often do have. It can therefore be expected that all firms do have a mixed motive as well. Nevertheless it is possible they have one dominant motive.
Overall, CSR motivation received enough attention in academic literature to recognize the existence of different types of motivation. Based on the literature mentioned in this chapter the first sub question of this study can be answered as follows; two motives can distinguished in literature, namely self-serving versus other-serving. Literature further implies different influences of CSM on CSP and CFP. Also can be concluded from this literature review that different indicators of CFP behave differently and that different conclusions can be drawn from different CFP indicators. Therefore this study tests the following hypotheses:
H1: Companies with a different CSR motive have different return on assets in different points in time.
H2: Companies with a different CSR motive have different efficiency rates in different points in time.
17
3.
Research design
The main objective of this study is to investigate whether CSM has influence on CFP. Therefore it tests whether firms with different motives indeed perform different financially. In order to do so it is necessary to split this study into two consecutive parts.
The first part of this research investigates CSM solely. Literature review indicates the existence of two types of CSR motives. This first part investigates CSM in a real market setting, namely the Dutch market. Therefore an empirical method is developed to classify companies. This method needs to measure a company’s CSM. This answers the second sub questions of this research; How can different CSM be distinguished empirically?
Subsequently, once companies are classified based on CSM, the second part of this research investigates CFP and tests whether companies with different CSM differ in CFP. This second part elaborates on the sample of selected companies from the first part this study.
18
4.
Methodology Part 1: CSR motivation
This first part of the research focuses on CSM. The amount of research on CSR motivation is limited, compared to the amount of research that is performed on topics like financial performance. Nevertheless various authors do investigate CSR motivation (Brønn & Vidaver-Cohen, 2009; Ellen, Mohr, & Webb, 2002; Graafland & Mazereeuw-Van der Duijn Schouten, 2012; Graafland & van de Ven, 2006; Vries, Terwel, Ellemers, & Daamen, 2013). The methodology is in part inspired on their work.
4.1
DataFor this research a database was built. This section explains how the database has been established, how the sample is selected and what requirements had to be met.
This study uses The Transparency Benchmark 2012 as starting point for the sample selection. This list facilitates selection of companies that 1) are social responsible, 2) are transparent and 3) operate on the Dutch market. These requirements are further explained in next paragraphs. This study selected the top 150 listed companies of the benchmark in order to have enough companies for valid results in further analyses and at the same time selecting only the companies that can be qualified as highly transparent.
Since study investigates CSR motivation solely, instead of CSR performance as such, the first requirement for selecting companies for this study is; companies need to act, at least at a minimum level, social responsible. Violation of this requirement could lead to inclusion of companies that cannot be classified based on CSM due to lack of overall motivation. Eccles et al. (2012) indicate the strong relationship between transparency and the degree of sustainability. Since the strong reliance of this study on companies’ reporting, the second requirement is; companies need to be as transparent as possible. Violation of this requirement could lead to delusive representation of companies’ motivation. This can e.g. be the case for companies that greenwash their reporting (Laufer, 2003; Ramus & Montiel, 2005). A third requirement for selecting companies is their operation on the Dutch market, which reduces the so called country effects. Adams, Hill, & Roberts (1998) found that the country of establishment can influence several characteristics of sustainability reporting. It is not the aim of this paper to totally exclude country effects, but it rather tries to minimize them this way. Consequently potential country effects, e.g. due to firms operating contemporaneously at foreign markets in countries outside the Netherland, are neglected.
Sustainability reports for the years 2008-2012 are downloaded by hand from the corporate websites of all the 150 top listed companies. In the cases there is no individual sustainability report available, an alternative report is created by selecting the sustainability chapter of the particular year annual report. Expert opinion of the researcher is involved by deciding whether sustainability chapters of the annual report are representative enough to be treated as sustainability report. Although variances between individual sustainability reports and those specific parts of the annual reports is not expected,
19 there is controlled for this, as described in the next section. Besides, the report needs to be written in English in order to perform the further analysis. As summarized in Table 1 (appendix III) this results in a database of 441 usable annual sustainability reports representing the top 150 of the Transparency Benchmark 2012. In order to control later in this research for changing motivation, due to e.g. inconsistent strategy, a minimum of three yearly reports in the period 2008-2012 per firm is required to keep the company into the sample. This results in a selection of 90 companies which published a total of 396 reports usable for further analysis.
4.2
MethodologyThe goal of Part 1 of this study is to classify the firms on CSM. This section describes and explains the various methods that are used. Three methods are going to be discussed. The first method is content analysis, including the creation of word lists and use of ratio’s. Secondly, the use of rankings and thirdly the motivation determination, including the final classification.
4.2.1 Content Analysis
Like in many CSR research, content analysis is seen as “a technique for gathering data that consists of codifying qualitative information in anecdotal and literary form into categories in order to derive quantitive scales of varying levels of complexity” (Abbott & Monsen, 1979, p. 504). Investigation of CSR in general has seen a variety of methodological approaches. Vourvachis (2007) focuses on the use of content analyses in CSR research, and by doing so he highlights the following alternatives used in literature; case studies (Adams, 2004; Cormier & Gordon, 2001), interviews (Babiak & Trendafilova, 2011), surveys (Brønn & Vidaver-Cohen, 2009; Graafland & Mazereeuw-Van der Duijn Schouten, 2012), longitudinal studies (Guthrie & Parker, 1989), experiments (O’Donovan, 2002) and theoretical investigations (Unerman & O’Dwyer, 2006). Nevertheless Vourvachis (2007) argues, based on a quote of (Milne & Adler, 1999), that “… the research method that is most commonly used to asses organisations’ social and environmental disclosures is content analysis.” (Milne & Adler, 1999, p. 237). These content analyses can either have an index approach or a volumetric approach. The volumetric approaches can be subdivided further by their coding units; words, lines, sentences, themes, standardised proportion of pages and page size. Vourvachis (2007) argues that although the use of words as unit of analysis for content analyses is criticized (Milne & Adler, 1999; Unerman, 2000), it has been employed in a number of previous studies (see for example Deegan & Gordon, 1996; Deegan & Rankin, 1996; Wilmshurst & Frost, 2000). Deegan & Gordon (1996) argue in favour of this method that “By counting words, which are the smallest possible unit of analysis, maximum robustness to error in calculating quantity is achieved.” (Deegan & Gordon, 1996, p. 189). In line with this Wilmshurst & Frost (2000) argue that “Words are the smallest unit of measurement for analysis and can be expected to provide the maximum robustness in assessing the quantity of disclosure.” (Wilmshurst & Frost, 2000, p. 16). Next to that they argue “words are a preferred
20 measure when it is intended to measure the amount of total space devoted to a topic and to ascertain the importance of that topic” (Wilmshurst & Frost, 2000, p. 17). Given these arguments, this study makes use of an volumetric approach with words as coding unit.
The methods used in the working paper of Eccles et al. (2012) are somewhat comparable to the methods that part 1 of this study uses and can therefore be seen as validation of the methods. In their study on CSR performance, Eccles et al. (2012) measured the balance between financial and nonfinancial discussion in conference calls by classifying the words of that particular conference call into either financial or nonfinancial. This method makes use of wordlists can be referred to as content analysis. Eccles et al. (2012) also construct a ratio that measures the amount of nonfinancial keywords over financial keywords. This enables them to conclude that one group of firms discusses more frequently about nonfinancial aspects. Additionally they test financial reporting of companies on integration of social information and find both higher social- and environmental information in high sustainability companies compared to low sustainability companies.
This study applies the formerly discussed method of Eccles et al. (2012) to classify companies, based on their motivation. Via content analysis companies’ sustainability reporting on a five year period is analysed. Besides their similarities, it is important to stress the difference in subject between the two studies. Where Eccles et al. (2012) focus on sustainability performance, this research focuses on the motivation behind their sustainable actions. Besides, Eccles et al. (2012) use content analysis to analyse conference calls, while this study uses it to analyse sustainability reporting. Although it was not their main focus, Eccles et al. (2012) also study sustainability reporting, but do this with a different purpose and in a manner that is not relevant for this study. Literature review of this study already stress the credibility of annual reporting. Like other researchers (Deegan & Rankin, 1996; Hedberg & von Malmborg, 2003), this research focuses on companies’ sustainability reporting. Additionally to the methods Eccles et al. (2012) use, this research continuous thereafter with the use of a ranking methodology in order to classify the companies.
The first step of the content analyses is creation of the so called wordlists. Each list contains words that are specifically related to one motivation type. The degree to which a company’s motivation is self- versus other-serving is measured by how many times those words are counted. An iterative process by a three-step-approach is used to create two wordlists. The first step is a literature study where academic articles (Cerin & Karlson, 2002; Graafland & Mazereeuw-Van der Duijn Schouten, 2012; Porter & Kramer, 2011) are analysed to find words that are specifically related to one type of motivation. Cerin & Karlson (2002) is analysed since it focuses on business incentives for sustainability. Graafland & Mazereeuw-Van der Duijn Schouten (2012) is analysed because this study focuses on the two types of motivations on the Dutch market. Porter & Kramer (2011) is analysed because this study is much-discussed in recent literature and focuses on the reason to perform CSR.
The second step is expertise selection, were two sustainability reports (Nestlé’s 2012 CSR report and Unilever 2012 CSR report) are analysed to find words that are specifically related to one
21 type of motivation. Nestle’s 2012 report is analysed since it is often referred to in relevant academic literature (Porter & Kramer, 2011). Unilever 2012 CSR report chosen randomly from the sample of firms operating on the Dutch market.
At the third step all collected reports (396) in the database are analysed by the use of text analyse program Nvivo. For this third step the words found at the first and second step are used as input to run text search queries. These queries not only check the text of all the reports on the input words, but also indicate stemmed forms of the words and synonyms of the words. This way the wordlists are expanded. Next to that a word frequency analysis shows the most frequently used words, including stemmed words and synonyms, and gives the possibility to check for additional words. The final wordlists can be found in Table 2 (appendix III).Although the length of the two lists differ this does not influence the research, since this study focus on the difference between firms and all firms are tested on the same wordlists.
The created wordlists are used for the content analysis by the use of the search queries-option of the Nvivo program. This way the software indicates per report how many times words from one wordlist are present in the report. This counting is referred to as references. Next to that it gives the coverage rate, which indicates the percentage of the references compared to the total report. By calculating this coverage rate, Nvivo excludes the words indicated as stop words in order to receive a more virginal representation of the texts’ content. The list of stop words used can be found in Table 3 (appendix V). Per report this results in a reference and coverage ratio for both motivation types; self-serving and other-self-serving. The use of coverage rate is preferred here since it controls for the length of the report. This way, the database extends with the coverage rate per firm of both the motivation for every year in 2008-2012 and a mean for all the five years.
4.2.2 Ranking
In order to classify companies based on their CSR motivation, this study compares all the companies in the database with each other. Therefore different rankings based on coverage ratio, are created. For each year and for the mean coverage rate of five years the self-serving (SS) ranking assort all the companies based on their self-serving coverage rate, while similarly the other-serving (OS) ranking assort the companies based on their other-serving coverage rate. This way, the database extends by twelve variables; six positions (each year and the mean over 2008-2012) on SS-rankings and similar six positions on OS-rankings to each firm.
When every company is ranked per year for its coverage rate on both types of motivation, the next step is the creation of a delta ranking. For each year and for the mean a delta ranking is created and this ranking orders all firms of the sample. To determine a firms position, its position on the OS-ranking is subtracted from its position on the SS-OS-ranking, separately for each year and for the mean over five years. E.g. a firm positioned third on the 2008 SS-ranking and positioned fiftieth at the 2008 OS-ranking ends up with a score of minus forty-seven at the 2008 delta ranking. This results in a delta
22 ranking score (the value of the subtraction) and a delta ranking position (the position on the delta ranking, based on the delta ranking scores). This way, the database extends by twelve variables; six delta ranking scores and six delta ranking positions for each firm. The benefit of this delta ranking is that it controls for a firm’s openness concerning its CSR motivation. E.g. relative more open firms are positioned higher in both initial rankings and end up with a score around zero in the delta ranking. The same counts for relative not so open companies that are low on both initial rankings. Consequently, only firms that have a relative high score on solely one of the two initial motivation rankings end up with a high absolute value on the delta ranking, either positive or negative. This way the usage of this delta ranking method maximizes the differences between firms.
4.2.3 Motivation classification
Classification of the firms occurs via analyses of their delta ranking position per year in combination with their mean delta ranking position over the five years. First all the companies are classified for each of the five years and for the mean over five years separately. They receive a ‘label’ (SelfServing (SS) and OtherServing (OS) for each year, which indicates how the company was motivated in that particular year, and a label for their motivation based on the mean over five years. Secondly they receive a final label for final classification as is described in the next paragraph. All six delta rankings are split into three equal parts. For each delta ranking the companies with the highest negative delta ranking score are labelled as OS. These firms have a high negative delta ranking score because they have a high position on the OS-ranking, while at the same time a low position on the SS-ranking. Similarly one third of the companies with the highest positive delta ranking score are labelled as SS. These firms have a relatively high position on the SS-ranking, while at the same time a relatively low position on the OS-ranking. For all six delta rankings, the remaining one third that neither is labelled as SS nor as OS is labelled as MixedMotive (MM). Due to this labelling per year, it occurs that some companies are labelled differently over the years. Besides, due to the minimum requirement of only three reports over the five years, some companies cannot be ranked for one or two year(s). Those missing positions lead to a varying amount of firms in each year’s ranking. Visual inspection of the missing positions checks whether there is a relation between position in the ranking and the missing of reporting. Consistent application of the ‘one third rule’, when classifying all six delta rankings, deals with the varying amount of total firms and resulted in different amount of firms that are classified as having the same motivation. This way, the database extends with a maximum of six labels for each firm.
After the firms received six labels they are checked on consistency over time, in order to determine the final label. As mentioned, some companies have various motivation labels over the five year period. Such an inconsistency of motivation labels could indicate a change in CSR motivation or lack of clear (CSR) strategy. Regardless the reason of this inconsistency, it is important to exclude the cases that are very inconsistent from this research. For this reason the company has to meet the
23 following criteria in order to determine the final motivation label; 1) the labels of the three most recent years (2010-2012) need to be the same as the mean label and 2) labels of the first two years (2008 & 2009) may not be the opposite motive type (self-serving or other-serving). Not only this method results in a selection of companies with a consistent strategy only, it also controls for possible variances by the use of annual report sections instead of standalone sustainability reports (as indicated in the section data collection). Additionally an even stricter criteria is formed; all five year labels need to be identical. Selection based on this most strict criteria is not leading in the second part of this study, since its strong negative impact on the sample. Nevertheless the classification with the most strict criteria is used to check for possible differences.
24
5.
Results on CSR Motivation
This section discusses the results of part 1 of this study and is focussed on classifying companies based on their CSR motivation.
The content analysis results in a reference and coverage rate per company for each of the five years and a mean over all the five years. A summary of these results, expressed by minimum, maximum and average, can be found in Table 4 (appendix V). The minimum reference score is 2 for SS and 0 for OS. The maximum reference score is 708 for SS and 456 for OS. The average reference for all companies’ mean over five years is 170 for SS and 62 for OS. The minimum coverage rate is 0.03 percent for SS and 0.00 percent for OS. The maximum coverage rate is 2.14 percent for SS and 2.58 percent for OS. The average coverage rate for all companies’ mean over five years is 0.81 percent for SS and 0.41 percent for OS.
Creation of the Delta Ranking resulted in an absolute score per company for each of the five years and for the mean over all the five years. A summary of these results can be found in Table 5 (appendix VI). The lowest minimum score is minus 71 and the highest maximum score is 72. The average delta ranking score for all companies’ mean over five years is 0.7.
Missing values, due to missing reports, did occur. Table 5 (appendix VI) shows the amount of firms that is ranked per year, this varies from 57 to 81. Visual inspection of the missing positions indicates that there is no relation between position in the ranking and the missing of reporting. It is therefore assumed that the missing values are equally spread over the rankings.
Application of the criteria for creating a final label results in the labelling of eight companies as SS while they have a MM label in 2008 and/or 2009. This is done because they have a SS mean label as well as a SS label consistently for the period 2010-2012. Four companies received the final label OS while having a mixed motive in 2008 and/or 2009. This is done because they have a OS mean label as well as having a OS label consistently for the period 2010-2012. There are three companies labelled as MM while having either a SS or OS label for 2008 and/or 2009. This is done because they have a MM mean label as well as having a MM label consistently for the period 2010-2012. Overall, as Table 6 (appendix VIII) shows, seventeen companies were labelled as SS, sixteen companies were labelled as OS and seven as MM. The final label defines how the companies are classified for further analyses.
Part 1 of this study shows the possibility to classify and label companies based on their CSR motive. By comparing the sample three kinds of companies showed up. First there is a group of companies found that is consistent over time in the region of being the most self-serving motivated, compared to the other companies. Second there is a group companies found that is consistent over time in the region of being the most other-serving motivated, compared to other companies. Although the third group is smaller compared to the other two, there is a third group of companies found that is equally self- and other-serving motivated over time, compared to the other companies. This third group is therefore classified as having a mixed-motive.
25
6.
Methodology Part 2: CSM & CFP
For the second part of this research, quantitative analyses are used in order to analyse CSM and CFP. This second stage of the research tests whether differently motivated companies indeed have different financial performances over time. Here the three hypotheses, as stated in the literature review, are tested by the use of quantitative methods. This is tested by the use of statistical software (SPSS). The use of different indicators is preferred since each variable might reveal different aspects of performance. Possible subsequent variation in outcomes can be used to inform interpretation of the results.
6.1
MethodologyAll three hypotheses are tested by the use of a Mixed-design ANOVA (ANalysis Of VAriances) method (Keselman et al., 1998). The mixed-design ANOVA compares the mean differences of a sample that is split in two independent variables. The primary purpose of a mixed ANOVA is to understand if there is an interaction between the two independent variables on the dependent variable. Due to the use of an ANOVA these independent variables are called factors.
In order to perform a Mixed ANOVA the following is required; one between-subjects factor that is categorical, one within-subjects factor that is categorical and one dependent variable that is continuous. To be able to provide a valid result the following five assumptions are advised to hold (Keselman et al., 1998):
1. There are no outliers in any group (or overall);
2. The data (or residuals) are approximately normally distributed; 3. There is homogeneity of variances;
4. There is homogeneity of covariances; 5. There is sphericity.
These assumptions are tested so that the result from the Mixed ANOVA is as valid as possible.
Prior to the Mixed ANOVA procedure there is tested for outliers. The use of a boxplot is preferred here since there is only a relative small number of companies in this analysis. The use of boxplots enables removal of outliers without losing feeling with the data (Pfannkuch, 2006). Besides normal distribution is tested by performing a Shapiro-Wilk test.
In SPSS the ‘Repeated Measures General Linear Regression’ option is used to perform the Mixed ANOVA method. This method determines whether there are statistically significant differences between the groups in CFP over the years.
Some more complicated tests, such as a regression with time as moderating variable, can be used in order to test for causality in the case their exist correlation. Since this research solely investigates the difference between differently CSR motivated companies, those additional tests are not used here (Nelson, Nelson, & Zaichkowsky, 1979).
26
6.2
DataThe result from Part 1 of this study is a sample of forty companies that are labelled based on their motivation. Seventeen companies are labelled as SS, sixteen as OS and seven as MM. For this forty companies financial data is collected for the period 2008-2012 by both the use of Compustat as well as by reviewing the yearly annual reports. Due to this research-by-hand it was not necessary to deal with missing values in the database.
Similar to Barnett & Solomon (2012, this study uses Return on Asssets (RoA) as indicator for financial performance. This continuous ratio variable is defined as the net income of a firm divided by its total assets. This indicator can be classified into the Domain of Financial Performance of (Venkatraman & Ramanujam, 1986) as discussed in the literature review.
Next to RoA, similar to Barnett & Salomon (2012), this study uses net income as indicator for financial performance. This continuous variable is defined as the earnings (after interests, taxes, depreciation and amortization) of a firm. This indicator also falls into the domain of financial ferformance (Venkatraman & Ramanujam, 1986).
The third indicator of financial performance that this study uses is the efficiency ratio. Efficiency is here defined as operational profit divided by revenue. This ratio provides insights into mechanism behind a change in net income. As It indicates the reliance of operational profit change on both change in revenue and change in costs. This can be relevant since it is argued that CSR can influence both (Porter & Kramer, 2011). Different from the previous indicators of CFP, this indicator falls into the Domain of Organizational Effectiveness (Venkatraman & Ramanujam, 1986).
Two independent ordinal variables are used in this study, namely time and CSR motivation type. Time consisted of five categories representing the time period 2008-2012. CSR motive consisted of three categories representing the classification of CSR motivation; SS (self-serving motivated), MM (mixed-motivated) and OS (other-serving motivated). All companies with the same motivation category are also referred to as group.
