English versus Dutch media reporting on CSR efforts of MNEs
A longitudinal multiple case study on the differences between English and Dutch media reporting on corporate social responsibility efforts of multinational enterprises
Amsterdam Business School
MSc. Business Administration: International Management
Supervisor: Arno Kourula
Author: Tristan Randy Wesenhagen
Student number: 10464794
Version: Final draft
1
Abstract
The main objective of this thesis is to expose the similarities and differences between the English and Dutch media regarding the way they report about the corporate social responsibility (CSR) efforts of multinational enterprises (MNEs). As the literature on CSR is very diverse, the second objective is to generate a comprehensive conceptual model that incorporates conventional and recent theories about CSR. A thorough study of the business literature reveals that the institution-based view and recent financial crisis influence CSR efforts of firms, and their relevance in relation to CSR efforts of MNEs is therefore tested in the model as well. This research tests the conceptual model, and corresponding assumptions, by means of a longitudinal multiple case study based on a vast amount of news articles from LexisNexis. Unilever and Shell, two vastly different types of MNEs, are selected as the cases. This choice provides the opportunity to research what the shared patterns are in media reporting between a firm that is renowned for exceptional CSR efforts and a firm that has the opposite reputation, and in which areas the good performer flourishes to distinguish itself from the bad performer. As many multiple case studies compare firms within the same industry, this research offers fresh insights. The most notable finding is that the English and Dutch media have an overall high degree of similarity in the way they report about CSR efforts of both Unilever and Shell. Secondly, both the English and Dutch media indicate that Unilever and Shell score quite similarly for seven out of eleven assumptions; both firms use management control systems (MCSs), satisfy their respective employees and local communities, alleviate environmental problems, provide transparent sustainability reports, and receive awards for sustainable innovations. Also, both Unilever and Shell are not necessarily proactive in weak institutional climates, and the Dutch media argue that Unilever occasionally reacts to accusations and scandals with resistance, just like Shell. However, Unilever mainly distinguishes itself from Shell by having a strong CSR fit, a CSR savvy CEO, and maintaining the belief that sustainable operations lead to a steady financial performance during the financial crisis.
2 Statement of originality
This document is written by Tristan Randy Wesenhagen who declares to take full responsibility for the contents of this document
I declare that the text and the work presented in this document are 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
Acknowledgements
First and foremost I want to thank Arno Kourula, my supervisor, for supporting me throughout the process of writing the thesis. The first hurdle of writing the thesis –finding a research gap that is both relevant and appealing- is always difficult. Thus, most of all I want to thank Arno for pitching the idea of doing a bilingual content analysis regarding corporate social responsibility and multinational enterprises. It really helped me in the right direction. The result lies before you. Then my parents, who always support me, deserve my gratitude and respect. It is because of them that I am privileged enough to follow this study. A finished thesis is probably the best gift I can give them in return.
4
Table of contents
Chapter 1. Introduction 7
Chapter 2. Literature Review 9
2.1 Corporate Social Responsibility (CSR) 9
2.1.1 The Economic Dimension of CSR 10
2.1.2 The Social Dimension of CSR 14
2.1.3 The Environmental Dimension of CSR 15
2.2 Institution-based view 16
2.3 CSR Related to Scandals and the Financial Crisis 17
2.4 Conclusion 20
2.5 Conceptual Model 22
Chapter 3. Methodology 24
3.1 Qualitative Research Criteria 24
3.2 Research Strategy 26
3.3 Case Selection 27
3.4 Data Collection and Data Reduction 28
3.5 Data Analysis 31
3.6 Conclusion 34
Chapter 4. Results 35
4.1 Company profiles 35
4.2 English Unilever findings 36
4.2.1 The triple bottom line 36
4.2.2 Institutions and CSR 38
4.2.3 Scandals, the financial crisis and CSR 40
4.3 Dutch Unilever findings 45
4.3.1 The triple bottom line 45
4.3.2 Institutions and CSR 48
4.3.3 Scandals, the financial crisis and CSR 50
4.4 English Shell findings 56
4.4.1 The triple bottom line 56
5 4.4.3 Scandals, the financial crisis and CSR 62
4.5 Dutch Shell findings 68
4.5.1 The triple bottom line 68
4.5.2 Institutions and CSR 71
4.5.3 Scandals, the financial crisis and CSR 72
4.6 Conclusion 79
Chapter 5. Discussion 80
5.1 Within-case analysis Unilever 80
5.2 Within-case analysis Shell 86
5.3 Cross-case analysis 89
Chapter 6. Conclusion 92
Chapter 7. Implications, limitations and recommendations 97
Chapter 8. References 98
Chapter 9. Appendices 116
Tables and figures
Table 1: Overview of assumptions and corresponding descriptions 21 Table 2: Data collection ‘CSR Unilever’; each phase of data reduction 30 Table 3: Data collection ‘CSR Shell’; each phase of data reduction 31
Table 4: Codebook 33
Table 5: Economic dimension of CSR (Unilever), English media quotes 41 Table 6: Social and environmental dimensions of CSR (Unilever), English media quotes 42 Table 7: Institution-based view (Unilever), English media quotes 43 Table 8: Impact of scandals and the financial crisis (Unilever), English media quotes 44 Table 9: Economic dimension of CSR (Unilever), Dutch media quotes 52 Table 10: Social and environmental dimensions of CSR (Unilever), Dutch media quotes 53 Table 11: Institution-based view (Unilever), Dutch media quotes 54 Table 12: Impact of scandals and the financial crisis (Unilever), Dutch media quotes 55 Table 13: Economic dimension of CSR (Shell), English media quotes 64 Table 14: Social and environmental dimensions of CSR (Shell), English media quotes 65 Table 15: Institution-based view (Shell), English media quotes 66
6 Table 16: Impact of scandals and the financial crisis (Shell), English media quotes 67 Table 17: Economic dimension of CSR (Shell), Dutch media quotes 75 Table 18: Social and environmental dimensions of CSR (Shell), Dutch media quotes 76 Table 19: Institution-based view (Shell), Dutch media quotes 77 Table 20: Impact of scandals and the financial crisis (Shell), Dutch media quotes 78 Table 21: Comparison of the English and Dutch findings (Unilever) 84 Table 22: Comparison of the English and Dutch findings (Shell) 85 Table 23: Economic dimension of CSR (Unilever), untranslated 117 Table 24: Social and environmental dimensions of CSR (Unilever), untranslated 118 Table 25: Institution-based view (Unilever), untranslated 119 Table 26: Impact of scandals and the financial crisis (Unilever), untranslated 120 Table 27: Economic dimension of CSR (Shell), untranslated 121 Table 28: Social and environmental dimensions of CSR (Shell), untranslated 122
Table 29: Institution-based view (Shell), untranslated 123
Table 30: Impact of scandals and the financial crisis (Shell), untranslated 124
Figure 1: Conceptual model 23
7
1. Introduction
The business literature extensively reports on corporate social responsibility (CSR). Especially the notion that it is ‘good for business’ can be read in wide variety of academic articles (Ambec and Lanoie, 2008; Pitta, Guesalaga and Marshall, 2008). While a more precise meaning of this term is discussed in the literature review, one should know that it does not encourage firms to be involved in scandals, in fact quite the opposite. Still, firms are caught in scandals and
misconducts quite regularly, for instance when they do not seem to have the best interest in the environment. Indeed, the recent Volkswagen scandal, concerning software that is designed to cheat on emission tests, is a good example (The New York Times, 2016).
Such actions are hard to condone, but there is one thing that can be said in their defense; the literature on CSR is quite diffuse and open to interpretation. Some argue that the concept has intertwined dimensions (Gimenez, Sierra and Rodon, 2012), while others put less emphasis on interconnectedness (Carroll, 1991). Therefore, it is possible that firms use this uncertainty to their benefit; to hide in the loopholes of CSR. Indeed, a comprehensive framework that integrates multiple theories and discusses the prevailing thoughts on CSR might give more insight. In fact, creating such a framework is one of the goals of this research. However, it has to bring
something new to the table as many comprehensive frameworks already exist. Ambec and Lanoie (2008) and Yuan, Bao and Verbeke (2011), for instance, have written articles that bring together multiple dimensions of CSR.
Fortunately, insights on the recent financial crisis can be included in this theoretic framework as well. Although some research has been conducted on the relationship between the financial crisis and CSR, the amount is still quite limited and, to my knowledge, not yet
incorporated in a framework that encompasses multiple perspectives on CSR. Garcia-Benau, Sierra-Garcia and Zorio (2013), however, do mention the financial crisis with regard to CSR, but also point to the fact that they performed ‘a pioneering study on the actual effects that the
financial crisis had on CSR reporting’ (Garcia-Benau et al., 2013: 1528). Therefore it would be rewarding if this research could contribute to a somewhat better understanding of that
relationship.
What is more, to ensure that this research has a unique contribution, the above mentioned framework is reflected upon through the eyes of the media. Sure, interviewing employees and corporate executives with regard to CSR might deliver valuable insights, even when it has been
8 done so many times before. However, if it is possible to get access to a vast amount of articles in multiple languages (LexisNexis, 2016), why not use those to look at CSR from a fresh
perspective? It comes with the added advantage that one does not have to be concerned with biased answers from respondents (Summers, 1969). Indeed it would be interesting to cross-analyze a whole range of articles in multiple languages, but alas, not everyone is affluent in many languages. Therefore, this research focuses on the English and Dutch media. It seems that such a bilingual combination has not yet been applied in CSR research, and certainly not in combination with the conceptual model that is presented at the end of the literature review (see chapter 3).
Although a research gap has been found, one might still wonder why it is necessary to fill this particular gap (Pratt, 2009). As the media have an important role to play in the creation of perceptions (Borah, 2011), firms should be rather interested in how the media report on their CSR efforts. Firms could use these insights to improve particular CSR efforts where the media report that they are lacking. In particular this research focuses on how English and Dutch media report on the CSR efforts of firms with regard to (1) the triple bottom line, (2) the surrounding institutional climate, and (3) scandals and the financial crisis. Again, these concepts are discussed in the literature review.
The findings of this multilingual research could urge firms to make tailor-made changes to their CSR efforts, as the different media might have contrasting points of attention. Similarly, a strong signal is sent to firms if both the English and Dutch media criticize particular CSR efforts. This might trigger firms to reevaluate their CSR efforts. Finding out what the similarities and differences are in media reporting on this topic, and building a comprehensive conceptual model from a wide variety of insights on CSR to facilitate this, represent the research objectives and main contributions of this research. Therefore, the following research question is maintained:
9
2. Literature Review
This chapter discusses the theoretic foundation of this research. The main theories are introduced by providing a general overview of their most important characteristics, before going into the specific comparisons and contrasts that are mentioned by different experts. By implementing this ‘funnel’ system, it is avoided that the theoretic foundation of this study is just an uncritical listing of readings (Webster and Watson, 2002). Moreover, it makes the chapter accessible for the reader, which ultimately contributes to a better understanding of the thesis. To conclude the chapter – at the end of the funnel- a table with assumptions and a conceptual model are presented as a logical outcome of the upcoming literature review.
2.1 Corporate Social Responsibility (CSR)
Despite a vast amount of research that has been done on CSR, there is still no universally acknowledged definition of the concept (Lindgreen et al., 2012). Some twenty years ago it was believed to be about ‘development that meets the needs of the present without compromising the needs of future generations’ (Gimenez et al., 2012, p. 150). Later on, it was mentioned that CSR encourages actions ‘that further some social good, beyond the interests of the firm and what is required by law’ (McWilliams and Siegel, 2001, p. 117). Other scholars build on this notion by stating that CSR is not just about addressing legal and economic commitments, but that it also suggests that firms should meet ethical standards and fulfill responsibilities to society such as taking care of employees and being good to the environment (Lindgreen et al., 2012). This perspective is maintained in this research as well; CSR as a concept with multiple dimensions.
As the previous definitions placed emphasis on different things, and in general provide little guidance for MNEs to engage in CSR, there is a need for a solid definition that addresses multiple key points that deserve attention. Pitta et al. (2008) mention a double bottom line: social goals combined with the business objective of profit. Gimenez et al. (2012) concur, but add to this definition by mentioning that there is also an environmental dimension. Carroll (1991) also mentions philanthropy as an additional dimension. In this research, though, philanthropy is believed to be part of the three previous dimensions (Gimenez et al., 2012). Therefore, to come to a concrete definition: CSR is about meeting the standards of a triple bottom line; sustainable behavior in an economic, social and environmental manner (Pitta et al., 2008; Gimenez et al., 2012).
10
2.1.1 The economic dimension of CSR
In years past the conventional line of thought was that CSR comes at an additional cost imposed on firms that has a negative impact on their global competitiveness. Taking care of the
environment or local communities, for example, would only cost money and serve philanthropic means (Ambec and Lanoie, 2008). Over the last decade there was a shift, however, toward a more positive view of the impact of CSR on business activities. Some scholars argued that improving a company’s environmental footprint can actually lead to better economic or financial performance. Therefore, the objective of firms should be twofold: maintaining the conventional goal of making a profit, but doing this in a socially and environmentally responsible manner (Ambec and Lanoie, 2008). The question that then remains is: ‘What are the economic benefits of CSR and how can be assessed whether a firm successfully incorporates CSR in its business activities?’ Ambec and Lanoie (2008) and McWilliams and Siegel (2001) argue that the answer lies within the business activities of the firm, while Pitta et al. (2008) point out that the notion of ‘bottom of the pyramid’ (BOP) is of concern as well.
With regard to the first notion, firms can reduce their environmental impact without compromising their economic performance by developing new innovation strategies. The benefits of such strategies are twofold: they can increase revenues and they can reduce costs (Ambec and Lanoie, 2008). Firstly, CSR should be seen as a form of investment (Arjaliès and Mundy, 2013). The firm can leverage its resources to create products with CSR attributes, or to enhance socially or environmentally responsible products even further. Thus, one way to increase revenues is by differentiating products. This enables them to enter more
environmentally conscious markets. The thought behind this is that, even if these products come at a higher price, consumers are willing to pay a premium price because of the socially or
environmentally responsible characteristics of the product (Ambec and Lanoie, 2008;
McWilliams and Siegel, 2001). Secondly, solid environmental performance is becoming more of a prerequisite for firms, because within the value chain suppliers or buyers often demand a certain standard, otherwise the deal is off. The same can be said of governments and consumers. Even more so, firms compete with each other to develop socially and environmentally
11 On the other hand, costs can be reduced for a number of reasons as well. For instance, one might argue that pollution is related to a waste of resources; raw material that has not fully been used or energy that could have been used more efficiently. Indeed, investing in innovations that can lead to a more closed cycle of production could result in lower expenses, as less material or energy is needed for the production of more products. In some cases even what appears to be waste can be reused (Ambec and Lanoie, 2008). While Gimenez et al. (2012) do not refute that these practices can lead to a cost reduction, they argue that these improvements should be labeled as ‘environmental sustainability’ (see paragraph 2.1.3). The cost of labor can also be reduced, albeit in a more indirect way, as an improved image of the company can improve the ambiance within the company and motivate employees to work harder (Ambec and Lanoie, 2008).
However, for a firm to be both profitable and good for the society or environment, there needs to be a ‘fit’ between CSR initiatives on the one hand, and the overall strategy and currently prevailing practices of the firm on the other (Yuan et al., 2011). If a firm is able to achieve such a balance, that firm has incorporated CSR in its core business. Maas and Reniers (2014) agree that successfully intertwining CSR in the core business is the best way for a firm to be engaged in CSR, otherwise it is just window-dressing; engagement in CSR to cover up harmful practices of the core business (Cai et al., 2012). The above mentioned information leads to the following assumption:
A1: MNEs are inclined to incorporate CSR in their core business, through product
differentiation and other forms of innovation, if they believe it can help them improve their financial performance or gain a competitive advantage.
Arjaliès and Mundy (2013) acknowledge that CSR can be an essential element of a firm’s core business, and indicate that a firm should nurture incentives and mechanisms to further the intertwinement of CSR within its core business strategy to enhance long term value. Therefore, CSR needs to be managed with the help of management control systems (MCSs). Management control systems are formal, information-based, routines and procedures managers use to maintain or change patterns in the firm’s activities (Simons, 1995). They act as a guideline and evaluation method to ensure that a firm assimilates social and environmental practices within the
12 that define and manage key performance indicators (KPIs) for CSR, to documents that
communicate core values within the firm. The latter method is especially useful to create a shared vision of CSR throughout the firm. This is in line with the notion of Torres, Garcia-French, Hordijk and Nguyen (2012) that a firm can use codes of conduct to prescribe values and principles which employees should aspire to follow. Additionally, interactive processes make up for a third method to increase the compatibility of CSR within the firm; through discussions between senior and operational managers (Arjaliès and Mundy, 2013). Thus, if used correctly, these measures can propel a firm to become more sustainable.
A2: MNEs use management control systems (MCSs) to evaluate and stimulate CSR intertwinement within core processes of the firm.
The management control systems are tools that can lead to useful insights, but the effectiveness also depends on the management itself. According to the upper echelon theory, strategic
decisions, including the ones that address sustainability, are significantly influenced by corporate executives, and especially by the chief executive officers (CEOs) (Tang, Qian, Chen and Shen, 2015). Tang et al. (2015) discuss that the psychological characteristics of CEOs affect the firm’s CSR participation. They found that CEO hubris, for instance, seems to hinder long-term
sustainability, because it leads to lower levels of participation in socially responsible activities. In addition, Slater and Fowler (2009) found that CEOs with international assignment experience engaged more rigorously in socially responsible activities. These psychological and
demographical characteristics go beyond the scope of this research, but the general conclusion that can be drawn from these insights is that, for a firm to embrace sustainability within its operations, the right persons with executive power have to be chosen. Spar and La Mure (2003) explain that charismatic and strong CEOs can press their personal preferences to address
sustainable issues, even if from a cost-benefit perspective they do not have to. With an increased pressure on firms to act socially and environmentally responsible (Waddock, 2008), it would seem logical for corporate executives to have a positive view on CSR, and see it as a prerequisite for doing business.
13 Most CSR minded managers are bound to have heard of the notion ‘bottom of the pyramid’ (BOP), as over the last couple of years it gained considerable attention. It mentions that firms can make profits in unconventional and unserved markets in lower income countries. In line with the previous mentioned definition of CSR, firms can enter these markets to make a profit while helping to alleviate the poverty in such a region (Pitta et al., 2008). Even though these consumers are poor, they are numerous and interested in quality products. Through meaningful product differentiation it is possible to reach consumers from the BOP. Firms can remove premium features from products, use different forms of packaging, or even reduce the size of products (Pitta et al., 2008).
Karnani (2005) challenges this line of thought and believes that international firms should be wary of the BOP. He indicates that consumers from the BOP do not have the means to buy luxury products as their needs differ from traditional consumers. The poor spend most, if not all, of their income on products that satisfy their basic needs such as food, clothing and fuel, and afterwards cannot afford anything else. Karnani (2007) also mentions that the opposite can be the case as well. Based on one of his case studies he argues that a firm might do well at the BOP, but not do any ‘good’. By this he means that firms might be profitable in some instances, but
unfortunately with negative implications for public welfare.
However, according to (Kochhar, 2014) this should not discourage large firms to explore these new markets. They argue that firms even have the moral obligation to enter these markets, as according to them the economic dimension is not limited to actions that are economically beneficial for only the firm itself, but also to the communities surrounding the firm. Product differentiation might be a good start, but firms should also put effort in trying to understand those new markets and especially what the needs of local consumers are. These needs might vary across different regions, as diseases and natural disasters frequently have a significant impact on the lives, and therefore priorities, of those consumers. For instance, a firm that is active in the business of personal health could market soap in regions where people struggle with hygiene (Kochhar, 2014). This leads to the following assumption:
14
2.1.2 The social dimension of CSR
Gimenez et al., (2012) bring up the concept of social sustainability. However, just like the umbrella concept of CSR, they state that it is not always clear what is meant with ‘social sustainability’. Their explanation of this concept is that it is about the firm taking care of, and acting responsibly toward, both the internal community and the external communities. With the internal community the employees are meant, and the external communities are the stakeholders outside of the company (Gimenez et al., 2012). For instance, a firm can provide fair
opportunities to all employees, look after their safety both in- and outside of the firm, invest in their studies, and create a stimulating workspace (Matten and Moon, 2008). Additionally, there are multiple external stakeholders such as the government, local communities (the people in the villages for instance), non-governmental organizations (NGOs), partners in the value chain, and shareholders (Waddock, 2008). The firm needs to take in account the wide range of interests of these parties as well. Remember that these considerations come on top of the ones from the economic dimension. Thus, to link this to the story of the previous paragraph, a firm that has entered one of the markets of the BOP, and engaged in meaningful product differentiation, might still be criticized by the local community and government for not buying local and sustainable raw materials for their production process (Pullman et al., 2010). Jensen (2001) states that also from a financial point of view it is in the best interest of the firm to take all stakeholders into account. ‘In order to maximize value, corporate managers must not only satisfy, but enlist the support of all corporate stakeholders; customers, employees, managers, suppliers and local communities’ (Jensen, 2001: 9).
Whereas Cai et al. (2012) mention the distinction between CSR as core business and CSR as window dressing, they argue that there is an in-between option as well, namely when firms just use CSR as socially neutral activities without any intention of deceiving stakeholders or value generation (Cai et al., 2012). Some scholars argue that every type of organization, whether it is for-profit or non-profit, is in fact a social organization. They say this because an organization can even indirectly ‘do good’, for instance by generating tax revenues and facilitating
employment (Dacin, Dacin & Matear, 2010). However, more concrete philanthropic actions and gestures can be deployed as well. According to Korosec and Berman (2006) firms have the responsibility to identify and address important social problems in their communities. To help alleviate the identified problems firms can develop new programs and services themselves, or
15 join efforts with other organizations. Examples of such efforts are the building of schools in regions with illiteracy, encouraging employees to do voluntary work a couple of days per year instead of their normal job, and organizing (sport)events for children in the local community (Ambec and Lanoie, 2008).
A5: MNEs try to satisfy multiple stakeholders in both their internal and external communities when engaging in CSR.
2.1.3 The environmental dimension of CSR
A firm has the moral responsibility to reduce its negative impact on the environment (Ambec and Lanoie, 2008). However, while the three dimensions –economic, social and environmental- are discussed separately, it should be emphasized once again that they are intertwined. The notion of Jensen (2001) from the previous paragraph, who stressed the importance of value maximization by taking into account the wishes of all stakeholders, is therefore also applicable to the
environmental dimension; there are stakeholders such as the government and local communities that often express their concern for the environment. Not taking them into account can lead to resistance, which can result in a bad reputation (Spar and La Mure, 2003). Furthermore, as mentioned in the explanation of the economic dimension, firms should strive to incorporate ‘being green’ in their core business, as that could result in a win-win situation of achieving cost reduction and limiting the environmental footprint (Ambec and Lanoie, 2008). Unfortunately, some firms can have a hard time finding such a fit, but those firms could still benefit from
initiatives outside of their core business in a more indirect way. For instance, leveraging financial resources to provide seeds to help regrow trees after a disaster could result in more goodwill as consumers, especially in nearby local communities, take kindly to such actions (Gimenez et al., 2012). However, sometimes the line between social and environmental initiatives is very thin (Dacin et al., 2010), as one could argue that in the previous example there is an indirect effect on the social wellbeing of local communities as well. Thus, the following assumption presents itself:
A6: MNEs identify and address environmental needs that are within their reach to improve or alleviate.
16
2.2 Institution-based view
In the previous paragraphs the importance of stakeholders and finding legitimacy were already mentioned. The institution-based view, though, gives some additional insight into the relevance of such concepts. The operations of firms are constricted by the institutions surrounding them (Peng, Sun, Pinkham & Chen, 2009). For instance, governments are able to provide MNEs that enter new markets a license to operate (Peng et al., 2009; Skare and Golja, 2014). At the same time, non-governmental organizations pay attention to the actions of firms and, if those actions are against their liking, can encourage other stakeholders to dissociate themselves from the firm, thereby decreasing the value of the license to operate. Indeed, it is almost a political game that is played between a great number of institutions, varying from governments and NGOs to
shareholders and consumers; all have some sort of interest in the firm (Spar and La Mure, 2003). As these institutions moderate the actions of the firm (Peng et al., 2009), it should be emphasized they also affect the CSR efforts of the firm. For instance, firms operate in multiple countries, each with their own institutions; formal institutions such as the rules of the
government, and informal institutions such as the customs and needs of the local culture. Therefore, firms need time to learn and adjust to the bidding of these institutions (Peng et al, 2009; Skare and Golja, 2014). Once a firm knows what is expected, it can employ the right CSR efforts (Gimenez et al., 2012). However, from a more skeptical perspective, firms could use this as an excuse to refrain from CSR efforts. Moreover, as firms often thoroughly analyze any market before they enter, it can even be the case that firms deliberately enter markets where the institutions are weak or work in their favor (Peng et al, 2009; Spar and La Mure, 2003). To give an example, certain countries maintain a weak policy when it comes to taxes, allowing firms to pay a low amount of taxes. It occurs that some firms solely want to make use of these loopholes by locating one of their subsidiaries there, without doing any meaningful operations to facilitate employment for the local community, or engage in any sort of CSR (Bame-Aldred, Cullen, Martin & Parboteeah, 2013).
However, due to globalization external stakeholders have the means to follow the actions of firms regardless of the distance; firms cannot wipe bad business under the carpet (Spar and La Mure, 2003). Waddock (2008) goes on to discuss that firms, in the absence of a global
governance structure, are still inclined to act in an accountable, transparent and environmentally sustainable way, because of emerging institutions that have a multi-bottom-line orientation that
17 encompasses social and stakeholder issues into their business models. Indeed, according to her there is a transition to a new context in which external stakeholders hold firms responsible for their impacts. Firms have to embrace CSR, because it is increasingly seen as a prerequisite for doing business. Peer pressure and visibility rankings trigger firms to embark on a sustainable mission sooner than later (Waddock, 2008). This is in line with the argument of Arjaliès and Mundy (2013) that stakeholders demand transparency regarding the CSR performance of the firm. Torres et al. (2012) add that the aforementioned codes of conduct offer guidance when there is a lack of clear rules set by institutions. On top of that, there are Global Reporting Initiatives (GRI) Guidelines which are created to stimulate socially responsible corporate behavior. These are only guidelines, but still firms voluntarily follow them because of peer pressure (Torres et al., 2012). Linked to the institution-based-view, it can be said that there is a new globally adopted informal rule that prescribes firms to act responsibly and sustainably (Peng et al., 2009; Waddock, 2008). Therefore, instead of using loopholes to achieve financial goals or back away from sustainable operations, firms are likely to leverage their resources to strengthen or complement weak local institutions (Waddock, 2008). Assumptions that can be made based upon these arguments are:
A7: MNEs are proactive and act responsibly when local institutions fail to pressure firms to do so.
A8: MNEs are transparent in the sense that they provide social and environmental reporting on a regular basis.
A9: MNEs are focused on performing well in sustainability rankings due to (in)formal rules and peer pressure.
2.3 CSR related to scandals and the financial crisis
In a nutshell, from the previous paragraphs it can be concluded that there are multiple triggers for firms to engage in CSR, ranging from business opportunities and intrinsic motivation to external institutional pressures (Ambec and Lanoie, 2008; Tang et al., 2015; Peng et al., 2009). To complement this framework, it is interesting to point out how CSR relates to more incidental
18 phenomena such as scandals and the financial crisis. While many scholars discuss the evolution and gaining significance of CSR, it is often underexposed that it comes at a cost that various firms have to pay (Torres et al., 2012). Indeed, the development of the concept of CSR depends in part on trial and error from firms, and therefore missteps from firms with regard to the social and environmental dimension are especially interesting (Torres et al., 2012). While it is good for research, it can have a devastating impact on firms. For instance, public scandals involving environmental pollution and the violation of human rights can result in unwanted media attention, which in turn can lead to reputation damage (Tulder and Zwart, 2006).
Torres et al. (2012) conducted multiple case studies and found that scandals indeed have an impact on the CSR policies of firms, although they conclude that not all firms respond quickly and that there is no ‘one best way’ of dealing with scandals; it depends on what kind of scandal it is, the scale of the scandal, and how it is picked up by stakeholders. Spar and La Mure (2003) agree that there is wide variation in possible responses, and try to explain it. They argue that a firm can make a cost-benefit analysis, by weighing the benefits of obedience with its costs. For example, a firm that is responsible for the contamination of water should calculate the cost of filtering it. Moreover, in line with this rational approach, they mention that firms pick their battles strategically; complying with demands that are cheap to address, while ignoring protests that do not endanger the future of the firm (Spar and La Mure, 2003). Especially firms in sinful industries, like the tobacco or oil industry, show signs of window-dressing. These firms often use CSR activities that are unrelated to their core business (Cai et al., 2012).
However, Spar and La Mure (2003) go on to mention that the approach of preemption can be viable as well. By preempting NGO demands firms can boost their brand image, which in turn can have a positive impact on their competitive position. By preemption they mean that a firm proactively deals with problems it has caused, before NGOs confront firms with it. In light of the aforementioned peer pressure and global informal rules this latter approach seems to be the more logical choice for most firms these days (Waddock, 2008; Peng et al., 2009). Moreover, firms that abide to the codes of conduct and institutional rules are likely to not even face such troubles. Furthermore, Tulder and Zwart (2006) add that firms that have been involved in a scandal often grow out to be leaders in their sector concerning CSR issues. Thus, the firms that have been on thin ice are inclined to learn from their mistakes and rebound strongly. This leads to the following assumption:
19 A10: When a MNE is accused of social or environmental misconduct it will meet the demands of stakeholders as soon as possible.
The financial crisis is another phenomenon that might be difficult for firms to deal with.
According to Garcia-Benau, Sierra-Garcia and Zorio (2013) the notorious financial crisis started in 2007, while its effects were felt in Europe in 2008 due to exposure to the financial markets of the US. For many firms this meant a transition to an uncertain business environment. Critical management decisions were needed to ensure survival in the market (Garcia-Benau et al., 2013). However, the survival of the firm must be placed in a broader perspective, as a firm’s survival is in the best interest of other stakeholders as well. Shareholders and firms in the value chain have a faith that is intertwined with that of the firm in question (Jensen, 2001; Pullman et al., 2010). Thus, if a firm hypothetically would be troubled by the financial crisis, those stakeholders would like to be informed; transparency of firm performance is a necessity (Pullman et al., 2010).
Transparency is one of the manifestations of CSR, but Garcia-Benau et al. (2013) state that CSR can play a bigger role when a firm has to endure a financial crisis. A decrease in consumer spending leads to fewer sales for the firm, and partners in the value chain that are susceptible to the effects of the crisis represent a liability. Therefore firms need to rethink their operations. Simply terminating unprofitable business lines might help in the short term, but reinventing the company in terms of sustainability, and doing business with equally sustainable partners, can help the firm prosper in the long term (Garcia-Benau et al., 2013). Indeed,
incorporating CSR into the core business, as discussed previously by Ambec and Lanoie (2008) and Arjaliès and Mundy (2013), can lead to a competitive advantage that can result in a
dominant position in the market. A ‘greener’ business model can lead to innovative products while using fewer resources (Gimenez et al., 2012). Such critical management decisions are needed to stay competitive during tough financial times (Garcia-Benau et al., 2013).
20
2.4 Conclusion
The above mentioned paragraphs taken together offer a framework of what CSR entails. Some of it is contrary to the beliefs of some researchers. Carroll (1991), for instance, argues that the responsibilities of the firm take the shape of a pyramid where economic responsibilities come first, followed by legal, ethical, and philanthropic responsibilities. In other words, the main responsibility of firms is to provide goods and services to make a profit; CSR is no prerequisite of doing business (Carroll, 1991). As the aforementioned theories in this literature review are predominantly insights from recent years it can be argued, however, that the ‘pyramid view’ has lost value.
Instead, these theories together offer a solid framework of what CSR currently entails. For the sake of structure the main pillars of CSR –the economic, social and environmental dimension- were explained separately. However, contrary to what Carroll (1991) proposes, it should be stressed that in this research these dimensions are intertwined; companies should do business that is profitable and sustainable at the same time (Gimenez et al., 2012; Ambec and Lanoie, 2008; Arjaliès and Mundy, 2013). Many of the consulted research papers point out a number of things that firms need to keep in mind in order to engage in CSR successfully, often seen from one particular perspective. For instance, Ambec and Lanoie (2008) indicate that firms should aim to incorporate CSR in their business to gain a competitive advantage and summarize the potential benefits, while Arjaliès and Mundy (2013) complement this by discussing some of the methods that firms can adopt to actually implement CSR successfully. Similarly, CEOs have an important role to play as catalysts for CSR activities (Tang et al., 2015). Other theories touch upon CSR as well. Jensen (2001) and Pullman et al. (2010) emphasize that CSR needs to be taken into account by firms for the sake of fulfilling the wishes of important stakeholders, while the institution-based theory explains that there are institutions that can force firms to behave responsibly and sustainably (Peng et al., 2009). Finally, with regard to respectively scandals and the financial crisis, Torres et al. (2009) argue that firms are inclined to learn from scandals that they are involved in by setting up or supporting sustainable initiatives, while the financial crisis should motivate firms to incorporate and invest more in sustainable operations (Garcia-Benau et al., 2013).
21 In many of the previous paragraphs the terms ‘CSR’, ‘stakeholders’, and ‘institutions’ are mentioned, indicating how intertwined these terms are. Taken together, these theories provide a broad framework that lead to a number of assumptions, which can be seen in table 1.
Table 1: Overview of assumptions and corresponding descriptions Concept Assumption Description
Economic dimension of CSR
A1 MNEs are inclined to incorporate CSR in their core business, through product differentiation and other forms of innovation, if they believe it can help them gain a competitive advantage.
A2 MNEs use management control systems (MCSs) to evaluate and stimulate CSR intertwinement within core processes of the firm.
A3 MNEs have intrinsically motivated corporate executives that stress the importance of CSR.
A4 MNEs try to reach consumers from the BOP. Social dimension of
CSR
A5 MNEs try to satisfy multiple stakeholders in both their internal and external communities when engaging in CSR.
Environmental dimension of CSR
A6 MNEs identify and address environmental needs that are within their reach to improve or alleviate.
Institution-based view
A7 MNEs are proactive and act responsibly when local institutions fail to pressure firms to do so.
A8 MNEs are transparent in the sense that they provide social and environmental reporting on a regular basis.
A9 MNEs are focused on performing well in sustainability rankings due to (in)formal rules and peer pressure.
Scandals A10 When a MNE is accused of social or environmental misconduct it will meet the demands of stakeholders as soon as possible.
Financial crisis A11 The financial crisis urges MNEs to invest more in CSR. Source: Author
22
2.5 Conceptual model
The aforementioned theories all influence the CSR efforts of firms. However, when critically reviewed it can be argued that, even though they are intertwined to some extent as explained in the literature review, their degree of salience regarding CSR decisions varies. Triple bottom line decisions are the most salient, as these represent the decisions the firm makes on a frequent basis. It involves strategic decisions, for instance to implement CSR fully into the core strategy of the firm (Ambec and Lanoie, 2008; Arjaliès and Mundy, 2013), and also CSR decisions to satisfy stakeholders (Jensen, 2001; Pullman et al., 2010). Based on the literature it can be concluded that when firms incorporate CSR in their core business – and thus hit targets on all three dimensions- the influence of the two outer layers will be minimal, as the firms already ‘do well by doing good’ themselves (Cai et al., 2012).
The institutions always have at least some influence, though, as formal rules point out what firms are (not) allowed to do (Peng et al., 2009). However, the firms that have not integrated CSR fully into their business strategy will feel the influence of the institutions, as institutions will try to push those firms in the right direction (Peng et al., 2009). For instance, firms that are reluctant to engage in CSR efforts will be criticized by NGOs or their peers, with firms often succumbing to the pressure (Waddock, 2008). Similarly, large firms with preemptive CSR strategies can leverage their resources to complement weak institutions (Spar and La Mure, 2003).
Lastly, scandals and the financial crisis form the outer layer, because contrary to the institutions, they are not always around; they are occurrences. When they do occur, though, they can influence the triple bottom line decisions of firms and the institutions surrounding them (Garcia-Benau et al., 2013; Skare and Golja, 2014). To give an example, scandals and the financial crisis can cause firms to change their corporate strategy by implementing more sustainable operations (Garcia-Benau et al., 2013), while institutions such as governments and legislative forces are likely to impose additional rules on firms to prevent a reoccurrence of such events (Skare and Golja, 2014). This leads to the conceptual model that can be seen in figure 1.
23 Figure 1: Conceptual model
Source: Author
Scandals & Crisis
Institution-based view (IBV) Triple Bottom Line (TBL) CSR EFFORTS Environmental Dimension Social Dimension Economic Dimension The
institution-based view affects and complements
decisions for the triple bottom line as
(in)formal rules and (peer)pressure guide firms in their
CSR decisions
Scandals & economic crises form the outer layer
because they are incidental. However,
once they occur they affect both the
triple bottom line and the institutions
CSR laden firms can frequently engage in
CSR efforts through all three dimensions
of the TBL. Additionally, these firms complement weak institutions
24
3. Methodology
This chapter describes the way this research was conducted. First, the criteria of doing qualitative research, and how these are met in this research, are discussed. From this broader context the focus shifts to the research strategy that is used; a longitudinal multiple case study design. After explaining this strategy, the selection procedure of the cases is described. Subsequently, the method of data collection for these cases is brought to light, which in turn flows into a
specification of the data analysis. Finally, this chapter concludes with an operationalization of the codes that are used in the data analysis to give an answer to the assumptions mentioned previously in the literature review.
3.1 Qualitative research criteria
Starting off, the research goal is of an explorative and descriptive nature, meaning that it is suited for qualitative research (Yin, 2009). Thomas and Magilvy (2011) build on this notion by stating that qualitative research is about selecting a certain phenomenon that you deem interesting, and then focus on depth. By ‘depth’ they mean that one can enrich their understanding of that particular phenomenon by collecting large amounts of information about it. As qualitative research, however, often struggles to deliver any causal or hard evidence (Gibbert and Ruigrok, 2010), it is important for a researcher to invest in trustworthiness, to make their claims
believable. This concept –trustworthiness- can be broken down to four criteria; credibility, transferability, dependability and neutrality (Thomas and Magilvy, 2011; Tracy, 2010).
Credibility is about the representativeness of the data as a whole. The researcher checks for similarities and patterns across different pieces of gathered data. For instance, in case
different respondents were interviewed, the researcher should look for overlap in the answers of different respondents, and afterwards check with the respondents if they agree with the findings (Thomas and Magilvy, 2011; Altheide et al., 2008). Tracy (2010) builds further on this notion by claiming that interpretative analyses should be plausible and reflect a reality that seems true. She posits that qualitative credibility can be achieved through the incorporation of both thick
description and triangulation. The first term points out that the researcher should account for the complexity and circumstantiality of the data by explaining the findings in concrete detail and providing in-depth illustrations. Although Pratt (2009) agrees with this statement, he argues that limiting an analysis to solely a description of the findings –even when insightfully done- is not
25 enough. There has to be a transition from data description and data display to an interpretation of that data. This notion is respected in this research by providing insightful data displays with the inclusion of illustrative quotes in the chapter of ‘results’ and discussing those findings in the subsequent chapter of ‘discussion’ (Pratt, 2009).
The latter term, triangulation, refers to the usage of multiple sources of data or types of data collected, with the aim to diminish subjective bias as much as possible (Tracy, 2010). In this research, however, no interviews were conducted. Indeed, it could have been valuable for the purpose of triangulation, but then again, news articles are the most relevant data source to answer this particular research question, as it should be emphasized that the focus is on English and Dutch media. As Yin (2009) states, the substance and form of the research question indicate which research strategy and methods should be applied.
Therefore, it seems more appropriate to invest in a vast number of news articles than to add a small number of interviews with questionable applicability. Instead of looking for similarities across both news articles and transcripts, the approach is to make use of the variety within the news articles; for both the English and Dutch media a wide range of publications is incorporated in the process of data collection. Therefore, while there is no triangulation with regard to the types of data collected, there is triangulation with regard to multiple sources of data (Tracy, 2010). This notion, combined with the acknowledgement of thick description, explains how credibility is achieved in this research (Thomas and Magilvy, 2011).
Transferability is about the ability to use the methods that were used in one research for a similar research with a different context (Thomas and Magilvy, 2011). As in this thesis a
multiple case study is conducted where the same methods are applied for different cases, there should be no problems for future researchers to assimilate this research design for their research, for instance if they want to ascertain whether the findings from this research hold true for MNEs in different –or the same- industries. As was frequently mentioned in the researched articles: ‘No organization can afford not to be engaged in corporate social responsibility’ (Australian
Financial Review, 2009).
Dependability can be seen as an extension of the previous term, in the sense that here it is about being transparent regarding the decisions the researcher has made throughout the research. Concretely, a study is dependable when its purpose and methodological foundation, such as the
26 case selection, data collection, and data reduction are all explained (Thomas and Magilvy, 2011). In this research these criteria have all been met and are discussed in the following paragraphs.
According to Thomas and Magilvy (2011) neutrality is achieved when the researcher abides to the three foregone rules and on top of that keeps an open and reflexive mind regarding the analysis of the data. This is similar to what Tracy (2010) calls ‘sincerity’; to apply self-reflexivity about subjective values, biases and inclinations of the researcher. To address this point, a mix of both deductive and inductive coding was maintained, as the latter term means that unexpected findings in the data are translated into new codes that, in turn, are used to recode the already coded articles to expand the richness of the findings (Altheide et al., 2008). The process of coding is explained in more detail in paragraphs 3.5 and 3.6.
3.2 Research strategy
The research question is: ‘How do the English and Dutch media report on CSR efforts of MNEs?’ To answer such a ‘how-question’, a case study is a proper type of research strategy (Yin, 2009; Pratt, 2009). Case studies are especially suited to understand a certain decision or phenomenon as they combine different collection methods such as documents, interviews and observations (Eisenhardt, 1989). This research, however, does not make use of the latter two methods and therefore the main research method is document analysis. It can be used to find in-depth knowledge about a certain subject, by critically searching for patterns, contexts and themes within the collected data (Yin, 2009; Altheide et al., 2008). This method is explained in greater detail in paragraph 3.5.
Still, since multiple MNEs are selected as cases and data is analyzed over a period of multiple years, it seems appropriate to classify the research strategy as a longitudinal multiple case study (Yin, 2009). Eisenhardt (1989) points out that when this strategy is chosen the
researcher must determine both the level(s) and specific unit(s) of analysis. This research focuses on the firm-level as from the collected data it became apparent that most media articles discussed CSR-efforts at this level. The specific units of analysis are MNEs from different industries, and just one from each industry. This broader approach of choosing MNEs from different industries was chosen because it matches the explorative intent of the research question to map out how the different media report about CSR efforts of MNEs in general; there is no intent to focus on MNEs within one industry (Altheide et al., 2008). However, the findings of this study might
27 inspire future researchers to pick one of these industries and select multiple MNEs within it to do an in-depth study.
To address the longitudinal aspect of the research strategy, a period of eight years (2008-2015) was chosen for the data collection. First of all, this is in harmony with the sub-goal of finding out how the different media report on the influence of the financial crisis on CSR efforts of MNEs, as some argue that the financial crisis had its biggest impact in 2008 (Garcia-Benau et al., 2013). Secondly, when reflecting back on the research question, in order to find any
meaningful patterns it is necessary to analyze a large amount of data. By collecting data over the course of eight years this criterion was fulfilled as well (Pratt, 2009; Yin, 2013). Paragraph 3.4 discusses the process of data collection in more detail.
3.3 Case selection
There are a couple of strategies to select the cases that are to be used for research. First of all, the cases have to meet predetermined criteria of importance (Palinkas et al., 2015). As the research question focuses on the difference between English and Dutch media reporting, the selected companies should preferably have a Dutch background, as that increases the chance of finding enough relevant Dutch news articles. Secondly, from a theory-based view, the constructs mentioned in the literature review should be applicable to the selected cases. Therefore, the selected companies should be active in CSR, although that criterion is not hard to fulfill as a brief search to a company website shows whether a company addresses this issue (Palinkas et al., 2015). Lastly, the cases are selected on the basis of maximum variation in the hope of finding important shared patterns that cut across cases with distinct differences, while at the same time selecting critical cases that allow for logical generalization. The latter part of the sentence means that if something is true in a critical case, it is likely to be true for similar cases (Palinkas et al., 2015) Therefore, leading MNEs within their respective industry are chosen as the findings from this research might say something about the other companies that are active in that industry. This line of reasoning, however, is challenged by Yin (2009), who remarks that case studies are in-depth studies that clarify questions about selected cases, but provide little basis for scientific generalization. One way to address this problem is to select multiple cases within the same industry, so that findings might be generalizable to that particular subgroup (Yin, 2009; Eisenhardt, 1989). Unfortunately, because of time constraints it was impossible to add more
28 cases to this longitudinal multiple case study, and therefore the selection of critical cases was maintained (Palinkas et al., 2015).
Based on the above mentioned criteria two companies were selected. First, Unilever was selected because of its number one ranking in the recent 2015 Dow Jones Sustainability Index (DJSI) in the Food, Beverage and Tobacco industry group, and its overall high performance over the last couple of years. The DJSI is a globally recognized independent benchmark that measures company performance across economic, social and environmental dimensions (Robeco, 2016), thereby matching the CSR perspective of this research. Unilever is also a British-Dutch MNE with its headquarters in Rotterdam, and therefore likely to be frequently covered in both English and Dutch news articles (Robeco, 2016; Unilever; 2016). Thus, based on its top ranking in the DJSI, company profile information from its website, and solid reputation concerning CSR Unilever was chosen as the first case.
Conversely, Shell –selected to be the second case- has been more topsy-turvy in the DJSI in the Oil and Gas industry group. Just as Unilever this company was included in the index since it started in 1999, and remained within the top 10 per cent in its industry until 2010. The next two years, however, it was removed from the index because of violations against the
sustainability guidelines (Robeco, 2016; Shell, 2016). This, combined with the company profile from the Shell website, proves that the company is involved with CSR. Also, as a company of Anglo-Dutch origins (Shell, 2016), it should turn up often in both English and Dutch news articles. The contrasting journey through the DJSI, compared to Unilever, is what makes Shell particularly interesting (Robeco, 2016). As the two MNEs are active in different industries and have contrasting appreciations in the DJSI, the criterion of maximum variation is achieved. It would provide a chance to find out if there seems to be any consistency in how English and Dutch media cover both companies despite the differences (Palinkas et al., 2015).
3.4 Data collection and data reduction
The news articles were collected via the LexisNexis website. This website is known for its wide range of publications over a longer period of time with a collection of different languages (LexisNexis, 2015). For the sake of dependability an instruction manual is included in the appendix on how all the news articles were collected (Thomas and Magilvy, 2011). This is
29 necessary because slight differences in the query can result in a vastly different collection of news articles.
Huberman and Miles (1983) point out that it is necessary to reduce the raw and unstructured collection of data in order for the researcher to make sense of it. In this case it means trimming down the complete raw data collection until only the relevant news articles are left. As shown in figure 2, this can be done in a systematic manner that is both efficient and effective (Altheide et al., 2008).
First, filter out all redundant articles. These are the documents that are duplicates
resulting from the usage of multiple search terms to collect the data (Huberman and Miles, 1983; Altheide et al., 2008). For instance, the Dutch articles were collected using both ‘corporate social responsibility’ and its Dutch equivalent ‘maatschappelijk verantwoord ondernemen’ as search terms, of course combined with respectively ‘Unilever’ and ‘Shell’ as additional search terms. However, as a result, articles that held both search terms appeared in the data collection twice. Fortunately, when the articles are stored as Word-documents the search option, Control F, can be used to quickly find and delete articles with the same heading. Then, a second step is needed to filter out the irrelevant articles; articles that do not contribute to the answering of the research question in any way (Altheide et al., 2008). This required a quick scan of the articles to search for important terms from the literature review, and whether they were related to the selected MNEs (Adcock and Collier, 2001). For instance, some articles misinterpreted the search terms; not the company Shell, but the noun ‘shell’. Another benefit of this process is that it represents a chance to capture redundant articles that were not filtered out in the previous step, because sometimes the headings differ but the body of text is completely the same. If this is done
correctly, what remains is a collection of relevant articles that can be used for the process of data analysis (Altheide et al., 2008). As an aside, upon request the Word-documents containing all news articles can be provided for inspection; they have been stored per stage of data reduction.
30 Figure 2: Process of data reduction
Source: Author
The use of the LexisNexis website alone was sufficient to gather a large amount of articles on CSR-efforts of Unilever and Shell in both languages, as can be seen in tables 2 and 3. In order to find meaningful patterns, however, such a large amount of news articles is necessary (Yin, 2009). While in qualitative data analysis there is no such thing as a magic number that is required for analysis, it is acknowledged that the larger the amount of analyzed material, the more value can be derived from the findings, on the condition that the analysis was done in rigorous fashion (Pratt, 2009; Gibbert and Ruigrok, 2010).
Table 2: Data collection ‘CSR Unilever’; each phase of data reduction Stage Year English Raw English Filtered English Relevant Dutch Raw Dutch Filtered Dutch Relevant 2015 86 41 32 37 30 18 2014 72 53 32 23 19 11 2013 55 36 22 43 23 11 2012 66 42 22 33 28 15 2011 52 40 18 67 31 15 2010 39 31 17 24 21 13 2009 31 20 9 43 30 16 2008 20 17 13 51 43 22 Total 421 280 165 321 225 121 Source: Author Raw Data •Gather all English and Dutch news articles for the last eight years from LexisNexis and store them per year; 'Unilever Dutch 2015' etc. Filtered Data •Remove all redundant news articles by deleting all copies that some articles might have; use 'Control F' to verify which headings have doubles. Relevant Data •Scan through the remaining articles looking for key terms from the literature review and delete any redundant articles with exactly the same bodies of text. Coded Data •Carefully read the remaining articles and use a codebook to code every single piece of text in both a deductive and inductive way
31 Table 3: Data collection ‘CSR Shell’; each phase of data reduction
Stage Year English Raw English Filtered English Relevant Dutch Raw Dutch Filtered Dutch Relevant 2015 129 70 29 38 18 13 2014 137 72 22 31 21 12 2013 79 42 15 42 25 12 2012 78 33 15 21 20 8 2011 61 40 10 101 54 27 2010 61 35 14 66 32 12 2009 46 30 17 57 47 20 2008 34 27 18 44 40 12 Total 625 349 140 400 257 116 Source: Author 3.5 Data analysis
Huberman and Miles (1983) argue that the process of data reduction not only allows data analysis, but it already is a form of analysis. It helps the researcher to get acquainted with the data and to gain an evolving sense of how the data fits in with the theoretic framework and how it could possibly help to answer the research question (Huberman and Miles, 1983). Adcock and Collier (2001) build on this by stressing that research stands or falls with measurement validity. They argue that it is important to actually measure what you are supposed to measure. To avoid a discrepancy between these two things a few steps need to be taken. First, a vague background concept should be molded into a concrete systemized concept. Then, that systemized concept should be operationalized into a number of smaller measures (Adcock and Collier, 2001). This line of reasoning is also applied in this research. For instance, in the literature review the broad concept ‘CSR’, which can be interpreted in various ways (Lindgreen et al., 2012), has been discussed and led to the formulation of a clear definition of CSR that is maintained in this
research. Thus, the transition from a vague concept to a concrete concept has already taken place. Now it is necessary to divide this concrete definition into smaller measures (Adcock and Collier, 2001).
These smaller measures, known as codes and sub-codes (Pratt, 2009), help with the qualitative content analysis of the news articles. Qualitative content analysis is a research
32 orientation with an emphasis on discovery and description by searching for patterns between two or more variables (Altheide et al., 2008). Now that the stages of concept development (see literature review) and data collection (see previous paragraph) have been dealt with, the method of qualitative content analysis prescribes that the data should be coded, analyzed and interpreted. The researcher should be systematic and analytic, but not rigid; the systemized concepts and smaller measures should initially help the researcher navigate through the data, but new ones are expected and allowed to emerge from the data (Adcock and Collier, 2001; Altheide et al., 2008).
To link this to the thesis, while scanning through the articles during the second data reduction stage, some new codes were added to the coding scheme (see table 4) as a first draft of inductive coding. Afterwards, when the relevant data collection was completed, constant
comparisons across articles were made to allocate inductive codes to already coded news articles when necessary. The inductive codes that are the result of this iterative process, are showcased in italic font in table 4. It was a good way to produce richer findings (Altheide et al., 2008; Burla et. al., 2008).
Before the findings are set forth, however, all relevant pieces of text need to be inventoried. It would not be very insightful, though, to describe each of those pieces of text separately. Instead, a small amount is showcased in the upcoming tables in chapter 4. It should be stressed that these pieces, or quotes if you will, are not randomly chosen. As recommended by Rynes and Gephart (2004), they need to represent or capture key concepts and be selected on conceptual or methodological grounds. To address this concern the upcoming tables reflect the structure of the coding scheme, so that the selected quotes are insightfully linked to the
appropriate theme. Moreover, by grouping the quotes that belong to one theme in NVivo -a helpful qualitative tool- beforehand, it became easier to compare the quotes within one theme, and to select the ones that yield the richest descriptions. Even when stripped from their context these quotes provide insightful and self-explanatory information, but through the use of thick description and an explanation of how these quotes relate to the broader corpus of data a more holistic understanding can be achieved (Tracy, 2010; Rynes and Gephart, 2004).
33 Table 4: Codebook
Assumption Code Description
CSR as core business
CSR Fit There is a fit between CSR and the firm strategy.
CSR I &PD Firm engages in CSR related innovation and product differentiation.
Window-dressing Firm engages in CSR to cover up harmful practices of the core business. Use of MCSs KPIs Firm measures CSR through key performance indicators.
Doc & Coc Use of documents and codes of conduct to create a shared vision of CSR. CSR CEO CSR CEO CEO is a promotor and catalyst of CSR.
Reaching BOP BOP Fit Firm differentiates products especially for the BOP.
MNE satisfies stakeholders
Employees Firm takes care of its employees and acts in their best interest. Local community Firm leverages resources to help the local community.
Government Firm complies with government rules and has good relations with them. Value chain Firm takes its responsibility within the value chain and helps partners.
Reciprocity Firm wants something in return for good behavior. Shareholders Firm looks after the best interest of its shareholders. Environmental
needs
Water purification Firm looks after the environment by restoring the quality of the water.
Other Firm looks after the environment through various means.
MNE acts proactive when institutions fail
Urge institutions Firm urges institutions to facilitate a strong institutional environment. Maintain own high
standards
Firm maintains its own high standards. Especially when they feel that the institutions come up short.
Hesitance& Pressure
Firm hesitates to take sustainable initiatives and/or institutions put pressure on them to do so
Transparent
Social &
environmental rep.
Firm publicly publishes social & environmental reports and clearly shows how financial goals and sustainability are intertwined. Upfront about
problems
Firm communicates openly about the problems and threats that it faces. This also encompasses the failure to achieve (sustainable) goals. Focus on high
CSR performance
Sustainability rankings, awards, and role model
Firm is adamant to perform well in sustainability rankings. The firm wins awards, is a role model for other firms, and propagates high ambitions.
Rebound from scandal
Accusations, scandals & misconduct
Firm is accused of not behaving in a sustainable (enough) manner, or is involved in a scandal or misconduct.
Preemption or compliance with demands
Firm is quick to comply with stakeholder demands after being
responsible for a scandal or misconduct. Or the firm preemptively deals with foreseeable problems with stakeholders
