Conflict & Collaboration: the mediating effect of NGO-corporate
collaborations on the relationship between NGO activism and
Corporate Social Performance
Master Thesis
MSc Business Administration – Strategy Faculty of Economics and Business Supervisor: Panikos Georgallis
Lara Jongejans 11414987 21-06-2018
Statement of originality
This document is written by Student Lara Jongejans who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
Abstract
The relationship between NGOs and corporates is one that ranges from conflict to collaboration. This study aimed to establish a mediating effect that NGO-corporate collaborations have on the relationship between anti-corporate activism by NGOs and corporate social performance. In addition, the study tried to respond to the influential discourse that has taken place in social movement literature that states that activism is something of the past and collaborations are the future. Using data of activism attacks on 260 U.S. based companies from the S&P 500 list, several ordinary least squares regressions were conducted to assess the mediating influence of NGO-corporate collaborations on the relationship between anti-corporate activism and corporate social performance. Contrary to the expectations, results of the study show that anti-corporate activism and NGO-corporate collaborations both lead to a negative change in CSP. The study did not find significant support for the proposed mediating effect. The study does confirm that anti-corporate activism positively and directly leads to more collaborations between corporates and NGOs and thus shows that activism is still fruitful.
Table of Content
Abstract ... 3
Table of Content... 4
1. Introduction ... 5
2. Literature review and Conceptual Framework... 8
2.1 Corporate social performance ... 8
2.2 NGO’s and Corporates... 9
2.3 Anti-corporate activism by NGO’s ... 10
2.4 NGO-corporate collaborations ... 12
2.5 Anti-corporate activism and NGO-corporate collaborations ... 14
3. Methodology ... 18
3.1 Research design and sample... 18
Figure 2: Time frame of the collected data ... 19
3.2 Independent variable ... 19 3.3 Dependent variable ... 20 3.4 Mediating variable ... 21 3.5 Control variables ... 22 4. Results ... 24 4.1 Descriptive statistics ... 24 4.2 Correlation analysis ... 26
4.3 Multiple regression analysis ... 27
4.3.1 Assumptions ... 27
4.3.2 Regression results ... 28
5. Discussion ... 35
5.1 Theoretical and practical implications ... 37
5.2 Limitations ... 38 5.3 Future research ... 40 6. Conclusion ... 42 7. Reference List ... 43 Appendix A ... 48 Appendix B ... 49
1. Introduction
On the 17th of March 2010 non-profit organization Greenpeace launched a social media
attack on Nestlé’s Kit Kat brand in which they shed light on Nestlé’s use of unsustainable forest clearing in the production of palm oil (Ionescu-Somers & Enders, 2012). The video shows a bored office worker eating a Kit Kat, which turns out to be the finger of an orangutan, highlighting the severity of the situation for this endangered species. The video was seen 1.5m times and what followed was 200,000 protest emails. In response Nestlé partnered with Forest Trust, a Swiss charity, to provide them an independent review and advice of its palm-oil supply chains (The Economist, 2010). A few years later, Nestlé signs a partnership with NGO World Animal Protection and commits to adopting higher animal welfare standards (Strom, 2014). The example illustrates that the relationship between NGOs and corporates is one that ranges from conflict to collaboration. In the current heat of environmental issues such as global warming, deforestation and sea level rise, NGOs are starting to play an even larger role in pressuring companies to lead environmental strategies, as is illustrated by the Greenpeace Kit Kat case.
The role of NGOs has increasingly changed from confrontational to a stronger collaborative player, opposed to confrontation and criticism, in the global field of corporates (Kourula & Laasonen, 2010). According to Heap (2000) this is because globalization and the multinational nature of business nowadays has given rise to more power of the business community and a reduction in power of the state. NGOs now understand the increasing power of the corporates and they either go for confrontation or collaboration. NGOs can therefore be divided into two camps; the dark greens that have adversarial relationships with corporations and confront them to seek radical change; versus the bright greens that work in close collaboration with corporations to address social and environmental issues (Hoffman, 2009).
An increasing stream of literature looks at the different responses from firms in their CSR activities in reaction to anti-corporate activism (Spar & La Mure, 2003; Arenas, Lozano & Albareda, 2009; King & McDonnell, 2012; Den Hond & De Bakker, 2007) and how these adversarial relationships can generate positive impacts for society. The collaborative relationship between NGOs and corporates is also increasingly gaining more attention amongst scholars (Kourula & Laasonen, 2010). NGOs and corporates can complement each other with resources, create partnerships that more effectively use the knowledge and thus achieve better environmental protection (Rondinelli & London, 2003). These partnerships between corporates and non-profit organizations such as NGOs are defined as cross-sector collaborations and partnerships (Googins & Rochlin, 2000).
The increase in NGO-corporate collaborations is sometimes associated with a decline in adversarial activism (Utting, 2005). Utting (2005) emphasizes that an influential discourse has emerged that suggests that confrontational activism is something of the past and that the future lies in NGO-corporate collaborations. On the other hand, research shows that activism is still a fruitful way to bring societal change and other scholars (Van Dyke, Soule & Taylor, 2004) show that activism towards corporations is actually increasing. Literature on anti-corporate activism and NGO-anti-corporate collaborations has developed largely separately. Therefore, there is no evidence whether these trends are related to each other or not.
Although the influence of confrontational NGO campaigns on firms is widely researched, this cannot be said about the link between NGO-corporate collaborations and the effect on corporate social performance (Van Huijstee & Glasbergen, 2010). There is limited research exploring the relationship between social movements, for example NGOs and activists, and corporate social initiatives (Georgallis, 2017). There is little to no empirical evidence that shows the effect of NGO-corporate collaborations on corporate social performance in comparison to confrontational relationships. Do firms that have experience with NGO
collaborations respond differently to anti-corporate activism? Are they more likely to increase their corporate social activities when under pressure? Could those collaborations influence the relationship between anti-corporate activism and the CSR agenda of the firm?
According to Googins & Rochlin (2000) cross-sector collaborations and partnerships have been poorly understood in the management literature. Much of the empirical work is prescriptive, describing the key success factors in NGO-corporate collaborations, or exploratory and case-based (Spar & La Mure, 2003; Argenti, 2004; Stafford, Polonsky, Hartman, 2000). This has provided interesting insights into the interactions between business and NGOs but has left space to assess the effect of NGO-corporate collaborations on corporate social performance. This thesis will attempt to close this literature gap and will propose a mediating effect of NGO-corporate collaborations on the relationship between
NGO activism and corporate social performance.
This thesis strives to create an answer whether firms that experience activism have higher corporate social performance when they are involved in NGO-corporate collaborations; if these collaborations really lead to higher corporate social performance. This thesis aims to shed light on the different relations between NGOs and corporates; should NGOs prefer collaborations over activism to instigate changes in CSP? The contributions of the thesis are threefold. Firstly, this thesis brings together two streams of literature, namely the one on NGO-corporate collaborations and the one on anti-corporate activism. These literatures have developed largely separately and this thesis provides empirical evidence to show that these two trends are not actually independent from each other. Secondly, this thesis will provide new insights for managers of NGOs and corporates on the consequences of NGO-corporate collaborations and activism attacks and how NGOs can instigate changes in corporate practices concerning CSP. Thirdly, the results will deepen our empirical understanding of NGO-corporate relationships, both collaborative and confrontational.
2. Literature review and Conceptual Framework
The following section contains a brief review of the existing literature on the topic. Firstly, the focus is on the emerging presence of corporate social responsibility (CSR) and corporate social performance (CSP). Secondly, both confrontational and collaborative interactions amongst NGOs and corporates are discussed. Thirdly, the literature review ends with describing the relationship between anti-corporate activism and NGO-corporate collaborations and the effect on CSP.
2.1 Corporate social performance
Increasing environmental and social issues pressure corporates into addressing the environmental impacts of their activities (Rondinelli & London, 2003) and with this their CSR agenda (Utting, 2005). Emerging standards related to corporate social performance and the fact that these performance ratings become more transparent to society, also increase pressure on firms to engage in CSR activities (Deckop, Merriman, Gupta, 2006). The academic literature contains many definitions of CSR. In line with Aguinas & Glavas (2012, p.933) CSR is defined here as: ‘context-specific organizational actions and policies that take
into account stakeholders’ expectations and the triple bottom line of economic, social, and environmental performance’. Stakeholder theory suggests that businesses should respond to
the concerns of multiple stakeholders, including NGOs, consumers, environmentalists and community activists (Eesley & Lenox, 2006). From the stakeholder view, a firm has the ethical responsibility to choose corporate social performance over maximizing shareholder wealth (Deckop, Merriman, Gupta, 2006). The company’s interaction with these stakeholders, such as societal groups and NGOs, is an important part of CSR (Pedersen, 2006). There is a growing body of research that looks at how social activists pressure and influence corporations to change the firm’s behavior regarding social issues (Briscoe & Gupta, 2016). Following Wood (2010), CSR-activities and CSR responsiveness are two elements of the
corporate social performance of a firm. CSP is considered to be the outcome of a range of business activity that focuses on the impacts for society, stakeholders and the firm (Wood, 2010). All these impacts and outcomes can be measured and evaluated and nowadays this becomes increasingly more important for corporates.
2.2 NGO’s and Corporates
In the last decade interactions between corporates and NGOs concerning issues of corporate social responsibility and corporate social performance have increased immensely (Rondinelli & London, 2003). There is no universal agreement in the literature on what an NGO specifically is; this thesis uses Martens’ (2002, p. 282) definition: ‘NGOs are formal
(professionalized) independent societal organizations whose primary aim is to promote common goals at the national or international level’. As this includes a broad range of NGOs,
the focus within this thesis is on NGOs that Arenas et al. (2009) define as ‘social purpose NGOs’. These NGOs include different human rights organizations, groups dedicated to fight hunger and most importantly environmental groups (Arenas et al., 2009). Simmons (1998, p. 84) states that NGOs can affect national and multinational corporations in four ways: setting
agendas, negotiating outcomes, conferring legitimacy, and implementing solutions. Whilst the
character of the interactions between NGOs and corporates started off being overly confrontational, it is becoming more collaborative in recent years (Van Huijstee & Glasbergen, 2010). NGOs can either confront or be collaborative towards corporations and are therefore considered to be influential stakeholders of firms (Den Hond, De Bakker & Doh, 2015). Collaborations between NGOs and corporates are considered to be cross-sector alliances, as the governance structures, visions and missions of both parties fundamentally differ from each other (Rondinelli & London, 2003). The current corporate demand in establishing credentials for CSR has led to an increase in interactions between corporates and NGOs (Baur & Schmidtz, 2012).
2.3 Anti-corporate activism by NGO’s
Over the years several environmental disasters, such as the burst of the tailings dam in a Brazilian mining state (Paddison, 2015), Shell’s activities in Nigeria (Croft, 2014) and the explosion of BP’s oil platform in the Gulf of Mexico (Gabbatt, 2010), linked multinational corporations to social and environmental issues. Activists and NGOs took advantage of the publicity and high-profile cases. Companies are increasingly becoming the target of social movement organizations and NGOs’ activist attacks (Burchell & Cook, 2012). One way for NGO’s to exert influence, that has increased since the 1980s, is the so-called watchdog activism in which NGO’s ‘name and shame’ companies for their malpractice (Utting, 2005). Whether it be environmental, labor or religious activists; all have engaged in campaigns to pressure corporates in the United States and Europe (Bartley & Child, 2011).
NGO activism has spurred major changes in corporate behavior and governance over the last decade (The Economist, 2003; Den Hond & De Bakker, 2007; Spar & La Mure, 2003). This activism is performed to instigate changes in the behavior of these corporates, in particular to establish changes in undesirable practices (King & Soule, 2007). Examples of actions are boycotts, protests and letter-writing campaigns (Eesley & Lenox, 2006). Activists display banners on rooftops of buildings, chain themselves to oil platforms or march the streets with protest songs. The rise of the Internet, and with that the rise of social media, has only given NGOs even more powerful tools to attack corporates and their reputations (Argenti, 2004). These attacks can lead to direct economic damage on firms as they lower potential revenue, but more importantly, they can lead to a loss in reputation (King, 2008). By using the media to gain unwanted attention for the firm, boycotters challenge the firm’s practices and try to influence the public image of the firm.
Several scholars have studied how activism influences corporate social performance and reputation. Firstly, through case studies, Spar & La Mure (2003) investigated what accounts
for the variation in how firms respond to different activist attacks. They found that corporates respond differently to attacks because every company has its own agenda and every company has its own constraints concerning activities and costs thereof. Secondly, in a case study on multinational Starbucks, Argenti (2004) shows how the firm changes its CSR-activities in response to several activist attacks by NGOs. Thirdly, another study that investigated the relationship between activism, or social pressures, CFP and CSP was performed by Baron, Harjoto & Jo (2011). Their results confirm that indeed an increase in social pressures, for instance instigated by NGOs, leads to an increase in corporate social performance. Fourthly, a study done by McDonnell, King & Soule (2015) shows how corporations become more receptive to social movement activism when they are targeted multiple times over a period of time. The authors argue that firms adopt new practices and social management devices for managing activist threats and over time develop better ways for showing their responsible and social behavior.
In the studies mentioned before, NGOs have had a significant role in putting pressure on corporates leading to an increase in CSP and CSR. Whether it is with boycotts or demonstrations, the NGOs gained media attention and put the corporate in a bad spotlight. Companies often answer these NGO activism attacks by starting a dialogue with the NGO in question or by investing in CSR related activities. These attacks lead to the company’s attention being shifted to the social or environmental issue targeted and more specifically to their own CSR agenda. Companies are made aware of their actions and as a reaction they are more likely to invest in their CSR activities. These activism attacks therefore result in a positive increase in the corporate social performance of firms. Therefore, the following hypothesis can be derived:
Hypothesis 1. Anti-corporate activism by NGOs leads to an increase in the target firm’s
2.4 NGO-corporate collaborations
In trying to reduce the risk of exposure and negative activism and campaigning by NGOs, firms increasingly engage in dialogue or partnerships with NGOs (Philips, 2002). There are, however, a lot more reasons why firms and NGOs collaborate. Firstly, according to alliance theory and building on the knowledge-based view, corporates and NGOs enter into collaborative relationships to acquire, otherwise impossible to obtain, tacit knowledge (Rondinelli & London, 2003). Secondly, businesses can use NGOs’ assets to gain a competitive advantage over other companies (Yaziji, 2004). NGOs have four strengths or assets that corporations can benefit from, namely legitimacy, awareness of social forces, elaborate networks and networking skills and specialized expertise (Yaziji, 2004). Hence, corporates can take advantage of the reputation benefits that arise from working alongside with an NGO. This is particularly useful as investors are more focusing on risk factors related to reputation (Heap, 2000). Thirdly, collaborations can raise awareness on social and environmental responsibility issues (Utting, 2005) and help engage in CSR activities (Peloza & Falkenberg, 2009). Lastly, Den Hond et al. (2015) performed a study to look at different factors influencing firms to collaborate with NGOs. They demonstrate that understanding factors, such as fit between the firm’s and NGOs resources and past experiences with NGOs, are important to be able to explain how firms deal with social issues in their competitive markets and why they collaborate with NGOs.
NGOs have other reasons to engage in collaborations with corporates. First, NGOs can design and exert influence on the CSR standards as well as the monitoring and auditing processes of these (Arenas et al, 2009). This process is actually beneficial for both partners. Whilst corporates often know exactly how to measure business benefits on output and input levels, NGOs have more know-how of social performance and hence serve as good partners in social performance evaluation (Hansen & Spitzeck, 2011). Secondly, one of the drivers for
NGOs to collaborate with corporations is the financial support the corporates can provide them with, sometimes simply with cash donations (Peloza & Falkenberg, 2009). Heap (2000) argues that the potential of NGO-corporate collaborations is huge, as NGOs can force business to do the right and social thing; work towards the objectives of the NGO itself. Therefore thirdly, NGOs might want to engage in collaborations with corporates as it provides them the opportunity to pursue their social goals.
Scholars are trying to define what NGO-corporate collaborations actually are and what stages they can have. Austen (2000) developed a framework of different NGO-corporate collaborations. He defines three types or stages of collaborations; philanthropic, transactional and integrative. The philanthropic stage is the type of collaboration in which the corporate provides its financial resources as a simple donation to the NGO. The transaction stage is a type of collaboration in which the NGO and the corporate engage more and exchange resources on specific activities. Last is the integrative collaboration; the highest strategic level of collaboration. This is the type of collaboration in which people of both partners share the same values, vision and mission; all activities and resources begin to merge into collective action and the two partners depend on each other for activities and outcomes. This continuum allows for collaborating parties to assess what type of relationship they have, how it is changing and to see how they would want their collaboration to develop. Peloza & Falkenberg (2009) extend this framework with the option of collaboration with multiple NGOs; all for the sake of achieving corporate social responsibility objectives.
Although the topic of NGO-business collaborations is gaining more attention amongst scholars, most studies are case-based and exploratory into how the collaborations emerge and the success factors thereof (Spar & La Mure, 2003; Argenti, 2004; Hansen & Spitzeck, 2011). There are however some studies that focus on the results of NGO-corporate collaborations on CSR. Firstly, one study that was performed to try to empirically understand the conditions
under which NGO-business collaborations lead to improved CSR was done by Van Huijstee & Glasbergen (2010). They studied the interactions between the Dutch bank Rabobank and NGO FoEN on the subject of palm oil and soy. They evaluated the different roles of the stakeholder context in collaborations and demonstrated that the collaboration indeed led to improvements in corporate social responsibility. Secondly, Albino, Dangelico, & Pontrandolfo (2012) tested whether inter-organizational collaborations positively affect environmental performance. By testing their model on the largest publicly traded U.S. companies, they found that NGO-corporate collaborations are indeed beneficial for a company’s overall environmental performance, overall reputation, and positive for the management of the company’s environmental footprint (Albino et al., 2012).
These studies show that collaborations, environmental ones in specific, between corporates and NGOs have a positive influence on the environmental performance of firms. In collaborations with corporates, NGOs can exert influence on the CSR activities of corporates. NGOs can make corporates more aware of their environmental footprint and thus influence corporates in this way to invest in CSR activities. Corporates on the other hand can use NGOs to help them measure, communicate and evaluate their social performance. The more corporates engage with NGOs, the more attention will be focused on social and environmental issues previously only addressed by the NGOs. This will lead to a positive change in CSP. Resulting from this reasoning, the following hypothesis can be derived:
Hypothesis 2. The more collaborations a firm has with NGOs, the more it will increase its
corporate social performance.
2.5 Anti-corporate activism and NGO-corporate collaborations
From the perspective of corporates, collaboration-based partnerships seem to be more attractive than adversarial-based relationships in which the corporate is being accused (Nijhof,
De Bruijn & Honders, 2007). Adversarial experiences with NGOs in the past can help firms understand how to deal with NGOs and what factors lead to successful collaborations (Den Hond et al., 2015). By building on and referring to past experience with NGOs, firms can more easily understand what the NGOs demand and the reasoning behind their perspectives. Past experience can provide firms with insights into what factors NGOs see as important in collaborations.
Responding to activism and environmental challenges, many corporates focused on CSR strategies (Burchell & Cook, 2012), hence negative pressures of activism attacks can turn into behavior and promises by the firm to do better in the future (Den Hond et al, 2015). It is therefore possible that corporates that have been under pressure by NGO activism in the past are more likely to have collaborations with NGOs to prevent them from being under attack again. Argenti (2004) investigated how multinational Starbucks went from being pressured by NGOs and having its annual shareholder meeting disrupted to working with multiple NGOs in collaborations. NGO Global Exchange targeted and criticized Starbucks for not investing in Fair Trade coffee. The NGO realized that they could increase awareness around the social issue by focusing its attack on the most reputable brand in specialty coffee (Argenti, 2004). By assessing the previous attacks Global Exchange had posed on Nike, Starbucks realized that the NGO would pursue the campaign and turn it into a nationwide campaign. This could heavily damage the firm’s reputation until Starbucks would capitulate (Argenti, 2004). Another heavily studied activism attack is the Shell Brent Spar case. In June 1995, NGO Greenpeace heavily protested against Shell’s plan to dispose the Brent Spar oil platform in the ocean (Neale, 1997). After the heavy attack, Shell changed its approach towards stakeholders and understood that acceptance by society is of high importance (Neale, 1997). By analyzing the actions and behavior of Shell before and after the incident, Neale (1997) shows how the activism attack in the end led to a more open-minded and collaborative approach of the
corporate towards NGOs. The Brent Spar incident led Shell to focus more on its corporate social activities and environmental responsibilities (Zyglidopoulos, 2002) and led to a transnational ‘Brent Spar Dialogue Process’ in which Shell invited other parties, such as NGOs, to work together on alternative on-shore disposal options (Grolin, 1998). These studies show that firms that have experienced anti-corporate activism in the past are more likely to engage in NGO-collaborations and dialogues, as they know what threats they are up against. They are thus more likely to give into the demands and focus on their CSR agenda and their CSR related activities. Increasing investments and attention in this area will lead to a positive increase in the overall corporate social performance. Therefore, the following hypothesis can be derived:
Hypothesis 3. The more anti-corporate activism by NGOs a firm has experienced, the
more collaborations it will have with NGOs.
The previous hypotheses suggest that corporates that experienced activism by NGOs in the past will actively try to avoid new activism by either engaging in collaborations with NGOs or by improving their CSR activities. NGO-corporate collaborations therefore positively mediate the relationship between anti-corporate activism and CSP. Therefore, the final hypothesis can be derived:
Hypothesis 4. The positive relationship between anti-corporate activism and corporate
social performance is positively mediated by NGO-corporate collaborations.
Figure 1 displays the expected relationships between the variables as proposed in the 4 hypotheses. This conceptual model shows NGO-corporate collaborations as a mediating variable in the relationship between anti-corporate activism and corporate social performance.
3. Methodology
This chapter provides a description of the research design, the chosen databases and the sample of companies. Subsequently, the independent, dependent and mediating variables are explained. Lastly, the chapter concludes with a description of the control variables.
3.1 Research design and sample
To answer the proposed research question and examine the proposed effects of NGO-corporate collaborations on CSP, a quantitative, empirical research design is applied. The study is based on data of U.S. corporations that is accessible via publicly available databases. The list of companies used is the S&P 500. This index includes the 500 leading US companies on the stock market, with a capture of approximately 80% coverage of available market capitalization (S&P Dow Jones Indices, 2015). The reason for choosing these firms is that they are large and successful firms, which are exactly the ones that are prone to activist, NGO and social movement attacks. These firms have a higher potential of being targeted by activists, as King (2008) found that activists are more likely to target large firms that have strong reputations. As Yaziji (2004, p. 111) says it: ‘The wealthier and better known the
company is, the juicier the target it makes’.
In order to test the proposed hypotheses and find out if anti-corporate activism and NGO-corporate collaborations actually lead to a change in CSP, data from different years is needed. The proposed hypotheses suggest that anti-corporate activism likely leads to firms engaging in collaborations with NGOs. In order to test this, data on the independent variable is taken from a time period (2010-2014) explicitly chosen before the time period (2015) of the NGOs collaborations. This way the effect of the first explanatory variable can be measured on the latter. In addition, to measure if either activism or collaborations have the proposed effect on corporate social performance, data on the change in CSP is taken from a time period directly following the collaborations. Data on the change in CSP is needed for two years after the
firms have collaborations, to see if the proposed interactions between NGOs and corporates establish a significant change.
The final time frame of the collected data, shown in figure 2, is from 2010 until 2016. Firstly, data on anti-corporate activism is retrieved from the period between 2010 and 2014. Secondly, the information on NGO-corporate collaborations is taken from CSR reports stemming from 2015. Thirdly, the data on corporate social performance is retrieved from the database focusing on the years 2015 and 2016. These years are chosen as 2016 is the most recent year for which most ESG ratings are available and 2015 is the most recent year for which most CSR reports are available.
Figure 2: Time frame of the collected data
3.2 Independent variable
Following prior research (King, 2008; Bartley & Child, 2011; King & Soule, 2007; Van Dyke, Soule & Taylor, 2004) the independent variable anti-corporate activism is measured using newspaper articles. Newspaper articles are often used as a source of data for social movement attacks, such as protests and boycotts (King, 2008). Newspaper-based event data provides scholars with the opportunity to examine different types of collective action and newspapers cover all the relevant events, effects and outcomes (Earl, Martin, McCarthy & Soule, 2004). As Lipsky (1968, p. 1151) states: ‘If protest tactics are not considered
significant by the media, or if newspapers and television reporters or editors decide to overlook protest tactics, protest organizations will not succeed. Like the tree falling unheard in the forest, there is no protest unless protest is perceived and projected.’
2010 – 2014 Anti-corporate activism 2015 NGO-corporate collaborations 2015 – 2016 Change in CSP
The LexisNexis news search database is used to collect articles from different newspapers. The time period of 2010-2014 is chosen in which corporates experience activism. Firstly, a pilot search was done to test different combinations of search terms. More information on the structure, an example of a search result to illuminate the process and how I arrived at the final keywords through the pilot search can be found in Appendix A. The final search terms used to find cases of anti-corporate activism are ‘activism’, ‘activist’, ‘NGO’, ‘demonstration’, ‘boycott’, ‘tension’, ‘nonprofit’, ‘non-profit’ and ‘pressure’ in combination with the name of the company.
3.3 Dependent variable
Data on the corporate social performance of each firm is collected via Thomson Reuters Asset4. The database, in particular the Asset4 part, has an extensive amount of data on the environmental and social performance of these firms, including firm size, performance numbers, various CSR measures, other characteristics of the firms and data on the CSP and corporate environmental performance. Following other scholars (Chatterji, Durand, Levine, & Touboul, 2016; Cheng, Ioannou, & Serafeim, 2014; Hawn & Ioannou, 2016), indicators in Asset4 will be the ESG ratings, namely the environmental performance, the social performance and the corporate governance performance. Thomson Reuters divides the overall performance of a company into these three categories (figure 3).
To be able to gain a coherent overview of the corporate social performance of the companies, several aggregate (or category scores) measures are used. For each firm the following scores are collected: environmental score, corporate governance score and the social score. The data is collected for the years 2015 and 2016. Following other scholars (Cheng et al., 2014; Barnett, Hartmann & Salomon, 2017) the annual environmental, social and corporate governance scores are used to construct a composite CSP index for the years 2015 and 2016 for each focal firm. The CSP variable is constructed by giving equal weights to all three pillars. Finally, the difference in CSP for each firm between 2015 and 2016 is calculated as I am interested in studying whether relations with activists affect changes in firms CSP’ behavior.
3.4 Mediating variable
The mediating variable NGO-corporate collaborations is measured as the number of collaborations the corporate has with NGOs. Data is collected from publicly available CSR and sustainability reports.a The GRI Sustainability Disclosure Database, published by the
Global Reporting initiative, is used to find the CSR reports. This database contains CSR and GRI reports of over 12,000 organizations (Global Reporting Initiative, 2016). When a firm isn’t available in this database, the CSR report of the year 2015 is retrieved from the company’s website. Following Shumate & O’Connor (2010), the CSR report is found in different sections on the company websites including ‘about us’, ‘community’, ‘partners’ and/or ‘social responsibility’. b
a While secondary data on NGO-corporate collaborations is available from ASSET4, I chose to manually code this variable from CSR reports to capture not just whether a firm has collaborations with NGOs but also the number of collaborations. The data in ASSET4 only captures the first. The latter is more consistent with my hypothesis and offers the opportunity to exploit more variation in the data. I also dummy coded the variable into 0 (no collaborations) and 1 (all other amounts of collaborations). A correlation analysis revealed that NGO-corporate collaborations is highly correlated with this dummy version of the variable (.64, p-value < 0.01). I chose to work with the original variable to include the variation of the amount of collaborations.
b After analyzing CSR and annual reports of a random sample of firms from the S&P list, it became clear that NGO-corporate collaborations were not always explicitly mentioned in the annual report. Therefore, I assume
Collaborations in different fields will count, ranging from environmental innovations, preserving the environment, to fields as stimulating education or health & safety. By researching collaborations between NGOs and S&P 500 publicly listed U.S. firms, Shumate & O’Connor (2010) found a range of different industries for NGOs. These include animal, child welfare, civic organization, civil rights, consumer protection, corporate foundations, diversity, education, elderly, environment, family welfare, fine arts, health, hobby, literacy, poverty, religious, science, sports violence prevention and women’s welfare. The amount of collaborations of corporates with NGOs in these areas is counted for the year 2015.
After carefully reading and analyzing 15 random CSR reports from the sample of firms, a collection of words was chosen that were most used to describe the collaborations. Collaborations are found in the CSR reports by using several search terms; ‘NGO’, ‘nonprofit’, ‘non-profit’, ‘partner’, ‘partnered’, ‘collaborate’, ‘collaboration’ and ‘collaborated’. Only the collaborations explicitly mentioned using those terms were counted. Relationships that were described with ‘funded’ or ‘sponsored’ were not incorporated, as they hint towards corporate philanthropy and simply giving money and donations.
3.5 Control variables
The four control variables included in the model are firm size, firm performance, industry and reputation. Each of these variables has potential links to CSP and anti-corporate activism.
Firm size is included as a control variable because larger firms face more public pressure
concerning their social and environmental impacts and are thus more likely to invest in CSR (Wickert, Scherer & Spence, 2016). Firm size is measured by the number of total assets in 2014. Following previous research (Marano & Kostova, 2016), industry is included as a control variable because it might influence CSR initiatives. Secondly, it is controlled for because industries face different combinations of stakeholders and thus experience differing degrees of activism than others (Griffin & Mahon, 1997). From the 10 categories, 9 have been
dummy-coded and the category ‘Health Care’ is left out as a reference category. Thirdly, firm
performance is included as a control variable and measured by the company’s Return on
Assets in 2014. This data is also retrieved from the Thomson Reuters database. Firm performance is controlled for, as bigger companies with more financial resources are expected to have a higher corporate social performance than smaller firms as they have more resources to invest in CSR activities. Lastly, the fourth control variable included is reputation. This is controlled for, as firms with strong reputations are more likely to be targeted by activists (King, 2008). Following other scholars (Perrini & Vurro, 2014; King & McDonnell, 2012)
reputation is measured by using Fortune’s ‘World’s Most Admired Companies’ list in 2014
(Mehta & Fairchild, 2014). Fortune annually identifies and ranks the Most Admired Companies in partnership with Hay Group.
The final list used consists of 265 companies from the S&P 500 list. The 235 companies not used were eliminated either due to a lack of corporate social performance data in the Thomson Reuters Asset4 database or a complete lack of CSR information and no CSR report of the fiscal year 2015 available. I discuss the implications of this for the results in the discussion section.
4. Results
This chapter provides the results of the analyses. It includes the descriptive statistics of all the variables and the results of the Pearson correlation analyses. Lastly, it ends with the results of several regression analyses, including the mediating role of NGO-corporate collaborations on the relationship between anti-corporate activism and CSP.
4.1 Descriptive statistics
Table 1 provides the descriptive statistics of the variables included in the conceptual framework of the study. It provides the definition, mean, standard deviation, minimum and maximum of each of the variables. The control variable industry is discussed in detail below.
Table 1
Descriptive statistics
Minimum Maximum Mean Std. Deviation
Anti-corporate activism 0 6 0,44 0,999
NGO-corporate collaborations 0 13 1,89 2,458
Corporate social performance -1 0,956 0,039 0,143
Firm size (in Mill.) 1,2 2573,1 83,3 269,6
Firm performance -9,03 43,54 7,985 6,002
Reputation 0 1 0,45 0,498
The average company has a positive change in CSP score of approximately 3.9% and has on average two collaborations with NGOs. This indicates that on average the CSP positively increases from 2015 to 2016. The independent variable anti-corporate activism ranges between 0 (207 companies) and the maximum of 6 (1 company) attacks. In total 21.89% of all the companies, 58 companies, experienced anti-corporate activism between 2010 and 2014. This amount of activism attacks is fairly consistent with findings in other research; Durand & Georgallis (2018) found that more than two-thirds of companies in their sample were never attacked by environmental activists and Bartley & Child (2011) found that 23.8% of all firms in their sample were attacked at one point. The number of NGO-corporate collaborations for a
single firm ranged from 0 (106 firms) to 13 (1 firm). Interesting to see in Figure 3 is that the majority of the firms had either zero, one or two collaborations with a majority of 40% (106 firms) that had no collaborations with NGOs at all.
The 265 companies are classified into 10 different industries according to their ICB Industry Name. Table 2 displays the industries and frequencies, with a majority of the
companies in 5 industries, namely Consumer goods (13,2%), Consumer Services (13,6%), Financials (15,8%), Industrials (18,1) and Technology (10,9%).
Table 2:
Industry types & frequencies
Industry Frequency Percent
Industrials 48 18,1 Basic Materials 14 5,3 Consumer Goods 35 13,2 Consumer Services 36 13,6 Financials 42 15,8 Health Care 22 8,3
Oil & Gas 17 6,4
Technology 29 10,9
Telecommunications 3 1,1
Utilities 19 7,2
Total 265 100
Verification for a normal distribution for each of the metric variables shows that actually none of these are normally distributedc. Univariate analysis of the variable CSP revealed a
few extreme outliers (z > 3) that significantly alter the distribution. Therefore, these 5 cases are excluded from the analyses. The standard deviation of the variables firm performance and
firm size both indicate an undesirable high spread so in order to make these variables more
cBoth firm performance and firm size are positively skewed to the right. Firm size is non-normally distributed, with skewness of 6,625 (SE=0.150) and kurtosis of 48,637 (SE=0.298). Firm performance is also non-normally distributed, with skewness of 1,184 (SE=0.150) and kurtosis of 4,430 (SE=0.298). The results of the Kolmogorov-Smirnov test are for both these variables highly significant (p<0.05). This reflects the skew in the histograms.
Figure 3: Amount of NGO-corporate collaborations
reliable to work with and facilitate an appropriate data analysis, log transformations are performed.
4.2 Correlation analysis
The Pearson correlation coefficients are used to check the strength of the linear relationships between the continuous variables anti-corporate activism, NGO-corporate collaborations, CSP and the control variables. Table 3 below presents the results.
Table 3:
Pearson correlation analyses
1. 2. 3. 4. 5. 6.
1. Anti-corporate activism
2. NGO-corporate collaborations .293**
3. Corporate social performance -.220** -.180**
4. Firm performance .017 -.033 -.047 5. Firm size .367** .195** -.155* -.390** 6. Reputation .225** .169** -.209** .004 .310** 7. Industrials -.137* -.085 -.061 .129* -.153* .037 8. Basic Materials -.054 .001 -.088 -.028 -.103 -.113 9. Consumer Goods .136* .203** -.140* .137* -.109 .039 10. Consumer Services .078 .059 .082 .065 -.042 .107 11. Financials -.016 -.002 .363** -.232** .332** -.040 12. Healthcare -.102 -.059 -.012 .158* -.025 -.013
13. Oil & Gas .142* -.069 -.084 -.134* .096 -.039
14. Technology -.067 -.059 -.096 .054 -.130* .085
15. Telecommunications .096 .062 -.073 -.070 .149* .047
16. Utilities .009 -.044 -.038 -.179** .075 -.165**
*p<.05 **p<.01
The independent variable anti-corporate activism shows a significant, positive correlation with NGO-corporate collaborations (r = .293), firm size (r = .367) and reputation (r = .225) at the 0.01 significance level. This indicates that when anti-corporate activism is high, it is likely the firm has more NGO-corporate collaborations than firms who have not been attacked. In other words, high anti-corporate activism is associated with high NGO-corporate collaborations. In addition, the larger firms with a strong reputation also are likely to have
more activism attacks by NGOs than smaller firms in terms of the total number of assets. This is in line with earlier research (King, 2008) that found that activists mostly focus on large-sized firms with strong reputations. The results also strikingly indicate a weak negative correlation between anti-corporate activism and corporate social performance (r = -.220). Another weak negative correlation (r = -.180) is found between NGO-corporate collaborations and CSP. To establish which variable is the cause and which is the effect, regression analyses need to be conducted.
4.3 Multiple regression analysis
Before starting with the regression analyses, several assumptions of linear regression need to be met. Hayes (2013) states that (1) the residuals of the regression should follow a normal distribution, (2) there should be homoscedasticity and (3) we should assume linearity. Field (2013) lastly states that we must assume absence of multicollinearity, meaning that the independent variables are not highly correlated with each other.
4.3.1 Assumptions
Firstly, linearity and the normal distribution are checked. The histogram of the standardized residuals of the dependent variable CSP shows an approximate normal distribution. By looking at the P-P plot of the standardized residuals of the dependent variable CSP, we can assume normality although there are slight deviations. To check linearity, the dependent variable anti-corporate activism is plotted against the independent variable corporate social performance in a scatter-plot of the residuals. As there was an approximate linear spread and not a curved one, linearity can be assumed.
Secondly, homoscedasticity was checked by looking at the scatterplot of the residuals. The scatterplot showed severe violation of this assumption (Appendix B) as there is a wider distribution to the right of the plot, almost like a cone shape. Violation of this assumption can
invalidate statistical inferences so in order to correct for this violation, White-Huber robust standard errors are used in the SPSS extension PROCESS by Andrew F. Hayes.
Lastly, multicollinearity exists when variables are highly correlated. Values above 0.8 indicate these strong relationships. The highest correlation, presented in table 3, is -.391. This confirms that multicollinearity does not pose a threat to the study.
4.3.2 Regression results
To test the hypotheses, a series of path analyses, also called regression analyses, are performed with SPSS macro analysis called PROCESS by Hayes (2013). This OLS regression-approach uses bootstrapping to generate a 95% confidence interval to measure the effect of the mediating variable on the relationship between the independent and the dependent variable. Hayes (2013) states that mediation occurs if zero is not included within the lower and upper bound of the confidence interval. The PROCESS extension provides the option to correct for heteroscedasticity with Huber-White robust standard errors, which are used in this study.
This thesis uses mediation model 4 (Figure 4) by Hayes (2013). Baron & Kenny (1986) state that in order for mediation to occur the following conditions must hold: the independent variable must affect the mediator (ai), the independent variable must also significantly affect
the dependent variable (c’), and thirdly the mediator must significantly affect the dependent variable (bi). If mediation occurs, then the effect of the independent variable on the dependent
variable is stronger through the mediating variable then when this variable is left out. Baron & Kenny (1986) also state that the independent and mediating variable should be correlated, because assumed is that one leads to the other. As can be seen in Table 3, this applies to this thesis as anti-corporate activism and NGO-corporate collaborations are significantly correlated.
Figure 4: conceptual and statistical diagram (Hayes, 2013)
Indirect effect of X on Y through Mi = ai bi Direct effect of X on Y = c'
Table 4 below presents the results of the regression analyses in PROCESS. Firstly, the relationship between the independent variable and the mediating variable is tested. The second model tests the relationship between the mediator NGO-corporate collaborations and the dependent variable CSP. Thirdly, the right column represents the total effect model which provides information about the effect of the independent variable on the dependent variable without taking the mediator into account. In all three models the control variables are added as covariates at once, as PROCESS doesn’t work in a hierarchical way.
Hypothesis 1 proposed that anti-corporate activism by NGOs leads to an increase in the target firm’s corporate social performance. From the results in table 4, the model of this hypothesis is significant with F(13,246)=6,9317 p < 0.001 and explained 27.9% variance in CSP. From the results in table 4 it can be derived that, contrary to what was hypothesized, a negative significant relationship (b=-.006, p < 0.05) is found between anti-corporate activism and CSP. This indicates that firms that experience anti-corporate activism by NGOs actually have a negative change in their CSP in comparison to firms who do not experience this. We can conclude that no empirical evidence is found in support of hypothesis 1, however
significant contradictory results are found. The implications hereof are discussed in the discussion section.
Hypothesis 2 proposed that the more collaborations a firm has with NGOs, the more it will increase its corporate social performance. The model is significant with F(14,245)=6,2797 p < .001. The results in table 4 show that the relationship is significant and negative (b=-.002, p < 0.05), contradicting the hypothesis. These results mean that firms with more collaborations with NGOs actually have a negative change in CSP in comparison to firms that have fewer collaborations. Thus, we can conclude that no evidence was found in support of hypothesis 2. The implications of the contradictory findings are discussed in the discussion section.
Hypothesis 3 proposed that the experience of anti-corporate activism has a direct and positive influence on the number of NGO-corporate collaborations; the more anti-corporate activism by NGOs a firm experiences, the more collaborations it will have with NGOs. As stated by the first model summary, the model of the third hypothesis is significant with F(13,246)=2,7772 p < 0.001. The results in table 4 show that activism indeed positively influences the number of collaborations corporates have with NGOs (b=.532, p<0.05). This relationship is moderately strong with a significance level of 0.05. This means that firms that have been attacked by NGOs have more collaborations with NGOs than firms that do not experience anti-corporate activism by NGOs. It can be concluded that empirical evidence is found in support of hypothesis 3.
The aim of the study is to investigate the role of NGO-corporate collaborations on the relationship between anti-corporate activism and CSP. Accordingly, the fourth hypothesis proposes that the relationship between anti-corporate activism and CSP is positively mediated by the presence of NGO-corporate collaborations. When the effect of the independent variable, in this case anti-corporate activism, on the dependent variable, in this case CSP, is best explained with another variable, there is sign of mediation (Baron & Kenny, 1986). Table
5 presents the results of the analysis. The mediating effect, in the table presented as the indirect effect, is negative, but not significant as zero falls in-between the two interval bounds (-.0039 and .0001). This means that the supposed mediating effect of NGO-corporate collaborations on the relationship between anti-corporate activism and corporate social performance is not supported by significant results. In conclusion, there is no significant empirical evidence found to support hypothesis 4.
Concerning the control variables, firm performance and reputation are not significantly related to NGO-corporate collaborations and CSP, meaning that there is no statistical evidence found that firms with a higher performance and a stronger reputation have more collaborations with NGOs or a stronger change in CSP in comparison to firms with a lower performance and a less strong reputation. Firm size, on the other hand, significantly and negatively relates to the overall change in CSP (b=-.029, p<0.01). This suggests that the smaller the firm size, the stronger their change in CSP is.
Looking at the results for the different industries, interesting to see is that concerning the relationship between anti-corporate activism and NGO-corporate collaborations, a significant effect is found for firms in the Consumer Goods industry (b=1.422, p<0.05). This means that firms in that sector have significantly more collaborations with NGOs when they have experience with anti-corporate activism than firms in the reference sector. In addition, the results also show that a significant effect is found for firms in the Basic Materials (b=-.030,
p<0.05) and Financials (b=.069, p<0.01) sector on the relationship between anti-corporate activism and CSP. This means that firms in the Financials sector have a significantly higher change in CSP when they experience activism by NGOs in comparison to firms in the reference category. Results for the industry Basic Materials show a significant negative effect, meaning that firms in this sector have significantly less change in CSP when experiencing activism than firms in the reference category.
From a mediation analysis conducted using ordinary least squares path analysis, it can be concluded that there is no significant evidence found that NGO-corporate collaborations positively mediate the relationship between anti-corporate activism and corporate social performance. There is however evidence found that anti-corporate activism significantly influences the amount of collaborations corporates have with NGOs. In addition, contradicting results are found regarding the two hypotheses on CSP; both anti-corporate activism as well as NGO-corporate collaborations are negatively associated with change in CSP. This indicates that both these interactions between NGOs and corporates do not lead to a positive change in CSP.
Table 4
Results regression analyses in PROCESS
Consequent
NGO-corporate collaborations (M) CSP (Y) CSP (Y)
Antecedent Coeff. SE p Coeff. SE p Coeff. SE p.
Anti-corporate activism (X) ai .532* .249 .033* -.005 .003 .127 c' -.006* .003 .041*
NGO-corporate
collaborations (M) --- --- --- bi -.002* .001 .038* --- --- ---
Firm performance (ROA) -.663 1.605 .679 -,027 .027 .314 -.026 .029 .370
Firm size (total assets) .500 .382 .192 -.029* .007 .000** -.030** .007 .000**
Reputation .326 .343 .342 -.012 .008 .134 -.013 .008 .114
Industrials .037 .474 .936 -.009 .009 .327 -.009 .009 .322
Consumer Goods 1.422* .692 .040* -.015 .010 .121 -.019 .010 .056
Consumer Services .481 .552 .384 .021 .013 .109 .019 .013 .131
Financials -.019 .522 .970 .069** .017 .000** .069** .017 .000**
Oil & Gas -.870 .574 .131 -.011 .011 .301 -.009 .011 .401
Technology -.007 .520 .988 -.019 .010 .058 -.019 .010 .060 Telecommunications .636 1.806 .724 -.007 .019 .710 -.019 .016 .593 Utilities -.221 .634 .727 -.005 .013 .673 -.005 .013 .705 Basic Materials .550 .704 .435 -.030** .009 .002** -.032** .009 .001** constant i1 -1.527 4.143 .712 i2 .303 .077 .000 -.307 .080 .000 R2 = .150 R2= .286 R2=.279 F (13, 246) = 2,7772 p<.001 F (14,245) = 6,2797 p<.001 F (13, 246) = 6,9317 p<.001 *p<.05 **p<.01
Table 5:
Direct, total & indirect (mediating) effects
Effect SE (HC0) p LLCI ULCI
Direct effect c1’ -.0051 .0033 .1271 -.0116 .0014 Total effect c1 -.0064 .0031 .0419 -.0126 -.0002
Boot SE Boot LLCI Boot ULCI
5. Discussion
The following chapter includes the discussion of the results, theoretical implications, practical implications and an overview of limitations related to methodology and results. The chapter ends with future research avenues.
The aim of the study was to examine the mediating influence of NGO-corporate collaborations on the relationship between anti-corporate activism and corporate social performance. As little is known about the effect of NGO-corporate collaborations on CSP (Van Huijstee & Glasbergen, 2010) the study was aimed to provide new insights. Do firms that have collaborations with NGOs respond differently to anti-corporate activism? Was the influential discourse Utting (2005) talked about right to say that activism is something of the past and collaborations with NGOs are the future? Hypothesized was that firms under attack are more likely to have collaborations with NGOs which in turn leads to a higher positive change in CSP. The results of the study are however strikingly different and contradictory.
Firstly, concerning hypothesis 1, the relationship between anti-corporate activism and corporate social performance is explored. Against expectations, this relation was found to be significantly negative and therefore no support was found to support hypothesis 1. This means there is a direct negative relationship between anti-corporate activism and CSP, which is surprising as previous literature stated that activism positively influences the corporate behavior of firms (Den Hond & De Bakker, 2007). Qualitative, case-based approaches describe the relationship and positive results of NGO activism on the behavior of corporates (Argenti, 2004; Spar & La Mure, 2003), however previous literature has fallen short to establish the empirical direct influence of activism on CSP. The negative association found is contradictory to what was hypothesized. There are several reasons for this. Firstly, an explanation for this finding can be based on the nature of the companies and their capabilities of adaptation. McDonnell, King & Soule (2015) show that corporations become more
receptive to social movement activism over time because they adopt new practices and devices for how to manage activist threat. This could indicate that firms do not necessarily respond with investing in their CSP practices right away. This could be a potential reason why the results defied expectations. However, more research is needed to confirm this and I will elaborate on this in the section on future research. Secondly, another reason why anti-corporate activism is negatively associated to anti-corporate social performance could be that change in corporate social performance requires investments and time. Investing in CSP activities, shortly after the firm experiences activism, might take a little longer than a few years to reap the desired results and turn into a positive change in CSP. I will discuss more on these limitations in the Limitations section.
Secondly, hypothesis 2 proposed that the more collaborations a firm has with NGOs, the more it will increase its corporate social performance. No significant evidence was found to support this hypothesis. Surprisingly, the results show significant evidence that the relationship between NGO-corporate collaborations and CSP is actually negatively associated. This is not in line with literature that shows examples of positive changes in corporate behavior resulting from NGO-corporate collaborations (Van Huijstee & Glasbergen, 2010; Albino et al., 2012). Most studies on the subject are case-based and exploratory into the success factors of the collaborations (Spar & La Mure, 2003; Argenti, 2004; Hansen & Spitzeck, 2011). These contradictory findings might spur research interest into what it actually is that NGOs establish with collaborations. I will elaborate on this in the section on future research avenues.
Thirdly, the effect of anti-corporate activism on NGO-corporate collaborations is explored. Significant evidence is found in support of the hypothesis that firms that faced activism by NGOs are more likely to engage in collaborations. This is in line with the previously described cases of Starbucks (Argenti, 2004) and Shell (Neale, 1997) that also showed that
activism leads to the corporate being more open-minded towards NGOs and possible collaborations. These findings confirm that firms that have experienced anti-corporate activism in the past are more likely to engage in NGO-collaborations and dialogues, as they know what threats they are up against.
The fourth hypothesis proposed that NGO-corporate collaborations mediate the relationship between anti-corporate activism and corporate social performance. The results showed that no significant evidence was found to support this hypothesis.
In summary, the relations between NGO-corporate collaborations on CSP and anti-corporate activism on CSP are both found to be negative at a significance level of 0.05. The relation between anti-corporate activism and NGO-corporate collaborations is found to be positive statistically significant. Potential reasons why the results defied the proposed expectations can be the adaptability of corporates and the fact that CSP requires investments and time. Other reasons, such as characteristics of the population, outside factors that could possibly explain the findings, and limitations are discussed in the limitations and future research section.
5.1 Theoretical and practical implications
There are a few implications of the findings to the area of research. First of all, this thesis brings together two streams of literature, namely the one on NGO-corporate collaborations and the literature stream on anti-corporate activism. As these literature streams have developed largely separately, there was no evidence to believe these trends are related to each other. This thesis provides empirical evidence to show that these two trends are not actually independent from each other. The results show that one actually leads to the other; anti-corporate activism significantly leads to NGO-anti-corporate collaborations.
Secondly, this thesis does not provide full confirmation of the influential discourse Utting (2005) describes and thus can’t provide a complete answer if activism is something of the past
and collaborations are the future. The results do show that activism is still fruitful as it leads to collaborations between NGOs and corporates. However, the results do not specifically show if activism or collaborations lead to a higher CSP. On the contrary, the results actually indicate that both relationships between NGOs and corporates lead to a negative change in CSP. As such, these results deepened our empirical understanding of NGO-corporate relationships, whether collaborative or confrontational. Thirdly, this thesis contributes to the social movement literature (Bartley & Child, 2011; King, 2008; De Hond & Den Bakker, 2007; King & Soule, 2007) as it finds that anti-corporate activism significantly leads to more collaborations between NGOs and corporates.
Concerning NGOs, an important managerial implication of the results is that activism and the pressure they exert on firms indeed results in more collaborations with these corporates. Thus, managers of NGOs can see activism as a path towards changing these firms into becoming more collaborative and open for dialogue. It is naïve to believe that they are able to completely alter the social behavior of the corporate, but they could introduce slight alterations concerning corporate activities and expenditures. The results show however that activism nor collaborations directly positively affect the corporate social performance of the corporates. Managers of NGOs might therefore still opt for activism as this will lead to more collaborations, and thus ways to influence corporates.
5.2 Limitations
This thesis has some limitations concerning the methodology and the results that have to be acknowledged. Firstly, with regard to the mediating variable, the data on collaborations is extracted from publicly available CSR reports. Therefore, the data is restricted to what is found in these reports and is highly susceptible to differences in reporting. Some companies were very elaborate in their explanations of their collaborations with NGOs, whilst others only briefly mentioned them. Companies could also collaborate with NGOs, but simply not
