The Influence of Core-stigma on Corporate Social Responsibility Reporting
Student: Robin Britt Danielle Bouma / Student № 10203265 MSc. Business Administration, Strategy Track
University of Amsterdam, Faculty of Economics and Business
Supervisor: Hesam Fasaei
Erasmus University Rotterdam, Rotterdam School of Management 22nd of June, 2017
STATEMENT OF ORIGINALITY
This document is written by Robin Bouma 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.
TABLE OF CONTENTS STATEMENT OF ORIGINALITY ... 2 TABLE OF CONTENTS ... 3 ABSTRACT ... 4 1. INTRODUCTION ... 5 2. THEORY DEVELOPMENT ... 9 2.1 ORGANIZATIONAL STIGMA ... 9 2.1.1 Core-Stigma ... 10
2.2 CORPORATE SOCIAL RESPONSIBILITY ... 14
2.3 CORE-STIGMA AND CORPORATE SOCIAL RESPONSIBILITY ... 17
2.4 ORGANIZATIONAL VISIBILITY ... 20
3. METHOD ... 23
3.1 SAMPLE AND DATA COLLECTION ... 23
3.2 MEASURES ... 24
3.2.1. Dependent Variable: Corporate Social Responsibility ... 24
3.2.2 Independent Variable: Core-stigma ... 24
3.2.3 Moderating Variable: Organizational Visibility ... 25
3.2.4 Control Variables ... 25 3.3 METHOD OF ANALYSIS ... 26 4. RESULTS ... 28 4.1 DESCRIPTIVE STATISTICS ... 28 4.2 HYPOTHESES TESTING ... 31 5. DISCUSSION ... 35 5.1 ACADEMIC RELEVANCE ... 35 5.2 MANAGERIAL IMPLICATIONS ... 37
5.3 LIMITATIONS AND SUGGESTIONS FOR FUTURE RESEARCH ... 38
6. CONCLUSION ... 41
REFERENCES ... 44
APPENDIX ... 49
List of Figures Figure 1. Conceptual Model. ... 22
List of Tables Table 1. Summary statistics per sample group. ... 30
Table 2. Correlations between variables. ... 30
Table 3. Cross-tabulation of stigma and CSR reporting. ... 31
Table 4. Results of the Logistic Regression Analysis. ... 33
Table 5. Results of the Additional Logistic Regression Analysis. ... 33
ABSTRACT
The current study investigates the relationship between core-stigmatization and corporate social responsibility (CSR) reporting. Core-stigma occurs when an organization is evaluated negatively by a social audience because it conflicts with certain social norms. This study aims to identify if core-stigmatized organizations are more likely to issue a CSR report than their non-stigmatized counterparts. Building on impression management theory it is theorized that core-stigmatized organizations will use CSR reporting as a way to cushion the negative effects experienced when facing this type of evaluation. Furthermore, the moderating effect of organizational visibility on this proposed relationship is studied. A sample of 123 US publicly listed firms is used, which are involved in alcohol, firearms, gambling, military, nuclear power and tobacco activities in the year 2012. These firms are chosen as these are in perceived violation of social norms. A matched sample of 123 control firms is also gathered for the year 2012, matched on size and industry. After testing the hypotheses, no difference in CSR reporting was found between the core-stigma group and the control group. Subsequently, no moderating effect was found for organizational visibility. These findings highlight the importance of more research into the influence of core-stigma, as they are not in line with earlier work. Furthermore, it questions formerly accepted views that CSR reporting gets used as a strategic tool to improve a firm’s reputation, as this does not hold for stigmatized organizations.
1. INTRODUCTION
“It takes twenty years to build a reputation and five minutes to ruin it.”
- W. Buffet (1995) Stigmatization of an organization is an example of something that can eradicate any existing positive reputation an organization might have (Mishina & Devers, 2012), thereby ruining years of hard work. Organizational stigma has been described as the “label that evokes a collective stakeholder perception that an organization possesses a fundamental, deep-seated flaw that deinidividuates and discredits the organization” (Devers, Dewett, Mishina, & Belsito, 2009). Interest in the stigmatization of organizations has been growing as a result of a rash of corporate scandals and failures (Devers et al., 2009). An example would be the publication of the book Silent Spring, that developed into an issue of critical proportions for the chemical industry in the United States (Hoffman & Ocasio, 2001). The book argued that the chemical industry was poisoning the whole food chain and thus ultimately ourselves, by using a certain pesticide. The publication of the book and subsequent media attention resulted in major changes in the industry and increased governmental control on pesticide application. Although research initially focused on the effect of stigma on individuals in organizations (Semadeni, Cannella, Fraser, & Lee, 2008; Wiesenfeld, Wurtmann, & Hambrick, 2008), recent work has begun to focus on organizational-level stigma and outcomes (Devers et al., 2009; Hudson, 2008; Hudson & Okhuysen, 2009; Mishina & Devers, 2012; Wolfe & Blithe, 2015).
A distinction has been made between event-stigma and core-stigma. Event-stigma is said to be the result of an unusual or anomalous event (Hudson, 2008). The second type of stigma, namely core-stigma, occurs when a social audience evaluates an organization negatively because of some core attribute of the organization, such as outputs, customers or routines that are in perceived violation of social norms (Hudson, 2008; Hudson & Okhuysen, 2009). Although both types of stigma are hard to overcome, a core-stigma involves an attribute
so fundamental for the organization that it cannot change this attribute without seizing to exist as it is now known.
When faced with core-stigma, organizations have to revert to systematic strategies in order to minimize the impact of the deeply-rooted negative evaluations (Grougiou, Dedoulis, & Leventis, 2016). Some of these strategies include isolation and hiding strategies (Hudson, 2008; Hudson & Okhuysen, 2009) that help the organization to minimize unwanted attention. Research has however shown that using these strategies alone is an insufficient way of dealing with core-stigma. Wolfe & Blithe (2015) showed that brothels manage their stigma by distancing themselves from identities that they perceive as socially undesirable, drawing analogies to non-stigmatized industries and constructing selectively permeable organizational boundaries to build familiarity with external community members. Other ways of dealing with the need to stay hidden, whilst also being visible have remained unexplored.
The above-mentioned strategies seem to be a way that core-stigmatized organizations try to manage the impression of their audiences. Through these impression management strategies, they try to lessen the stigma they are facing, seeing that stigmas reside in the eye of the beholder (Hudson, 2008). A possible and understudied impression management strategy that could be adopted by core-stigmatized organizations is the use of Corporate Social Responsibility (CSR) and more specifically CSR reporting. Previous research has shown that CSR is used as a strategic tool through which managers secure broader stakeholder support and as a way to improve their reputation (Aguinis & Glavas, 2012), but research has mostly focused on firms with favourable reputations (Grougiou et al., 2016). Taken together this leads to the first research question:
What is the influence of core-stigma on CSR reporting?
Organizational visibility as a predictor of social and environmental responsiveness has received much attention (Aguinis & Glavas, 2012; Bowen, 2000; Brammer & Millington, 2006;
Chiu & Sharfman, 2011; Yu, Lo, & Li, 2016). Organizational visibility can broadly be defined as how visible an organization is to its constituents (Bowen, 2000). Although there is a well-accepted view that higher organizational visibility leads to more pressure on the organization and thus more responsiveness (Henriques & Sadorsky, 1999; Sharma & Nguan, 1999; González-Benito & González-Benito, 2006 in Yu et al., 2016), there are also studies that have supported the view that higher visibility leads to more resistance and thus less responsiveness (Lefebvre et al., 2003; Darnall et al., 2010; Udayasankar, 2008 in Yu et al., 2016). It is possible that organizational visibility does not have a direct effect, but rather is a moderator. Core-stigmatized organizations that are very visible to their constituents might have a bigger incentive to counter negative evaluations by adopting CSR practices. By using CSR as an impression management tactic these organizations aim to protect their social license to operate, which stakeholders can easily revoke (Banerjee, 2012). Social license to operate has been defined as “an ongoing acceptance or approval from the local community and other
stakeholders who can affect profitability” (Parsons, Lacey, & Moffat, 2014, p. 84). This
pressure to act as a good corporate citizen might increase with higher visibility and when an organization has been stigmatized. Taken together this leads to the second research question:
What is the influence of organizational visibility on the relationship between core-stigma and CSR reporting?
This research contributes to the existing literature in multiple ways. Firstly, it furthers the nascent literature on stigma by trying to identify an impression management technique that gets implemented by core-stigmatized organizations. Furthermore, it contributes to the body of research on CSR and extends its scope by investigating stigmatized firms rather than firms that have a favourable reputation. Thirdly, by studying the moderating effect of organizational visibility, this research furthers the debate about the effect of visibility on corporate social responsiveness by treating it at as a moderator rather than a direct effect. Finally, this study tries
to inform managers and business participants about the strategies used in organizations facing core-stigma. As such, it can give insight into the motivations for issuing a standalone CSR report.
2. THEORY DEVELOPMENT
2.1 ORGANIZATIONAL STIGMA
The concept of stigma originated in ancient Greece, where the label was used as a social control mechanism to represent an individual’s inferior moral status (Devers et al., 2009; Goffman, 1963). It signalled that an individual was in some way tainted and that contact should be avoided. The most used definition of stigma is the one proposed by Goffman (1963), which describes stigma as an “attribute that is deeply discrediting” and reduces the bearer “from a whole and usual person to a tainted, discounted one”. Stigmatization can lead to devastating social and economic outcomes and ultimately threaten survival, due to its collectively held nature (Devers et al., 2009). The research on stigma has its roots in the labelling theory from the sociology of deviance and mental illness (Mishina & Devers, 2012). This theory states that the behaviour of deviants does not necessarily make them deviants as opposed to non-deviants, but rather that the label given to this behaviour by conventional and conforming members of society is what transforms persons into deviants (Kitsuse, 1962). An important characteristic of stigma thus is that it resides in the eye of the beholder (Hudson, 2008), as a label given to a certain individual by others.
Research into organizational stigma compared to individual level stigma is more nascent (Mishina & Devers, 2012). Organizational stigma has been defined as a “label that a certain social audience affixes to an organization through a socio-political process (Pozner, 2008) that evokes a group-specific, collective perception that the organization possesses a fundamental flaw that deinviduates and discredits the organization” (Devers et al., 2009). This negative social evaluation by stigmatizing stakeholder groups tends to conflate individual stigmatized organizations with all other organizations carrying that stigma (Wolfe & Blithe, 2015). Research into organizational stigma has focused on differences between individual- and
organizational level stigma, strategies used to survive when facing organizational stigma (Hudson, 2008; Hudson & Okhuysen, 2009; Mishina & Devers, 2012; Wolfe & Blithe, 2015) and the influence of stigmatization on certain actors in an organization (Pozner, 2008; Wiesenfeld et al., 2008)
2.1.1 Core-Stigma
Recently a theoretical distinction has been made between two types of stigma, being event-stigma and core-event-stigma (Hudson, 2008). It is argued that an event-event-stigma is the result of an anomalous or episodic event such as bankruptcy (Neu & Weight, 1992; Sutton & Callahan, 1987), industrial accidents (Hoffman & Ocasio, 2001) and significant product defects (Hudson, 2008). Research has shown that event-stigma negatively affects both organizational outcomes and actors associated with the firm (Hudson, 2008; Sutton & Callahan, 1987; Wiesenfeld et al., 2008). Furthermore, research has shown that organizations respond to event-stigma by using public statements, excuses, justification, apologies, denials and attacks (Grougiou et al., 2016) and by attempting to regain lost social support (Hudson, 2008). In a recent example, the company Uber came under attack after letting their drivers pick people up from airports in New York City during a strike by other cab drivers. The strike was called after a controversial immigration ban was imposed by President Trump. In a reaction, the company opposed the immigration ban and even sent messages to customers deleting the app stating: “We wanted to
let you know that Uber shares your views on the immigration ban: it’s unjust, wrong and
against everything we stand for as a company.”1
On the other hand, Hudson (2008) defined core-stigma as “an evaluation held by some social audience that on organization or set of organizations is discounted, discredited, and/or
1 Furfaro, D. (2017). Uber paying the price for not supporting driver strike. New York Post. Viewed 09-03-2017 < http://nypost.com/2017/02/01/uber-paying-the-price-for-not-supporting-driver-strike/>
tainted in some way owing to some core attribute or attributes.” These organizations experience social condemnation as a result of the nature of their work, products or clientele (Wolfe & Blithe, 2015). Social audiences evaluating these organizations as inappropriate, devalued and tainted not only withdraw their social support, but view it with mistrust and hostility (Hudson, 2008). Examples of core-stigmatized organizations include men’s bathhouses (Hudson & Okhuysen, 2009), abortion clinics, arms dealers and tobacco companies (Grougiou et al., 2016). These so called “sin” industries revert to multiple different strategies to survive without broad social support and approval (Hudson & Okhuysen, 2009) and as a way to lessen the impact of core-stigmatization (Grougiou et al., 2016).
Amongst these strategies are so called specialist, hiding and challenging strategies (Hudson, 2008). The first type of strategies used by stigmatized organizations are specialist strategies. This means that stigmatized organizations typically operate in a single line of work as specialist, rather than in multiple areas as generalists. The reason for stigmatized organizations to adopt these strategies, is to minimize attention drawn to the organization (Hudson, 2008). When operating in more than one line of business, organizations face difficulties due to the increased attention to their operations. As a result, the expressions of stigma will probably intensify leading to increased negative consequences (Hudson, 2008). These difficulties arise when diversifying into another stigmatized domain as well as into non-stigmatized domains (Hudson, 2008). Hiding strategies are used in the face of greater stigma and include the use of discrete locations, signage and architecture and limited or target advertising all aimed at minimizing attention gained from (potentially) stigmatizing audiences (Hudson, 2008). Abortion services are an example of organizations that use these hiding strategies when social hostility is high and are often located in low-traffic areas. The final group of strategies entail the challenging strategies which are the opposite of the hiding strategies
(Hudson, 2008). Sometimes organizations choose to widely advertise their existence in order to challenge the stigmatization of their core attributes. These three type of strategies all aim at reducing the impact of core-stigma by either minimizing attention or by challenging the stigma altogether (Hudson, 2008).
Another group of processes adopted by core-stigmatized organizations was identified by Hudson and Okhuysen (2009) when researching men’s bathhouses. These boundary management processes included isolating, integrating, dramaturgy, associational and conventional processes that allow these bathhouses to survive in the face of core-stigmatization. The isolation processes are similar to the hiding strategies presented by Hudson (2008) and include the use of physical boundaries and the use of discrete locations and signage in order to hide from stigmatizing audiences, but also to protect the bathhouses’ relationships with for example customers and network partners (Hudson & Okhuysen, 2009). The integration processes were used to make outsiders into organizational insiders. Preferably, contact with outside suppliers in these bathhouses was limited in order to avoid unwanted attention. As a result, a lot of the tasks were performed in-house. If the use of outside help was unavoidable, trusted partners were selected and made into insiders. By integrating the supplier into the bathhouse, the boundary of the bathhouse is extended (Hudson & Okhuysen, 2009). Dramaturgy processes are the third classification of processes used by bathhouses (Hudson & Okhuysen, 2009) and these processes are the processes in which the bathhouses and its regulators “act out their roles with precision, matching one another’s needs.” (p.144). Regulators, such as building inspectors, face the possibility of stigma transfer when operating in the bathhouses. These dramaturgy processes provide both the bathhouse and the regulators the control over their mutual boundaries. An example includes the use of gym-like characteristics in the bathhouses, which gave the opportunity for regulators to assess the
bathhouse as a gym (Hudson & Okhuysen, 2009). The associational processes were processes that gave the partner the ability to define the association and the boundary with the bathhouse (Hudson & Okhuysen, 2009). By having control over the boundary and the ability to acknowledge or deny their inclusion with the bathhouse, the partners had the ability to minimize stigma transfer. Hudson and Okhuysen (2009) found that the most used associational process was the use of a discrete membership card, which had little or no information about the bathhouse. As a result, customers can deny any relation with the establishment. The final boundary management processes identified by Hudson and Okhuysen entail the conventional processes. These processes were used in accepting environments and are similar to processes used by non-stigmatized organizations (Hudson & Okhuysen, 2009). In these environments, bathhouses were able to deviate from the use of discrete locations and buildings and could act in a more open manner.
It is evident from the analysis above that most of the strategies used by core-stigmatized organizations include isolation, concealment and hiding strategies (Hudson, 2008; Hudson & Okhuysen, 2009). With the exception of challenging strategies and the use of conventional boundary processes in accepting environments, all the strategies used are aimed at minimizing unwanted attention. Research has shown however, that using these strategies alone is an insufficient way of dealing with core-stigma. Wolfe & Blithe (2015) showed that although brothels tend to lay low in the sagebrush (p. 548) as a way to protect customers and their right to exist as a legal business, many of them are increasingly becoming aware of the need and value of being visible and known in their local communities. Their research showed that core-stigmatized organizations work to promote recognizable organizational images with the use of the stigma-management strategies of passing and enactment. The passing strategies are aimed at distancing themselves from identities that they perceive as socially undesirable and drawing
analogies to stigmatized industries (Wolfe & Blithe, 2015). By drawing analogies to non-stigmatized industries brothels intend to normalize the deviant practices performed in their organizations. By comparing sex work with caregiving and therapy and drawing the analogy to industries of sales, an attempt is made to legitimize discourse and normalize the deviant practice (Wolfe & Blithe, 2015).
Furthermore, enactment strategies are used in order to challenge the core-stigma. In the brothels, this happened through constructing selectively permeable organizational boundaries to build familiarity with external community members (Wolfe & Blithe, 2015). By inviting outsiders into their brothels in controlled boundary crossings, brothel owners attempted to dispel myths about prostitution based on misinformation and misperception. Another enactment strategy adopted by brothel owners was that of promoting community-engagement activities as a way to become more visible in their communities. Strategic donations entailed an important part of these visibility efforts. These strategies combined are a way in which the brothels attempt to construct a more socially positive organizational image. Other ways in which core-stigmatized organizations try to construct a positive image have remained largely unexplored.
2.2 CORPORATE SOCIAL RESPONSIBILITY
The concept of Corporate Social Responsibility (CSR) has been evolving since the early 1930s (Carroll, 1979). For decades, the concept faced large ambiguity, as some addressed the legal responsibilities an organization has, others voluntary activities and again others the economic activities of an organization. It was Carrol (1979) who proposed that a definition of CSR needed to encompass all the obligations an organization has to society and as such needed to incorporate economic, legal, ethical and discretionary aspects of business performance. The economic responsibility of an organization is to produce goods and sell them at a profit or in other words, to maximize profits. The legal responsibilities of an organization are adhering to
the laws and regulations whilst conducting business. The third dimension of CSR are the ethical responsibilities of an organization, which are behaviours that are not written in law but are expected of organizations from society. Finally, the discretionary responsibilities are those concerning responsibilities that are left to the individual judgement and choice of organizations and thus consist of philanthropic activities. The definition of CSR following this reasoning is “The social responsibility of business encompasses the economic, legal, ethical and discretionary expectations that society has of organizations at a given point in time” (Carroll, 1979, p. 500). A pyramid is proposed in which the base of the pyramid are the economic responsibilities, followed by the legal, ethical and finally discretionary responsibilities (Carroll, 1991).
A more recent definition of Corporate Social Responsibility (CSR) was proposed by McWilliams & Siegel (2001) as “actions that appear to further some social good, beyond the interest of the firm and that which is required by the law” (p.117). This definition focusses on the ethical and discretionary aspects of Carrol’s pyramid (Carroll, 1979, 1991) and is often used by researchers to conceptualize CSR. Furthermore, a distinction has been made in the type of CSR that an organization demonstrates being philanthropy, the integration of corporate responsibility (CR) and CR innovation (Kourula & Halme, 2008). Philanthropy has an emphasis on charity, sponsorships and employee voluntarism. The principal idea here is “doing good” by for example donating money or encouraging employees to do volunteer work. In the case that an organization shows CR integration, the emphasis is on conducting existing business operations in a more responsible manner by for example modernizing existing processes in order to reduce pollution. The final form of CSR performed by organizations is CR innovation, in which organizations develop new business models to solve social and environmental problems. In sum, although different conceptualizations of CSR have been used it highlights the fact that organizations have responsibilities that go beyond profit maximization.
Interest in CSR has been big due to an increasing number of firms trying to incorporate the practice into their businesses (Cai, Jo, & Pan, 2012). As a result, the number of articles on CSR has nearly doubled since 2005 (Aguinis & Glavas, 2012). According to Porter and Kramer (2006) the heightened corporate attention to CSR has “not been entirely voluntary” (p. 80) and claim that it was the public who focussed organizations attention to issues that they previously had not even seen as part of their business responsibility. CSR was initially viewed as a joke and a contradiction in terms by the businesses community (Lee, 2008) and some viewed CSR as something that would only put an unfair and costly burden on shareholders (Friedman, 1970). In his classic critique on CSR, Milton Friedman (1970) described how the only social responsibility of organizations was to increase its profits. From the late 1990s onward however CSR became something that was promoted by all constituents in society (Lee, 2008). Activists and other stakeholders have since become more adept at bringing public pressure on corporations. These pressures indicate that stakeholders are trying to hold organizations accountable for their actions and highlight the risks that come with being labelled as irresponsible. Organizations seem to have become more aware about these risks, resulting in actual efforts at improving for example pollution reduction, but also in public relations and media campaigns with CSR reports as the centrepiece (Porter & Kramer, 2006).
It has been shown that CSR improves a firm’s reputation (Aguinis & Glavas, 2012), financial performance (Oh, Bae, & Kim, 2016; Porter & Kramer, 2006) and helps to obtain social legitimacy (Oh et al., 2016). Porter & Kramer (2006) identified four prevailing justifications for adopting CSR activities accordingly as being moral obligation, sustainability, license to operate and organizational reputation. The first – moral obligation – argues that organizations have to “do the right thing” and as such need to be good citizens. The sustainability justification means that organizations need to operate in a way that does not jeopardize the ability of future generations do meet their needs. Thirdly, the license to operate
is often put forward as a reason to operate in a responsible way, as organizations need tacit and explicit permission of stakeholders to conduct their business. Finally, reputation is commonly used as a justification by organizations for engaging in CSR initiatives at it would improve a company’s image and strengthen the brand. Especially the last two justifications seem relevant in the case of core-stigmatized organizations.
2.3 CORE-STIGMA AND CORPORATE SOCIAL RESPONSIBILITY
Most of the research into CSR has focused on organizations with a favourable reputation (Grougiou et al., 2016) leaving core-stigmatized organizations largely unexplored. Although some research has looked into the effect of CSR in “sin” or “controversial” industries on reputation (Cai et al., 2012; Vanhamme & Grobben, 2009; Yoon, Gürhan-Canli, & Schwarz, 2006), it is not clear whether having a stigma actually drives organizations to use more CSR than their non-contested counterparts. Core-stigma is the result of perceived severe value incongruence by a critical mass of stakeholder group members (Devers et al., 2009) that is of a permanent nature. Consequently, it is important for these organizations to minimize the effects of the negative evaluations and lessen the impact of the endured stigma.
By using the strategies and processes described in the section “core-stigma”, stigmatized organizations attempt to construct a more socially positive organizational image. It seems as though core-stigmatized organizations engage in impression management. Impression management theory is a theory originating from social psychology that is concerned with how individuals present themselves to others in order to be perceived favourably (Hooghiemstra, 2000). Impression management can take on multiple forms, either verbally or non-verbally. Although it originally studied individuals, the theory has extensively been applied to organizations. Organizational impression management (OIM) has been described as “any action that is intentionally designed and carried out to influence an audience’s perceptions of
the organization” (Elsbach, Sutton & Principe, 1998 in Bolino, Kacmar, Turnley, & Gilstrap, 2008, p. 1095). Impression management tactics can be divided into assertive tactics, which are proactive and are used to improve an organization’s image and defensive tactics, which are reactive and are used in the case of threatening situations that could damage the organization in some way (Bolino et al., 2008).
The tactics used in the face of an event-stigma fall into the reactive category, as these are for example the use of excuses and apologies. These tactics are aimed at minimizing the negative repercussions of the events (Bolino et al., 2008). Elsbach and Sutton (1992) showed how the adoption of defensive impression management tactics could be used to draw attention away from illegitimate actions and in effect focus attention on more favourable and accepted goals. As organizational stigma and thus core-stigma is in the eye of the beholder (Hudson, 2008), seeing it is an evaluation by social audiences external to the organization, it is possible that organizations try to proactively influence these perceptions. The disclosure of corporate social responsibility (CSR) reports are a way in which organizations can attempt to influence the perceptions of their audiences. It can act as a form of impression management that can contribute to an organization’s reputation (Hooghiemstra, 2000).
In their study, Kilian and Hennigs (2013) found that organizations in controversial industries are more active in CSR communications than organizations in non-controversial industries. They studied German organizations grouped into high-controversial, middle-controversial and low-middle-controversial groups. The high-middle-controversial group included organizations involved in chemicals, pharmaceuticals and energy, the middle-controversial group contained organizations in the automotive, transportation and tourism sectors and the low-controversial group were organizations involved in consumer goods, financial services and manufacturing. It was found that the controversial organizations, which were believed to have a negative influence on environmental issues, reported more on CSR than the non-controversial
groups (Kilian & Hennigs, 2013). These controversial industries used CSR communications as way to counter negative attitudes and to comply with higher expectations given to them by their stakeholders (Kilian & Hennigs, 2013).
It is argued that a similar effect will be found for core-stigmatized organizations, as engaging in CSR reporting is a way in which core-stigmatized organizations try to distract attention from their stigmatized activities and cushion the impact of negative evaluations (Elsbach & Sutton, 1992; Grougiou et al., 2016). These proactive impression management tactics also help in drawing analogies to non-stigmatized organizations, one of the strategies identified earlier on (Wolfe & Blithe, 2015). By engaging in CSR, stigmatized organizations try to keep disapproval at a minimum in order to maintain their social license (Grougiou et al., 2016; Parsons et al., 2014; Porter & Kramer, 2006). Organizations’ social license to operate is becoming increasingly important as society is putting more pressure on companies to act as good corporate citizens (Banerjee, 2012). In the case of core-stigmatized organizations, part of society condemns the practices of the organization so keeping this license to operate from other stakeholders is of great importance. By distracting attention from their stigmatized activities and highlighting their efforts at being a good corporate citizen, stigmatized organizations are using CSR as an impression management tactic in order to keep their license to operate.
Taken together, by identifying that hiding strategies are insufficient in dealing with a core-stigma and by identifying that core-stigmatized organizations use proactive impression management strategies in order to improve their reputation, it is hypothesized as follows;
Hypothesis 1. Core-stigmatized organizations are more likely to engage in CSR reporting than non-stigmatized organizations.
2.4 ORGANIZATIONAL VISIBILITY
Visibility has been defined by Bowen (2000) as “the extent to which phenomena can be seen or noticed” (p.93). A distinction has been made between organizational visibility and issue visibility (Bowen, 2000; Brammer & Millington, 2006), where organizational visibility refers to the extent to which organizations are seen by important constituents and issue visibility to the extent to which issues are noticeable to groups inside and outside the organization. A further distinction can be made between generic and domain-specific visibility (Yu et al., 2016). Organizations with high generic visibility possess characteristics that make them widely known in society, like high reputation, status and prominence, whereas organizations with high domain-specific visibility possess some features that expose the firm to greater scrutiny regarding this particular domain, like for example worker safety (Yu et al., 2016). In this research, generic organizational visibility is studied.
Society is increasingly putting pressure on organizations to be more legitimate and as argued before, organizations respond by engaging in more corporate social performance (CSP) in order to be perceived as better corporate citizens (Chiu & Sharfman, 2011). An important area research has focused on is identifying the underlying reasons for corporate sensitivity to these growing pressures (Brammer & Millington, 2006). Visibility has been one of the characteristics of organizations that has received a lot of attention with regards to motives for engaging in CSR (Aguinis & Glavas, 2012; Bowen, 2000; Brammer & Millington, 2006; Chiu & Sharfman, 2011; Yu et al., 2016).
There are two views with regards to the effects of organizational visibility on CSR. The first states that as an organization becomes more visible, it becomes more likely to attract more attention from its stakeholders (Yu et al., 2016). Stakeholders are a primary source of legitimacy, but take it away just as easily as providing it (Chiu & Sharfman, 2011). As a result, highly visible organizations are more likely to be vulnerable to public scrutiny and more
exposed to stakeholder pressures to be socially and environmentally responsible (Bowen, 2000; Brammer & Millington, 2006; Chiu & Sharfman, 2011; Yu et al., 2016). This view, that greater visibility leads to more environmental responsiveness, is well-accepted (Henriques & Sadorsky, 1999; Sharma & Nguan, 1999; González-Benito & González-Benito, 2006 in Yu et al., 2016). Chiu and Sharman (2011) for example found that a firm’s visibility to stakeholders is a key driver behind managerial decisions regarding corporate social performance. They proposed based on their results that “corporate social performance can be conceptualized as (at least in part) an instrumental response by managers to legitimacy pressure on the firm based on its visibility” (Chiu & Sharfman, 2011, p. 1577).
There are other studies though, that are more supportive of the second view that states that organizational visibility tends to make organizations more resistant to outside influences and as a result less environmentally responsive (Lefebvre et al., 2003; Darnall et al., 2010; Udayasankar, 2008 in Yu et al., 2016). Darnall, Henriques, & Sadorsky (2010) found that less visible firms tend to be more responsive to stakeholder pressures than their highly visible counterparts. It is argued here that these mixed results could be the result of looking at organizational visibility in a wrong way. It might be the case that organizational visibility does not have a direct effect, but rather acts as a moderator in the proposed core-stigma and CSR reporting relationship.
As was mentioned before, core-stigmatized organizations are constantly trying to minimize unwanted attention (Hudson, 2008; Hudson & Okhuysen, 2009). It could be the case that for a core-stigmatized organization that is very limitedly visible, issuing a CSR report could bring more attention to their firm than there was in the first place. As such, it is argued that these types of firms will be less likely to revert to using CSR as a defensive mechanism in the face of organizational stigma. Furthermore, it seem likely that highly visible stigmatized organizations have the biggest motivation to counter negative attitudes about their practices
when compared to organizations that are less visible and stigmatized (Hudson, 2008; Hudson & Okhuysen, 2009). In the case of a highly visible stigmatized firm, the chance of having your social license to operate revoked becomes a big possibility. The chance that this license to operate will get revoked by society leads to organizations behaving in a way that is perceived as “good” by society (Banerjee, 2012). As a result, it is argued that these organizations will increase their CSR reporting in order to protect that license and minimize scrutiny from the public. Durand and Vergne (2014) showed that stigmatized firms under attack in the media divest assets from the stigmatized industry. This response portrays a way in which stigmatized industries that are visible try to counter this negative publicity. It is argued here, that reverting to CSR is another way in which highly visible core-stigmatized organizations try to improve their reputation.
Hypothesis 2. Organizational visibility moderates the relationship between core-stigma and CSR, in such a way that higher visibility will strengthen the relationship between core-stigma and CSR.
Figure 1. Conceptual Model.
Core-stigma
Organizational visibility
3. METHOD
3.1 SAMPLE AND DATA COLLECTION
In order to test the hypotheses, data of stigmatized firms’ CSR engagement was compiled by using secondary quantitative data. This data was obtained from multiple databases like the Morgan Stanley Capital International (MSCI), formerly known as the Kinder Lydenburg Domini (KLD) and the Global Mapping International (GMI). This database lists the annual CSR engagement of US publicly listed firms over multiple dimensions of CSR (Oh et al., 2016). The KLD ratings have widely been used by studies in order to select samples and measure CSR (e.g. Backhous, Stone & & Heiner, 2002; Hillman & Keim, 2001; Hull & Rothenberg, 2008; Mattingly & Berman, 2006; Schnietz & Epstein; 2005; Scholtens & Zhou, 2008 in Parastoo Saeidi, Sofian, Saeidi, Parisa Saeidi, & Alireza Saaeidi, 2014). Although using only US publicly listed firms provides a way of controlling for the effect of culture differences on CSR reporting (Grougiou et al., 2016), it also brings the limitation of not being able to generalize the results to firms outside the US.
The sample was determined by using the exclusionary screens provided by the KLD in order to discriminate between the core-stigma category and the non-stigma category (Forehand & Grier, 2003; Grougiou et al., 2016; Oh et al., 2016). It labels those companies as stigmatized that are related to the manufacturing, licensing, trading, funding and ownership of alcohol, firearms, gambling, military, nuclear power and tobacco activities (Cai et al., 2012; Forehand & Grier, 2003; Grougiou et al., 2016; Oh et al., 2016), also known as “sin” industries. An initial sample of core-stigmatized firms and control firms was determined for the year 2012. After this, matches were made between the core-stigma group and a control group through case-control matching based on size and industry. The final sample consisted of 123 core-stigmatized firms and 123 control firms whose full data could be found in the KLD, Compustat and
LexisNexis database. The year 2012 was chosen as year to base the sample due to data availability.
The KLD scores for core-stigma and the control-match represent scores in 2012. The scores for visibility represent average scores for the years 2010 through to 2012. The scores on CSR reports represent scores obtained in 2013. The independent variables were lagged one year, to allow the effect of experienced stigma and visibility to take place on subsequent CSR reporting (Chiu & Sharfman, 2011).
3.2 MEASURES
3.2.1. Dependent Variable: Corporate Social Responsibility Reporting
The dependent variable, CSR reporting (CSR), was measured by looking whether or not an organization issued a standalone CSR report (Grougiou et al., 2016). This information was hand collected through multiple sources including the websites CorporateRegister, SocialFunds and company websites. A dummy variable was computed, in the case that an organization released a CSR report a 1 was given and 0 if otherwise.
3.2.2 Independent Variable: Core-stigma
Core-stigma was operationalized using the exclusionary screens incorporated in the KLD database. The KLD provides an “involvement” criteria using a binary rating (0/1) and assigns a 1 when the company is involved in alcohol, firearms, gambling, military, nuclear power and tobacco activities and a 0 when this is not the case (Oh et al., 2016). These industries are core-stigmatized because they are not only known to potentially threaten the health of people, society and environment, but also are known to be against social norms (Oh et al., 2016). A dummy variable was made for core-stigmatized organizations (STIGMA), where a 1 was given in the case of involvement in one of these criteria and a 0 if this was not the case. This measurement has previously been used to discriminate between stigma and non-stigma firms (Grougiou et al., 2016; Oh et al., 2016).
3.2.3 Moderating Variable: Organizational Visibility
The moderator organizational visibility (ORG_VIS) was operationalized by using secondary data from the LexisNexis database, a database containing news from multiple sources over the world (Brammer & Millington, 2006; Chiu & Sharfman, 2011). As the sample included US publicly listed firms, it was determined to use the amount of annual new hits in US newspapers and newswires. An average number of hits was measured for the years 2010, 2011 and 2012 in order to determine organizational visibility. This average was used as a way to reduce sensitivity to fluctuations in media visibility (Brammer & Millington, 2006) in the analysis. A large variability was found in this variable, with the smallest number being an average of 0.667 hits (Applied Industrial Technologies Incorporated) and the largest 4062,667 (Microsoft Corporation).
3.2.4 Control Variables
Following from previous literature there were some control variables that needed to be taken into account being: size, financial performance, CSR performance and industry type (Aguinis & Glavas, 2012; Grougiou et al., 2016; Oh et al., 2016). The first control variable, size, is controlled for as it has been said to influence how organizations engage in and report on CSR (Grougiou et al., 2016; McWilliams & Siegel, 2001; Oh et al., 2016). Bigger organizations tend to disclose more CSR reports than smaller organizations, because they generally have a bigger impact on society (Grougiou et al., 2016). Size (SIZE) was measured as the natural logarithm of total assets (Bansal & Clelland, 2004; Chiu & Sharfman, 2011; Khan, Muttakin, & Siddiqui, 2013).
Another variable that needed to be controlled for was the financial performance of firms. Organizations with better financial performance tend to have more resources with which they can engage in CSR activities and subsequent reporting (Grougiou et al., 2016). Financial performance was measured as the ratio of an organizations’ net income to total assets (RoA).
Furthermore, CSR performance (CSP) is controlled for as it has been shown that a higher performance leads to more disclosures (Clarkson, Li, Richardson, & Vasvari, 2008). CSR performance is obtained from the KLD. The KLD rates organizations on strengths and concerns in the following list of characteristics: corporate governance, community, diversity, employee relations, environment, human rights and products (Statman & Glushkov, 2009). It uses a binary rating (0/1) and assigns a score of 1 when a company demonstrates strength on an indicator and a 0 if it does not (Grougiou et al., 2016; Statman & Glushkov, 2009). The same applies for the concerns; a 1 is assigned when a company demonstrates a concern and a 0 when this is not the case. Due to problems with combining strengths and concerns in research, it was chosen to solemnly look at the strength indicators (Mattingly & Berman, 2006). An aggregate strength construct was composed by adding the scores on the different strength ratings.
Finally, the effect of industry type needs to be controlled (Grougiou et al., 2016). Industry was measured as a two-digit SIC code (Brammer & Millington, 2006; Grougiou et al., 2016; Hong & Kacperczyk, 2009; Oh et al., 2016). Dummy variables were created for industry (IND) that equal 1 if an organization belongs to the industry and 0 otherwise. The reference group was left out of the analysis in order not to fall in the dummy trap. Apart from the score on CSP, which was obtained from the KLD, all control variables were obtained from the Compustat Database.
3.3 METHOD OF ANALYSIS
IBM’s SPSS is used for the statistical analysis. A test of frequencies was performed in order to test for errors, but none were found. Furthermore, it was checked if any data was missing, which also was not the case. The following model was specified in order to test the first hypothesis;
!"#$%&= )*+ )&",-./0 + )1"-23 + )4#50 + )6!"7 + ∑)9-:; + <9
Where, CSR = CSR reporting, equals 1 if a report was issued and 0 if not, STIGMA = core-stigma, based on KLD equals 1 in case of stigma and 0 if not, SIZE = natural logarithm of total assets, RoA = return on assets, measured as the ratio of net income to total assets, CSP = CSR performance, based on the aggregate score on strengths extracted from the KLD and IND = dummy variables for industries based on two digit SIC codes. An overview of organizational representation per industry was made (see Appendix). In order to test the first hypothesis, a binary logistic regression was used, as the dependent variable is categorical. It was chosen to use the enter method in the analysis, as stepwise analysis should be used for exploratory purposes only (Burns & Burns, 2008). Before performing the logistic regression, a cross-tabulation was performed. A cross-cross-tabulation depicts the frequencies of combinations between the two variables, in this case core-stigma and CSR reporting (Field, 2009). This gives an initial picture about whether or not core-stigma leads to more CSR reporting.
The second hypothesis tried to establish whether organizational visibility acts as a moderator in the relationship between stigma and CSR reporting. This is known as a simple moderation, which was analysed by using the Process macro developed by Andrew F. Hayes (2012) for SPSS. The model used in Process was number 1; for a simple moderation.
4. RESULTS
4.1 DESCRIPTIVE STATISTICS
Table 1 provides the summary statistics for both the core-stigmatized firms and the control firms. The mean (median) for the CSR variable for the control firms is 0.370 (.000) and 0.360 (.000) for the core-stigmatized firms, suggesting that there is no difference in CSR reporting between the two groups (.895). Furthermore, there do not seem to be differences in size between the core-stigmatized firms, 15.281 (15.059), and the control firms, 15.275 (15.177), indicative of the fact that the case-control matching was performed properly (.976). Organizational visibility shows a small, though non-significant difference between the two groups with core-stigma firms having an average of 179.281 (83.667) annual media hits, compared to 185.252 (68.333) hits for the control firms (.909). Interesting to note is the fact that the range in visibility seems to be much larger for the control firms in comparison with the core-stigmatized firms. The minimum for the control group is an average of annual news hits in US newspapers and wires of 0.667 and maximum of 4062.667. The core-stigmatized group on the other hand, has an average minimum of 13.667 and maximum of 1776.000. Financial performance and CSP do not seem to differ between both of the groups; with RoA being 5.421 (5.049) for the stigmatized firms and 5.594 (5.295) for the control group (.810) and CSP being 3.020 (1.000) for the core-stigma group and 2.500 (1.000) for the control group (.239).
Table 2 presents the Pearson correlations and p values from the two-tailed test. Stigma is not significantly correlated to CSR reporting (r = -0.008, p > 0.05). This could be indicative of the fact that there is no relationship between core-stigma and CSR reporting, but analysis through logistic regression will investigate this further. Organizational visibility on the other hand seems to be significantly related to CSR reporting (r = .280, p < 0.01), likely indicating that as an organization becomes more visible it will engage in more CSR reporting. What is interesting to note is the fact that some of the control variables and the moderating variable are
significantly correlated. For example, organizational visibility and size are significantly correlated (r = .481, p < 0.01), likely indicating that the bigger a firm is the more visible it will be. Furthermore, organizational visibility and CSP are significantly correlated (r = .561, p < 0.01), likely indicating that more visible firms tend to have a better performance regarding corporate responsibility activities. Furthermore, the control variables CSP and size are significantly correlated, (r = .674, p <0.01), probably indicating that with bigger size comes better CSP.
All the variables were checked for multicollinearity, which can exist when high levels of correlations exist (Field, 2009). Although some of the predictors are significantly related, none of the correlations are above 0.7, which is a first indication that multicollinearity probably is not an issue in this model. According to Field (2009) this is “a good ball park method” (p.224), but more subtle forms of multicollinearity can be detected by looking at the variance inflation factor (VIF) and the related tolerance statistic. Acceptable levels for VIF and tolerance levels are respectively lower than 10 and higher than 0.2. VIF and tolerance levels for all the predictors were calculated indicating no multicollinearity issues, as all these VIF levels ranged between 1.011 and 1.930 and tolerance levels between 0.518 and .989.
Table 3 through to table 6 show the results of the regression analysis and moderation analysis. Tables 4 and 5 show the results of the logistic regression testing the effect of core-stigma on CSR reporting. Table 6 shows the results of the moderation analysis testing the effect of organizational visibility on the relationship between core-stigma and CSR reporting.
Table 1. Summary statistics per sample group.
Variables Core-Stigma Control Mean difference
Variable Mean Median Std.Dev Min Max Mean Median Std.Dev Min Max p-value
CSR .360 .000 .481 0 1 .370 .000 .484 0 1 .895
ORG_VIS 179.281 83.667 286.901 13.667 1776.000 185.252 68.333 502.742 .667 4062.667 .909
Size 15.281 15.059 1.589 12.069 19.626 15.275 15.177 1.438 12.128 19.266 .976
RoA 5.421 5.050 6.070 -12.167 40.153 5.594 5.295 6.605 -18.41 26.38 .831
CSP 3.020 1.000 3.688 0 16 2.500 1.000 3.207 0 13 .239
Table 2. Correlations between variables.
Variable 1 2 3 4 5 6 1. Stigma 1 2. CSR -.008 1 3. Org_VIS -.007 .280* 1 4. Size .002 .552* .481* 1 5. RoA -.014 -.041 .081 -.030 1 6. CSP .075 .561* .510* .674* .114 1
Hypothesis 1 predicted that core-stigmatized organizations are more likely to disclose a CSR report. First, the cross-tabulation was performed to see the frequencies in variable combinations. The result can be found in table 3. The results show that out of the core-stigma group 44 of the 123 organizations disclosed a CSR report. In the control sample, 45 out of the 123 organizations disclosed a CSR report. This seems indicative of there being no difference between the two groups.
Table 3. Cross-tabulation of stigma and CSR reporting.
CSR Report No CSR Report Total
Core-Stigma 44 79 123
Control 45 78 123
Total 89 157 246
In order to test the significance of the relationship, a binary logistic regression was performed of which the results can be found in table 4. A test of the full model against a constant model gave a statistically significant result, which indicates that together the predictors correctly distinguished between cases that would disclose a CSR report against that who would not (Wald χ2 = 134.352, p < 0.01). Nagelkerke’s R2 of .577 indicates a moderately strong relationship between what was predicted and what was grouped in the model. The prediction success for no report was 90.4% and for disclosing a CSR report 69.7%, averaging 82.9%. Furthermore, the model adequately fits the data (Hosmer-Lemeshow = 12.921, p = .115).
The Wald criterion demonstrated that stigma was not a significant predictor of CSR reporting (Wald χ2 (1) = .551, p = .448), indicating that core-stigmatized organizations are not more likely to disclose a CSR report compared to the matched control group. Hypothesis 1 thus does not seem to be supported. Financial performance also did not seem to be a significant predictor of CSR reporting (Wald χ2 (1) = .034, p = .665), indicating that organizations that
have a better financial performance in this sample do not disclose more CSR reports than organizations with a lower financial performance.
Size on the other hand was a significant predictor of CSR reporting (Wald χ2 (1) = 6.452, p < 0.05), indicating that bigger organizations tend to disclose significantly more CSR reports than smaller organizations. Finally, CSP also was significantly related to voluntary CSR disclosing (Wald χ2(1) = 10.741, p < 0.01) and seems to be the strongest predictor. This indicates that firms that have a better score regarding their CSP are more likely to disclose a CSR report. It should be noted that the Exp(B) is relatively low for both size and CSP, with size having an Exp(B) of 1.122, indicating that with an increase of 1 unit for size the odds ratio is 1.122 times as large and therefore organizations are 1.112 times more likely to disclose a CSR report. In the case of CSP an increase of 1 unit leads to an increase of 1.106 of the odds ratio, indicating organizations are 1.106 times more likely to disclose a CSR report.
Lastly, the casewise list produced a list of 10 cases that did not fit the model well. This could be indicative of the fact that additional variables need to be added to the model (Burns & Burns, 2008). This will be further addressed in the limitations section. Furthermore, in the case that excessive outliers are present, these should be removed. Burns & Burns (2008) recommend to consider to remove cases for which the standardized residuals (ZResid) exceed 2.58, as this can affect the results significantly. In this dataset six cases had a ZResid of more than 2.58. In order to see if this influenced the results, these cases and their matched control cases were removed and an additional logistic regression was performed. The results of this additional regression can be found in table 5. The overall model is significant (Wald χ2 = 151.814, p < 0.01) and fits the model well (Homer-Lemeshow = 7.188, p = .516). Nagelkerke’s R2 of .656 indicates a moderately strong relationship between what the model predicted and what was grouped in the model. Furthermore, the prediction success for not disclosing a CSR report improved to 92.1% and for disclosing a CSR report to 74.7%, averaging 85.9%.
The Wald criterion in the additional regression demonstrates that stigma again was not a significant predictor of CSR reporting (Wald χ2 (1) = 1.705, p = .192), although it seems to have improved in comparison to the initial analysis. Again, no support for hypothesis 1 was found. Financial performance was again not a significant predictor of CSR reporting (Wald χ2 (1) = .149, p = .700). In the additional analysis, size again significantly predicted CSR reporting (Wald χ2(1) = 7.114, p < 0.01), as did CSP (Wald χ2 (1) = 13.826, p = <0.01). The Exp(B) for size in this model is 1.832 and for CSP 1.391; which are minor improvements in comparison with the first model.
Table 4. Results of the Logistic Regression Analysis.
Variablesa B SE Wald χ2 Exp(B) 95% Conf.
Interval (Constant) -9.158 3.058 8.966 .000 Stigma -.275 .371 .551 .759 .367 1.570 Size .504 .198 6.452* 1.655 1.122 2.440 RoA .015 .034 .188 1.015 .950 1.085 CSP .251 .077 10.741** 1.286 1.106 1.494
n = 246; Wald χ2 = 134.352, p < 0.0; Hosmer Lemeshow χ2 = 12.921, p = .115 aIndustry dummies are included in the analysis.
* p < 0.05, **p < 0.01
Table 5. Results of the Additional Logistic Regression Analysis.
Variablesa B SE Wald χ2 Exp(B) 95% Conf.
Interval (Constant) -10.721 3.498 9.393 .000 Stigma -.549 .420 1.705 .578 .253 1.317 Size .605 .227 7.114** 1.832 1.174 2.859 RoA .016 .040 .149 1.016 .938 1.099 CSP .330 .089 13.826** 1.391 1.169 1.655
n = 234; Wald χ2 = 151.184, p < 0.0; Hosmer Lemeshow χ2 = 7.188, p = .516 aIndustry dummies are included in the analysis.
Although no direct effect was found between stigma and CSR reporting, the second hypothesis was still tested. In order to test hypothesis 2, which predicted that the relationship between stigma and CSR reporting will be moderated by organizational visibility, a moderation analysis was performed. The results regarding the moderation analysis can be found in table 6. The model summary indicates a moderately strong relationship between what was predicted and what was found, with Nagelkerke’s R2 being .548. It can be identified if a moderation is taking place by looking at the interaction effect provided by the analysis. It was however observed that there was no interaction effect p = .732. Therefore, there seems to be no support for the second hypothesis. Furthermore, no direct effect of organizational visibility was found (p = .657). Finally, the moderation analysis provided further evidence for the effect of size and CSP on CSR reporting.
Table 6. Results of the Moderation Analysis.
Variablesa Coeff SE Z p LLCI ULCI
(Constant) -.625 .266 -2.350 .019 -1.147 -.104 Stigma -.447 .378 -1.184 .236 -1.187 .293 Visibility -.176 .396 -.444 .657 -.953 .600 Interaction -.176 .515 -.342 .732 -1.185 .833 Size 1.236 .287 4.301 .000 .673 1.799 RoA -.016 .207 -.076 .939 -.422 .390 CSP 1.160 .271 4.274 .000 .628 1.692 n = 234; Nagelkerke’s R2 = .656
5. DISCUSSION
The purpose of this study was to add to the nascent literature about core-stigma and its effects on CSR reporting and the moderating effect of visibility. Through testing multiple hypotheses this research sought to find an answer to the two research questions; What is the influence of core-stigma on CSR reporting? and What is the influence of organizational visibility on the relationship between core-stigma and CSR reporting?
The first hypothesis, which stated that core-stigma will lead to more CSR reporting, was not supported in the current research. These results are not in line with former initial research into the subject (Grougiou et al., 2016; Kilian & Hennigs, 2013). There could be multiple explanations for this; firstly, it could mean that the proposed relationship does not exist; secondly, it could mean that the proposed relationship does exist, but was not found in the current study due to the research design. The first possibility will be addressed in the academic relevance section. The second possibility will be addressed in the limitations section, together with the limitations of the current study. The second hypothesis, which stated that this relationship is moderated by organizational visibility was also not supported. As is the case with the first hypothesis this could have several reasons, while the most important one is the present study is the lack of the direct relationship between core-stigma and CSR reporting. As such, a moderation effect was not likely to be found.
5.1 ACADEMIC RELEVANCE
This research contributes to the literature on organizational stigma, core-stigma, organizational visibility and CSR in multiple ways. Seeing the literature on organizational stigma is nascent and in full development, any contribution is a potential valuable one. As the results of the current study do not seem in line with earlier research, it is possible that subtle nuances are present in the concept of core-stigma and related outcomes that have not yet been discovered.
Although previous research has found results indicative of the relationship existing and has provided arguments used for the current study (Grougiou et al., 2016; Oh et al., 2016), it could be possible that research not finding this result simply has not been published. This tendency to publish significant findings has extensively been researched, as this so called “publication bias” seems to effect experimentation in multiple fields of work (Stanley, 2005). Especially in an area of research that is so new, finding no significant result is just as important as finding significant results. As such, the current study contributes to the existing literature on core-stigma by highlighting that organizations that face core-stigma might not revert to CSR as a way to cushion the negative effects of having a stigma. An explanation that no relationship might exist between core-stigma and CSR reporting, could be that organizations realize that CSR reports can potentially be viewed as hypocritical by the stakeholders giving them the stigma-label (Yoon et al., 2006). A well-known example of CSR campaigns backfiring is that of Philip Morris, a huge tobacco company that decided to support a campaign trying to prevent the youth from taking up smoking (Yoon et al., 2006). In this case, not only critics ridiculed the campaign, but also consumers. It could thus be that core-stigmatized organizations are afraid of attracting extra unwanted attention by publishing a CSR report, instead of seeing it as a way to distract attention away from their core-stigmatized activities.
Furthermore, the current study contributes to the literature on CSR reporting by identifying that although it was generally assumed that CSR was used as a strategic tool in order to improve an organization’s reputation, this might not hold for stigmatized organizations. It could be the case that more relationships found between CSR and possible outcomes and predictors, don’t apply to this specific group of organizations. As such, this thesis highlights the importance of studying well-known effects for this particular sample, as the results obtained for stigmatized firms are likely to be very different than for their non-stigmatized counterparts.
Lastly, this thesis extends the research on organizational visibility as it finds that visibility does not have a moderating effect on the proposed relationship, nor a direct effect on CSR reporting. The debate about organizational visibility has mostly been about whether organizational visibility increases or decreases environmental responsiveness (Bowen, 2000; Brammer & Millington, 2006; Chiu & Sharfman, 2011; Yu et al., 2016). In the current study, no direct effect of visibility was found on CSR reporting at all. It was argued here that visibility would act as a moderator in the proposed relationship, but due to not finding a direct relationship this was not found. As such, the current study contributes to the literature on visibility by highlighting the possibility that visibility does not directly affect environmental responsiveness and other ways of influencing possible outcomes and relationships should be investigated.
In sum, this thesis has multiple academic contributions as it opposes previous research into the relationship between core-stigma and CSR reporting, portrays that results regarding CSR might not hold for stigmatized organizations and furthers the research on organizational visibility by questioning if there is a direct relationship at all.
5.2 MANAGERIAL IMPLICATIONS
Apart from academic contributions, the current study also has managerial implications. CSR has quite often been viewed as something that gets practiced by organizations as a strategic tool in order to gain support from important stakeholders, with CSR reports being the centrepiece of these public relation campaigns (Grougiou et al., 2016; Porter & Kramer, 2006). Although it can be viewed as an important part of a larger impression management strategy, the current study highlights that previous mentioned claims about the use of CSR reports might not always hold. In the past, organizations and business participants were warned to be cautious when factoring CSR engagement into their financial decision making, as CSR reports might not reflect the genuine thoughts and attitudes in an organization (Grougiou et al., 2016). As no
relationship was found between core-stigma and CSR reporting, this could mean that these types of organizations do not revert to CSR reporting more or less than their non-contested counterparts. As such, when dealing with core-stigmatized organizations no extra caution is necessary in comparison to doing business with an organization that does not have this label. 5.3 LIMITATIONS AND SUGGESTIONS FOR FUTURE RESEARCH
As mentioned before, it could also be the case that no significant effect was found in this study, but that this relationship does actually exist. This could be the result of problems in the research design. A possible explanation for not finding a significant result could be the operationalization of core-stigma. As a result of using the “involvement” criteria provided by the KLD database, core-stigma might not have been properly operationalized. Involvement in alcohol, firearms, gambling, military, nuclear power and tobacco activities does not necessarily mean that the organization is only active in this particular industry and could even be active in an industry that does not carry a stigma at all. Furthermore, following from previous literature and the theory development in the current study, severely core-stigmatized organizations tend to be specialists operating in a single line of work rather than in multiple sectors (Hudson, 2008). The organizations included in the analysis were mostly organizations that were active in multiple industries, as the inclusion criteria was “involvement”. It could thus be the case that the organizations included in the current sample are not stigmatized severely enough to drive them to adopt CSR reporting as a defensive mechanism. Future research should therefore look at other ways to operationalize core-stigma, in order to see if a relationship is present. By studying core-stigmatized organizations that operate in a single line of work, for example men’s bathhouses as studied in Hudson’s research (2008), or arms dealers as studied in Durand and Vergne’s research (2014), a more severely stigmatized group is studied. This could lead to finding the proposed relationship between core-stigma and CSR reporting.
