The effect of CEO targeting and CEO characteristics on the success of social issues through social media

(1)

M.Sc. Business Administration – International Management

Master Thesis

“THE EFFECT OF CEO TARGETING AND CEO

CHARACTERISTICS ON THE SUCCESS OF SOCIAL ISSUES

THROUGH SOCIAL MEDIA”

Submitted by:

Selina Mia Schwetz

Student number: 11627174 Leopoldstraße 61e 76344 Leopoldshafen Germany First Supervisor: Dr. Michelle Westermann-Behaylo Second Supervisor: Dr. Markus Paukku

26 January 2018

(2)

I

Statement of Originality

This document is written by Selina Mia Schwetz who declares to take full responsibility for the contents of this document.

I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it. The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

(3)

II

List of Abbreviations

CEO chief executive officer

CSR corporate social responsibility

CSP corporate social performance

et al. et alii/ et aliae/ et alia (= and others)

NAICS North American Industry Classification Scheme

p. page

(5)

IV

Abstract

Social issues pose significant challenges at major corporations while threatening their reputation. Due to the rise of social media, these reputational threats are aggravated and social media itself can be argued to constitute a new way on which these social issues arise. The CEO, being the symbol of the company, thereby often represents the direct target of the social issue. This study investigates the effect of targeting the CEO on the subsequent success of the issue, displayed in a substantive firm response. It is hypothesized that social issues that target the respective CEO will be more likely to result in a success compared to issues not targeting the CEO. Furthermore, the effect of tenure of a targeted CEO is expected to influence the subsequent success of the issue in a way, such that targeting longer tenured CEOs will decrease the likelihood of success compared to a newer CEO. As a second characteristic, the nationality of the CEO is hypothesized to influence the success by moderating the above-mentioned relationship. Thereby, targeting CEOs from collectivist countries is expected to increase the likelihood of a substantive firm response.

The analysis of petitions directed at the Fortune Global 500 companies reveals that targeting the CEO increases the likelihood of success in isolation. However, this effect diminishes by taking control variables into account such that the number of petition supporters exerts a significant effect on the success of the issue. This effect also holds true while investigating the impact of the specific CEO characteristics. In addition, no support was found for the effect of CEO nationality. However, tenure does have a significant effect on the success such that targeting longer tenured CEOs increases the likelihood of success, opposing the hypothesized direction. This study enriches existing literature by analyzing data from 2009 up to the year 2017 while taking into account social media as a new medium on which these issues arise.

(6)

1

1. Introduction

In April 2017, United Airlines suffered severe consequences after a business scandal that went viral through social media within minutes. After having violently dragged off a

passenger of an overbooked plane at Chicago airport, videos of the officers mistreating the respective passenger were available on social media platforms shortly after the incident (Victor & Stevens, 2017). Thereby, the outrage of society was fueled, and the issue took on a dramatic scale which led to the classification of the issue as a business scandal (Tybout & Roem, 2009). Afterwards, the company was accused of having responded inadequately to the issue, and the reaction and weak excuses of United Airlines CEO Oscar Munoz were heavily criticized (Rushe & Smith, 2017). CEO Munoz, symbol of the company, constituted the main target of further accusations and related issues, and took on full responsibility for the

incident. In addition to financial consequences, the scandal led to promises to improve customer satisfaction and policies in general. Furthermore, personal consequences for the CEO resulted from the incident, among others, that Munoz will no longer become chairman of the board (Carey, 2017; Reed, 2017; Rushe & Smith, 2017; Zhang, 2017).

This example substantiates the increasing importance and popularity of social media in recent years and hence demonstrates its undeniable influence in the business world (Aula, 2010; Drobocky, 2013; Kaplan & Haenlein, 2010). According to Kaplan and Haenlein (2010), social media can be defined as “a group of Internet-based applications that build on the ideological and technological foundations of Web 2.0, and that allow the creation and exchange of User Generated Content” (p. 61). Social media nowadays consists not only of social networking platforms like Facebook but also of many webpages that allow users to post self-generated content regarding social issues that are of public interest as it was the case in the United Airlines scandal. As argued within this work, social media constitutes a new way through which social issues are put forward compared to traditional ways. These

(7)

2 traditional ways consist of tactics that bring up issues outside of the company like boycotts or within the firm like shareholder resolutions (David, Bloom & Hillman, 2007; McDonnell, King & Soule, 2015).

Social issues can be seen as a major concern to the whole society (Clarkson, 1995;

Friedman, 1970). Many scholars have investigated the responsiveness of a firm to such social issues which determines the degree of success of the forwarded issue (Bundy, Shropshire & Buchholtz, 2013; Eesley & Lennox, 2006; McDonnell et al., 2015; Mitchell, Agle & Wood, 1997; Waldron, Navis & Fisher, 2013). However, no prior scholarly work has investigated the relationship between scandals that arise out of social media as a new phenomenon and the likelihood of these issues to be successful. This is especially important as social media is proposed to increase reputational risk (Aula, 2010) which in turn determines the firm’s responsiveness (Waldron et al., 2013).

As the United Airlines case illustrates, the CEO plays a vital role within the firm’s

response to social issues by being responsible for securing the image and designing the firm’s strategy (Waldmann, Siegel & Javidan, 2006). Based on the upper echelons theory

(Hambrick & Mason, 1984), strategic decisions of CEOs are determined by their individual characteristics and consequently these characteristics will influence the decisions regarding the outcome of social issues. Prior studies have investigated the effects of specific

characteristics on outcomes of social issues (Lewis, Walls & Dowell, 2014; Manner, 2010; Rehbein, Logsdon & Van Buren, 2013; Slater & Dixon-Fowler, 2009). However, no study has analyzed the effects of CEO characteristics on the success of social issues that arise through social media. Researching this missing piece within current literature is especially interesting for academic research as well as offers practical implications because it will shed light on the role of the CEO and his characteristics in determining the success of forwarded social issues through the current predominant medium.

(8)

3 To fill the demonstrated gaps in the current literature, this work will first investigate

whether social issues that arise through social media are more successful if the CEO of the

respective company is targeted; and in a second step will analyze if specific characteristics of

the targeted CEO affect the likelihood of success of social issues that arise through social

media. Both questions will be empirically analyzed with a sample of social issues that target

Fortune Global 500 companies. Data regarding the content as well as the targeted person of the social issues and data to determine the subsequent success will be gathered from the international petition platform ‘Change.org’. Regarding the specific characteristics of the CEOs, tenure (Lewis et al., 2014; Rehbein et al., 2013) and nationality based on Hofstede’s cross-cultural differences (1984a; 2001) will be statistically investigated.

Thereby, this work will contribute to the leadership literature by reinforcing the

importance of the CEO with regard to social issues directed at the respective company, and by demonstrating the effects of specific leader characteristics. In this way, it will enrich the existing literature by taking social issues that arise through social media into account and establishing a link between the CEO characteristics and the outcomes of these issues. Consequently, the results will be, on the one hand, valuable for social activists in improving the effectiveness of forwarded issues. On the other hand, the results will have managerial implications by raising awareness of the specific characteristics that determine the outcome of the issues and the role of the CEO in general, which in turn can be leveraged when appointing CEOs.

This master thesis will first give an overview in chapter two of the existing literature covering social issues and ways in which these social issues arise as well as firm’s

responsiveness to these issues. Furthermore, the importance of social media will be reviewed and the role of managers and their characteristics within the decision making of the firm and especially the role of the CEO will be illustrated. Afterwards in chapter three, the conceptual framework with specific hypotheses for answering the research questions will be given. In

(9)

4 chapter four, the sample as well as the main variables which will be used to test the

established hypotheses will be discussed and the resulting methodology will be explained. The following chapter five will illustrate the resulting statistical outcomes of the hypotheses testing. In chapter six, the results will be discussed and implications of this work as well as suggestions for future research will be given. This work will close with acknowledgements in chapter seven.

2. Literature Review

Scholarly work has investigated topics that are of special interest to the research questions put forward in this work. Many scholars have researched when and how firms, that are targeted by stakeholder groups regarding specific social issues, respond (Bundy et al., 2013; Eesley & Lennox, 2006; McDonnell et al., 2015; Mitchell et al., 1997; Waldron et al., 2013). Moreover, the difficulty of narrowly defining social issues is of vital interest when dealing with those concerns (Clarkson, 1995). Furthermore, these social issues are put forward in different ways (McDonnell et al., 2015), and the role of social media within these processes constitutes a newly evolved phenomenon. However, when investigating the responses of firms to social issues or even business scandals, the role of the personal interpretation and non-objectivity of responsive managers cannot be neglected (Cyert & March, 1963; Hahn, Preuss, Pinkse & Figge, 2014; Waldron et al., 2013). In the following sections, the state of the current scientific literature will be reviewed and crucial links between the previously mentioned main topics will be established.

(10)

5

2.1. Social Issues

Social issues that are able to pose significant challenges to the respective target company are omnipresent and cannot be ignored by managers (McDonnell et al., 2015). Within the existing literature, a wide spectrum of issues is classified as a social issue and a narrow definition is difficult due to the broad meaning of the term (Clarkson, 1995). However, Friedman (1970) interprets social issues as non-business issues fostering a clear distinction between businesses and society. It can be inferred that social issues are issues that are of major concern to the whole society and do not only address business aspects (Clarkson, 1995; Friedman, 1970).

Clarkson (1995) points out that a distinction should be made between social and

stakeholder issues. While social issues are of concern to the whole society as defined above, stakeholder issues are relevant to one or more stakeholder groups but not necessarily to the whole society (Clarkson, 1995). Moreover, to classify an issue as a social one, the availability or need for legal regulations can serve as a guideline. However, social and stakeholder issues are not necessarily the same, but these categories can overlap, and both need to be addressed by the targeted corporations (Clarkson, 1995). Within this work, the broader category of social issues will be investigated.

When those broached social issues cover events or actions of a company that are negatively perceived by the public and gain an increased popularity within a relevant audience, one can classify these issues as a business scandal (Tybout & Roem, 2009). To reach the dimension of a scandal, a public response to the issue is necessary and the

development depends on the content of the issue itself. In case the issue is relevant to central attributes of a firm’s identity, the negative issue is more likely to reach the dimension of a scandal (Tybout & Roem, 2009). This is of vital interest as it is often argued that a firm’s responsiveness increases when their identity is threatened by the reputational risk of a

(11)

6 contrasting, publicly raised opinion (Waldron et al., 2013). This aspect will be further

discussed in the following section.

2.2. Firm Responsiveness and Success of Forwarded Issues

When social issues arise, and are put forward by stakeholder groups, the decisive criteria for their success is whether their actions are able to provoke the desired response from the targeted company. Generally, stakeholders can be defined as any group of people or single individuals that are able to affect a firm’s behavior or are categorized as being affected by it (Freeman, 1984; Mitchell et al., 1997). To clarify the terms, these stakeholders can be subdivided into primary and secondary stakeholders (Clarkson, 1995; Mitchell et al., 1997). While the continuing participation of primary stakeholders is necessary for the corporation to survive, the participation of secondary stakeholders, who are not involved in corporate transactions, is not crucial for the company’s survival (Clarkson, 1995). This distinction already reveals the varying importance of stakeholder groups for the company and hints at a differing degree of firm responsiveness. However, actions taken by both stakeholder groups can significantly harm the company (Clarkson, 1995).

A multitude of scholarly work has investigated how and when firms respond to social issues and stakeholder concerns (Bundy et al., 2013; McDonnell et al., 2015; Waldron et al., 2013). Whether firms respond in general to social issues can be subsumed under the term of firm’s responsiveness which is defined as “the degree to which a firm is willing to provide a thoughtful response to stakeholder concerns and commit to continued work on the issue" (Bundy et al. 2013, p.352). Consequently, when the targeted firm responds positively meaning in “a way consistent with the stakeholders’ request” (Eesley & Lennox, 2006, p 767), it will be interpreted as a victory or success within this work. Thereby, the company is

(12)

7 willing to take respective actions of change rather than a purely symbolic response (David et al., 2007; Waldron et al., 2013),

In determining what causes firms to respond favorably to stakeholder concerns or

specifically to social activists, the literature offers many different influencing factors. Some scholars focus on the stakeholders themselves as already outlined previously (Eesley & Lennox, 2006, Mitchell et al., 1997). They claim that stakeholders’ power, legitimacy, and urgency determine whether the firm responds positively to the issue as these attributes heighten the stakeholders’ saliency (Mitchell et al., 1997). Eesley and Lennox (2006) extend this framework by proposing that the interaction between the firm and the stakeholder group is critical in determining their saliency. Consequently, the attributes necessary to elicit a positive firm response, put forward by Mitchell et al. (1997), arise out of an interdependency of the nature of the stakeholder, the firm, and the specific request (Eesley & Lennox, 2006). In contrast to that, other scholars focus on the salience of the issue as a primary antecedent of firm responsiveness (Bundy et al., 2013). The issue is defined as being more salient when it matters to a large degree to the management and is therefore prioritized (Bundy et al., 2013). Consequently, an issue must be perceived and interpreted by the management as a threat to the strategic goals of the company in order to provoke a positive firm response (Bundy et al., 2013). Moreover, Waldron et al. (2013) also point out that the reputational pressure executed by activists raising social issues creates a threat to the organization’s identity and thereby increases the likelihood of a positive firm response.

The aforementioned scholars assume a rather static relationship between the outlined antecedents and the targeted firm’s responsiveness. This constitutes a contrast to the study of McDonnell et al. (2015) which reveals that the firm’s responsiveness is part of a dynamic process. According to this study, the responsiveness of a targeted company increases by being continually targeted by activist groups (McDonnell et al. 2015). These theoretical perspectives not only explain how and when targeted companies respond to social issues but

(13)

8 also explain the indisputable variation in firms’ responses (McDonnell et al., 2015) based on differing antecedents.

2.3. Ways Through Which Social Issues Arise

It must be pointed out that a distinction should be made between the ways through which social issues, that can potentially transform into corporate scandals, arise (McDonnell, 2016; McDonnell et al., 2015). Social issues can arise by means of tactics that bring up issues outside of the company which need to be differentiated from institutional tactics that bring up issues within the company itself (McDonnell et al., 2015). As companies have become more prominent in the global economy, they have become major targets of activism (Spar & La Mure, 2003). Thereby, social activists often make use of tactics like protests, boycotts, or media campaigns to get attention for social issues outside of the company (King & Pearce, 2010).

In contrast to that, institutional tactics often comprises shareholder resolutions which can be defined as “written proposals presented for a vote at a company’s annual meeting” (McDonnell et al., 2015, p. 656). Shareholder resolutions can be classified as a traditional way of bringing social issues to the management’s attention and highlighting the need for change (David et al., 2007; McDonnell et al., 2015; Rehbein et al., 2013). These shareholder resolutions are prevalent in major companies and can include a variety of corporate social responsibility (CSR) related issues (Porter & Kramer, 2006). However, managers are not legally bound to implement changes proposed in the resolutions and therefore their effectivity is questionable (David et al., 2007). David et al. (2007) propose, on the one hand, that

shareholder resolutions can serve a disciplining role forcing managers to improve corporate social performance (CSP). On the other hand, the authors point out that resolutions signal to the public that the CSP of a firm is inadequate. This in turn hints at managers’ incapability to

(14)

9 implement and to transform social demands into adequate improvements within the company. This argument proposes that managers tend to maintain the status quo mitigating the effects or success of shareholder resolutions (David et al., 2007). Shareholder salience is expected to increase the managers’ responsiveness resulting in a settlement of the shareholder resolution. Thereby, the stakeholder salience framework of Mitchell et al. (1997), which plays an

important role in the firm’s responsiveness as outlined in the preceding section, is leveraged. However, David et al. (2007) find that shareholder resolutions do not improve subsequent CSP. This indicates that managers mostly refuse to take significant actions and rather react symbolically after a shareholder resolution. According to the previously illustrated definition, these actions cannot be interpreted as a success. Hence, this finding heightens the importance of investigating the effects and the likelihood of victories of social issues that arise outside of the firm in contrast to internal tactics like shareholder resolutions.

Activist campaigns can be classified as one category of tactics to bring up social issues from outside of the company. As campaigns can provoke substantive as well as symbolic responses (Waldron et al., 2013), there is no clear tendency of being a more successful tactic arising outside of the company compared to the aforementioned shareholder resolutions. McDonnell and Werner (2016) note that increased media attention supports the effectiveness and likelihood of the success of boycotts. King and Pearce (2010) generalize the effect of media attention for tactics that bring up issues outside of the company. The media attention reinforces the reputational pressure on the targeted company which in turn increases the firm’s responsiveness as explained earlier in this work (McDonnell & Werner, 2016; Waldron et al., 2013).

This argumentation hints at the importance that media plays in influencing the outcomes of social issues. Hence, it can be proposed that social media, as a current media phenomenon that has gained increasing popularity and importance in the business world, will also have a significant influence in the evolution and process of social issues (Aula, 2010; Drobocky,

(15)

10 2013; Kaplan & Haenlein, 2010). Elaborating this argumentation further, one can propose that social media even constitutes a new way of how social issues arise and are put forward outside of the targeted company. The relationship between social issues arising out of social media compared to previously outlined traditional ways and the likelihood of a success of the issue has not been investigated in previous studies. Therefore, this relationship constitutes a gap in the existing literature and will be addressed within this work.

2.4. Social Media

The increasing relevance of social media in the private as well as in the business environment is undeniable (Aula, 2010; Drobocky, 2013; Kaplan & Haenlein, 2010), as illustrated with the entering example. Five years after its launch in 2004, Facebook had already about 175 million registered active users which grew to 1,37 billion active daily users in September, 2017, reinforcing the importance of social media (Facebook Company

Information, 2017; Kaplan & Haenlein, 2010). Besides indisputable effects of social media in the private as well as in the business world, it is possible that both dimensions overlap

(Drobocky, 2013). Social media, as the currently predominating way to connect, is not only used to acquire personal relationships but also to gather job opportunities, which reinforces its importance in business settings (Drobocky, 2013).

Furthermore, constant connectivity, especially between employees, blurs the lines between the private and business world, alters privacy concerns in general and thereby potentially facilitates misuse (Mainiero & Jones, 2013). Consequently, these developments pose threats to companies by influencing the working atmosphere and calling forth for new policies (Mainiero & Jones, 2013). Thereby, legal risks for companies are likely to arise through social media (Mainiero & Jones, 2013).

(16)

11 This already hints at the duality of social media for companies. On the one hand, social media offers a way to directly engage with consumers and stakeholders as a low-cost

communication channel and thereby increases efficiency (Kaplan & Haenlein, 2010). Hence, it provides a medium through which an organization can communicate publicly with several stakeholder groups besides the consumers (Aula, 2010). This in turn indicates that, vice versa, a communication of social issues directed to the targeted company by diverse stakeholders is also possible through the medium.

On the other hand, social media poses several threats to a company by expanding

reputation risk and increasing risk dynamics (Aula, 2010). Reputation risk can be defined as “the possibility or danger of losing one’s reputation” (Aula, 2010, p. 44). These risks are reinforced through social media by being less controllable and manageable than conventional media types (Aula, 2010). This is based on several reasons.

Nowadays, not only most actions of a company are public, but also all stakeholder groups are able to post self-generated content about a company without previous verification or approval. Generally, the sum of all ways users make use of social media can be subsumed under the term user-generated content (Kaplan & Haenlein, 2010). Consequently, activists or stakeholders can put forth ideas or beliefs about a company, which potentially differ

dramatically from the image the company wants to have or the identity it intends to represent. The risk to lose reputation increases with the gap that arises between the company’s intended reputation and the user-constructed reality offered through social media (Aula, 2010).

Moreover, the responsiveness of the company depends on the degree of reputational risk as outlined previously. Therefore, social media, by reinforcing or even aggravating reputation risk, has to play a critical role in the success of the discussed social issues (Waldron et a., 2013).

Thereby, companies are inevitably forced to react on issues that are put forward by stakeholder groups through this channel. This is substantiated by Bhattacharya and Polman

(17)

12 (2017) who point out that in order to limit reputational risk, a company needs to engage more actively and to offer transparency with stakeholders having increased access to

communication. As a result, the specific responses of firms to such issues within social media, as a way of communication, are critical factors for mitigating the risk (Aula, 2010). Furthermore, gaps between the expectation of a company, and those of public stakeholders reinforced by social media, can have important consequences for organizations by affecting among others competitiveness and firm outcomes (Aula, 2010).

2.5. The Role of Managers

After having illustrated the different factors influencing the firm’s responsiveness as well as the importance and ways on which social issues arise, it is important to acknowledge that these decisions of how and whether a firm positively responds are finally taken by managers of the targeted firm (Waldron et al., 2013). Generally, managers cannot be assumed to be wholly rational as their decisions depend on individual interpretations of the threat and identity risk an issue poses to the company (Cyert & March, 1963; Hahn et al., 2014; Waldron et al., 2013). Furthermore, social issues put forward by stakeholders can only be successful if they are able to get to managers’ attention and if managers are willing to comply (Eesley & Lennox, 2006; Mitchell et al., 1997; Waldron et al., 2013). Thereby, managerial preferences as well as personal values can have an important impact on the subsequent success of an issue (Spar & La Mure, 2003). Hence, focusing solely on a firm’s

responsiveness by classifying the firm as the decision-making body regarding social issues is overly simplistic and underestimates the rather complex dependencies.

The firm should be displayed as a coalition of multiple individuals, among others the managers of the firm (Cyert & March, 1963). These individuals pursue different goals and interests and not only their attention is directed to differing issues, but also their ability to

(18)

13 address them is restricted (Cyert & March, 1963). Hence, the role of the managers’ attention and their individual interests are supposed to be of critical importance in addressing social issues. This view of the firm is adopted and supported by Pfeffer and Salancik (1978) who propose that companies are “other-directed” (Pfeffer & Salancik, 1978, p.257). Furthermore, all outcomes, therefore also the responses to social issues, are influenced by

interdependencies between agents and the firm (Pfeffer & Salancik, 1978). Thereby,

managers are also influenced by the environment. Consequently, the firm can be displayed as a coalition of differing interests that is even dependent on the environment, and managers constitute the center of this network. Hence, the role and perspective of the manager is critical for the success of and response to social issues (Cyert & March, 1963; Mitchell et al., 1997).

Furthermore, as managers constitute the symbol of the targeted organization, they often constitute the target of external pressures (Pfeffer & Salancik, 1978). Therefore, it is unclear whether social issues that are put forward are more successful when managers of the firms are directly targeted. This constitutes an aspect that will be investigated further within this work. Moreover, referring to the illustrated antecedents of firm responsiveness like

stakeholder salience mentioned above, these antecedents cannot be seen as independent explanatory factors on their own. Instead, their effectiveness depends on the managers’ attention which is in turn influenced by their individual interpretation of the issue (Mitchell et al., 1997). These managerial interpretations result from individual perceptions that arise out of their differing characteristics (Mitchell et al., 1997). This reinforces the need to take specific characteristics of managers into account.

The statement that specific characteristics of managers determine their perception, and hence their strategic decisions, is based on the upper echelons theory established by

Hambrick and Mason (1984). This relationship is grounded on the argument that individual cognitive values and frames serve as predictors for the choices managers make. According to

(19)

14 the model of the upper echelons theory, a manager brings his individual background, which is composed of values and experiences, to the decision-making. These values and experiences serve as a screen between the perceived setting and the objective situation (Hambrick & Mason, 1984). Furthermore, the manager is only able to direct his attention to restricted areas which results in an even more limited perception. Subsequently, the information gathered, and the individually perceived situation are interpreted through a filter composed of the manager’s mindset and values. Hence, the personalized interpreted situation based on the specific characteristics will lead to certain behaviors and distinct strategic choices which then have an impact on the company’s outcome (Hambrick, 2007; Hambrick & Mason, 1984). Taken together, the investigation of the general role of the targeted manager and the impact of the specific characteristics is critical in evaluating the likelihood of successful outcomes of social issues.

2.6. The CEO and CEO Characteristics

The upper echelons theory holds true for single managers as well as top management teams (Hambrick & Mason, 1984). The general importance of the CEO in the literature is often controversial (Ling, Simsek, Lubatkin & Veiga, 2008). While some researchers propose that the characteristics of the top management team are more influential than the ones of the CEO in strategic decisions (Papadakis & Barwise, 2002), other scholars regard the CEO as being the main actor within the top management team by being most powerful (Arendt, Priem & Ndofor, 2005; Hambrick, 2007; Hambrick & Mason, 1984; Ling et al., 2008). The latter view, assuming a higher importance of the CEO, will be adopted within this work. According to this view, the characteristics of the CEO are supposed to dominate the team’s

(20)

15 disproportionate power and a dominating influence within the team (Hambrick, 1994 in Ling et al., 2008).

Furthermore, the CEO is to a large extent responsible for designing a firm’s strategy and securing a positive image of the firm (Waldmann et al., 2006). This image is severely

threatened by the reputational risk arising through social issues as stated previously (Waldron et al., 2013). Therefore, the role of the CEO within the investigation of outcomes of social issues is of vital interest. Prior studies have confirmed the influence of CEO characteristics on the outcomes of social issues that were brought up in traditional ways (Lewis et al., 2014; Manner, 2010; Rehbein et al., 2013; Slater & Dixon-Fowler, 2009). Especially prior

leadership literature has acknowledged that characteristics and values of leaders, hence the CEO, influences the decision making with regard to social responsibility decisions

(Waldmann et al., 2006).

A CEO’s type of education and prior functional experience, gender as well as salary are supposed to impact the firm’s corporate social performance (Manner, 2010). Moreover, besides functional experience, international assignment experience of a CEO is expected to heighten the CEO’s awareness of societal expectations as well as his or her motivation to act in accordance with them, which in turn leads to a higher corporate social performance of the respective company (Slater & Dixon-Fowler, 2009). Huang (2013) confirms that

demographic characteristics of CEOs like age, gender, tenure, and nationality are proposed to affect the corporate social performance of the respective company. Hence, these

characteristics will also have an impact on outcomes of social issues which at the end constitute the level of social performance. Thereby, a higher level of social performance indicates a greater receptivity of managers to scandals which arise on traditional ways like shareholder resolutions. Specifically, the power and duality of the CEO are supposed to result in differing outcomes of shareholder resolutions (Rehbein et al., 2013).

(21)

16 Furthermore, in the studies previously outlined, it becomes apparent that CEO tenure is a heavily analyzed characteristic that exerts significant influence. Lewis et al. (2014) confirm that CEOs’ tenure, besides educational background, impacts the likelihood that a firm discloses environmental information, which can be classified as a social issue requested by shareholders. This can be based on the fact that CEOs gain increased power and hence are able to better resist external pressures with a longer time in office (Lewis et al., 2014; Miller, 1991). According to the study of Miller (1991), CEOs with a longer tenure are less likely to match the strategy with a changing environment. This also suggests that longer tenured CEOs are less likely to implement changes that are asked for by stakeholders regarding social issues. Furthermore, a longer tenure of the CEO results in lower responsiveness to

shareholder resolutions according to Rehbein et al. (2013). Consequently, tenure of the CEO determines the outcomes of social issues that arise on traditional ways like shareholder resolutions. Thus, it can be hypothesized that tenure also influences the outcome of social issues that arise through social media. Therefore, the effect of tenure as an important CEO characteristic will be conceptualized and analyzed within this work.

Besides tenure, the effects of the nationality of the CEO are often investigated (Huang, 2013). Hambrick (2007) points out that CEOs differ in their decision making according to the upper echelons theory and that these differences may also depend on the society, and

therefore the country, the CEO comes from or operates in. Nationality of the CEO was

proven to significantly influence the firm’s compensation and capital structure (Jalbert, Chan, Jalbert & Landry, 2007). Moreover, Weber (2010), focusing on the moral integrity of CEOs in the automobile industry, found that differences in CEOs’ moral reasoning can be attributed to differences in nationalities. Thus, it can be inferred that nationality, by affecting moral behaviors of CEOs, will also play an important role within the outcomes of social issues, as manager’s personal values and demographic backgrounds influence the decision making (Hambrick & Mason, 1984; Spar & La Mure, 2003).

(22)

17 However, Huang (2013) does not prove a significant effect of CEO nationality on

corporate social performance which can be related to the outcome of raised social issues. Nevertheless, this finding can only be generalized with reservation as the data used was biased towards American and European companies (Huang, 2013). Besides, Riviera and De Leon (2005) find no significant correlation between CEO nationality and firm participation in voluntary environmental programs which can be classified as an outcome of a social issue. This finding is surprising and calling for further research as the authors expected a significant positive correlation. Thereby, assuming that CEOs from industrialized countries would be more likely to participate in voluntary programs due to prior claims that these CEOs are more proactive and environmentally aware due to norms in their home country (Riviera & De Leon, 2005). Given the limitations and contrasting results of prior studies, and regarding the theoretical foundations that argue for an impact of CEO nationality, it seems important to further investigate the nationality of CEOs with reference to social issues.

However, responding to social issues in general is determined by the prioritization of managers as explained above (Bundy et al., 2013; Waldron et al., 2013). According to the upper echelons theory, this prioritization of managers is determined by their demographic background and values (Hambrick & Mason, 1984). Furthermore, Hambrick and Mason (1984) point out that the decision making of managers, in accordance with their upper echelons theory, is influenced among others by the socioeconomic roots which can be associated with the individual’s nationality.

Taking the argumentation to a broader level, the nationality of the CEO can be seen as a proxy for his cultural background (Jalbert et al., 2007). The cultural background of an

individual varies according to the national culture the individual stems from which comprises a specific value system. Moreover, an individual’s values are heavily influenced by the value systems in which the individual grows up as the specific individual values are programmed in early life phases (Hofstede, 1984a; Hofstede, 2001). Furthermore, studies proved the link

(23)

18 between national culture and individual personality (Hofstede, 2011). Taken together,

nationality is an important determinant of an individual’s, and hence a CEO’s, values. Consequently, as the values of the CEO determine the perception and subsequent decision making (Hambrick & Mason, 1984) which in turn determines the degree of attention and responsiveness of the CEO towards social issues (Eesley & Lennox, 2006; Mitchell et al., 1997; Spar & La Mure, 2003; Waldron et al., 2013), nationality can be argued to play a critical role regarding the success of social issue through social media.

Accounting for the previously explained national differences, Hofstede’s (2001)

distinction between individualist and collectivist countries can serve as an antecedent of the receptivity of CEO towards social issues. In individualist countries, people are motivated by self-interest while collectivist cultures prioritize group interests (Hofstede, 1984b). As

defined earlier in this work, social issues are of major concern to the whole society (Clarkson, 1995; Friedman, 1970), and hence can be classified as a group interest within the society the issue arises. Following the argumentation above, whether the CEO comes from a country with high individualism or collectivism can be argued to determine his responsiveness to the forwarded issues. This is based on the causality that inherit values will lead him to a distinct prioritization of business self-interests versus collective interests depicted within forwarded social issues. Thereby, hinting at a tendency which CEOs will respond more favorably towards social issues. As a result, when investigating the success of social issues arising through social media, it is important to take nationality of the CEO as a moderating characteristic into account.

Taken together, CEO characteristics play a vital role in influencing the success of social issues that arise on traditional ways (Lewis et al., 2014; Manner, 2010; Rehbein et al., 2013; Slater & Dixon-Fowler, 2009). By adopting the view that the CEO is the most influential player within the management team as explained above (Arendt et al., 2005; Hambrick &

(24)

19 Mason, 1984; Ling et al., 2008), this work will solely focus on the specific characteristics of the CEO of targeted companies and will leave the impact of the characteristics of other TMT members for future research. While acknowledging the importance of the CEO, no prior study has investigated his or her role as a targeted person specifically in social issues that arise through social media as a new way of scandals to arise. This also holds true for the impact of the CEO characteristics on the likelihood of a success of social issues that arise out of social media as a new phenomenon as illustrated in the previous sections. This constitutes a gap in the existing literature that will be addressed within this work. Furthermore, no study analyzing the success of social issues has investigated the dependency with recent data up to the year 2017 which will be used in the empirical analysis of this work. Hence, the following research questions will be addressed in order to fill the identified important gap in the

scientific literature.

First, the question “Are social issues that arise through social media more successful if

the CEO of the respective company is targeted?” (Figure 1) will be addressed. Second, the

research question: “Do the characteristics of the targeted CEO affect the likelihood of

success of the social issue that arises through social media?” (Figure 2).

The following figures visualize the relationships that will be investigated with the research questions of this work.

(25)

20 Figure 2: Illustration Research Question Two (Own representation)

3. Conceptual Framework with Hypotheses

As identified in the preceding section, the previous literature does not investigate the effect of CEO targeting within social issues arising by means of social media, and the effect of specific CEO characteristics. However, these can be argued to have an impact on the success of social issues that arise through social media. Hence, a conceptual framework with hypotheses to answer the outlined research questions will be established in this section. Thereby, the previously highlighted characteristics of CEO tenure and CEO nationality will be investigated.

3.1. Effect of CEO as Targeted Person on Success of a Social Issue

The first research question refers to a hypothesized relationship between the direct targeting of the CEO of the respective firm and the success of the social issue that was forwarded through social media. As managers, specifically CEOs, are perceived as the symbol of a corporation, they often constitute the target of social issues that are directed at the company (Pfeffer & Salancik, 1978). Following the argumentation of the previous

chapter, the CEO can be regarded as the main actor within the top management team (Arendt et al., 2005; Hambrick, 2007; Hambrick & Mason, 1984; Ling et al., 2008). Hence, the CEO

(26)

21 not only constitutes the frequent target of social issues but also embodies the one determining the company’s response. The firm’s responsiveness is supposed to result in the firm taking actions, hence a success of the issue, when the CEO interprets the issue as threat to the company’s reputation and identity (Cyert & March, 1963; Hahn et al., 2014; Waldron et al., 2013).

Besides, the reputation of a company is closely related to the image and reputation of its CEO. Moreover, both the company’s reputation and the CEO’s image not only have a significant impact on the general firm outcome but can also be harmed by negative media coverage (Jin & Yeo, 2011). Hence, social media even intensifies the reputational risk and thereby the threat a social issue poses at the respective company. Consequently, it can be argued that a social issue that arises through social media will also constitute a threat to the CEO’s image (Aula, 2010; Waldron et al., 2013; Jin & Yeo, 2011).

As a result, it can be hypothesized that when the CEO is directly targeted within the social issue, this threat is even aggravated. As the reputational risk determines the firm’s

responsiveness (Waldron et al., 2013) and the CEO is the main decision-making body

(Arendt et al., 2005; Hambrick, 2007; Hambrick & Mason, 1984; Ling et al., 2008), the CEO is supposed to react more favorably to the issue when he is targeted directly because of an increased reputational risk for him or her and the company. This argumentation leads to the following first hypothesis visualized in Figure 3.

H1: When the CEO is directly targeted within a social issue that arises through social media,

(27)

22 Figure 3: Illustration Hypothesis One (Own representation)

3.2. Effect of CEO Characteristics on Success of a Social Issue

The second research question raised at the beginning of this work deals with a

hypothesized relationship between specific CEO characteristics and the likelihood of success of the social issue raised through social media. The two characteristics investigated will be CEO tenure and CEO nationality.

3.2.1. Tenure

Hambrick and Mason (1984) point out within their upper echelons theory that tenure is one of the main characteristics that influence the manager’s, and hence the CEO’s, decision-making. Within the literature, a longer tenure is generally associated with a greater

commitment to the status quo (Henderson, Miller & Hambrick, 2006, Miller, 1991; Wiersema & Bantel, 1992). This tendency is confirmed by an inverted U relationship between CEO tenure and firm performance. Thereby, CEOs first engage in improvements to increase the fit with the environment which also plays a role within social issues. However, with a longer tenure, the CEO becomes increasingly committed to the status quo, so that it becomes less likely that he or she engages in implementing and executing change (Henderson et al., 2006). This argument is supported by Lewis et al. (2014), who point out that CEOs are supposed to take major actions within their first two and a half years after their appointment. As

(28)

23 previously outlined, a social issue can only be counted as successful when the targeted firm engages in actions of change (David et al., 2007; Waldron et al., 2013). Consequently, the tenure of the CEO, by influencing the propensity of implementing changes, is expected to determine the success of a forwarded social issue through social media.

Therefore, longer tenured CEOs are supposed to be less eager to implement changes fostered in forwarded social issues. Thereby, a longer tenure will be associated with a decreased likelihood of success of the social issue. The argumentation that longer tenure results in decreased firm responsiveness is confirmed by Rehbein et al. (2013) for shareholder resolutions. This finding can be based on the fact that longer tenured CEOs are seen as

having more informal power which enables them to resist external pressures (Lewis et al., 2014)

Besides, a longer tenure indicates that the CEO has greater experience in dealing with strategic risks and therefore encounters less uncertainty (Simsek, 2007). As social issues arising through social media pose risks on the targeted company and the CEO as illustrated previously (Aula, 2010; Waldron et al., 2013; Jin & Yeo, 2011), longer tenured CEOs can be seen as more willing to take these risks. Hence, they are less likely to elicit positive firm responses in order to mitigate the risk. These arguments lead to the following hypothesis visualized in Figure 4.

H2: A longer tenure of the CEO decreases the likelihood that a social issue put forward

(29)

24

3.2.2. Nationality

As illustrated in the preceding sections, the socioeconomic roots, values, and experiences of the CEO determine his decision-making (Hambrick, 2007; Hambrick & Mason, 1984). Hence, a CEO’s values are crucial within the decision on how to react to a social issue that arises through social media and targets the respective company. The values of an individual can be seen as determined by demographic characteristics, among others the nationality, and therefore the society within which an individual has grown up (Hofstede, 1984a; Hofstede, 2001). Consequently, it is highly probable that the nationality of the CEO significantly determines his strategic decisions, and in turn constitutes an important factor within the responses to forwarded social issues.

With regard to Hofstede’s (1984a; 2001) distinction between individualist and collectivist cultures, individualism describes the relationship between the individual and prevalent group interests. People from countries high in individualism tend to prioritize self-interest and hence make decisions independent from group interests (Hofstede, 1984b; Triandis, 2001). In contrast to that, people from collectivist cultures strive for group success as they see

themselves as part of the group. Consequently, they shape their behavior according to norms that are prevalent in the respective society (Hofstede, 1984b; Triandis, 2001). As social issues are of major concern to the whole society (Clarkson, 1995; Friedman, 1970), it can be argued that CEOs stemming from countries that are high in collectivism prioritize the interest of the whole group and therefore societal interests. Hence, collectivist CEOs are expected to be more likely to positively respond to raised social issues through social media.

This argumentation is substantiated by the fact that the distinction between individualism and collectivism also affects the notions of morality of different nationalities (Triandis, 2001). Consequently, people, and hence CEOs, from collectivist cultures put more emphasis on ethics and social responsibility (Vitell, Paolillo & Thomas, 2003). This leads to the

(30)

25 conclusion that CEOs from countries that are high in collectivism will value social

responsibility more than their individualist counterparts, and as values affect their decision making (Hambrick & Mason, 1984), these CEOs will tend to be more responsive to forwarded social issues through social media. This argumentation leads to the following hypothesis visualized in Figure 4. Thereby, the degree of individualism is expected to

moderate the outlined relationship between tenure and subsequent success in hypothesis two.

H3: Social issues forwarded through social media are more likely to be successful when the

targeted CEO stems from a country that is high in collectivism compared to a CEO that stems

from an individualist country.

(31)

26

4. Methods

The hypotheses established in the previous section will be quantitatively analyzed within this work. The following chapter will give an overview over the sample that is used as well as an illustration of the data collection process. Furthermore, it will be argued why the chosen sample is suited to reach consistent results. Moreover, the specific variables as well as the quantitative research design will be described.

4.1. Data and Sample

To analyze the effects of CEO targeting and CEO characteristics on the success of social issues, social petitions arising through social media in the years ranging from 2009 to 2017 and targeting companies listed in the Fortune Global 500 firms between the years 2012 and 2016 will be used (Ferris, Jayaraman & Sabherwal, 2013). The Fortune Global 500 is a list of the 500 largest companies worldwide, measured by revenue, which is published annually by the Forbes magazine (Forbes, 2017).

Due to their size, coverage, and power, the Fortune Global 500 companies constitute an appropriate sample because companies are more likely to be targeted when a broad visibility is given. This is based on the fact that activists prefer targeting companies which not only seem to have the ability of addressing the problem, but also have enough visibility and an image to maintain, so that public external pressure is expected to constitute a serious financial threat to the targeted company (McDonnell, 2016; Spar & La Mure, 2003). Moreover, using the Fortune Global 500 companies offers the opportunity to explore geographic differences (Muller & Whiteman, 2009). Hence, the effect of CEO nationality can be analyzed for a broad range of countries which leads to reliable results. The Fortune Global 500 firms are publicly listed companies and therefore respective data is publicly retrievable which

(32)

27 their companies as explained previously makes them more vulnerable to being targeted directly (Pfeffer & Salancik, 1978). This effect is expected to be more prevalent when the company, and hence their symbol, is publicly known as it is the case for the Fortune Global 500 companies, thereby substantiating this choice of the outlined data sample.

The international petition platform ‘Change.org’ will be used to retrieve data on social issues targeting the Fortune Global 500 companies in the respective years. By reviewing all petitions associated with the respective companies on the platform, data on the type of social issue, the success of the issue, the date the petition was created as well as the target person of the issue, if applicable, was coded. This preliminary dataset was supplemented with data on the industry using the 2-digit North American Industry Classification Scheme (NAICS) code, as well as enriched with data on targeted persons and their respective position by using external data sources to complete the dataset. Furthermore, the resulting dataset was cleaned by the supervisor so that only petitions consisting of a social issue that is relevant to more than one person remained. Companies in the Global Fortune 500 for which no petitions were coded from the online platform were excluded from the resulting dataset.

The cleaned dataset was then checked and completed by hand with data on the targeted CEOs, namely their date of inauguration and hence their complete years in office as of the date the petition was created. Moreover, their nationality with the respective individualism score according to Hofstede Insights (2017) was added. Thereby, a data sample of 2070 coded petitions was created which will be used for the quantitative analysis.

4.2. Variables

In the following paragraphs, the variables used to test the hypotheses quantitatively will be explained and their sources will be illustrated.

(33)

28

4.2.1. Dependent Variable

The dependent variable for both research questions is the success of the forwarded social issue. As illustrated previously, the success or victory of the forwarded issue is interpreted as a positive response of the firm, which means that the respective firm takes actions of change (David et al., 2007; Waldron et al., 2013). Hence, only substantive reactions are counted as a success within this work (N = 180; 8.7%). This success is evaluated according to the

descriptions of the specific social issue and respective data was gathered from the petition platform ‘Change.org’ as outlined in the previous section. The variable is a categorical, binary variable that will take the value of 1 for ‘success’ and 0 for ‘no success’. Hence, the dependent variable will constitute a dummy variable.

4.2.2. Independent Variable

For the first research question, the independent variable represents whether the respective company’s CEO constitutes the targeted person of the social issue under consideration (N = 359; 17,3%). This information was retrieved in the first coding process from the petition platform ‘Change.org’ within the category ‘targeted person’ as outlined before. The variable is again a categorical and binary variable with the values of ‘yes’ coded as 1, when the CEO is the targeted person, and ‘no’ coded as 0 in the contrary case.

For the second research question which investigates the effects of specific CEO characteristics on the success of the social issue, the independent variable consists of the tenure of the CEO (M = 6.22; SD = 7.054). Following Lewis et al. (2014), data to

complement the dataset on CEO tenure was collected by hand from executive profiles. Apart from official company information, Bloomberg executive profiles were used to retrieve the date of inauguration. Based on this data, CEO tenure was calculated as the total years a CEO has been in office until the issue arose, what is in accordance with the method used by

(34)

29 Henderson et al. (2006). Thereby, tenure is a numerical and continuous variable. The positive skewness of the tenure is accepted in the further analysis as it reflects accurate values.

4.2.3. Moderating Variable

As argued in the conceptual framework, CEOs from countries higher in collectivism are supposed to respond more favorably to social issues, that are per definition of interest to the whole society, due to their group success orientation (Clarkson, 1995; Friedman, 1970; Hofstede, 1984b; Triandis, 2001). Hence, nationality in terms of collectivist orientation is expected to moderate the effect of tenure on the success of social issues, thereby constituting the moderator variable (M = 85.48; SD = 13.791). To investigate the effect quantitatively, data on the nationality of the targeted CEO as well as the respective individualism score according to Hofstede (1984b; Hofstede Insights, 2017) was included in the dataset. Thereby, a lower individualism score is assigned to more collectivist-oriented countries.

Following Ferris et al. (2013), the nationality of the CEO will be determined by his place of birth within this work and respective data on the nationality was collected by hand. Thereby, most of the data was coded from Forbes Magazine what is consistent with the method used by Jalbert et al. (2007). However, if the nationality was not retrievable on this way, the data was complemented using respective company information, newspaper articles, among others from the ‘Financial Times’ as well as ‘The Guardian’, and from the online platform ‘boardroom insiders’.

The respective numerical individualism score was subsequently retrieved from the organization ‘Hofstede Insights’ (2017). Consequently, the moderator is a numerical,

continuous variable. As the values reflect a widely accepted scale representing real numbers in a specified range, the negative skewness, reflecting the presence of many CEOs from

(35)

30 countries with high individualism values such as the United States, was accepted for further analysis.

4.2.4. Control Variables

Within the quantitative analysis of both research questions, the industry will be a control variable (David et al., 2007). As the impact of the CEO varies between industries

(Wassermann, Nohria & Anand, 2001), and the role and characteristics of the CEO are vital within this research, it is indisputable to control for the industry. Data concerning the industry was collected by hand and retrieved as secondary data from company information within the first coding process. To classify the companies within the industries, the two-digit NAICS code is used. Hence, controlling for the industry is done with a categorical variable. In total, 21 different industry categories were coded (df = 20). To decrease the degree of freedoms (df = 16), the industry categories were clustered according to the United States Census Bureau (2017). Thereby, NAICS code 45 was coded in 44, code 49 was recoded in 48 and 32 as well as 33 were recoded each in code 31. Through this procedure, 17 different industry categories remained in the dataset. This data shows that companies conducting business in retail trade with the NAICS code 44 and 45 are overrepresented (N = 750; 36.2%). Second in terms of numbers is the manufacturing industry represented by NAICS codes 31 until 33 (N = 333; 16.1%). In order to use a logistic regression to further analyze the data, 17 dummy variables were created, one for each coded industry category.

Furthermore, the quantitative analysis will control for the number of petition supporters (M = 12185.25; SD = 50376.454). As explained earlier in this work, an issue is more likely to elicit a positive response when it is interpreted as a threat to the strategic goals of the

company, thereby increasing reputational risk (Bundy et al., 2013; Waldron et al., 2013). This threat and perceived risk is likely to be greater with more petition supporters meaning

(36)

31 constant targeting with every new supporter. Moreover, online petitions that are supported by more people are more likely to lead to actions of change (Veil, Reno, Freihaut & Oldham, 2015). The corresponding data was retrieved from the petition platform ‘Change.org’ within the first coding process as outlined before. Hence, for every petition the respective number of supporters was included in the dataset, constituting a continuous control variable.

However, the distribution of the number of petition supporters was extremely positively skewed with a skewness of 7.795 (SE = 0.054). Hence, petitions with less than ten supporters constituting a deminimus amount of support were excluded from further analysis. This resulted in a skewness of 7.519 (SE = 0.056) which still represents an extreme positive skewness. The variable was transformed with a logarithm resulting in a skewness of 0.777 (SE = 0.056) which can be accepted as nearly normally distributed. Thereby, a mean of 2.05 (SD = 1.17985) resulted.

4.3. Methodology

The established hypotheses will be statistically analyzed within this work. Having collected all necessary data as illustrated in the previous sections, 2070 observations remained in the final dataset. Thereby, each represents one coded social petition including 359 issues in which the respective CEO constituted the targeted person.

After performing preliminary tests like normality and correlation tests with the dataset, the first hypothesis will firstly be tested with a Chi-Square Analysis and secondly with a Binary Logistic Regression to include the effects of the control variables. Furthermore, the second as well as the third hypothesis will be tested with a Binary Logistic Regression Analysis. Thereby, the hypothesized relationship will be tested with and without the control variables using the PROCESS macro by Andrew Hayes (2013). These specific tests and their results will be further illustrated in the following section.

(37)

32

5. Results

For testing the outlined hypotheses with the previously illustrated variables, Correlation tests, Chi-Square as well as Binary Logistic Regression analyses were performed. The test results are explained in detail in the following section.

5.1. Correlation

To get first insights of the relationships between the variables in the model and to investigate the issue of multicollinearity, correlations between the dependent and the independent variables are tested. Furthermore, the moderator and one control variable are included. To allow a better focus, the industry control variable was not included in the correlation at first due to its high number of categories. The resulting Pearson Correlation Matrix is visualized in Table 1. The Correlation Matrix including the industry categories can be found in Table 2.

However, due to the aforementioned exclusion of petitions with a deminimus amount of support, hence less than ten supporters, the total cases under investigation amount to 1927 which are relevant for the first research question. As the second research question focuses solely on cases in which the CEO constitutes the target, these tests will investigate 344 cases of the dataset. This smaller set is also used to test correlations of tenure and individualism with the remaining variables.

(38)

33 Table 1: Mean, Standard Deviations, Correlations

(39)

34 Table 2: Mean, Standard Deviations, Correlations including Industries

(40)

35 First, the correlation matrix shows that no variable is too highly correlated (> .8) and consequently any issues of multicollinearity can be alleviated. The independent variable which represents whether the CEO constitutes the target of the petition, is positively and significantly correlated with the success of the social issue with p < 0.01. This correlation indicates that it is more likely that the targeted company reacts with a significant action of change when the social issue is directed at the CEO. Hence, this correlation suggests that there is a relationship between targeting the CEO and subsequent success, thereby substantiating the reasoning of the first hypothesis.

However, CEO Tenure shows no correlation with the dependent variable of Success. As hypothesized, a lower tenure is expected to increase the likelihood of substantive actions of change. The correlation analysis offers as a first insight that this hypothesized relationship is doubtful. Nevertheless, this relationship will be further investigated in additional tests including the moderator variable.

The moderator variable, represented by the individualism score of the CEO’s country of birth, shows no correlation with the dependent variable of petition’s success. This indicates that there might be no direct relationship as hypothesized in the third hypothesis. However, Individualism constitutes the moderator and hence is expected to moderate the relationship between CEO Tenure and the resulting Success. According to Table 1, Individualism shows a significant correlation to CEO Tenure with p < 0.05. Consequently, it can be expected that a moderation effect will occur in further analyses.

The number of petition supporters which constitutes a control variable shows a significant correlation with the success of the issue with p < 0.01. Furthermore, it is significantly

correlated to the independent variable of CEO Target with p < 0.01. This indicates that a success is more likely with a higher number of supporters. Besides, it seems that petitions with more supporters target the respective CEO more often compared to petitions with a lower number of supporters. This substantiates the importance to control for the number of

(41)

36 supporters when further analyzing the effect of CEO targeting on the resulting success of the issue.

Regarding the industry control variable, some industry categories show marginal or significant correlations to the dependent variables as well as to the independent variable of Success which can be seen in Table 2. This substantiates the importance to control for the industry effects in both models. To investigate the hypothesized relationships more in detail and check for causality, further tests, explained in the following, will be conducted.

5.2. Analysis of Research Question One

According to the first hypothesis, targeting the CEO will increase the likelihood that the respective issue will be successful. As both variables, dependent and independent variable, constitute a dummy variable, this hypothesis can first be tested with a Chi-Square Analysis using SPSS. The results are shown in Table 3.

Table 3: Chi-Square Analysis

In total, about 9% (N = 177) of the social issues under investigation (N = 1927) yield a substantive firm response. When the CEO is not the target person, roughly 8% of the issues result in a success. In contrast to that, 15% of the issues in which the CEO constitutes the main target elicit a positive firm response. This already indicates a clear relationship between CEO Target and Substantive Firm Response.

(42)

37 Furthermore, there was a statistically significant association between CEO Target and

Substantive Firm Response, χ2(1) = 17.659, p < .001. Besides, based on the ϕ statistic on a scale from 0 to 1, the association seems to be of small size (ϕ= .10). Based on the frequency distribution, this result seems to represent the fact that there is a higher chance of a

substantive firm response among issues targeting the CEO compared to issues not targeting the CEO.

As this hypothesized effect does not occur in isolation, the relationship needs to be further tested by including the control variables. As the outcome variable of Success is a

dichotomous variable and the predictor variable of CEO Target also constitutes a categorical variable, a logistic regression is suited to investigate the first hypothesis (Peng, Lee &

Ingersoll, 2002). In Model 1, the uncontrolled effect of CEO Target on the dependent

variable of Success is tested to verify the results from the previous Chi-Square Analysis. In Model 2, the control variable of the number of supporters is included in the model as it is expected to be of great importance based on the observed significant correlations with the dependent as well as the independent variable. Moreover, in Model 3, the industry control variable is included and hence this model offers complete insights for evaluating the first hypothesis. The results of the Logistic Regression Models can be found in Table 4.

(43)

38 Table 4: Logistic Regression Models

The effect of CEO targeting and CEO characteristics on the success of social issues through social media