The relationship between an allegation of a human rights abuse and corporate reputation : the influence of allegation severity and response type

(1)

The relationship between an

allegation of a human rights abuse

and corporate reputation

The influence of allegation severity and response type

Master Thesis

Author: S.A. de Koomen Student number: 6141773

Study: Msc Business Studies - International Management Academic year: 2013-2014

Supervisor: Dr. M.K. Westermann-Behaylo Second reader: Prof. S.B. Rodrigues

(2)

1

List of tables and figures

Tables

Table 1 Descriptive statistics: Corporations during the period 2006 till 2012. p. 35 Table 2 Results T-test mean reputation of corporations with CAA. p. 37 Table 3 Difference in absolute reputation for corporations with CAA. p. 38 Table 4 Difference in relative reputation for corporations with CAA. p. 39 Table 5 Difference in relative difference in reputation for companies with CAA p. 40

compared to companies without CAA.

Table 6 Reputation for companies with environmental CAAs. p. 41 Table 7 Reputation for companies with development and poverty CAAs. p. 41 Table 8 Coefficients and SDs with dRep as dependent variable. p. 43 Table 9 Coefficients and SDs with RdRep as dependent variable. p. 45 Table 10 Coefficients and SDs with DRdRep as dependent variable. p. 48 Table 11 Descriptive statistics of the regression sample. p. 62

Figures

(4)

3

Abstract

This study aims to examine the relationship between allegations of human rights abuses and corporate reputation. It is hypothesized that severity of the allegation has a negative influence on this relationship and response type of the company influences the relationship in either a positive or negative way, depending on the type of response. Results suggest that a Corporate Abuse Allegation (CAA) has a negative effect on reputation. A higher severity of the CAA did not show any significant influence on the relationship between a CAA and reputation. A surprisingly counterintuitive result was suggested related to the influence of response type. Previous research suggested a positive influence on the reputational impact after an apology or plan for change. However, an apology or plan for change showed a negative influence on reputation in this empirical sample, similar to the findings for denial, acknowledgement of the allegation and justification of the action. Whereas a negative influence of denial, acknowledgement and justification were expected, apology and plan for change surprisingly showed the most negative influence on the reputational impact. Implications of these findings and limitations of this study are discussed and suggestions for future research are made.

(5)

4

1. Introduction

Over the past two decades, human rights have become an increasingly pressing issue in international business (Hamann, Sinha, Kapfudzaruwa and Schild, 2009; Cernic, 2010). More and more human rights violations have been recorded from all over the world. Take for example the oil spills from Shell and other oil companies in Nigeria which destroy the environment and pollute the water and ground (Kaufman, 2011). This makes it dangerous for people who live there, because their water and food might be polluted. Another late scandal was the collapse of a factory in Bangladesh, which killed over a thousand people (Burke, 2013). Or do you remember the 90s with Nike’s boot camps in Vietnam in which young girls and women were working for only $1,60 a day (Herbert, 1997)? Since these scandals came to light, people got more concerned with human rights and how corporations act regarding those rights.

In 2005 Professor John Gerard Ruggie (hereafter Ruggie) was named special representative on the issue of human rights and transnational corporations and other business enterprises by the United Nations Secretary-General Kofi Annan (United Nations, 2005). Ruggie was responsible for creating the “protect, respect and remedy” framework; the state’s duty to protect against human rights, the corporation’s duty to respect human rights and the need for effective access to remedy (Ruggie, 2009). His work is known worldwide and was of big influence on the relationship between business and human rights and the awareness of this relationship.

Many scholars have investigated this relationship in a number ways. For example the relationship between human rights and foreign direct investment (FDI) was examined by Blanton and Blanton (2006), who found that human rights are a significant determinant of

(6)

5

the amount of FDI inflows. Others looked at the influence of Corporate Social Performance (CSP) on financial performance and the other way around (Makni, Francoeur and Bellavance, 2008).

According to Drimmer (2010), an allegation of abusing human rights can cause negative publicity that harms operations and future business. He argues that companies should protect themselves against human rights problems by taking some simple steps. These steps include incorporating voluntary sustainability initiatives such as corporate responsibility programs or developing a code of conduct on ethical matters. But stating on paper that you will “do good”, doesn’t always imply that it actually happens.

Hafner-Burton and Tsutsui (2007) did research on country level and found that after countries ratified human rights treaties, the levels of abuse actually increased. Other scholars say that ratification of such treaties is used merely for show, to let the (international) community know you are “doing good” (Vreeland, 2008; Hathaway, 2003). And again others state that countries ratify such treaties purely to align themselves with others who have done the same, to keep up with a legitimated model (Keith 2002; Hafner-Burton and Tsutsui, 2007). Even though evidence shows that the ratification of treaties does not improve the performance on human rights, countries keep using them. That is why they are often referred to as “myths” and “ceremonial treaties” (Cole, 2012; Clark, 2010).

But if the treaties are to show the community that you are doing good, and thereby improve your reputation, then what happens if you sign a treaty as a country and still get accused of human rights abuses? As Cole (2012) argues: “if human rights violations are effectively monitored or enforced, the legitimacy that accrues to disingenuous treaty ratification evaporates”. He is saying that once you get caught on a human rights abuse the

(7)

6

legitimacy of the treaty ratification no longer exists. Then how does this influence a country’s reputation?

Human rights treaties are a country-level commitment to social responsibility, whereas a code of conduct is a company-level instrument of human rights regulation. This is voluntary self-regulation which companies can use to show commitment to Corporate Social Responsibility (CSR) (Kolk and van Tulder, 2002; Campbell, 2006). However, multiple studies on code effectiveness have found either a weak or no significant effect at all (Erwin, 2011), thereby implying that codes are comparable to human rights treaties on the account that they might just be used as ‘window dressing’. Are companies actually becoming more socially responsible or are codes of conduct used as communication strategies (Kolk and van Tulder, 2005)? Erwin (2011) argues that codes of conduct may be merely a symbol of CSR awareness to create a better corporate reputation and legitimize the public image. With a drafted and signed corporate code of conduct in place, what happens if a company is accused of a human rights abuse, does this damage the corporate reputation?

Reputation is an important performance indicator for companies. According to Drimmer (2010) the perceived reputation of a company helps to attract and retain customers and investors. That is why it is important to look at the relationship between allegations of human rights violations and reputation. This relationship might be influenced by other factors. Multiple studies have shown how the severity of a crisis can influence the impact on reputation (Coombs, 1998; Claeys, Cauberghe and Vyncke, 2010). Others showed how a company response strategy might help to decrease the reputational impact (Dutta and Pullig, 2011; Coombs, 2007). The main question leading this thesis will therefore be; how does an allegation of a human rights violation influence a company’s reputation? Is this relationship moderated by the severity of the allegation and the response of the company?

(8)

7

2. Literature review

2.1 Human rights

In the 1990s there was a worldwide expansion of the private sector and a rise in transnational economic activity. This attracted the United Nations’ attention, especially with regard to the impact of businesses on human rights (Ruggie, 2011), because during this time, an ever increasing flow of human rights abuses were reported. These abuses were mainly in the extractive and apparel industries (Ruggie, 2007). A subcommittee from the Commission on Human Rights (the latter hereafter referred to as the Commission), a subsidiary body of experts on human rights, got appointed to create proposals regarding business and human rights. In 2003 they proposed a draft called “Norms on the Responsibilities of Transnational Corporations and Other Business Enterprises with Regard to Human Rights” (Ruggie, 2007). These norms, referred to as the “UN Norms”, were non-voluntary and treaty like regulations set at international level. Businesses were opposed to these legal duties to protect human rights. The Commission did not adopt the initial proposal from its subcommittee, but rather it requested the secretary-general Kofi Annan to appoint a special representative. In 2005 John Ruggie was appointed as the special representative of the secretary-general (SRSG). Ruggie’s mandate was “to “identify and clarify” international standards and policies in relation to business and human rights, elaborate on key concepts including corporate “complicity” and “spheres of influence”, and submit “views and recommendations” for consideration by the Commission” (Ruggie, 2007).

At first Ruggie was asked to stay for two years only, but eventually he has served as Special Representative from 2005 till 2011 (Ruggie, 2011). During his time, he created a framework which rests on three pillars. It is called the “Protect, Respect, and Remedy”

(9)

8

framework. As the name signifies, the first pillar is based on protection. This protection comes from the State, who is obligated to protect the people against human rights abuses by third parties. The second pillar is directed at the corporate society. All business enterprises have the responsibility to respect human rights and they should act with due diligence to avoid human rights abuses. The last pillar is based on the concept of remedy. Although the framework is set out to prevent human rights abuses from happening, the efforts by the states and businesses can unfortunately not prevent all abuse from happening. That is why victims need access to effective judicial and non-judicial remedies (Ruggie, 2011). The “protect, respect and remedy” framework was implemented using recommendations that address how governments and businesses can improve the prevention of human rights abuses. These recommendations are referred to as “Guiding Principles” (Ruggie, 2011).

2.2 Due diligence

The focus of this thesis lies at the firm-level. The firm-level within human rights is embodied by the pillar of respect. All businesses, i.e. national, international, large and small have to respect human rights. Respect however is an ambiguous concept. It is hard for companies to confidently claim or objectify they respect human rights. There was a need for a support system. This is where the process of due diligence was introduced. Due diligence is “a process whereby companies not only ensure compliance with national laws, but also manage the risk of human rights harm with a view to avoiding it” (Ruggie, 2008).

The process of due diligence consists of three parts; awareness of the business’ impact on human rights, prevention of human rights abuses and addressing the adverse human rights impacts. Each part explains the steps a company has to take to fulfill its responsibility to respect human rights (Ruggie, 2008). Ruggie points out four basic elements

(10)

9

that should be included in the process: policies, impact assessment, integration and tracking performance. A firm needs its own human rights policy to give meaning to the specific commitments and provide guidance for its employees. Impact assessment is needed for a company to understand how their processes and activities affect human rights. Integrating the human rights policies is however likely the hardest part. A firm needs not only leadership, but training for the policies to reach the whole company. Last but not least, a firm has to track its performance through monitoring processes. Tracking performance is important for the improvement of the firm’s policies and integration (Ruggie, 2008).

The scope of this process depends on the circumstances a company is in. Ruggie (2008) argues that three sets of factors need to be taken into account; country context, a company’s own activities in that context and the relationships connected to the company’s activities. The country context relates to the (host) country where the company’s activities take place. Different countries may pose different human rights challenges. The second factor is based on the company’s own activities and the impact this might have on the specific context they are in. The company’s presence in a specific context can be of great influence on the employees and the community around them. The last factor focuses on the company’s relationships with third parties such as suppliers or other business partners. Through these relationships a company can influence the third party to also carry out due diligence (Ruggie, 2008).

2.3 Reputation

Before being able to explore the relationship between human rights abuses and corporate reputation, both concepts need to be apprehensible. On that account, what exactly is corporate reputation? Barnett, Jermier and Lafferty (2006) discuss the field of research on

(11)

10

reputation in order to find one unambiguous definition of corporate reputation. In their study they identified three different clusters of definitions of corporate reputation; asset, assessment and awareness. In these clusters they found commonly used terms to define corporate reputation. For example, ‘perceptions’ was a word often used in the awareness cluster. For assessment the main idea is that stakeholders are constantly assessing the status of a firm, in which case corporate reputation is seen as an evaluation or judgment. The last cluster, asset, is different than the others since it acknowledges reputation as a form of significant value to the firm. With this last cluster they do question the applicability of this definition to reputation as they mention it might be more in line with the consequences of reputation (Barnett et al., 2006). Barnett et al. (2006) define corporate reputation in the language of assessment; “Observers’ collective judgments of a corporation based on assessment of the financial, social, and environmental impacts attributed to the corporation over time”. This definition is in line with an earlier research by Fombrun and van Riel (1997). In an overview of the literature on reputation, Fombrun and van Riel point out six different views on reputation; economic, strategic, marketing, organizational, sociological and accounting. Integrating these views, they find that “reputations constitute subjective, collective assessments of the trustworthiness and reliability of firms”.

Multiple studies base reputation solely on economic performance, but this is not the only way to assess firms (Fombrun and Shanley, 1990). This is based on the idea that reputation is a collective assessment of a firm. Economic performance is mainly important for investors of a firm, but there are also other stakeholders that a firm serves. For example its employees, consumers and the overall society it affects. Taking this into account, the reputation of a firm is based on the assessment of multiple stakeholders, and thus multiple

(12)

11

performance measures. The eventual reputation is a cumulative judgment of the different publics (Fombrun and Shanley, 1990).

An important aspect of reputation is social and environmental responsibility. Fombrun and Shanley (1990) argue that being socially responsible can create goodwill from all stakeholders, which can enhance long-run profitability for the firm. In their research they find that being socially responsible has a positive influence on your reputation. Fombrun (2005) even points out that nowadays firms are expected to meet certain standards and that meeting these standards are often beneficial for your reputation. If you do not meet the standards on the other hand, you fail to fulfill the expectations of the stakeholders and it can become a risk to your reputation.

As stated before, Drimmer (2010) noted that the perceived reputation of a company helps to attract and retain customers and investors. This makes reputation an important performance indicator for a company. Not only is a company’s reputation informational, it also functions as a backer of contracts and actions (de Quevedo-Puente, de la Fuente-Sabaté and Delgado-Garcia, 2007; Fombrun and Shanley, 1990). This is based on the idea that when a company establishes a reputation, it has to keep up with the expectations of the stakeholders. If the company fails to do so, it might lose its investors or consumers who are usually attracted to companies with good reputations (de Quevedo-Puente et al., 2007). So if a firm values the reputation they have built, this will be reflected in the actions and reactions of its managers when dealing with social and environmental events (Fombrun and Shanley, 1990). To conclude, reputation is an important driver of success. But, a corporation’s reputation is hardly ever noticed until it is at risk in which case a negative reputation comes to light (Barnett et al., 2006; Fombrun and van Riel, 1997)). The question that then rises is; what should a company do to rectify this negative reputation?

(13)

12

According to de Quevedo-Puente et al. (2007), corporate reputation (CR) is strengthened by an uninterrupted series of positive Corporate Social Performance (CSP). CSP is defined as “the legitimate behavior of the firm with every stakeholder by the standards of the institutional context in each moment of time” (de Quevedo-Puente et al., 2007). They argue that CSR is measured objectively by accounting for a firms past actions and that CR on the other hand is perceptual, which is in line with the definition of CR by Barnett et al. Furthermore, they say that a legitimation process links these two concepts in a way that the past actions of a firm (CSP) are transformed into future expectations (CR). According to Suchman (1995); “Legitimacy is a generalized perception or assumption that the actions of an entity are desirable, proper, or appropriate within some socially constructed system of norms, values, beliefs, and definitions”. This means that society is influenced by the institutional context and expects a firm to behave in a way that matches these expectations.

In trying to fulfill these expectations, firms in the same society become more homogeneous. This phenomenon can be explained by isomorphism. Organizational isomorphism is defined as the similarity between different companies that are operational within the same environment (DiMaggio and Powell, 1983; Deephouse, 1996). DiMaggio and Powell (1983) identify three types of isomorphism; coercive, mimetic and normative. Coercive isomorphism results from formal and informal pressures that are applied by organizations on which the focal company depends and by the society around it. Mimetic isomorphism is driven by uncertainty. When the focal company is unsure about processes, technologies or goals, they often look at the companies around them and mimic their business models. The last one, normative isomorphism, derives from professionalization. Members from a specific profession aim to define the settings of their work in order to establish legitimation for their professional independence (DiMaggio and Powell, 1983).

(14)

13

Different studies show that legitimacy is increased by organizational isomorphism (Deephouse, 1996; Deephouse and Carter, 2005; DiMaggio and Powell, 1983). Deephouse and Carter (2005) take a step further and look at the relationship between isomorphism and legitimacy and the influence of reputation. They find that isomorphism has a positive effect on legitimacy when an organization has a lower reputation, but this is the other way around for companies with better reputations. In that case, isomorphism has a negative effect on legitimacy and reputation.

2.3.1 Corporate Social Responsibility

Over the years, sustainability and social responsibility have become an important debate in many industries. With the extractive industry labeled as “one of the most environmentally and socially disruptive activities undertaken by business” (Jenkins and Yakovleva, 2006), the public targeted the industry and doubted their legitimacy. Due to these pressures and negative publicity for extractive industries, CSR became an important ‘tool’ to objectify their good intentions. As Gilberthorpe and Banks (2012) argue; “the emergence of CSR in the extractive industries represents a bid to legitimize the sector after decades of environmental disasters and the trampling of indigenous rights”.

According to Jenkins and Yakovlova (2006), the progress of CSR has three dimensions; economic development, environmental protection and social cohesion. Economic development can be reached through investments in the host-country to improve the living standards of the communities and ensure future development. Environmental protection is needed especially in the extractive industries. Companies need to make sure not to exploit natural resources and protect the land around their operations to allow successive use. Social cohesion can be attained when companies keep their operations transparent and use

(15)

14

a stakeholder approach. Companies should minimize the disturbances of social and cultural life of the communities (Jenkins and Yakovleva, 2006). However, one cannot forget the ever present need of companies; making a profit.

A company has to find a balance between its shareholders and its stakeholders. In most developing countries there is a lack of governance regulations to protect human rights, this means it is up to the companies to include policies within their corporate governance systems to protect their stakeholders. However, as research points out, corporate governance systems are still mainly focused on shareholders (Lauwo and Otusanya, 2013; Jenkins and Yakovleva, 2006). With this shareholder view in mind, CSR initiatives of companies can be linked to a potentially lucrative opportunity which leads us to the ‘business case for CSR’ (Gilberthorpe and Banks, 2012; Lauwo and Otusanya, 2013). The business case for CSR indicates that as long as it enhances shareholder value it makes sense to initiate CSR. This creates a gap between the company’s alleged commitments to human rights and the actual action it takes to enforce those commitments (Lauwo and Otusanya, 2013). As a consequence it is the corporate sector that benefits from CSR instead of the local communities who need it the most.

The question then rises why initiate CSR processes at all, if you are not intending to improve the development of the local communities or protect the environment and human rights? One explanation might be peer pressure from the industry (Slack, 2012). As Ruggie introduced due diligence into the corporate world, many companies introduced human rights policies. The scope of due diligence reaches to third party relationships as well, so other companies in the industry might pressure the focal company to also carry out due diligence (Ruggie, 2008). With regard to CSR, Gilberthorpe and Banks (2012) argue that “it is adopted on voluntary grounds so it clearly seeks to improve corporate reputations and

(16)

15

influence the nature of the global debates surrounding sustainability and transparency”. Therefore another explanation for including CSR into the business process is to gain the support of the public and improve the corporate reputation.

2.3.2 Code of Conduct

With the emphasis on environmentally and socially just business, many companies implemented a code of conduct and countries signed treaties to show they are doing the right thing. But are they really doing the right thing? Many studies show that over the past few years, more and more treaties have been ratified, there is however no sign of improved performance on account of human rights (Cole, 2012; Hafner-Burton and Tsutsui, 2007). Also on the company-level there remain questions as to whether CSR efforts are a rhetorical rather than practical measure (Gilberthorpe and Banks, 2012). Cole (2012) evaluates four human rights treaties and its effectiveness. His study is on a country-level basis where he looks at the compliance of states on different levels of commitment to treaties, either signing or ratifying a treaty. He concurs that in many situations treaties are merely ceremonial, but he does find that in some cases human rights treaties can have a positive impact on practices. The remainder of this thesis will focus on the company-level of human rights regulation.

Firm self-regulation is the company-level of human rights regulation. Self-regulation is often realized with corporate codes of conduct (CCCs) as their main instrument (Christmann and Taylor, 2006; Kolk and van Tulder, 2002; Campbell, 2006). Self-regulation can be seen as voluntary rule-setting to show commitment and to control its own behavior regarding human rights. This goes beyond the requirements set by the law (Christman and Taylor, 2006; Kolk and van Tulder, 2002). A code of conduct is used by companies as a CSR

(17)

16

tool to establish sustainable and responsible business practices and to communicate these practices to the stakeholders of the company (Erwin, 2011; Kolk and van Tulder, 2002). They are designed to “explicitly detail an organization’s commitment to CSR and outline expected conduct from the organization’s employees” (Erwin, 2011).

Kolk and van Tulder (2002) study the effectiveness of codes of conduct on child labor. They find codes of conduct to be effective when they are specific, strictly implemented and thereafter monitored. Several limitations and side effects of CCCs are found, for example the debate about whether expectations of stakeholders become too high to the extent that corporations are taking over government responsibilities. Kolk and van Tulder (2002) argue however that these limitations are of no harm to the effectiveness of CCCs. Erwin (2011) takes a different approach with regard to code effectiveness. He investigates the influence of the quality and content of the codes on the final result of code effectiveness; ethical performance. His results showed a significant relationship between the quality of codes and CSR performance. Besides this he also finds a significant result which shows that high quality codes were linked with better CSR rankings. Companies with high quality codes were better represented in the top ranking systems for corporate citizenship, sustainability, ethical behavior and public perception (Erwin, 2011). These results indicate that the quality of codes can influence their effectiveness.

According to Campbell (2006), Voluntary Codes of Conduct (VCCs) often lack focus and are inefficient. The rules that are stated are merely seen as aspirations or guidelines and are obscured once concrete procedural issues come along. But he says that if you include human rights in VCCs, companies will be more willing to keep to those promises, because they don’t want the bad publicity that comes with human rights violations. As he notes; “in

(18)

17

moments of reflection people often do want to approve of their own behavior, but this is not carried through into actual conduct without the impact of actual or anticipated disapproval by others”. This means that the fear of disapproval will drive companies and workers to comply with the VCCs. He concludes that codes can work effectively if they are backed up by legal provisions, so that the company’s activities become public knowledge and their actions are open to public scrutiny. This way the VCC is socially pressured which will help the VCCs to be more efficient (Campbell, 2006).

The effectiveness of codes on socially responsible business behavior is still ambiguous. However, many companies have voluntarily adopted a code of conduct. The question is why do companies do this? The reason for this might be that a code of conduct is an easy CSR tool to communicate the good intentions of the company (Erwin, 2011). External organizations and institutions pressure companies to conform to normative standards and due diligence. If you do not, you may harm your legitimacy (Wright and Rwabizambuga, 2006). Adopting a code of conduct is an ‘easy’ way to show your good intentions; this can legitimate your image and improve your corporate reputation (Kolk and van Tulder, 2005; Wright and Rwambizambuga, 2006).

(19)

18

3. Theoretical development

This section will explain the theoretical relationships between specific types of human rights and corporate reputation and how severity of the human rights abuse and corporate response to an allegation can influence that relationship. The research model in figure 1. shows the hypothesized relations, and their possible effects.

(20)

19

3.1 Environmental human rights

In 1948, human rights were defined in the Universal Declaration of Human Rights (UDHR). The UDHR was formulated by the United Nations and includes 30 articles that represent the fundamental basis of human rights (United Nations, 2014). Back in 1948, no specific article on environmental human rights was included. During the 1990s, the relationship between environmental protection and human rights was an ongoing international debate (Shelton, 1991). Theorists and specialists debated whether human rights and environmental protection were based on the same values or whether their fundamental values conflicted. Supporters of the idea that the two concepts were based on the same values suggested that “environmental issues belong within the human rights category, because the goal of environmental protection is to enhance the quality of human life” (Shelton, 1991). On the other hand, opponents argued that “human beings are merely one element of a complex, global ecosystem, which should be preserved for its own sake … human rights are subsumed under the primary objective of protecting nature as a whole” (Shelton, 1991). A third argument, which links the two opposites, suggests that human rights and environmental protection have some overlapping values. In his article, Shelton (1991) offsets the two fields of environmental protection and human rights. He finds that in reality the survival of mankind depends on the protection of the environment and thus the two concepts cannot be separated. He comes to the conclusion that the overlapping values can be used to achieve the objectives of both concepts. He states; “A clearly and narrowly defined international human right to a safe and healthy environment, currently emerging in international law, can contribute to this goal” (Shelton, 1991).

Over the years arguments for the right to a clean environment can been traced back to the UDHR. Article 25.1 states: “Everyone has the right to a standard of living adequate for

(21)

20

the health and well-being of himself and of his family, including food, clothing, housing and medical care and necessary social services, and the right to security in the event of unemployment, sickness, disability, widowhood, old age or other lack of livelihood circumstances beyond his control” (United Nations, 2014). This article comes down to basic rights of subsistence. As Sachs (2004) argues; “If people do not have the basic capabilities to support themselves in dignity, their human rights are under threat”. People have the right to use the environment, so when they are dependent on the environment for their basic needs such as water and food, their subsistence becomes dependent on the environment. When the environment on which people depend is exploited in any way, the subsistence rights of that community are undermined (Sachs, 2004). Subsistence rights are part of basic human rights as written in the UDHR and that is why the right to a clean environment has become part of human rights.

This means that environmental degradation is in violation with human rights. As stated in the literature review, stakeholders expect companies to meet certain standards pertaining human rights. Companies use CSR efforts and codes of conduct to promise such standards to their stakeholders. If companies fail to uphold these promises and thus fail to meet the stakeholders’ expectations; it may harm the reputation of the company (Fombrun, 2005). The importance for companies to respect human rights (including environmental human rights) is reflected by the following hypothesis:

H1: An allegation of violating environmental human rights will have a negative influence on corporate reputation.

(22)

21

3.2 The human right to development

Companies in extractive industries mainly define CSR as “commitments to respect environmental and human rights standards and providing benefits to local communities” (Slack 2012). According to Hilson (2012), the extractive industries have one of the largest environmental footprints and capabilities to influence the wellbeing of society. Very often, the resource-rich countries are developing countries. Hilson argues that if extracting oil and mining projects are managed properly, this could bring wealth to the local community and raise their living standards. Lauwo and Otunsanya (2013) also acknowledge the potential benefits that transnational corporations (TNCs) can have on the lives of people in developing countries. However, they also point out that the operations from a TNC can harm the communities and environment.

Many resource-rich, developing countries are cursed with the so called ‘resource curse’. This implies that the host-countries have not been able to better position themselves to maximize returns from the activities of the extracting companies. They only get nominal taxation payments in return. A reason for this might be that developing countries are under pressure from competing for foreign investments. To get a competitive advantage they lower taxes and soften their laws and regulations (Lauwo and Otusanya, 2013). Furthermore, governments in developing countries are often weak and lack the ability to enforce legislation, which bring companies in a self-regulation position (Hilson, 2012). When firms set up their business abroad, they are no longer tied to their home jurisdiction. In developing countries they can take advantage of the weak social, environmental and human rights regulations and are able to produce at lower costs. This is referred to as exploitation of the governance gap (Lauwo and Otusanya, 2013).

(23)

22

There is no explicit article in the UHRD which defines the human right to development. However, the exploitation of the governance gap is in contradiction with article 29 from the UHRD which states that: 1) “Everyone has duties to the community in which alone the free and full development of his personality is possible”, and 2) “In the exercise of his rights and freedoms, everyone shall be subject only to such limitations as are determined by law solely for the purpose of securing due recognition and respect for the rights and freedoms of others and of meeting the just requirements of morality, public order and the general welfare in a democratic society” (United Nations, 2014). This means that companies who operate in an underdeveloped country should encourage the development (meeting the just requirements of the general welfare) of that country instead of using their underdevelopment to their own advantage.

Development can be defined in two ways: economic development and human development (Donnelly, 1999). Economic development is often defined as sustainable growth in per capita gross domestic product (GDP). Human development is reflected by five aspects: empowerment, co-operation, equity, sustainability and security. Empowerment brings people the freedom to make their own choices that will affect their lives. Co-operation concerns the interaction between people. It is supposed to bring personal fulfillment and a sense of meaning or purpose. Equity refers to more than just income, for example access to an educational system. Sustainability reflects to meeting the needs of this generation, without compromising the needs of the following generation. The final aspect is security. Security is aimed at the security of livelihood. People should be free from threats to their lives, for example in case of illness; people should have access to medical care (Donnelly, 1999). Sengupta (2004) defines development as “a comprehensive process, going

(24)

23

beyond economics to cover social, cultural and political fields and aiming at “constant improvement”, meaning progressive and regular improvement of well-being”.

The human right to development was officially recognized in 1986 when the United Nations adopted the Declaration on the Right to Development (Sengupta, 2004). Instead of exploiting the governance gap, companies should try to regulate their activities in such a way that it can help in the development of the host country and the community they affect, just as their CSR reports suggest. If they do not adopt such regulations, they are in violation of the human right to development. As a violation of human rights is not in line with expectations of stakeholders, the implication of a company with such a violation can impact the reputation of a company in a negative manner. This leads to the following hypothesis:

H2: An allegation of violating the human right to development will have a negative influence on reputation.

3.3 Severity of a crisis

A Crisis can be defined as “an unexpected event that threatens to disrupt an organization’s operations and poses both a financial and a reputational threat” (Coombs, 2007). A wide range of stakeholders can be harmed by a crisis in several ways; physically, emotionally or financially (Coombs, 2007). The severity of a crisis is defined as “the amount of damage generated by a crisis including financial, human and environmental damage” (Coombs and Holladay, 2002).

Zyglidopoulos (2001) investigates the relationship between accidents and corporate reputation. According to him: “Accident severity should be expected to play a crucial role in

(25)

24

the reputational impact of an accident”. His argument is based on an intuitive and emotional base; the greater the damage, or especially when human life is lost, people tend to behave in an intense emotional manner. In 1998, Coombs set up an analytical framework for crisis situations. This framework specifies different types of responses to a crisis. He argues however, that before one can understand the effectiveness of different responses to a crisis, one should analyze the crisis situation.

Coombs (1998) analyzes the crisis situation based on crisis responsibility. Crisis responsibility is affected by three elements: crisis attributions, organizational performance and severity of the crisis. Crisis attributions is about the control and causality of the crisis situation, organizational performance looks at the history of the firm with other crises and severity of the crisis looks at the damage the crisis inflicted. He argues that as the severity of the crisis damage increases, for example when injuries or even deaths are involved, the image of the firm becomes more negative. His findings do not show any significant results, but his assumptions are further analyzed by other scholars.

Claeys et al. (2010) built on Coombs’ work and replicated his study using perceived crisis severity (instead of the amount of property damage and related injuries that Coombs used) to uncover the impact of crisis severity on reputation. They do find support for the assumption that the higher the severity of the crisis, the greater the reputational damage.

Coombs and Holladay (2002) also found support for the severity of a crisis and the negative influence on reputation. They looked at 13 types of crises, two of which could be related to environmental human rights or rights to development; megadamage and organizational misdeed management misconduct. Megadamage is an accident that involves significant environmental damage. Organizational misdeed management misconduct is

(26)

25

spoken of when laws and regulations are knowingly violated by management (Coombs and Holladay, 2002). The severity of such crises leads to the following hypotheses:

H3a: The higher the severity of an environmental human rights abuse allegation, the more negative the impact on corporate reputation will be.

H3b: The higher the severity of an abuse allegation of the human right to development, the more negative the impact on corporate reputation will be.

3.4 Company response

Multiple studies have investigated the influence of company response to a crisis (Coombs, 1998; Coombs and Holladay, 2002; Coombs, 2007; Claeys et al., 2010; Dutta and Pullig, 2011) There are different types of response to a crisis. According to previous research, people intuitively see taking responsibility and promising corrective measures to be the most effective strategies (Dutta and Pullig, 2011). This can be linked to the remedy pillar from Ruggie’s framework. Since mistakes can be made and not all harm can be foreseen, remedy is an important part of the guiding principles. Another type of response is denial. Dutta and Pullig (2011) argue that even though the effectiveness of a response depends on the type of crisis, denial is always the least effective.

Coombs (2007) developed the Situational Crisis Communication Theory (SCCT), a framework that can help managers understand the dynamic of how communication can help to protect the firm’s reputational assets. The SCCT makes use of different ‘crisis clusters’ that represent different types of crisis. The first cluster is the ‘victim cluster’. The victim cluster includes the types of crisis where the firm itself is also a victim (e.g. a natural disaster). The

(27)

26

second cluster is the ‘accidental cluster’. In that cluster all crises that were on an unintentional basis are included (e.g. a technical-error accident). The last cluster involves crisis types in which the firm knowingly took risky actions or violated regulations. This cluster is called the ‘preventable cluster’ and the firm is deemed responsible for the crisis (Coombs, 2007). All crisis types have a negative effect on reputation, but the higher responsibility for the crisis, the more negative the effect; in the victim cluster the firm has no responsibility and in the preventable cluster the firm has full responsibility, the accidental cluster is in between. The different clusters need different crisis response strategies to repair the reputation. Coombs (2007) identified three groups of response strategies; denial, diminish and rebuild. Denial is self-explanatory. Diminishing strategies include justification of the action and excuses; trying to minimize organizational responsibility and perceived damage. Rebuilding strategies include apologizing and compensation responses. In that case the firm takes responsibility and tries to make up for their mistakes by reimbursing the victims for their losses, which can be monetary or in the form of other sorts of gifts (Coombs, 2007).

Human rights violations can be interpreted as a crisis in the preventable cluster. It is a violation of laws and regulations and the company is aware of these rules and thus responsible for respecting these rules. Coombs (2007) found rebuild crisis response to be the best strategy for a preventable crisis. Claeys et al. (2010) also used the SCCT and found that overall, rebuild strategies lead to more positive reputations than diminish strategies. However, they did not find a significant difference on the influence of reputations for the deny strategies.

Meznar and Nigh (1995) used a different approach and allocated reactions of managers to a crisis in to one of two groups; buffering or bridging. A ‘buffering’ response to a crisis would be to dissociate the firm from the blame and responsibility. This would include

(28)

27

responses such as denial and acknowledging the allegation without taking responsibility. A ‘bridging’ response includes accepting responsibility and trying to make up for their mistakes, which are in line with apologizing and making a plan for change. In case of a crisis or negative event (e.g. an allegation of human rights violation) research has shown that a bridging response strategy can reduce the negative effect such an event has on reputation. A buffering response on the other hand does not (Zyglidopoulos, 2001). This leads to the following hypotheses:

H4a: When alleged of an environmental human rights violation, a ‘bridging’ (rebuilding) response will have a positive effect on reputation and a ‘buffering’ (denial or diminishing) response will have a negative effect on reputation.

H4b: When alleged of violating the human right to development, a ‘bridging’ (rebuilding) response will have a positive effect on reputation and a ‘buffering’ (denial or diminishing) response will have a negative effect on reputation.

(29)

28

4. Method

4.1 Measures and data collection

This section will address the dependent, independent, moderating and control variables. It will discuss the different measures that were used for this research and how the data was collected. Furthermore, this section will address the different types of databases that were used, including the CHRD, Fortune’s GMAC list and the Orbis database, which is used for the financial data.

4.1.1 Dependent variable

The dependent variable in this model is the company’s reputation. According to Walker (2010), the most common method to measure corporate reputation is Fortune’s Most Admired Companies. Walker (2010) conducted a research on methods and measurements and found that 39% of the quantitative studies used Fortune’s measure. Most of these researches were based only on US firms and used the American Most Admired Company list. However, since 2006 they started the same concept but then for companies all over the world; Fortune’s Global Most Admired Company list. This is the list that will be used to gather the data on reputation for this research since the companies included are spread globally.

This research relied on the data gathered by Fortune. Since 2006, Fortune conducted a yearly survey on the global list. The respondents of this survey were senior executives, financial analysts and directors of hundreds of selected companies. For example for 2013, 692 companies from 30 countries were selected from the Fortune 1000 companies and Fortune’s Global 500 companies. These 692 companies had revenue of $10 billion or more

(30)

29

and belonged to the 15 or 10 largest companies in the international and U.S. industry respectively. A total of 3920 respondents were asked to rate the companies in their own industry on nine key attributes of reputation. They asked them only to rate companies in their own industry because they are specialized in these areas. The nine key attributes of reputation are; innovation, people management, use of corporate assets, social responsibility, quality of management, financial soundness, long-term investment value, quality of products and/or services and global competitiveness (Fortune, 2014).The average of these nine attributes make up the total reputation score (Flanagan and O’Shaughnessy, 2005). These scores are made publicly available at the beginning of each year. This implicates that the reputation of for example 2007 is influenced by everything that happened during the prior year, 2006.

4.1.2 Independent variable and moderating variables

All independent and moderating variables are retrieved from the Corporations & Human Rights Database (CHRD). The CHRD collects allegations to human rights violations. I was part of the coding team. The coding team was responsible for coding a specific part of the CAAs that are included in the database; oil and mining companies in Africa. The CHRD is a work in progress and once it is finished it will become available to the public. The coding team had access to the portion of the coding that has been completed so far. This made it possible to use that data in this exploratory study.

Corporate Abuse Allegation. The independent variable is the CAA. CAA’s are coded in

the CHRD using a Qualtrics Survey coding tool. Through this tool, the following information about each CAA was collected; company and allegation description, violation characteristics, affected parties, response, legal action, non-legal action, subsequent Business and Human

(31)

30

Rights Resource Centre (BHRRC) updates and sources. The BHRRC is a non-profit organization that draws attention to positive and negative human rights impacts. It has information related to human rights of over 5100 companies throughout the world. In the part about violation characteristics, CAA’s are assigned to one of the following groups; environment, health, labor, development & poverty or abuse. Only in case of an abuse allegation, there was a related violation of one of the four other types of abuse.

The allegations were grouped by country and region by the founders of the CHRD and coders were assigned to a series of allegations in order to find all the above information. Online links were followed to the original articles which stated an allegation was made. The coder was asked to fill in the survey using the original articles and as necessary to conduct internet searches for additional articles that can verify the information or provide updates to the original articles. Throughout the survey there are open questions and multiple choice questions where coders provided any additional data found in their searches. There was however no force response feature. This means that if for some parts of the survey there was no information available, the respondent could leave the question blank and move to the next part.

The CHRD contains a total of 716 allegations for the period 2002 till 2013. For this research, only the CAA’s of the previously stated 14 companies were selected. This limited the database to 160 CAA’s. After this, only the allegations that were made during the period of 2006 till the end of 2012 were selected, because these are the limitations of the reputational data. Also, this research only takes into account allegations that involve the environment or development & poverty, also including the abuse allegations related to one of those two. This left the dataset with 69 CAA’s. These 69 CAA’s are coded by year and

(32)

31

company in the dataset. This means that some companies can have multiple allegations in one year, while others or other years of the same company have zero allegations.

Severity of the CAA. One of the types of allegations is abuse. Abuse is coded

differently, because it has to do with the severity of an allegation. The CAA type abuse is linked to one of the other four types; environment, health, labor or development and poverty. For this research, only the abuse type allegations that are related to environment or development and poverty are included in the dataset. A CAA is coded as an abuse when the violation involves violence, injuries or even death. In this research, an abuse allegation will be coded as a CAAplus.

This variable will be used as a moderating variable, because it is expected that the effect of a CAA on reputation will be worse when the CAA is an abuse type, a CAAplus.

Response type. Another moderating variable is the response type. The BHRRC seeks

responses from companies who received allegations regarding human rights abuses. This ensures that the BHRRC has a balanced coverage of information. Furthermore it provides an opportunity for the companies to present their responses and it encourages them to address the alleged violations. The CHRD includes information on the different types of responses the companies had to the allegations. If a company responded to the allegation, the responses in the CHRD are coded as a denial, acknowledgement of allegation, justification of action, apology, plan for change or other. In case of a denial, there were four subcategories; the claim was exaggerated, the claim does not take into account all of the facts, the claim was blatantly false or other. In case of ‘other’ or ‘denial: other’, the coder was able to fill in extra information in a text box.

The responses are coded per company, per year. This means that for some companies who have multiple CAAs in one year, there can be multiple response types. It is

(33)

32

expected that the response type will influence the relation between a CAA and the reputation of the company. To reduce the complexity of the statistical tests, the response types are categorized in three groups, according to the sort of influence it will have on the reputation.

Response type A will include the ‘bridging’ response types; apology and plan for change. It is expected that when companies admit they were wrong and try to compensate for that wrongdoing, people are more forgiving. This implicates that a type A response can influence the relationship between a CAA and the company’s reputation in a positive way.

Response type B will include the different types of denials and no response at all. These types of response are worse for your reputation than type A responses. When companies do not respond to an allegation, it is like it does not exist to them. In other words; they deny it is there. It is expected that this will have a slightly negative influence on the relationship between a CAA and the reputation of the company.

The last and supposedly ‘worst’ response group is type C; the diminishing strategy. Type C includes the responses acknowledgement of allegation and justification of the action. These two response types acknowledge that they were violating human rights, but they are fine with it. They do not try to make up for their wrongdoings. With justification of the action a company even claims their actions were just. It is expected that type C responses will have a negative influence on the relation between a CAA and the company’s reputation.

The response type ‘other’ is treated differently. The information in the text box will provide a basis for which group it will be appointed to. The extra information will be compared with the different types of responses and be allocated to the type most suitable. There are also some companies that used multiple response strategies in one response to a single allegation. For example when a company first seems to deny the violation, but in the

(34)

33

end does present plans for better communications with the community or clean-up plans for the environment, this is coded as both denial and plan for change (the coder can select multiple response types per allegation). The ‘best’ response will be coded in the dataset, in this case a plan for change which is a type A response. A final remark is for the code UTD, which stands for unable to determine. When the response was coded with UTD, the coder did find something that indicated the company responded to the allegation, however was not able to find out what the response was. The code UTD will be treated as a no response. An overview of the response types can be seen in table 11 (Appendix).

4.1.3 Control variables

To account for other variables influencing reputation, the model is controlled for the size of the firm, financial performance and the industry. These are all commonly used control variables when dealing with research on reputation (Flanagan and O’Shaughnessy, 2005; Zyglidopoulos, 2001; Walker, 2010)

Firm size. According to Flanagan and O’Shaughnessy (2005), smaller firms have less

stable reputations than larger firms. This is mainly due to visibility. Reputation is built on the visions that people have about the firm. If this visibility is small, less people observe the behavior and this will result in a thin reputation which is more easily shattered than a thick reputation (Flanagan and O’Shaughnessy, 2005). In this model, firm size is measured by total net sales of the company in a specific year. This is measure has proven adequate in previous research (Roberts and Dowling, 2002).

Financial performance. Previous research has shown that the relationship between

financial performance and reputation works in two ways. A firm’s financial performance is influenced by its reputation, but reputation is also influenced by the firm’s prior financial

(35)

34

performance (Roberts and Dowling, 2002). Return on assets (ROA) is a common measure for financial performance in previous literature (Flanagan and O’Shaughnessy, 2005; Roberts and Dowling, 2002; Zyglidopoulos, 2001), and will be used in this research as well.

Industry. This research examines the reputations of firms in two industries; the oil

industry and the mining industry. A dummy variable will control for this difference (oil industry = 1, mining industry = 0).

All of the financial data was obtained from the Orbis database.

4.2 Sample

For this study, data was gathered from the Corporations & Human Rights Database (CHRD) and Fortune’s Global Most Admired Company (GMAC) list. The CHRD includes data starting from 2002, while Fortune only started the GMAC list in 2006. This research only takes into account companies from the oil and mining industry. First of all, for companies to be eligible for this research, reputational data had to be available for at least 5 consecutive years during the period of 2006 till 2013. The GMAC only includes data between 2006 and 2013, and for some companies data was missing for several years. To be able to compare the reputational data from one year to the next, consecutive years of reputational data was needed. This data was gathered from the online Fortune GMAC lists. Only twenty companies met these requirements and remained in the dataset. The next step was to cross reference the remaining companies from the Fortune GMAC list with the companies in the CHRD database that had an allegation during the period of January 2006 till December 2012. Any allegations beyond this date (in 2013) would not have had an influence on the reputation of 2013, since the value for reputation is formed at the beginning of each year. Fourteen companies met these requirements. The other six companies that did have enough reputational data, but

(36)

35

did not have an allegation during that specific period will be used as a control group. Table 1 shows the final sample and some descriptive statistics.

Table 1. Descriptive statistics: Corporations during the period 2006 till 2012

Corporation

2006-2012 Mean rep Min rep Max rep CAAs total CAAplus CAA ENV CAA D&P

Repsol 5,45 4,75 6,28 10 0 6 4

Royal Dutch Shell 7,43 7,05 7,98 8 2 5 1

ExxonMobil 7,88 7,36 8,24 5 0 3 2 Chevron 7,60 7,02 7,93 7 1 5 1 Total 6,80 6,62 7,03 3 0 1 2 Rio Tinto 6,40 6,03 6,81 2 0 2 0 Petrobras 6,00 5,70 6,95 9 3 4 2 Occidental 6,37 5,07 6,94 2 0 2 0 ENI 5,18 4,73 5,77 0 1 0 0 ConocoPhillips 6,60 4,99 8,17 1 1 0 1 BHP Billiton 5,10 4,81 5,85 9 2 6 1 Anglo American 6,80 6,31 7,73 9 4 1 4 EnCana 6,15 5,30 6,68 0 0 0 0 China National Petroleum 5,77 5,30 6,30 1 0 0 1 BP 7,74 7,27 8,34 1 0 1 0 Sinopec 6,44 5,90 7,09 1 0 1 0 Statoil 5,29 4,72 5,95 0 0 0 0 Valero 6,06 5,67 6,46 0 0 0 0 Devon 7,13 6,64 7,44 0 0 0 0 Oil&NaturalGas 5,37 4,19 6,00 0 0 0 0 Total 68 14 37 19

The sample is arranged as panel data. Panel data is a combination of cross-sectional data and time-series or longitudinal data. Cross-sectional data looks at multiple subjects at one point in time, whereas time-series data looks at a single subject but follows the changes

(37)

36

over a course of time. This study is combined and examines the changes of multiple subjects over the course of 2006 till 2013.

4.3 Data analysis

To test the hypotheses quantitative analysis will be used. Different statistical techniques will be applied to analyze the collected data. To start, multiple T-tests have been conducted, paired sample T-tests and one sample T-tests. This was to measure the first hypothesis. After this a regression analysis is used to test the moderator effects on the model. It is argued that both an ordinary least squares regression and a change score approach can be used (Flanagan and O’Shaughnessy, 2005). However, the SPSS build in guide notes that OLS regression can only be used in case of a bi-directional variable. This is not the case for this model, so a change score approach will be used in which case the dependent variable is the difference in reputation between years t and t-1, called dRep. The variable dRep is the absolute difference between two consecutive years, while another variable, RdRep is used to look at the relative difference. The next section will discuss the analyses and its results.

(38)

37

5. Results

5.1 T-test

Hypothesis 1 and 2 predict that a CAA will have a negative influence on a firm’s reputation. To be able to test this, the firms were divided into two groups per year; a group of firms that had a CAA against them that year (CAA=1) and a group of firms that did not (CAA=0). Using a paired sample T-test, the mean reputation in one year of all the firms that had a CAA could be compared with the mean reputation of those firms in the next year. This test was repeated for all 7 years (see table 2). Table 2 shows that 4 out of the 7 years the reputation dropped. However, only in the year 2006 was this drop significant (using p<0,05). This drop in reputation is measured using an absolute scale.

Table 2. Results T-test mean reputation of corporations with CAA

Year Corps (n=) Mean Rep

year x-1 Mean Rep year x Diff. mean rep p-value

2006 5 6,77 6,62 -0,156 0,011 2007 8 6,55 6,74 0,185 0,079 2008 4 6,53 6,80 0,270 0,319 2009 8 6,13 6,41 0,284 0,280 2010 7 6,52 6,48 -0,043 0,856 2011 6 6,84 6,65 -0,187 0,399 2012 5 6,72 6,69 -0,036 0,885

From the difference in mean reputation between the consecutive years, the variable dRep was constructed. Using this variable, a new T-test was conducted. A one sample T-test that compares the difference in reputation (dRep) of the companies with a CAA against the difference in reputation (dRep) for companies without a CAA. In Table 3 you can see the outcomes of the one sample T-test for the years 2006 till 2012. The variable dRep represents the difference in reputation for the companies with a CAA and the variable Repcoefficient

(39)

38

represents the difference in reputation for the companies without a CAA. In the table you can see a surprising indiction; for the years 2007, 2008, 2009 and 2011, the difference in reputation for companies with a CAA show a more positive result than for companies without a CAA. Again there are no statistically significant outcomes, only the year 2012 approaches the 0,05 p-value.

Table 3. Difference in absolute reputation for corporations with CAA

year n= dRep Repcoefficient p-value

2006 5 -0,156 -0,149 0,851 2007 8 0,185 0,147 0,686 2008 4 0,270 -0,227 0,116 2009 8 0,284 -0,103 0,155 2010 7 -0,043 0,366 0,120 2011 6 -0,187 -0,560 0,125 2012 5 -0,036 0,578 0,058

All of the previous tests used the absolute differences in reputation. Another way to examine the relationship between a CAA and reputation is by using the relative difference in reputation. Table 4 shows the relative difference in reputation (RdRep) for companies with CAA’s. A one sample T-test calculates the difference in reputation from the previous year with the year in question. When using relative numbers, this means that the previous year will be stated as the full 100% that has been achieved (RdRep yr. T-1) and the year in question (RdRep Yr. T0) shows the achieved reputation relatively to the previous year. The results of this test show a difference with table 2, where reputation dropped 4 out of 7 years. In Table D you can see that only 3 out of 7 years show a drop in relative reputation. Only the year 2006 shows a significant drop. This is consistent with the results shown in table 2.

(40)

39

Table 4. Difference in relative reputation for corporations with CAA.

Year Corps (n=) RdRep Yr.

T-1 RdRep Yr. T0 (mean) Diff. mean rep p-value

2006 5 1,00 0,976 -2,45% 0,025 2007 8 1,00 1,033 3,32% 0,064 2008 4 1,00 1,052 5,21% 0,319 2009 8 1,00 1,061 6,07% 0,168 2010 7 1,00 0,995 -0,48% 0,902 2011 6 1,00 0,969 -3,09% 0,349 2012 5 1,00 1,003 0,25% 0,953

Table 5 shows the results from a one sample T-tests for the years 2006 till 2012. This test compares the difference in relative difference in reputation for the groups with CAAs with the difference in relative difference in reputation for the groups without CAAs. The difference between these groups is referred to as the DRdRep. DRdRep is expressed in the percentage of growth (either positive or negative) in reputation for companies with CAAs compared to the ‘standard’ growth of the group without CAA’s. It is calculated as follows:

DRdRep= 1-[(RdRep CAA=0)-(RdRep CAA=1)]

The group without CAAs (CAA=0) functions as a control group. The results suggest that companies with CAA in the years 2006 and 2011 had a negative growth in reputation; 95,18% (p=0,024) and 95,48% (p=0,455) respectively. Only the year 2006 was significant (p=0,024). Companies in the years 2007, 2008, 2009, 2010 and 2012 again indicate growth in spite of the CAAs; 104,42%, 107,39%, 104,73%, 104, 68% and 103,60% respectively. However, none of these were statistically significant.

(41)

40

Table 5. Difference in relative difference in reputation for companies with CAA compared to companies without CAA.

year Corps (n=) Rep year x-1 DRdRep p-value

2006 5 1,00 0,9518 0,0240 2007 8 1,00 1,0442 0,2740 2008 4 1,00 1,0739 0,2690 2009 8 1,00 1,0473 0,4940 2010 7 1,00 1,0468 0,3870 2011 6 1,00 0,9548 0,4550 2012 5 1,00 1,0360 0,8550

All of the above tests show mixed results when examining the relationship between a CAA and a company’s reputation. There is little support for the overall prediction that a CAA will have a negative influence on reputation. To test hypotheses H1 and H2 specifically, all companies that did not have an allegation against them during the specific year were excluded from the sample. Companies that had multiple CAAs against them during one specific year, but these CAAs were not all of the same type, were also excluded from the sample. This left the sample with only a few cases per year (see tables 6 and 7 for the exact numbers).

First H1 was tested. Hypothesis 1 predicts a negative relationship between an environmental CAA and reputation. For this, only the corporations that had a CAA regarding the environment were used to conduct a paired sample t test. Table 6 shows the results; only in 2006 and 2012 the reputation of the companies dropped after an environmental CAA. For the remaining years, the reputation of the companies increased. For the years 2011 and 2012 there was only one environmental allegation and thus significance could not be calculated. For the years 2006 till 2010 none of the results were significant. This means that H1 is not supported.

(42)

41

Secondly H2 was tested using the same test. Hypothesis 2 predicts a negative relationship between development and poverty CAAs and reputation, so this time only the CAAs regarding development and poverty were included. Table 7 shows that for the years 2006, 2010, 2011 and 2012 there was a drop in reputation. However also this test shows no significant results and H2 is therefore not supported.

Table 6. Reputation for companies with environmental CAAs.

year Corps (n=) Rep year t-1 Rep year t Diff. mean

rep p-value 2006 3 6,367 6,207 -0,160 0,107 2007 5 7,296 7,368 0,072 0,542 2008 2 6,400 6,755 0,355 0,627 2009 3 6,437 6,830 0,393 0,569 2010 5 6,560 6,612 0,052 0,862 2011 1 7,240 7,930 0,690 x (n<1) 2012 1 6,360 6,330 -0,030 x (n<1)

Table 7. Reputation for companies with development and poverty CAAs.

year Corps (n=) Rep year t-1 Rep year t Diff. mean

rep p-value 2006 1 6,840 6,740 -0,100 x (n<1) 2007 2 4,925 5,325 0,400 0,228 2008 1 6,040 6,140 0,100 x (n<1) 2009 4 5,950 6,255 0,305 0,378 2010 1 5,950 5,220 -0,730 x (n<1) 2011 2 6,920 6,675 -0,245 0,377 2012 2 6,675 6,515 -0,160 0,817 5.2 Regression analyses

A regression analysis was conducted to examine the causal relationships between the variables. Different dependent variables were used, respectively dRep, RdRep and DRdRep.

The relationship between an allegation of a human rights abuse and corporate reputation : the influence of allegation severity and response type