Consumer responses to corporate social irresponsibility : the role of relatedness between the crisis and the company’s reputation

(1)

Consumer responses to corporate social

irresponsibility

The role of relatedness between the crisis and the company’s reputation

Master thesis Melody Baeriswyl

Supervisor: Marlene Vock Kennewegsteeg 4B

Second reader: 2312CZ Leiden

Date: 5882958

06 12051529

(2)

Abstract

Corporate social responsibility (CSR) has been researched a lot in the past few years. Recently there is a new phenomena which is becoming more important, corporate social irresponsibility. Behaving in a corporate social irresponsible manner, can have serious consequences for companies. This research investigates the possible negative effects as a result of behaving irresponsible and the role of the CSR reputation. A favorable CSR reputation is important for companies, because it leads to positive consumer attitudes, higher buying intentions, customer identification and positive word of mouth. In contrast, a negative CSR reputation becomes less and less acceptable in an era of critical, aware and informed consumers. It seems like creating a positive CSR reputation is the key to success. But what happens when a company with a good CSR reputation finds itself in a CSR crisis? Will the good reputation be able to protect the company from the damaging effects of the crisis? The outcomes of this research show, that a good CSR reputation can indeed protect the company. Furthermore the results will also give insights under what circumstances this ‘protection effect’ occurs and how it works for companies with a negative CSR reputations.

(3)

1. Introduction

1.1. Corporate Social Irresponsibility (CSI)

Corporations are increasingly encouraged by stakeholders to behave socially responsible. Today most researchers and also more and more managers agree that it is not whether corporate managers should consider the needs of society but to what extent they should do this (Bertels & Peloza, 2008; Lii & Lee, 2011; Wilson, 2003). Being accused of CSI is not something that only happens to organizations without any CSR activities or organizations that are considered irresponsible overall. Cases of corporate social irresponsibility (CSI) can happen to all companies, even if their operations in general are responsible. H&M is an example of an organization that had to deal with a lot of CSI accusations throughout the past few years. In 2010 the Volkskrant published an item filled with critique about the irresponsible practices of H&M like, the use of child labor, not paying enough taxes to their host countries, underpaying employees and the presence of serious safety issues (Volkskrant, 2010). Another example of an organization that had to cope with a great amount of negative media attention is BP. Think about the oil spill in 2010 and the skeptic news items that followed (het Financieele Dagblad, 2010).

Worldwide business scandals involving big high profile organizations represented in the news, like in the two examples above, led to organizations finding themselves in serious CSR crises. Eventually, a CSR crisis or CSI leads to bad consumer confidence (Leonard & McAdam, 2003). Consumers have more positive attitudes towards socially responsible brands than to brands with irresponsible behavior or brands with unknown CSR behavior. Furthermore, literature shows that consumers dealing with socially irresponsible corporate brands are less likely to reward and more likely to punish the brands (Sweetin, Knowles, Summey & McQueen, 2013). Past research showed that CSI results in negative emotional reactions like anger, sadness and disgust, which eventually leads to consumer actions like negative word of mouth and protest behaviors (Grappi, Romani & Bagozzi, 2013). But still, there is a lot more research and knowledge about positive consumer reactions and attitudes to CSR than negative consumer reactions and attitudes to CSI. Also the role of a company’s prior CSR reputation in a CSR crisis and its effect on consumers’ attitudes has rarely been addressed in existing research about CSI/CSR crisis. Thinking about the role of a company’s CSR reputation in a crisis raises the question if a company’s positive CSR reputation could serve as a buffer or protection when CSI occurs?

(6)

1.2. The role of a company’s CSR reputation

In general, a good reputation can lead to better customer loyalty, increased customer identification with the company, the ability to charge premium prices and attract better human resources (Lii & Lee, 2011). For this reason, the importance of developing a good reputation is well recognized by firms (Bertels & Peloza, 2008; Lii & Lee, 2011). Already more than 80% of companies that are in the Fortune 500 mention CSR issues on their websites and there are also the companies that are participating in CSR initiatives without mentioning it. This probably reflects the belief that being involved in CSR activities does lead to better a reputation (Lii & Lee, 2011).

According to Einwiller, Fedorikhin, Johnson and Kamins (2006) highlighting CSR aspects that appeal to stakeholders can enhance a favorable CSR reputation and enhance consumer’s identification with the organization. Companies like Starbucks and Tom’s are examples of companies that use a CSR strategy and highlight their strong environmental stances in order to achieve a good CSR reputation. But a good CSR reputation does not protect companies from getting into a CSR crisis. An example of a company with a good CSR reputation that was accused of unethical practices is Toyota. They are known for their environmental friendly cars, but still found themselves in a CSR crisis after ignoring safety issues (NRC Handelsbald, 2014) and conflicts with their employees in developing countries (Financiële Dagblad, 2014). Shell is an example of a company that has a bad CSR reputation despite of the fact that they are trying to improve this reputation by being involved in a lot of CSR initiatives. Their involvement in CSR crises and the constant stream of negative media attention caused them to have a bad CSR reputation. Even though, a good CSR reputation cannot prevent a CSR crisis, results showed that identification as a result of good CSR reputations can have a buffering effect in times of crisis, unless the conditions are extremely negative (Coombs & Holladay, 2006; Einwiller, et al., 2006). So a CSR reputation can protect companies in a crisis. However, there has not been a lot of research about the buffer effect yet and the ‘buffer’ effect has not been researched in the context of CSR related crises in connection with the company’s prior CSR reputation. If the CSR crisis is related/unrelated to the company’s prior CSR reputation could be topic related (like environmental CSR reputation and environmental crisis) or related in a sense that the CSR reputation is negative just like the crisis. In this thesis these contexts will be researched to find out how strong the ‘buffer’ effect is when the crisis is CSR related and when it is unrelated.

(7)

1.3. A related vs an unrelated CSR crisis

What if the crisis is related to the company’s CSR reputation? An example would be Starbucks that is known for their Fairtrade coffee which ‘guarantees’ fair treatment of the coffee farmers. If they would be in a crisis due to bad working conditions of their coffee farmers, the crisis would be directly related to their CSR reputation. Would the buffer effect of a positive CSR reputation still work in such a related crisis? Or would they be better off with a crisis unrelated to their reputation, like environmental pollution? Fennis and Stroebe (2014) show in their research that the buffer effect can occur in the case of self-disclosure. Their results show that a company’s self-disclosure can help companies with a poor reputation to promote positive effects on consumers’ judgment. However, Fennis and Stroebe (2014) only focused on the general reputation of companies. This general reputation was based on financial value and market value and not on CSR. Also they mainly looked at the halo/buffer effect of self-disclosure and not if a good reputation can serve as halo or buffer effect. The relatedness of Fennis and Stroebe (2014) is about relatedness to consumer expectations and not relatedness between the reputation and the crisis. In this thesis the focus will not be on the general reputation but the CSR related reputation which is dependent of a company’s CSR behavior and not their financial value or market value. Even though it is interesting that the buffer effect of the self-disclosure only worked when the prior CSR reputation was negative, this does not say anything about the buffer effect of the CSR reputation. In this thesis the buffer effect of the CSR reputation will be researched.

1.4. Problem definition

The purpose of this master thesis is to find out what the impact of an existing CSR reputation is on consumers attitudes in a CSR crisis, and how relatedness of the crisis to the existing reputation influences this effect. Also the mediating role of consumer emotions will be researched. Thus, this leads to the following main research question:

Do consumers’ responses to a CSR crisis differ for companies with a positive versus a negative CSR reputation? And do these responses depend on the level of relatedness between the CSR reputation and the crisis?

This will lead to an interesting research because it implies that the relatedness or un-relatedness of a CSR crisis and the existing CSR reputation can make a big difference on consumer attitudes. This would mean that it is not only about which crisis strategy or CSR

(8)

strategy is used but that a large part of the positive or negative outcome is due to the nature of the crisis and if this is related to the CSR reputation. If this is true, it would be no longer a given that a good prior CSR reputation leads to more favorable outcomes after a crisis. The goal of this research is to gain useful insights about the relatedness between reputation and the crisis. The following research questions can be addressed:

1. What is corporate social responsibility? And how do consumers respond toward CSR activities?

2. What is a CSR crisis? And how do consumers respond to irresponsible CSR behavior? 3. Can a positive CSR reputation serve as a buffer during a CSR crisis?

4. How can related versus unrelated CSR influence consumers?

5. What is the role of consumer emotions on consumer attitudes in response to CSR and CSI?

By conducting an experiment through online questionnaires the research questions will be answered and the results will show how the relatedness of CSR reputation and a CSR crisis influence consumer attitudes towards the company or brand.

1.5. Theoretical contribution

CSR in general but also CSR reputation, CSR crises and relatedness of CSR are popular topics in literature. A general conclusion is that a favorable prior CSR reputation can lower the negative impact of the crisis on the post-crisis reputation and consumer attitudes due to the so called ‘halo’ or ‘buffer’ effect (Coombs & Holladay, 2006; Einwiller, et al., 2006). According to Wagner, Bicen and Hall (2007), there is still a lot to be discovered about conumers’ negative CSR perceptions. So far, in spite of demonstrated significance of the consequences of a company’s irresponsible behavior, the literature has mainly focused on outcomes of responsible behavior (Lange & Washburn, 2012). Previous literature already contains research about CSR in relation to corporate reputation but so far, there has been no research on the effects of reputations that are based on CSR. Furthermore, the ‘buffer’ effect a relatively new research topic and has yet not been addresses in the current context. This research will focus on the possible negative and positive effects of related versus unrelated CSR crises and how relatedness influences the ‘halo’ or ‘buffer’ effect. Current literature will be completed by this new.

(9)

1.6. Managerial contribution

Since it is widely accepted by organizations that it is important to develop a good reputation (Bertels & Peloza,2008; Lii & Lee, 2011), it would be useful to know how the effect of a good or bad reputation can be influenced by the relatedness of a crisis. Managers need to protect the organization’s reputation from the possible harm a crisis can cause, hereby, the effect of the relatedness of the crisis should be taken in consideration when mapping out a strategy. This research will give managers an indication to what level of harm, on consumer attitudes, a crisis might lead. Also if the strong effect of a good CSR reputation still exists in the new context of this research is important to know for managers, it tells them how much they can trust on the effect of their favorable prior reputation. But it could also reveal that an unfavorable CSR reputation might not have worse effect on consumers attitude than a favorable reputation. With this information managers can adjust their strategies and decide how important their CSR reputation is in a crisis. The results will be useful for any company that is coping with crises or bad media attention.

Overview and focus of the rest of the thesis:

In chapter two to five the sub-research questions mentioned earlier in chapter 1.4 will be addressed. These three chapters represent the theoretical framework of this research. Previous research that addresses the relevant topics to answer these questions will be analyzed. Chapter two will start with some general information about corporate social responsibility and the way consumers react to it. In chapter three the same will be done but this time for corporate social irresponsibility. Chapter four will introduce the idea of the ‘halo/buffer’ effect of a favorable CSR reputation and the role of the related versus the unrelated crisis. Also in chapter four, the first few hypotheses will be presented. Chapter five, the final chapter of the theoretical framework, will elaborate on the role of consumer emotions. In the remaining chapters, six to nine, the method and the results of the pre-test and the actual experiment will be presented.

(10)

2. What is corporate social responsibility? And how do consumers respond

toward CSR activities?

2.1. What is CSR?

Only few notions have captured the corporate imagination in a way corporate social responsibility (CSR) has. Now, companies are supporting CSR initiatives more than ever and with greater financial and marketing input (Bhattacharya & Sen, 2004). Researchers argue that being socially responsible is not only the right thing to do, but it also leads to doing better (Bhattacharya & Sen, 2004; Lindgreen & Swaen, 2010). But what is meant by CSR? The meaning of corporate social responsibility can be different depending on the context in which it is used or explained and different associations people might have with it. CSR includes issues like human rights, environmental aspects, unfair business practices, community involvement, social development and workplace issues (Leonard & McAdam, 2003).

Carriga and Melé (2001) developed a classification system that groups the main CSR theories. They did this to bring some clarity and to show that theories share similarities which makes it able to group them. One group of theories falls under instrumental theories, here the corporation is viewed as an instrument for wealth creation. Only the economic aspect is considered and wealth creation is the only social responsibility, this means CSR is viewed as an instrument to reach the goal of better profits (Carriga and Melé, 2001). It is safe to say that this group of theories is probably outdated because today’s companies are highly pressured by their stakeholder to behave responsible. A second group of CSR theories falls under political theories where the social power of the corporation is emphasized, specifically power in relationship with society and its responsibility in the political field. As a result corporations accept social duties or participate in a social cooperation. The third group consists of integrative theories which is about social demands. These theories argue that business depends on society for growth and even for the existence of the business itself. The last and fourth group falls under ethical theories. Here, CSR is viewed as a relationship between business and society (Carriga & Melé, 2001). This fits the perspective of Moir (2001) who argues that business and society are independent of each other when it comes to their needs for a certain environment, in other words, business and society are interwoven rather than opposite entities. The interdependence between those two entities and their shared needs resulted in CSR becoming more and more popular. Now that business and society become

(11)

connected, for some companies the right thing for business and being ethical become one and the same (Leonard & McAdam, 2003). This is in line with the idea of Bhattacharya and Sen (2004) and Lindgreen and Swaen (2010) that being ethical is the right thing to do and the right thing for business because it leads to doing better.

To explain what is really meant by CSR, researchers developed many definitions. In this thesis Carroll’s definition will be used because it is considered as the most suitable in general and also for this thesis for two reasons: One, Carroll’s (1979) definition is used in many articles discussed in this thesis. And two, it is about the stakeholders’ expectations and it acknowledges the fact that those expectations and thus CSR is not static but always changing.

“The social responsibility of business encompasses the economic, legal, ethical, and

discretionary expectations that society has of organizations at a given point in time” (Carroll, 1979, p.500).

This definition is based on normative arguments and states that a company has to fulfill those four main responsibilities (Podnar & Golob, 2007). According to Carroll (1979), to have a definition that embodies the entire range of obligations businesses have to society it must include the economic, legal, ethical and discretionary categories. The four groups of CSR theories discussed before by Carriga and Melé (2001) can be recognized in the four categories Carroll (1979) distinguishes. The basis of businesses is economic and this is therefore an important social responsibility. A business is responsible to produce goods and services to serve a society’s wants and needs and sell these with profit. But there are also ground rules under which business is expected or has to operate. Society expects businesses to operate within the legal framework. Ethical responsibilities are not always codified into law but society expects businesses to behave according ethical norms. The last dimension of Carroll’s (1979) definition is discretionary responsibilities. Society has no clear message to business and discretionary responsibilities are left to individual judgment and choice. Even though they are at business discretion and not directly demanded by society they can still be called responsibilities because society expects businesses to take social roles above the previous three described. But it is voluntary and the decision to participate is left to businesses and their desire to engage in CSR. It seems like discretionary responsibilities offer therefore the best way to differentiate. The fact that CSR could offer ways to differentiate led to researchers increasingly examining the strategic role of CSR in organizations instead of simply identifying CSR activities (McWilliams, Siegel & Wright, 2006).

(12)

2.2 Consumers’ responses to CSR activities

Customers form one of the most important stakeholder groups and appear to be particularly susceptible to company’s CSR initiatives. General marketplace polls show a positive relation between CSR initiatives and consumers’ responses to the company (Sen & Bhattacharya, 2004). But research found that the relationship between CSR activities and consumer reactions is not straightforward and that there are several factors that affect this relationship. Sen & Bhattacharya (2001) found two factors as key moderators of consumers’ responses to CSR initiatives, company-specific like the CSR domain or product quality and individual-specific like CSR-related beliefs. The corporate social responsibility – corporate ability (CSR-CA) belief is an important moderator and is about consumers’ beliefs about the relationship between CSR and a company’s ability to make quality products. Their findings show that the beliefs about the trade-offs a company makes between its CSR and CA efforts can play a central role in the reactions to CSR. When consumers perceive the CSR initiatives as damaging for CA, the CSR initiatives will not be perceived as positive. Furthermore, the results of Sen & Bhattacharya (2001) highlight the mediating role played by consumers’ perceptions of relatedness between their own characters and that of the company, also called ‘consumer-company identification’. If consumers feel connected to the CSR of a company and feel that it relates their own characters they will respond more positively than in the case of low congruence between their character and a company’s CSR. Not only level of congruence plays a role in consumers’ attitudes but also positive or negative CSR associations in general. Positive CSR associations can enhance product evaluations and negative CSR associations have a detrimental effect on overall product evaluations. What is notable is that consumers’ company evaluations are more sensitive to negative CSR information. Only if consumers are highly supportive of the CSR issue they react positively to positive CSR information while all consumers react negatively to negative CSR information. This means that corporate social irresponsibility (CSI) can have a significant impact on consumers’ perceptions about a company and it is extremely important for companies to act carefully.

Green and Peloza (2011) found another influence on consumer responses. Consumer responses to CSR depend on the type of CSR or in other words how CSR is manifested (Green & Peloza, 2011). CSR can be manifested as environmental responsible or as product responsible and this leads to different outcomes on how the CSR actions are perceived by consumers. Bhattacharya and Sen (2004), use the ‘Socrates: The corporate social ratings

(13)

monitor’ database to divide different CSR actions undertaken by 600 companies into six broad domains:

- Community support: housing initiatives, education, support of arts and health programs.

- Diversity: family, race, gender.

- Employee support: employee involvement and protection, profit-sharing, job security. - Environment: waste management, pollution control, recycling.

- Non-US operations: overseas labor practices, operations in countries with human rights violations.

- Product: product safety, marketing controversies, innovation, R&D.

(Bhattacharya & Sen, 2004).

Each domain can have different outcomes depending on what consumers consider to be CSR-related activities. This means that there is a significant heterogeneity in consumers’ responses toward CSR initiatives and different CSR reputations companies have. What works for one consumer does not always work for the other. This shows that the type of CSR reputation influences consumers’ perceptions. The role of a CSR reputation will be further expanded in this thesis. Because consumers react so differently to different kinds of CSR reputations companies must ascertain clearly and concretely the value of the resources devoted to CSR activities and keep in mind that the desired outcomes might be difficult to achieve and depend on a lot of factors (Bhattacharya & Sen, 2004). Also, for initiatives to provide the desired returns to the company, the initiatives must first provide a return to individual stakeholders. Stakeholders responds to CSR initiatives based on the amount of personal benefits the individual can derive (Bhattacharya, Korshun & Sen, 2009). If firms want consumers to support their engagement in CSR, the consumer must receive value from the exchange. This is another example of how the ‘win-win situation’ between company and consumer arises, the ‘consumer win’ is received value and the ‘company win’ is the received reward by the consumer.

(14)

Table 1: Summary of important factors that influence consumer behavior

Important factors

Source

Sen & Bhattacharya (2001)

Company specific (like the CSR domain and product quality)

Consumer-company identification

Inidividual specific (like CSR-related beliefs)

CSR-CA belief

Green & Peloza (2011) Type of CSR

CSR reputation

Bhattacharya, Korshun & Sen (2009)

Personal benefits

As explained before, consumers are more sensible to negative CSR information. Sometimes even companies that do not intentionally behave irresponsible find themselves in a CSR crisis. This means that if company’s choose to participate in CSR activities driven by market considerations rather than just by ideology, they should select those initiatives that enjoy the most support among their consumers. Managers need to be particularly aware of the dangers of being perceived as socially irresponsible and eventually end up in a CSR crisis (Sen & Bhattacharya, 2001). Even good CSR initiatives can turnout negative for a company, there are conditions under which consumers can become suspicious about the true motives underlying CSR behavior. Companies do not need to be perceived as irresponsible in order for consumers to become suspicious or skeptical. Once consumers perceive CSR efforts as a way only to enhance a company’s image, the CSR activities become inefficient and may eventually backfire (Yoon et al, 2006).

3. Corporate Social Irresponsibility (CSI) and CSR crises

3.1 What is a CSR crisis?

In the past few years, there were numerous corporate crises and corporate social irresponsibility scandals (Kolk & Pinks, 2006). While the public might enjoy reading about a

(15)

good crisis in the media now and then, most companies fear the thought of becoming the center of negative attention (Fennis & Stroebe, 2013). Worldwide business scandals involving big high profile organizations have resulted in a lot of negative media attention and critical articles about companies’ socially irresponsible behavior. Corporate social irresponsible (CSI) behavior is bad for consumers confidence in business leaders and it creates concerns about business ethics (Leonard & McAdam, 2003). Eventually CSI can result in a corporate (CSR) crisis. A corporate crisis is “a sudden and unexpected event that threatens to disrupt an organization’s operations and poses both a financial and a reputational threat” (Coombs, 2007, p. 164). This is why a crisis goes hand in hand with high consequences, ambiguity and decision making time pressure (Hale, Dulek, & Hale, 2005).

The ongoing irresponsible behavior of many companies suggests that for a long period of time, companies did not take care of stakeholder interests. Until the crisis or the scandal, these companies chose to neglect their stakeholder demands. The businesses did not believe in their moral duty or CSR (Kolk & Pinkse, 2006). A corporate crisis can occur either because the accused is held responsible for an action or because the act is considered offensive. There will not be an unfavorable impression of a firm if there is not the belief that the company is responsible for the act or if the act is not considered offensive. This shows that perceptions play an important role in responsibility more than reality does. It does not matter whether the business is in fact responsible for the offensive act, but what matters is if the firm is thought to be responsible for the offensive act or not. Because as argued before, even companies with good CSR reputations can find themselves in CSR crises. If the firm is in fact responsible for the offensive act or not can be a component that influences the company’s response. But importantly, as long as the company is considered as socially irresponsible, the corporate (CSR) reputation is at risk (Benoit, 1997; Wagner, Bicen & Hall, 2007). The survival of a company when acting irresponsible depends partly on satisfying expectations from its environment (Lange & Washburn, 2012). Since companies are usually aware of their irresponsible behavior long before the crisis occurs they have to consider what kind of reaction stakeholders would expect from them. However, congruence or no congruence between consumer expectations and a company’s response, behaving in a socially irresponsible manner has an impact on consumers’ perceptions and leads to negative consequences for companies (Lange & Washburn, 2012

(16)

3.2. CSI and CSR crises and the effect on consumers

What is clear so far is that consumers decide whether behavior is irresponsible or not on the basis of their perceptions, once a company is perceived as irresponsible this can lead to multiple negative consequences. This part will focus on the negative consequences CSI and CSR crises have on consumers. Despite heavy investment by companies in order to publicize their good work, reported incidents of irresponsible corporate behavior abound (Skarmeas, Leonidou, 2013). Information consumers receive about firms has been long recognized to influence consumer’s attitudes (Eisingerich, Rubera, Siefert & Bhardwaj, 2011). Consumers are more often exposed to negative than positive CSR information about retailers through mass media and are also more likely to share negative than positive information with each other (Eisingerich, et al., 2011; Wagner, et al., 2007). According to Wagner et al. (2007), there are numerous factors that influence perceived CSI in the retail industry. These factors are: practices that are potentially harmful to the natural environment, the threat for local businesses, foreign economies, local employment, societal rules, employee benefits, employee wages, local working conditions, employee discrimination, foreign labor, sales practices, dishonesty, offensive material and pricing policies (Wagner et al., 2007). So just like the type of CSR reputation has a different effect on consumers so does the type of CSR crisis. In this thesis different types of CSR crises will be researched to see if they will have different outcomes in combination with positive or negative CSR reputations.

Only if consumers are highly supportive of the CSR issue they react positively to positive CSR information while all consumers react negatively to negative CSR information. This is in line with Lange and Washburn’s (2012) argumentation about negative CSR information. They argue that when consumers are confronted with negative behavior, they will spend more time thinking about this negative behavior than they would about positive behavior. This is because after being confronted with information about irresponsible behavior, consumers will search more extensively for information. So their reactions and behavior will be more extreme. However, expectations and evaluations about CSR are not objective realities in the firm’s environment but they are subjective and above all changeable (Lange & Washburn, 2012). Important in light of the current research, is that information about CSI weights heavier than positive information and has severe consequences for the company that is considered as irresponsible (Lange & Washburn, 2012; Sen & Bhattacharya, 2001). This makes it so important to research CSI.

(17)

Furthermore, what becomes apparent from previous literature is that consumers have a growing amount of skepticism toward CSR claims and that they are willing to punish CSI (Murphy & Schlegelmilch, 2013; Skarmeas & Leonidou, 2013). Grappi et al. (2013), investigated two types of consumer responses after corporate social irresponsibility. These consumer responses are in line with the will to punish CSI (Murphy & Schlegelmilch, 2013; Skarmeas & Leonidou, 2013). The first is negative word of mouth in which consumers share their emotions with others. Negative word of mouth can take the form of saying negative things, recommending against purchasing and discrediting the company. The second consumer response is protest behavior, actions against CSI. Logically, consumers experience emotions after CSI. The will to punish and the reaction of consumers in general is influenced by the type of CSI or CSR crisis but also by the type of emotions consumers experience. CSI can take different forms of moral transgressions which will result in different emotions. Following this explanation Bhattacharya and Sen (2010) assume that any discrepancies between stakeholder’s perceived motives and the company’s stated motives would trigger skepticism and feelings of deception (Bhattacharya & Sen, 2010). The reason why underlying motives are shortly discussed in this thesis is because they will eventually influence the way the firm is perceived. This will contribute to a company’s positive or negative CSR reputation, which will possibly make consumers react differently once there is a CSR crisis.

Table 2: The negative effect of CSI on consumer attitudes

CSI's negative effect

Source

Eisingerich, et al. (2011), Wagner, et al. (2007) Negative word of mouth intentions

Lange & Washburn (2012) More extreme reactions and behavior

Murphy & Schlegelmilch (2013), Skarmeas & Leonidou

(2013) Punishing intentions

Bhattachary & Sen (2010), Murphy & Schlegelmilch

(2013), Skarmeas & Leonidou (2013) Growing skepticsm

What can be concluded is that companies can find themselves in a CSR crisis like an environmental disaster that happened outside their control or they can be in fact responsible. Sometimes it was outside the companies direct control but consumers still hold the company

(18)

responsible so the reputation is always at risk. But the complexity lies in the fact that it is hard to control consumer responses because they are dependent of many different factors like products, industry, origin-related cues, consumers’ unconscious level, stereotypes, consumer perceptions, type of CSR reputation, type of CSR crisis and so on. Despite of the complexity, some companies are able to successfully change the negative effects into positive ones through their CSR activities, but for some the same strategy backfires (Yoon et al., 2006). An even better scenario arises when it is not even necessary to fight negative effects because the current reputation is so good that it protects a company from negative effects during a crisis. Under what circumstances does this happens? And how does it work?

4. CSR reputation, the ‘halo’ effect and the role of fit

4.1. CSR reputation

The importance of developing a good reputation is well recognized by firms (Bertels & Peloza,2008; Lii & Lee, 2011). A general common definition for corporate reputation is:

“A corporate reputation is a stakeholder’s overall evaluation of a company over time. This evaluation is based on the stakeholder’s direct experiences with the company, any other form of communication and symbolism that provides information about the firm’s actions and/or a comparison with the actions of other leading rivals” (Gotsi & Wilson, p. 29, 2001).

In this thesis the focus is not on general corporate reputations but corporate CSR reputations. Which means, companies that earned their good CSR reputation because of their ethical behavior and their negative CSR reputation because of CSI which can take many different forms.

As argued before by Wilson (2003), it is no longer the question if a firm should make CSR commitments but how they should do this best (Bertels & Peloza, 2008; Lii & Lee, 2011). CSR has become an integral component for firms’ reputations. This led to firms that have a corporate reputation based on CSR, like Starbucks, Ben & Jerry’s, and Tom’s shoes. One of the reasons CSR is so important for reputation are the earlier mentioned stakeholder expectations, this encourages firms to integrate CSR activities with core values and competencies to create a CSR focused reputation (Bertels & Peloza, 2008). According to Hillenbrand and Money (2007), antecedents of a good reputation include proper CSR standards, philanthropic giving and a good relationship with stakeholders. This probably

(19)

reflects the believe, that being involved in CSR activities does lead to better a reputation, in other words being good leads to a good CSR reputation (Lii & Lee, 2011).

It is known that the people who decide if a company has a good corporate CSR reputation or not are consumers and other stakeholders, as also became clear in the earlier mentioned definition of reputation, and dedicating a share of the revenues to social issues is expected by the public (Tixier, 2003). But for most consumers it is not easy to verify CSR claims made by organizations. If an organization claims good working for their employees in Bangladesh, consumers cannot easily go there and have a look themselves. This is why most consumers are skeptic when it comes to information directly obtained from firms, and search and experience are viewed as more provable (Pomering & Dolnicar, 2009). Most companies are aware of the pressure by public opinion and their desire for greater commitment to causes of public interest. Those companies spend a lot of time and resources to make sure they have a good CSR reputation (Tîxier, 2003). Once companies achieve the desired CSR reputation, they could experience a lot of potential benefits in times of crisis.

In chapter three it became clear that even though companies cannot control the diffusion of negative information, they can attempt to mitigate the damage through things like openness and fulfilling consumers’ expectations. But above all building a strong reputation can protect companies in times of crisis (Eisingerich et al., 2011). But the question is; how strong is the protection effect of a CSR reputation during a crisis and under what circumstances the effect of a good CSR reputation becomes stronger or weaker?

4.2. A good reputation leads to the ‘halo’ effect

Building a strong CSR reputation may eventually lead to consumers’ resistance to negative information, this is the so called ‘halo’ or ‘buffer’ effect. This means that companies with a good CSR reputation will experience less damage from a crisis because their reputation serves a ‘buffer’. So how does this work? The idea of the halo effect is based on the argument that CSR may enable firms to enjoy greater levels of goodwill with consumers and thus lower firm-idiosyncratic risk (Eisingerich et al., 2011). Thus, this suggests that a strong CSR reputation offers companies a kind of insurance against negative information about their irresponsible behavior in times of crisis. Resistance to negative information, the ‘halo’ effect, can be explained as consumers that do not allow negative information to diminish their view of a firm (Eisingerich et al., 2011). Peloza (2006) even argues that this can reduce the risk of

(20)

potential harm like boycotts. Eisingerich et al. (2011), conducted a research to find out how effective CSR is in insulating firms from scrutiny. Furthermore, their study compared the importance of the CSR protection effect compared to other important marketing measures, such as service quality orientation. The results show that CSR practices make consumers more resistant to negative information when confronted with negative information about the firm. CSR shields companies from negative information about CSR practices but not from negative information about a firm’s core service offerings. Eisingerich, et al. (2011), state that CSR offers less of a halo effect than assumed in previous research. The respondents in the research were asked to answer questions about their own telephone provider. One of the questions was perceived CSR of their telephone provider, this made it possible to view if CSR leads to higher resistance to negative information levels. However they did not look at a company where the reputation is actually based on CSR, a CSR reputation.

Klein and Dawar (2004) show another view on the ‘halo effect’. They view the ‘halo’ effect as a spillover effect. What has been demonstrated the past years is that CSR plays a role in consumer behavior that goes beyond economic and rational considerations. This means that CSR has a spillover effect on non-product associations’. Klein and Dawar (2004) researched how the ‘halo’ effect (or spillover effect) of consumers’ prior beliefs about a company’s CSR, influences attributions in a product-harm crisis. Klein and Dawar (2004) found that the consumers’ prior beliefs, or in other words the CSR reputation, has an important effect on consumers’ perceptions. If the company that is involved in the CSR crisis has a long-standing good reputation and is viewed as integer and competent, it is unlikely that the crisis can do serious damage to the company’s reputation. Even though the type of crisis and seriousness of the crisis play a role in consumers’ perceptions, in general the CSR reputation will diminish the negative outcomes. Fennis and Stroebe (2013) agree with the idea that a strong reputation can make up for negative impacts crises normally have. They argue, that when there is no shield of a positive reputation, the company will probably face more severe negative consequences. Klein and Dawar (2004) explain why CSR reputation plays such an important role in times of crisis. A crisis results in a context where consumers cannot rely on product evaluation anymore. The findings of Klein and Dawar (2004) suggest that CSR associations start playing a significant role in these non-product evaluation contexts when consumers rely on corporate associations to inform their judgments. It could be said that CSR has a type of ‘dormant’ effect that only becomes visible in crisis circumstances in which consumers rely on

(21)

non-product associations to base upon their judgments. Klein and Dawar (2004) call CSR a kind of insurance policy that is there when you need it. This leads to hypothesis one:

H1: Consumer attitudes after a crisis will be less negative for companies with a positive CSR reputation than for companies with a neutral or negative CSR reputation.

4.3 The role of related versus unrelated CSR

What is clear now is that a good CSR reputation is expected to limit the negative effects of a CSR crisis. But the effect might not be unconditional and for bad reputation companies it only occurs when it disconfirms consumers’ expectations. When companies communicate inconsistent CSR information they are basically lying because this would mean that there CSI is directly related to their CSR (Wagner et al., 2009). So this suggests that consumers find it less hypocrite if the CSI activity is unrelated to the previous CSR reputation. According to Wagner, et al. (2009), a CSR crisis can lead to corporate hypocrisy which is the belief that a firm claims to be something that it is not. A consumer’s attitude towards a company’s CSR is his or her overall assessment of the extent to which a company is socially responsible (Du, et al. 2007). Consumers’ exposure to CSR information leads to the consumers’ CSR beliefs. Some companies choose to actively promote their CSR efforts to create a reputation of social responsibility, examples are Starbucks and The Body Shop. Other companies only start promoting their CSR activities to protect their reputation after reported information about socially irresponsible actions. The results of Wagner, et al. (2009), show that inconsistent CSR information has a big impact on consumers. Inconsistent CSR information raises feelings of hypocrisy and affects the attitudes towards the CSR of the company. It is possible that if the CSR reputation and the CSR crisis are related people feel deceived and as a result view the company as hypocrite. In contrast, when the crisis is unrelated to the CSR reputation, people might feel less deceived but it still may raise skepticism. This leads to hypothesis 2a:

H2a: A positive CSR reputation will lead to more negative consumer attitudes when the crisis is related and less negative consumer attitudes when the crisis is unrelated.

When the CSR reputation is already negative relatedness of the crisis would probably lead to less skepticism. If Shell already has a negative reputation regarding the natural environment and now appears to be involved in another environmental scandal, we are not surprised. This means that there is no need for feelings of hypocrisy or skepticism. In the case of a negative 20

(22)

CSR reputation an unrelated crisis would probably lead to more negative consumer feelings. A possible explanation is that an unrelated crisis means that they were already doing wrong in one CSR area and now also in a second CSR area. This leads to hypothesis 2b:

H2b: A negative CSR reputation will lead to more negative consumer attitudes when the crisis unrelated and less negative consumer attitudes when the crisis is related.

5. Consumer emotions as mediating factor

In this thesis it is expected that emotions mediate the effect of the CSR reputation and the CSR crisis plus their relatedness on consumer perceptions. In previous research, emotions emerged as an important mediating role in several contexts such as complaining and product attitudes. The type of emotion a consumer experiences, influences the outcomes in consumer behavior, perceptions and attitudes. The article of Laros and Steenkamp (2005) proposes four negative and four positive basic emotions. They tested these emotion levels for different types of food and show that this is a better categorization of emotions than just looking at emotions as either positive or negative. The four negative emotions that fall under the negative affect are anger, fear, sadness and shame and the four positive emotions that fall under the positive affect are contentment, happiness, love and pride. According to Green and Peloza (2011), emotion is one of the three value drivers of CSR alongside social value and functional value. The emotions that arise from a CSR reputation are a result of a social or environmental attribute that makes consumers experience a ‘warm glow’. This means that good CSR can lead to value creation for consumers because they will experience positive emotions. However they did not research how this phenomenon works when the emotions are negative so value diminishing instead of creating. During a crisis, consumers can also develop all kinds of negative emotions. According to Jorgensen (1996), emotions play an important role in consumers perceptions of a company after the crisis. As said before, this eventually influences their purchasing behavior and their attitudes toward the company. Furthermore, emotions are highly linked to the amount of responsibility of a company in a crisis. Again, this is not about actual responsibility but perceived responsibility. Anger toward the company is higher when they are considered to be highly responsible. So emotions are influenced by the type of crisis and the prior reputation and influence consumers’ attitudes and perceptions. Angry consumers are for example more likely to share their negative experiences (Eisingerich et al., 2011).

(23)

Furthermore, Grappi et al. (2013) argue that emotions as a result of irresponsible corporate behavior motivates consumers to participate in protest behaviors and negative word of mouth. Next they argue that these emotions are moderated by the strength to which consumers hold certain ethical virtues. This means that emotions were stronger for consumers who were more ethically responsible themselves. After being moderated by consumer virtues, emotions have a direct effect on negative word of mouth and consumer attitudes.

There are (1) ethical transgressions which concerns harm done to other people such as employees and arises for example because of freedom violation. And there are (2) social transgressions, this is when a corporation does harm to a community. The three main emotions that were found in the study of Grappi et al. (2013), after consumers faced a social and a ethical transgression are, contempt, anger and disgust. In the experiment of this thesis the emotions of Grappi et al. (2013) will also be used because the results showed that they influence consumer attitudes. Therefore it is expected that these consumer emotions might mediate the effect of the CSR crisis and the CSR reputation on consumer attitudes. The emotions of Grappi et al. (2013) are complemented with the emotions researched by Laros and Steenkamp (2005). The emotions distinguished in their research are; anger, fear, sadness, shame, contentment and happiness. Because the experiment is about a CSR crisis the emotions contentment and happiness will not be used.

This leads to the following hypotheses:

H3: Consumer emotions will mediate the effect CSR crisis and CSR reputation have on consumer attitudes

H3a: The stronger the negative consumer emotions are, the more negative consumer attitudes will be

H3b: The more positive consumer emotions are, the more positive consumer attitudes will be

(24)

Figure 1: conceptual model

reputation

CSR crisis

Consumer

emotions Consumer attitudes relatedness

(25)

6. Methodology

6.1. Introduction

Next the methodology of the present study will be presented which is an experiment via online questionnaires. To answer the main research question of this thesis: “Do consumers’ responses to a CSR crisis differ for companies with a positive versus a negative CSR reputation? And do these responses depend on the level of relatedness between the CSR reputation and the crisis?”, the experiment is considered as the most suitable method. This is because to answer the main research question, causal hypotheses will be tested. With the help of an experiment the hypotheses can be tested under controlled circumstances. By manipulating the conditions in different ways, it is possible to demonstrate the different effects as a result of the manipulations. This means that the independent variables are differently manipulated so that eventually the possible differences in the dependent variable will show. Because the experiment is conducted via online questionnaires, the questionnaires contain manipulation texts to which the respondents will be exposed to.

The questionnaires are created and spread among the respondents with the use of Qualtrics. The study employed a 3 (neutral CSR reputation, positive social CSR reputation, negative social CSR reputation) x 3 (social crisis, community, environmental crisis) between subjects design. The respondents were informed that the questionnaire would be about a retail company and they were asked to give their personal opinion about the company. The retail company used in the presented scenario’s is the fictive company f+m. They could indicate their personal opinion by indicating whether they agreed or disagreed with statements using a 7-point Likert Scale.

6.2. Manipulations and variables

6.2.1. Manipulations

All respondents were exposed to a manipulation check and one of the three transgressions (crises). First respondents were informed about what they were going to read, this explained that they had to pretend to be real f+m customers browsing the internet and crossing information about the company. After the general information, the respondents continued by reading one of the three manipulation texts. The manipulation texts contain some general background information about f+m and information about f+m’s CSR reputation. The first

(26)

group ‘the control group’ was only exposed to the neutral information and did not read any information about f+m’s CSR reputation. The neutral information informs the respondents about what kind of retail company f+m is and about their attractive low prices. The people in the positive CSR reputation group also received neutral background information but also information about the positive social reputation of f+m. They received information about how well f+m treats its employees in developing countries. In contrast, the respondents in the negative CSR reputation group read information about how bad f+m treats its employees and the bad working conditions and salaries they offer. The positive as well as the negative CSR information is about the social CSR reputation of the company.

After the information about the CSR reputation, respondents were asked to give their opinion about the company’s CSR. This served as a manipulation check and shows the effect of the independent variable ‘CSR reputation’ on the dependent variable ‘perceived CSR reputation’, before the respondents were exposed to the crises. The scale for the manipulation check ‘perceived CSR reputation’ is based on Eisingerich, et al. (2011), view appendix B. After the manipulation check the respondents were exposed to one of the three transgressions, the crises. One of the crises was about a fire in a f+m factory in Bangladesh were many workers got hurt. This is a social transgression/social crisis and is therefore related to the negative and the positive social CSR reputation. The second crisis is about how f+m is polluting the Chinese rivers with toxic chemicals. This is an environmental crisis unrelated to the negative and the positive social reputation, and is therefore one of the two unrelated transgressions. And the third crisis is about how f+m scorns the homeless in New Delhi (see appendix B). This is a community crisis and is the second unrelated transgression.

Figure 2: Different experimental conditions

(27)

6.2.2. Dependent variables

All the experimental conditions were followed up by the same questionnaire where the respondents were asked to give their personal opinion about given statements with the use of a 7-point Likert Scale. The questionnaire includes questions about the respondents’ emotions, which will be tested as the mediator. The used scale to test consumer emotions is based on Grappi, et al. (2013) and Laros and Steenkamp (2005). Also the respondents were asked to indicate their attitude towards the company. The consumers’ attitude is the dependent variable in the conceptual model. The used scale to test consumers’ attitudes is based on Spears and Singh (2004). And respondents were asked to give their opinion about the relatedness between the CSR reputation and the CSR crisis, which is used to test the moderator in the conceptual model. The scale to test relatedness is based on Wagner, et al. (2009) and Mohr and Webb (2005).

Previous research showed that there will not be an unfavorable impression of a company if there is not the believe that the company was guilty or responsible for the crisis (Benoit, 1997; Wagner, et al. 2007). If there are no significant results on consumer attitudes the company’s responsibility might offer an explanation. The company’s perceived responsibility in the crisis will therefore serve as one of the control variables. The used scale is based on Jorgensen (1996). Furthermore, the respondents personal CSR involvement could offer an explanation for the differences between the experimental groups. Grappi, et al. (2013) argue that emotions can be influenced by the strength to which consumer hold ethical virtues. That means that emotions can be stronger for consumer who are more ethically responsible themselves. The CSR involvement of the respondents is therefore the second control variable and the used scale will be based on Mohr and Webb (2005). The control variables will only be included in further analysis if the descriptive statistics show that there are significant results between the respondents in the different conditions. Because if there are a lot more respondents in one group that are highly involved in CSR than in the other groups, this might influence the results. However, if there are no significant differences between the groups in CSR involvement and perceived responsibility they will not cause any skewness in the dependent variable consumer attitudes. Table 3 shows a summary of the scales, for the survey and the complete scales see appendix B.

(28)

Table 3: Research scales

Scale Source

Perceived CSR reputation Eisingerich, et al.

(2011) I consider f+m as a socially responsible firm

The company does not contribute to society in positive ways

This firm is more beneficial to society's welfare than other firms

This firm contributes to society

Consumer emotions Grappi, et al. (2013)

and Laros & Anger Surprise Disgust Sadness Shame Sympathy Aversion

Consumer's attitude Spears & Singh (2004)

Unattractive-attractive Bad-good

Unpleasant-pleasant Unfavorable-favorable Unsympathetic-sympathetic

Perceived relatedness Wagner, et al. (2009)

and Mohr & Webb F+m's recent behavior is consistent with its reputation

F+m acts hypocritically

CSR involvement of the consumer Mohr & Webb (2005) I try to buy from companies that help the needy

I try to buy from companies that pay their employees a living wage

When giving a chance to switch to a brand that gives back to community, I take it

When I am shopping, I try to buy form companies that are working to improve conditions for employees

Company's responsibility Jorgenssen (1996)

Responsible for the crisis To blame for the crisis Controllable role in the crisis

(29)

6.3. The selection of the respondents

The selection of the respondents happened through a snowball effect/convenience sampling method. First some easy accessible acquaintances were approached to fill in the questionnaire and asked if they could send the link to the questionnaire to family, friends and others like colleagues. The questionnaire was only accessible online and spread via Facebook and via e-mail. Since all people are consumers, everyone could participate in this experiment.

The respondents got assigned to one of the nine experimental conditions and in total 247 consumers participated:

1) Neutral reputation + social crisis: N=28 2) Neutral reputation + community crisis: N=31 3) Positive reputation + social crisis: N=26 4) Positive reputation + community crisis: N=25 5) Negative reputation + social crisis: N=27 6) Negative reputation + community crisis: N=29 7) Neutral reputation + environmental crisis: N=26 8) Positive reputation + environmental crisis: N=27 9) Negative reputation + environmental crisis: N=28

6.4 The pretest

A short pretest was conducted in order to test the stimuli that will be used for the actual questionnaire. The pretest’s aim was to find out which type of crisis scenarios/transgressions are suitable for the actual experiment, the eventual goal was to identify three suitable CSR domains. One that is related to the CSR reputation, which is a social issue, and two that are unrelated to the CSR reputation (i.e. different from a social issue). If the transgressions are suitable for the actual experiment is determined by how much their effect on credibility, perceived CSR and perceived relatedness differ. In order to be comparable in the actual experiment they cannot differ too much from one another. Furthermore, a manipulation check for CSR reputation is done in order to find out if there is a significant difference in how consumers perceive negative versus positive information about a company.

A quantitative pretest was conducted with 20 respondents via an online questionnaire. Ten respondents received information about a positive CSR reputation, followed by four types of crisis. And ten respondents received information about a negative CSR reputation, followed

(30)

by four types of crisis. The four crises in both the negative and the positive condition are; social crisis, community crisis, environmental crisis and an animal crisis. Just like in the actual experiment, the social crisis is the crisis related to the CSR reputation. The other three crisis are the unrelated transgressions.

Figure 3: Pretest conditions, positive CSR reputation (10 respondents)

And ten respondents were exposed to information about a company with a negative social CSR reputation that was linked to four different crises:

Figure 4: Pretest conditions, negative CSR reputation (10 respondents)

Table 4 shows the results of the manipulation check, in other words, the different effect of the negative and the positive CSR reputation on consumers’ perceived CSR of the company. There are big differences between the two groups, this means that the negative CSR reputation and the positive CSR reputation differ enough to use in the real experiment F(1;18)=31.95, p<0.001.

Negative social reputation

Environm

ental Community

Social Animal

Positive social reputationCSR

Environm

ental Community

Social Animal

(31)

Table 4: Perceived social responsibility of the company (1=negative perceived CSR to 7=positive perceived CSR)

Reputation Mean SD

Positive 4.97 1.13

Negative 1.44 0.64

Furthermore, the results show that all the respondents viewed the described scenario’s as credible and realistic, all the means are >4. This means that the scenarios are realistic enough to use in the experiment. The next tables 5 and 6 show that the pretest does not show big differences in the perceived relatedness between the CSR reputation and the crisis. This is probably because the respondents in the positive reputation scenario were exposed to all four crises as well as the respondents in the negative reputation scenario. It is possible that this makes it more difficult to rate relatedness.

Table 5: Perceived relatedness (reputation*crisis) in the positive reputation condition

Mean SD

Social 3,68 0,98

Environmental 4,00 1,55

Animal 3,60 1,35

Community 4,15 1,00

Table 6: Perceived relatedness (reputation*crisis) in the negative reputation condition

Mean SD

Social 4,38 1,41

Animal 3,56 2,13

Community 4,44 1,94

Table 7 shows that there are no big differences in hypocrisy, which makes the different crises comparable to each other. This means that overall the results show, that all four crises scenario’s are suitable for the real experiment. The decision was therefore randomly made to use the environmental and community crisis as the two unrelated transgressions and the social crisis as the related transgression.

(32)

Table 7: Perceived hypocrisy

Reputation Crisis Mean SD

Positive Social 4,45 1,37 Environmental 4,50 1,51 Community 3,90 1,60 Animal 4,60 1,65 Negative Social 4,25 2,27 Environmental 5,13 2,56 Community 3,88 2,30 Animal 4,63 2,20

7. Research findings

7.1. Reliability analysis

The findings were first tested with Cronbach’s Alpha to assure the reliability of the multiple-item scales. The first scale’s internal consistency was tested for the manipulation check variable ‘CSR reputation’. The Cronbach’s Alpha of CSR reputation is 0.76 which is above 0.70 which means that the scale’s reliability can be seen as acceptable.

Table 8: Cronbach’s Alpha coefficients for multiple-item scale (CSR reputation)

Mean St. _deviation Cronbach's Alpha if item deleted Social Responsibility 3,75 1,92 0,66 Adds no positive to society 4,94 1,49 0,84

Better than other

companies 3,33 1,60 0,59 Adds to society 4,12 1,57 0,64

Even though the reliability of the CSR reputation scale could be improved by deleting the item ‘Adds no positive to society’, this was not executed because the reliability of the scale was already acceptable.

(33)

Next the scale for the mediator ‘consumer emotions’ was tested. The Cronbach’s Alpha of emotions is 0.62, this is not a very high score but could not be improved by deleting an item.

Table 9: Cronbach’s Alpha coefficients for multiple-item scale (emotions)

Mean St. deviation Cronbach's Alpha if item deleted Anger 4,81 1,41 0,51 Surprise 3,82 1,82 0,59 Disgust 3,90 2,01 0,62 Sadness 3,84 1,63 0,56 Sympathy 5,60 1,54 0,63 Shame 4,31 1,78 0,50 Aversion 5,07 1,45 0,49

The Cronbach’s Alpha for the scale of the dependent variable ‘consumer attitudes’ is 0.86, which means that the reliability of this scale can be viewed as high. The scale cannot be improved by removing any items.

Table 10: Cronbach’s Alpha coefficients for multiple-item scale (attitudes)

Mean St. deviation Cronbach's Alpha if item deleted f+m is unattractive - attractive 3,08 1,33 0,84 f+m is bad-good 2,94 1,21 0,84 f+m is unpleasant-pleasant 3,16 1,23 0,83 f+m is unfavorable-favorable 3,20 1,35 0,85 f+m is unsympathetic-sympathetic 2,55 1,20 0,87

The Cronbach’s Alpha for the interaction effect of CSR reputation and crisis ‘relatedness between reputation and crisis’ is very low 0.28. The scale only consists of two variables which explains why the reliability is so low and no items can be deleted.

(34)

Table 11: Cronbach’s Alpha coefficients for multiple-item scale (relatedness) Mean St. deviation Cronbach's Alpha if item deleted Behavior

conform 3,59 1,56 Not possible Hypcrisy _3,45 _1,40 _{Not possible}

The scale of the control variable ‘CSR involvement of the consumer’ is highly reliable 0.93. Table 12 shows that it was not necessary to remove any items to improve the scales reliability. The scale for ‘perceived responsibility’ of the company has a low score on reliability, Cronbach’s Alpha is 0.61. However, table 13 shows that the reliability could not be improved by removing any items.

Table 12:Cronbach’s Alpha coefficients for multiple-item scale (CSR involvement of the consumer) Mean St. _deviation Cronbach's Alpha if item deleted Goodworkcondition 4,13 1,62 0,90 Goodpay 4,20 1,56 0,90 Helping 4,00 1,58 0,91 Change to responsible retailer 4,50 1,61 0,94

Table 13:Cronbach’s Alpha coefficients for multiple-item scale (perceived responsibility)

Mean St. deviation Cronbach's Alpha if item deleted Perceived responsibility 4,15 1,68 0,46 Perceived guilt 4,76 1,6 0,35 Perceived controllability 4,51 1,53 0,68 33

(35)

7.2. Descriptives

In this part of the thesis the descriptive statistics of this research will be explained. This research makes use of a 7-point Likert scale which means that the scores equal or higher than 4 can be viewed as neutral/positive scores for the scales, and 3 and lower can be viewed as negative scores for the scales. This goes for all the scales. The aim of the descriptive statistics is to summarize the datasets and give insights in consumer attitudes, consumer emotions, consumers’ CSR involvement, perceived responsibility, and perceived CSR reputation. With the results from the descriptive statistics it is easier to interpret the results from the rest of the analysis like ANOVA.

The following tables summarize the results of the descriptive statistics.

Table 14: Consumer attitudes before the crisis (manipulation check)

Mean SD

Negative 3,08 0,11

Neutral 3,82 0,11

Positive 5,27 0,11

Table 14 shows the descriptives of the manipulation check. The respondents in the positive condition are more positive than the respondents in the neutral and the negative condition. This means that consumer perceptions about the company’s CSR are influenced by the company’s CSR reputation.

Table 15a+b: Consumer emotions (mediator)

a: Consumer emotions in the b: Consumer emotions in the different different CSR reputation conditions CSR crisis conditions

Mean SD

Negative 4,45 0,99

Neutral 4,42 0,80

Positive 4,64 0,95

The tables 15 a and b show that there are no big differences between consumer emotions in the different groups. The respondents had to rank their emotions with the 7-point Likert Scale, 1=do not feel the negative emotion at all to 7=feel the negative emotion a lot. Most consumer ranked >4 which means that they all had negative emotions after the CSR crisis, whereby the

Mean SD

Community 4,71 0,89

Social 4,35 0,99

Consumer responses to corporate social irresponsibility : the role of relatedness between the crisis and the company’s reputation