The effects of responses to a CSR-related automotive crisis on perceived response credibility

The effects of responses to a CSR-related automotive crisis on perceived

response credibility.

Master’s thesis

By Bouke van Riel - 10420304 Final version

January 27th 2017

MSc Business Administration – Marketing University of Amsterdam

Statement of originality

This document is written by Bouke van Riel who declares to take full responsibility for the contents of this document.

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

Abstract

As corporate social responsibility (CSR) has become a main component of a company, a CSR crisis can have a huge impact on the their credibility. When facing a crisis, responses and actions can be taken in order to regain the companies’ CSR credibility. This study takes the “Dieselgate” crisis at Volkswagen (VW) as an example, whereas it examines whether several responses to the crisis have a different effect on perceived credibility. VW has misled their stakeholders by selling Diesel fuelled cars that far exceed the toxic emission standards. VW plans to introduce a new fully electric high-end car model to repair their CSR image, where this study expects introducing a more mainstream fully electric car model leads to a higher perceived credibility based on the signaling theory. From this quantitative survey research with a sample of 151 respondents from The Netherlands, it can be concluded that introducing an electric mainstream model after an automotive CSR crisis leads to more perceived

credibility than a high-end model. Additionally, it appears that the perceived motive why the company is involved with CSR is of main importance when evaluating credibility. No moderation effects of sustainability orientation of the respondent and perceived luxury product have been found. The main implications of this study are the theoretical contribution to the credibility/ethical approach to CSR, application of the signaling theory to this field of research and the formulation of a new response credibility scale.

Keywords: corporate social responsibility, Dieselgate, perceived credibility, crisis management, signaling theory.

Table of Contents

1.1 Introduction ... 5

1.2 Volkswagen and the Dieselgate ... 6

1.3 Importance of social responsibility ... 7

1.4 Responding to a crisis ... 8

1.5 Research gap ... 10

1.6 Research question ... 13

2. Literature review 2.1 Corporate Social Responsibility ... 14

2.1.1 Greenwashing ... 15 2.2 Crisis or a scandal ... 17 2.3 Crisis response ... 19 2.4 Perceived credibility ... 21 2.5 Signaling theory ... 23 2.6 Sustainability orientation ... 26 2.7 Perceived CSR motive ... 28 2.8 Luxury products ... 29 2.9 Conceptual model ... 31 3. Method 3.1 General ... 32 3.2 Pre-tests ... 34 3.3 Variables ... 35 3.4 Sample... 37 4. Results 4.1 Statistical analysis ... 39 4.2 Correlation matrix ... 43 4.3 Hypotheses testing ... 44 4.4 Additional results ... 45 5. Discussion 5.1 Theoretical implications... 48 5.2 Managerial implications... 52

5.2 Limitations and recommendations for further research ... 53

6. Conclusion ... 56

7. Bibliography ... 60

Introduction

1.1

When thinking about the automotive industry, there is a big chance that the first top-of-mind brand is Volkswagen. The mother brand called Volkswagen AG (also known as Volkswagen Group) is the biggest car manufacturer in Europe and the second biggest worldwide (Schmitt, 2016). According to the latest available year report of the Organisation Internationale des Constructeurs d’Automobiles (OICA) of 2014, Toyota is the biggest car manufacturer in the world with over ten million vehicles. Volkswagen AG comes in second place, with a little less than ten million vehicles, followed by General Motors on third place (OICA, 2015). Volkswagen AG owns twelve world known car brands such as Volkswagen, Audi, Porsche, Lamborghini and SEAT. In the following paper, the car brand Volkswagen (VW) will be discussed and not it’s mother brand Volkswagen AG.

VW is the most popular car brand manufactured by Volkswagen AG and is originated in Wolfsburg, Germany. The Nazi Deutsche Arbeitsfront founded the company in 1937, as it was produced in accordance with Adolf Hitler’s vision to be an affordable German family car (Bowler, 2015). To describe Volkswagen’s importance for the car industry; VW has

produced 2 out of the 5 best selling cars of all time (i.e., the Beetle and the Golf) (Daidone, 2013). Given VW’s size and popularity, VW is capable of making a big difference for the environment we live in. As VW is very focused on a sustainable future (Volkswagen, 2007), an investigation in 2015 stirred up the automotive industry.

1.2 Volkswagen and the Dieselgate

In September 2015 the U.S. Environmental Protection Agency (EPA) announced that they started investigations whether VW had violated emission tests by illegally installing software in some of their diesel-powered cars, which has led to fabricated emission results to pass the tests (Cue, 2015). There was a large discrepancy found between the laboratory- and road tests, executed by the International Council on Clean Transportation (ICCT), where the test executed on the roads showed NOx emissions (nitrogen oxide and nitrogen dioxide) to be 15 to 30 times the US legal maximum (Jaynes, 2015). To paint a clearer picture, NOx

emission contributes to smog forming and pollution. The software made by VW could detect that the car was being tested, so that the test results of NOx emission during the test condition would be minimal. When on the road, the emission control would be automatically switched off where it by far would not comply with the emission regulations minimum. The media widely covered the news whereas, partly due to media pressures, VW decided to admit cheating, and that their internal investigation revealed that the software affected 11 million cars worldwide (Foxnews.com, 2015). This so-called Dieselgate is the biggest scandal in the car industry in 78 years and the company stock value declined. VW also faced the potential of criminal charges (Cremer, 2015; Plumer, 2015).

Dieselgate caused a great international stir and the critics have not been taking the

incident lightly, Car and Driver for example calling it a “heinously bad, unethical and

trust-eroding decision that equally battered both its reputation and its market value” (Robinson,

2015). On the other hand, some people are defending VW, whereas according to them the reputation of the EPA is questionable. Claims that the EPA is “out of control” and “keeping

the small and efficient diesel motors for cars from the market by stricter NOx limits, because they themselves did not control the technology” were also spread (Domjahn, 2015; Neuerer,

2015; Peters, 2015). Whatever each of the stakeholder’s intentions were, the truth is that VW has made a deliberate mistake.

As corporate social responsibility (CSR) is part of VW’s corporate culture (Volkswagen.com), it is ironic to see that specifically VW sparked one of the greatest scandals in the automotive history. VW has started a campaign named Think Blue a few years ago, stating all carmakers carry a responsibility in; “creating environmentally friendly

products and solutions, encouraging more eco-conscious behavior and contributing to a sustainable future” (Volkswagen.com). If VW has actually acted according to this

proclaimed vision seems questionable. The main question now is how Dieselgate will affect VW’s CSR credibility.

1.3 Importance of social responsibility

As CSR is becoming more important for companies in general and specifically for VW, most companies are involved with CSR to make a contribution to society or to support their strategic goals (Skarmeas & Leonidou, 2013). On the other hand, CSR is often used to build favorable reputations, as reputation plays a great role in crisis communication

(Fombrun, 2005). CSR is no longer an activity by just a few companies; it has become a mainstream, highly visible, and commonplace practice (Skarmeas & Leonidou, 2013). Since more and more companies engage in CSR activities to increase reputational and financial benefits, incidents of socially irresponsible company behavior continue to increase (i.e., Carson, 2003; Wagner et al., 2009). Because of the conflicting information consumers

receive via several media channels (such as in the VW case), consumers tend to question why companies are involved with CSR and become skeptical about CSR activities in general (Forehand & Grier, 2003; Vanhamme & Grobben, 2009). Because these instrumental and business case approaches to CSR currently are most popular (i.e., Lee, 2008; Dahlsrud,

2008), other viewpoints should also be discussed. Next to business case approaches, research and businesses should also focus on credibility approaches by obtaining support for the moral motivations that businesses are involved with CSR (Garriga & Mele, 2004; Husted &

Salazar, 2006). The current research could extend the knowledge for these credibility approaches.

1.4 Responding to a crisis

The way companies respond to a crisis is very important for their future business performance. Brand-related adverse events (brand crises) are very common and generate a lot of publicity, whereas they have been heavily researched in the past years and it’s negative effects on brand confidence and reputation give importance to the way companies deal with this phenomenon (e.g., Ahluwalia et al., 2000; Dutta and Pullig, 2011). Next to the

importance of consequences of the crisis itself, it is equally important how the firm responds to the crisis. This eventually determines if, or to what extent, the consumer’s confidence in the brand is restored (Pearson & Clair, 1998).

Since the current crisis involves CSR, it is important to study what the effects of corporate responses are in a CSR context. VW’s reputation has been affected and the CSR claims that the company now makes can now be seen as less credible than before.

This study tries to research what an effective way of responding to a CSR crisis is. As Dean (2004) states: “To counter any loss of legitimacy, the organization must re-establish

congruency between the values implied by its actions and accepted societal norms” (p. 193).

A response is needed in order to restore the legitimacy and credibility of the organization. According to S. Bratzel, director of the Center of Automotive Management, ‘The diesel crisis

could be regarded as an opportunity to initiate a conversion to sustainable engine technology in the automobile industry’ (Schöttle & Bratzel, 2016, p. 12).

There are several studies that investigate different corporate responses to a crisis event, ranging from no response to accepting the responsibility (e.g., Bradford & Garrett, 1995). After the news emerged, VW’s CEO at that time, Martin Winterkorn, took

responsibility and admitted the admission cheating. Shortly after he resigned from the company (Cremer, 2015). VW has accepted their responsibility and have deliberately misled their stakeholders, who unknowingly polluted the environment way more than they thought they would. VW started compensating the stakeholders that were involved (Cue, 2015). Their objective now would be to re-establish the congruence between their values and the socially accepted norms. VW has to show that it does care about the environment and increase the credibility of their intentions, as Mr. Bratzel stated before. The strategy VW used in attempt to re-establish their image was introducing the new all-electric Phaeton model (Campbell, 2015). The VW Phaeton is a full-size luxury sedan and VW’s premium class vehicle, with prices starting at €86.390 in The Netherlands. The Phaeton is VW’s top model and can be compared to several luxurious cars from top brands such as the Audi A8.

The main reason for this research is to investigate whether VW’s response to

Dieselgate to recover their image, introducing fully electronic cars (starting with the most

premium class vehicle they have), is credible in the eyes of the consumer. This kind of strategy can be compared to Tesla, which also first introduced their high-end car in order to later achieve more mainstream success (to be further discussed in the literature review about luxury products). As the great body of research on crisis management (e.g., Bond &

Kirshenbaum, 1998; Coombs, 2014a; Mizerki, 1982), additional research on CSR crises and the way companies respond to the crisis in order to restore their credibility, reputation, or on how to rebuild the brand trust are needed (Wang & Yu, 2014). Wang and Yu (2014) give an advice for further research to experimentally research whether different types of CSR have different brand repair effects, because of a crisis or a bad reputation, the consumer can

become suspicious of the true motive for the CSR activity. When consumers sense that the CSR activity is solely to improve its image, CSR campaigns will not be effective and may actually backfire (Moratis, 2016a; Yoon, Gurhan-Canli & Schwarz, 2006). This forms an interesting gap for research to determine whether the used tactic is effective on the credibility of the company and the perception of their image in consumer’s eyes.

1.5 Research gap

One of the main theories on which this study is based on is the signaling theory, as reviewed by Connelly, Certo, Ireland and Reutzel (2011). The signaling theory, in this specific case also named the costly signaling theory, states that when parties are interacting with each other, the message they send will only be credible or honest when they are difficult or ‘costly’ to produce (this can be either financial or in terms of effort). This way the party makes clear that they are willing to carry the burden of these costs, which make them more credible (Moratis, 2016a). VW’s decision to make the high-end vehicle line fully electric can be seen as a signal that is not very ‘costly’ to produce. This decision by VW is the practical implication of the current research. The question now is whether this decision by VW is credible enough. As the costly signaling theory states, VW should make clear that they are willing to carry the burden. There are several other ways of a corporate strategic response that, according to the signaling theory, could be more effective and increase the credibility of the company even more. Introducing for example one of the most sold mainstream models fully electronic (such as the VW Polo or VW Golf car) would be way more “costly” for the company, which in turn may increase their credibility. Next to that, it will also send a signal as making a more popular car fully electric has a larger impact on the environment.

The goal for VW would then be to show that they are willing to carry the burden of the costs of responsive actions, in order to be perceived as being more credible (Moratis, 2016a). Instrumental and business case approaches to CSR (i.e., cutting costs, develop their business), according to the costly signaling theory, may potentially lead to harmful messages. As these approaches mainly communicate self-interest rather than a willingness to incur the costs of being socially responsible, they will possibly result in less credibility and

trustworthiness (Moratis, 2016a). Research has yet to test this empirically, which creates an interesting gap for further research. Moratis (2016b) gives an implication for further research in his paper to research signaling strategies and provide insights in the conditions under which signaling strategies are deployed in practice. Signaling strategies could depend on several conditions of the corporate environment such as media coverage of corporate

greenwashing. Greenwashing is disinformation spread by an organization as to portray an

environmentally responsible public image (Ramus & Montiel, 2005).

There are several variables that can influence the effect that is studied in the current research. Bogel (2015) argues that involvement with CSR topics leads to greater cognitive efforts in evaluating the communicated information, whereas Sen and Bhattacharya (2001) conclude that personal investment with CSR issues moderate the response to CSR

information. Personal CSR investment is formulated as sustainability orientation of the respondent in this study, as it may affect the relationship that is studied. Furthermore, as described by Moratis (2016a), when the CSR information is being perceived as out of self-interest, the credibility will be lower. As Becker-Olsen, Cudmore and Hill (2006) state, when consumers are exposed to CSR communication, they are likely to assign it to either self-serving purposes or public self-serving purposes. Past research shows that these motives are likely to influence evaluations of the message and the company (Campbell & Kirmani, 2000; Ellen et al., 2000), which is the reason why the CSR motive is also important to take into account.

Finally, research by Janssen, Vanhamme, Lindgreen and Lefebvre (2012) shows that the perception of a luxury product plays an important role in the fit with CSR. They give a specific further research advice to focus on luxury products by a brand that has been

negatively affected by an ethical scandal, which will have a lower perceived fit with CSR. As the current research does investigate this topic, the perceived luxury product is expected to have some sort of influence.

This study fills the gaps stated by giving implications of how corporate responses can help retain CSR credibility and trustworthiness, specifically by analyzing and testing VW’s response and the possible alternatives based on the costly signaling theory. As the use of CSR claims may be an effective way of countering the negative impact of a crisis, studies on the effectiveness of these claims have been conducted in previous researches (Vanhamme & Grobben, 2009). What is missing in the current literature is the effectiveness of these claims when the crisis itself emerged from CSR misconduct. The effectiveness of different responses to a CSR crisis or scandal will become clearer and also in which way it will affect the

perceptions in consumers’ eyes. Next to that, instrumental- and business case approaches dominate the current debate of CSR topics (i.e., Lee, 2008; Dahlsrud, 2008). This study tries to shed light upon the credibility and ethical approach to CSR practices, as these approaches seem to become more important and are currently less researched (Moratis, 2016a).

On the other hand this study could be helpful for managers in decision-making positions when facing a corporate crisis. Results of this study could prove that the costly signaling theory could be helpful in applying it to corporate marketing issues (originally the theory was formulated in biology/psychology), where it could facilitate better decision-making. Furthermore, more credibility oriented instead of business case approaches could lead to better insights into the moral motivations of being involved with CSR and how the

company’s stakeholders perceive this. To fill these gaps introduced the following research questions are formulated:

1.6 Research question

RQ: How would the effectiveness differ between introducing a new fully electric mainstream car model and a fully electric high-end car model as a response to a corporate social

responsibility automotive crisis on the perceived response credibility?

Sub RQ: To what extend will the sustainability orientation of the consumer, perceived CSR motive and the perceived luxury product perception moderate the relationship between the

Literature review

2.1 Corporate social responsibility

CSR has been a booming topic in the recent couple of years. Due to subjects such as global warming and climate change, companies nowadays are expected to be involved with CSR and care for the environment they operate in. The reasons for companies for being involved with CSR plays an important role in the perception of the consumers, which can differ from emerging from a company’s own values to responding to the desire of its

stakeholders. The intention of using CSR as a tactic rather than actually believing in building a better future makes a difference in the perception of the company in consumer’s eyes, which can lead to skepticism towards the brand (Ellen, Webb & Mohr, 2006). Despite the body of literature on CSR topics, the definition of CSR is still a debate among scholars (Matten & Moon, 2006). According to the article by Matten and Moon (2006), CSR consists of “clearly articulated and communicated policies and practices of corporations that reflex

business responsibility for some of the wider societal good” (p. 405). They do mention that

the manifestation and directions of the responsibility depend on the organization itself. An important addition to this definition is the one given by McWilliams and Siegel (2001), who explain CSR as “actions that appear to further some social good, beyond the interests of the

firm and that which is required by law” (p. 117). This questions the older definitions of CSR,

for example given by Friedman (1970), which states that CSR is differentiated from pure business goals of profit and the social responsibilities orchestrated by governments. The difference between these older and newer definitions could be seen as an example of the growing importance of social responsibility nowadays, as CSR nowadays goes way further than just obeying with social responsibilities required by law.

There has been a lot of work on linking CSR involvement with the firm’s (financial) performance. An example of one of these studies is the one by Margolis and Walsh (2003),

where this (as most of the studies) finds a positive link between CSR and financial

performance. Further research on this topic will not be discussed, as it is beyond the scope of this current study. More important is the connection between stakeholders and an

organization’s CSR involvement. According to the research by Akerlof and Kranton (2010), stakeholders want to affiliate with a company that they can identify with, where the observed values of the company matches the stakeholders values. If these observed values are not the actual values as in the VW case, the stakeholder could be left out and feel betrayed.

CSR involvement is nowadays considered as necessary for companies to point out their responsibility to the society as well as social and ethical standards (Lichtenstein et al, 2004). Because the involvement with CSR has become necessary for firms, it is interesting to research the effects of CSR involvement and what the consequences would be when a CSR crisis emerges. Literature on this subject states that in the occurrence of a crisis and the negative media attention that goes with it, the organization is pressured to take action and restore their image. Disclosing information about working against a certain issue could be an example of this (Islam & Deegan, 2010). As discussed in the introduction, the field of CSR is currently dominated by business case approaches. As more credibility approaches to CSR are needed (Garriga & Mele, 2004; Husted & Salazar, 2006), the current research tries to extend the knowledge on moral motivations by businesses of being involved with CSR.

2.1.1 Greenwashing

When affected by a crisis, corporate reputation and image are likely to be damaged. Literature predicts that CSR operations can enhance the corporate reputation (Brown et al., 2010). The problem occurs when organizations claim to be socially responsible when they actually are not. When this happens there is a possibility of counterproductive effects, which again can lead to negative reputational outcomes (Brown et al., 2010; Cheng & Chang,

2012). The last decades, there has been a lot of research on the topic of greenwashing (e.g., Delmas & Burbano, 2011; Laufer, 2003; Ramus & Montiel, 2005). As cited in Ramus and Montiel (2005), “Greenwashing is disinformation disseminated by an organization so as to

present an environmentally responsible public image” (p. 377). This happens to strengthen

the own brand and to create an environmentally responsible public image (Furlow, 2010).

Greenwashing has shown that it can have negative effects on consumer confidence in green

products, hurting the whole green consumer market (Furlow, 2010). Since there is a difference in what companies say and actually do, skepticism has been generated among customers (Bowen & Aragon-Correa, 2014). Furthermore, companies involved with

greenwashing have faced lawsuits for engaging in false environmental advertising (i.e.,

Honda has settled a lawsuit for false statements regarding the fuel efficiency of a hybrid car) (Lane, 2010).

The question is why companies engage in greenwashing. Companies are not lawfully required to have environmental policies and to communicate their environmental

performance to the public (Mitchell & Ramey, 2011). Delmas and Burbano (2011) conclude that because of the negative effect greenwashing has on stakeholder confidence, stakeholders become reluctant to reward companies for environmentally responsible performance. This has further impact on the companies itself, which then gain incentives to engage in

potentially hurtful social behavior and will negatively impact social welfare. This

self-fulfilling prophecy clearly illustrates the problem that companies face nowadays. In addition, Polonsky et al. (1998) found that greenwashing has led to companies that are actually telling the truth and are trying to make a positive environmental change, losing their competitive advantage. On one hand many consider CSR involvement necessary for a organization to define their role in society and apply social and ethical standards (Lichtenstein, Drumwright & Braig, 2004), on the other hand there is the risk, as described with the greenwashing effect,

to not live up to the communicated CSR responsibilities by an organization, which in turn could potentially damage the reputation even more.

The tradeoff and the consumer skepticism leads to difficult decisions for managers, where in the case of VW greenwashing could be one of the main drivers of the current crisis. As greenwashing threatens the trustworthiness component of corporate credibility

(Musgrove, Choi & Cox, 2016), VW should be careful with these kinds of claims in the future.

2.2 Crisis or a scandal?

Corporate crises and its consequences have been popular subjects in past research, as well as attempting to define a crisis. One of the definitions that comes closest to the core of a crisis is given by Coombs (2007b): a crisis is “the perception of an unpredictable event that

threatens important expectancies of stakeholders and can seriously impact an organization’s performance and generate negative outcomes” (p. 2-3). In the event of a crisis, publicity will

bring the crisis to the public’s attention, whereas publicity is regarded to be more credible and influential than a company’s own communications (Bond & Kirshenbaum, 1998). Negative information, which is the case when discussing a crisis, has a huge potential of damaging the corporate image and reputation. The reason for this, as discussed by Mizerki (1982), is that due to its credibility and negative effect there is a tendency to focus more on negative information than positive information for evaluation.

The reason that organizational crises are as heavily researched as they are is because of the impact a crisis and the response may have on the company. The cost of badly managed crises can be way more influential than the, mostly negative, counsel from the public

that cannot be predicted, but are expected” (Coombs, 2010, p. 20). Managers should

anticipate that crises could occur, especially given the VW case.

In the event of a crisis the company in question has to change their tactics and/or respond to their current situation in order to prevent or minimize the damage to the company. As seen in the previous chapter, crises have a potential to damage the corporate image. According to the Image Restoration Theory, the nature of the crisis should be considered, whereas a crisis consists of two components (Benoit & Dorries, 1996):

i. The accused is held responsible for an action. ii. The act is considered offensive.

For both conditions it is very important to note that the perceptions are more important than reality. Whether the company is in fact responsible for the crisis is not relevant, it is about whether the company is thought to be responsible for the crisis (Benoit, 1997).

The question is whether we are talking about a crisis or a scandal. De Maria (2010, p. 69) describes the difference as: “An organization can be in a crisis but not scandalized and

an organization can be scandalized without (any longer) being in crisis”. When talking about

a scandal, it involves actions that create public critique because it is considered illegal or immoral, where members of the organization intentionally undertake these actions (Carroll, 2016). Scandals can therefore be seen as an antecedent of a crisis where in the case of VW the scandal (i.e., not complying with the emission law) lead to the organization being in a crisis (i.e., facing criminal charges and declining stock value). One very important aspect is how stakeholders perceive the event or scandal, which in turn depicts if the scandal will become a crisis. The definition of a crisis given acknowledges the stakeholders’ concerns as they co-create the meaning of a crisis (Coombs & Holladay, 2011). This has also been the

case with VW, as they not only did not comply with the emission law, but also consciously violated the stakeholders’ expectations.

Because of VW’s previous reputation of being involved with different dimensions of CSR, the crises on these specific dimensions can be extra damaging. Janney and Gove (2011) conclude that when a firm possesses a good reputation on CSR dimensions, violations

regarding these dimensions are more heavily sanctioned. If the firm in crisis is known for doing well environmentally, a scandal related to the environment will lead to greater negative market reactions, in comparison to firms without this reputation. Investors will see

organizations involved with these types of scandals as being hypocrite (Tillman, Lutz & Weitz, 2009) or decoupling (Zajac & Westphal, 2004), where the actual behavior differs from the stated behavior. Like the literature depicts, this is the type of scandal that VW is involved with. Interesting here is to see how companies come out of a crisis. As the companies (CSR) reputation is damaged, responding to a crisis can become crucial in the way companies are perceived after the crisis (Janney & Grove, 2011).

2.3 Crisis response

The antecedents of corrupt and unethical business practices in which they can lead to an organizational crisis, have been heavily researched (e.g. Baucus, 1994; Simpson, 2002). Nevertheless, there is not a lot of research focused on the organization’s behavior after the occurrence of a scandal or transgression (Pfarrer et al., 2008). An important concept here is crisis management, which is defined as “a set of factors designed to combat crises and to

lessen the actual damage inflicted” (Coombs, 2007b, p. 5). As seen in the previous chapter,

market reactions to a scandal are not only influenced by its previous reputation, but also by the responses to the scandal (Janney & Gove, 2011), which are part of the crisis management set. The study of Pfarrer et al. (2008) focuses on ways of restoring the organizations

legitimacy and the reintegration with their stakeholders after a public transgression (i.e. a corrupt or unethical act that places the organizations stakeholders at risk). Stakeholders are defined as those individuals and groups that affect or are affected by organizations’ actions (Freeman, 1984; Grunig, 1992). Organizations that are most successful in restoring the legitimacy and reintegration with their stakeholders, are the companies that specifically focus on- and respond to the demands of salient stakeholder groups (or the ones that have the most legitimacy, power and urgency of claims) (Mitchell, Agle & Wood, 1997).

This paper analyzes VW’s current position according to the stage model of reintegration as originated by Pfarrer et al. (2008). This model focuses on the actions an organization needs to take in particular stages in response to stakeholder demands to try and reestablish the organizations legitimacy. The model is based on the assumption that the shareholders’ demands change in each different stage, which means that the organizations actions should also change in each different stage.

The first stage is the occurrence of a transgression and that there is public awareness of the event (also called the discovery stage). The conclusion is that the more actions an

organization takes to unveil all the information regarding the transgression, the faster the organization will complete this stage.

The second stage regards the explanation process. They state that the more actions the organization undertakes to explain the events (e.g. admitting wrongdoing, accepting

responsibility, expressing regrets), the quicker it will finish this stage.

The third stage is the punishment stage, which predicts that the more actions the organization takes to accept the punishments given (e.g. verdict acceptance, limited resistance) the quicker they accomplish this stage. There is a difference made between official and unofficial punishment. Official punishments are authority judgments such as court fines. Unofficial punishments are for example stakeholder group setting up boycotts

and negative press coverage (with the purpose of giving additional stakeholder feedback through shaming and revenge for the damage inflicted to stakeholders). The model depicts that the amount of unofficial punishment will increase when the official punishment is deemed to be insufficient.

The final stage of the model is the rehabilitation stage, where the consistency between the organizations internal and external rehabilitative actions is needed to retain legitimacy. This means that when someone responsible in the company for the wrongdoing is fired, this will give an internal signal that they are doing what they can to focus on renewal. These internal signals should also be signaled to the external stakeholders, which can be done through charitable giving or implementing new CSR measures. Inconsistency between these internal and external actions may distort the renewal process, and likely will lead to a decrease in stakeholder trust (Turner, 1976). This is exactly the phase where VW is in at the moment, where the Phaeton tactic could be one of these signals to the external stakeholders.

Pfarrer et al. (2008) claim that to which: "extent the transgressor organization

completes each of the four stages (i.e., concurrence among stakeholders is reached, meaning their demands have been met), the more likely it is to regain stakeholder support, restore its legitimacy, receive access to critical resources, and return to profitability” (p. 740). The

authors give a more extensive overview of the actions that can be taken by the organization (including moderating effects), but including the model more extensively would not

contribute to this study.

2.4 Perceived credibility

Perceived brand credibility refers to the credibility of a company’s brand as a signal, which is the extent that consumers believe in the company’s trustworthiness and expertise (Erdem, Swait & Louviere, 2002). Expertise can be seen as the competence and capability

level of a company in making and delivering the advertised product and/or service. Trustworthiness in turn is if the company can be relied upon and if it is honest and not misleading (Newell & Goldsmith, 2001). As the expertise component is less important in the current study involving a crisis, this will not be included in the current research. Companies use different marketing mix elements to signal their product, services and CSR practices for example. In the case of CSR, credibility is important for consumers as the identity of a company becomes more attractive and trustworthy when they are more similar to their own values and preferences (Dutton et al., 1994; Schneider et al., 1995).

Brand credibility has proven to be effective on the consideration process of

consumers, as well as on brand choice (Erdem & Swait, 2004). Previous research concludes that CSR has a significant positive effect on corporate brand credibility (Hur, Kim & Woo, 2014), in which this study will further research the specific situation when dealing with a scandal.

Interestingly, the same relationship has also been found the other way around, in which a brand’s credibility has a positive effect on the social responsibility image (Alcaniz, Caceres & Perez, 2010). As mentioned, a crisis threatens the legitimacy of the company and could damage the reputation and how stakeholders perceive the company. Legitimacy and reputation is often conceptualized by using the core elements of credibility and trust

(Coombs, 1998; Coombs & Holladay, 2001; Fombrun, 1996; Massey, 2001). In the research by Hur et al. (2014), credibility is a crucial variable in explaining the relationship between CSR and corporate reputation. As this variable is found as a crucial variable in the CSR relationship, the current research tries to find the same effects when a company is in a CSR crisis. This study could be seen as an implication for further research that focuses more on corporate reputation.

As shown, credibility is an important concept in crisis CSR situations. In the current study a specific form of credibility will be measured: response credibility. Measuring response credibility would be more suitable in the current study, as brand credibility will be difficult to measure in the current experimental setup. As mentioned by Reynolds and Seeger (2005), credibility is important to the believability and effectiveness of the message, whereas credibility is seen as a fundamental persuasive attribute in crisis communication. Previous research points out that credibility perceptions should be included in crisis response research (Fediuk, Pace & Botero, 2010; Kim & Cameron, 2011). Subsequently, credibility of the communicator is seen as critical to influencing individual judgments on the quality of crisis communication (Heath, 1997). Heath (1997) also mentions that: “A company suffering a

crisis must be able to tell a credible story, one that has factual fidelity that can withstand the scrutiny of reporters, governmental investigators, and concerned citizens” (p. 319).

Response credibility will be used as the dependent variable in this research, as it has proven to be one of the main aspects of crisis communication.

2.5 Signaling theory

As mentioned in the introduction of this paper, the signaling theory forms the basis for the main hypothesis in the current study. There has not been a lot of research done on the credibility of the response to a failing CSR. It is interesting to research whether several tactics are effective on the credibility of the company and the perception of their image in consumers’ eyes. To conduct research on this matter, this study is based on the signaling theory, as reviewed by Connelly et al. (2011).

The theory has originated from biology, from where it has emerged in different fields of research. The signaling theory states that when parties are interacting with each other, the message they send (the signal) will only be credible or honest when they are difficult or

‘costly’ to produce (this can be either financial or in terms of effort). This way the party makes clear that they are willing to carry the burden of these costs, which make them more credible (Moratis, 2016a). As the costs of these signals are such an important aspect of the signaling theory, some refer to the theory as the “theory of costly signaling” or the “costly signaling theory” (e.g., Bliege-Bird & Smith, 2005; Moratis, 2016).

The theory consists of four different aspects. First is the signaler, which is the person or party that is the sender of the message. Next is the actual signal that is sent, whereas the signaling theory primarily focuses on intentional communication of positive signals about a specific attribute of the individual or the company (the signaler). If the signaler does not have the specific attribute or quality associated with the signal, but still believes signaling this quality would result in some sort of advantage, it could lead to false signaling. To maintain the effectiveness and sincerity of signals, the cost of signals should be structured in a way that dishonest signals do not lead to an advantage for the signaler (Connelly et al., 2011). The final two aspects are the receiver and the feedback. In order for signaling to occur, the

signaler should benefit by a specific action by the receiver, so the signaling has a strategic effect. The current situation VW is involved in could be analyzed according to the signaling theory.

This research will analyze VW’s response to the Dieselgate scandal by introducing the new fully electric Phaeton class. VW is because of their status and performance, seen as a big international player and brand. According to the signaling theory, the signaler (VW) should send signals to their stakeholders (consumers) in order to be perceived as credible. The theory states that these signals should be costly for the company, in which the company would become vulnerable and the signal would be honest. VW’s decision to make the high-end vehicle line fully-electric can be seen as an easy choice to make, as the cost of making the luxury line electric are probably easy to earn back by integrating it in the product cost.

This could lead to the ‘cost’ of producing this tactic to be seen as low, which could lead to less credibility and perceived honesty. The willingness to be vulnerable by a large corporate company will increase the credibility and can influence the effect of the chosen strategic choice that VW has made. The question is if this tactic is effective enough and the signal by VW is perceived as “costly” enough. In the current research, an alternative to the Phaeton based on the described theory has been included, where a car response with a more

mainstream model was tested. Introducing this more mainstream model (VW Polo) would be way more “costly” for the company, which could send a more honest signal and in turn may increase their credibility.

According to research, transparency and accountability are needed to explain that ‘steps are being taken to help ensure that a similar incident will never happen again’ (Fortunato, 2008, p. 117). VW has already apologized for the mistakes and published a full-size apology called Schaamrood, in which they explain that they are not proud of their new red color (implies blushing) and have damaged the trust of the consumer (Hilderink, 2015). This ad was published in Dutch national daily papers, whereas they have also apologized in other countries with different kinds of formats (Mogg, 2015). The image was VW’s first public response to the crisis in which they apologized, which, according to research by Lyon and Cameron (2004), will sooner lead to more favorable responses than a defensive responses for example. Research outcomes about the best way to move forward from a crisis is for organizations to become ‘redemptive organizations’ that choose to rebuild (Coombs, 2007b; de Maria, 2010). VW seems to have made the right decision to apologize and rebuild the organization, but the question is whether the path they have chosen with the Phaeton is the right one.

When going back to the costly signaling theory, it can be argued that introducing a full-electric luxury line as response to a crisis can be not as credible they think it would. VW

should make clear that they are prepared to carry the burden of the costs of responsive actions, in order to be perceived as being more credible (Moratis, 2016a). As mentioned in the introduction of this paper, instrumental and business case approaches to CSR (i.e., cutting costs, develop their business) may lead to potentially harmful messages. These approaches primarily communicate self-interest rather than a willingness to incur the costs of being socially responsible; which will possibly result in less credibility and trustworthiness (Moratis, 2016a).

The costly signaling theory and the implications mentioned by Moratis (2016a), together with the previously given conclusion of Mitchell et al. (1997), form the basis for hypothesis 1. As a reminder, Mitchell et al. (1997) state that organizations that heavily weigh in and respond to the demands of salient stakeholder groups, or the ones that have the most legitimacy, power and urgency of claims, are better at restoring their legitimacy in times of transgression. This leads to the following main hypothesis of this study:

H1: In the case of the VW Dieselgate scandal, responding to a CSR scandal will be more effective on the perceived credibility when introducing a fully electronic mainstream model

(VW Polo) instead of a high-end model (VW Phaeton).

2.6 Sustainability orientation

As sustainability emerged as a hot topic over the last decades, consumer’s interest and orientation regarding this subject also grew (Follows & Jobber, 2000; Shaw & Clarke, 1999). Sustainable consumption regards a decision making process that next to individual needs and wants includes consumer’s social responsibility and the needs of future generations (Tukker et al., 2006). This trend has resulted in the occurrence of the so-called ethical consumer, who feels responsibility towards society and expresses these feelings in terms of his or her

purchase behavior (De Pelsmacker, Driesen & Rayp, 2003). Based on the theoretical background of the Elaboration Likelihood Model (ELM), as originated by Petty and

Cacioppo (1986), involvement of a person influences the processing of CSR communication. The ELM argues that people that are involved or interested in a certain topic, will process the information via the central route. This means that they will cognitively elaborate the

information and be more critical towards it. On the other hand there is the peripheral route (persons who are less involved), in which people will elaborate the information less

thoroughly and pay more attention to superficial aspects (i.e., images). Studies on this topic have found that the processing of information and the resulting evaluations depend on personal values of stakeholders (Bigne-Alcaniz, Curras-Perez & Sanchez-Garcia, 2009). Sen and Bhattacharya (2001) for example concluded that people’s personal investment in a CSR issue and their general CSR believes moderate the responses to CSR. More importantly, Bogel (2015) looks at whether a subject with high involvement with CSR spends more time with the CSR information than low involvement subjects, thus investing greater cognitive efforts. The hypothesis is supported, but was not significant. They also state that an initially negative company image might reduce motivation to process CSR information, leading to processing via the peripheral route (Pomering & Johnson, 2009) and thus a low level of cognitive effort.

As shown, the literature does not point in the same direction, but that there could be some effect of CSR involvement or sustainability orientation on the consumer is evident. For this reason sustainability orientation is included as a variable in the current research. As the literature shows, people that are more involved with the CSR topic will elaborate the message via the central route. Being involved with sustainability-oriented topics could lead to further elaboration of the message, where it is expected that the Polo car response will be seen as a signal more “difficult” to produce. This is expected to result in a higher perceived credibility

for the Polo car response when the respondent will have a higher level of sustainability orientation than people that are less sustainably oriented. These indications form the basis for the second hypothesis of this study:

H2: Responding to a CSR scandal by introducing a fully mainstream model instead of a high-end model will lead to a higher perceived credibility when the respondent will have a higher

level of sustainability orientation.

2.7 Perceived CSR Motive

As the article by Moratis (2016a) stated, CSR communication after a crisis because of self-interest rather than willingness to incur the costs of being socially responsible will lead to less credibility. This shows that the motivation of responding to a CSR scandal is

important for the potential outcomes. Further research on company motivation confirms the importance. When consumers question a company’s motivation of CSR involvement, it could elicit more persuasion knowledge, which in turn ends up in greater cognitive elaboration in the evaluation of the outing (Friestad & Wright, 1994, 1995). Skepticism can occur because the consumer could believe that the CSR involvement is primarily motivated by corporate self-interest (Speed & Thompson, 2000; Webb & Mohr, 1998). An important concept regarding research on this topic is the attribution theory (Kelley, 1972). The favorability toward a CSR program depends on certain attributions consumers make regarding the motives of the organization (Walker, Heere, Parent & Drane, 2010).

As Gilbert and Malone (1995) state: “people care less about what others do than

about why they do it” (p. 21). The attribution theory together with the persuasion knowledge

model provides a basis that consumers will try to determine the company’s motives involved with marketing communications (Becker-Olsen et al., 2006). So when consumers see a CSR

marketing communication, they are likely to elaborate on it and assign one of the two most common types of motives to the company; self-serving (i.e., to increase profits) or public serving (i.e., to raise awareness) (Becker-Olsen et al., 2006). Past research has shown that these motives are likely to influence evaluations of the company (Campbell & Kirmani, 2000; Ellen et al., 2000).

Although these researches study the effect of motivation for CSR involvement, specific research regarding motivations in CSR responses as an answer to a scandal has not been sufficiently researched. Becker-Olsen et al. (2006) do find that self-interest motivation (such as profit) does not lead to a reduced perceived credibility. This makes it interesting for this study to see how it would affect firms that just had to deal with a CSR scandal. Because the actual motive by a company to be involved in CSR activities is very difficult for outsiders to determine, the perceived CSR motive was used in this study. As the Polo car response is expected to be perceived as a more public serving CSR motive than the Phaeton, it will probably lead to a higher perceived credibility. This and the article by Moratis (2016a) are used as the foundation for the third hypothesis of this paper:

H3: Responding to a CSR scandal by introducing the electric Polo car (mainstream) instead of the electric Phaeton car (high-end) will lead to a higher perceived credibility when the

respondent will perceive the CSR motive as public serving instead of self-serving.

2.8 Luxury products

Another (and the last) important topic to discuss within this paper is luxury products. The term ‘responsible luxury’ seems a contradiction in terms, whereas the fit between luxury and CSR is an interesting phenomenon. Luxury brands and products may take a risk when being involved with CSR, in that recent research suggests that consumers do not regard

luxury and CSR compatible (Janssen et al., 2012). There seems to be little consensus in literature about the definition of luxury: “what is luxury to one may just be ordinary to

another” (Phau & Prendergast, 2000, p. 123). One of the most important defining

characteristics of luxury is their scarcity or limited availability (e.g. Catry, 2003; Dubois & Paternault, 1995). In turn, the scarcity principle posits that scarcity increases the perceived value of products, thus are more desirable and high priced than readily available ones (Cialdini, 1987).

A good example of a company using luxury pricing as marketing strategy is the car brand Tesla, which is often associated with CSR because of its environmental focus and solely electric cars. As described in the work of Mangram (2012), Tesla started out as a premium brand by using new technologies to introduce a high-end, relatively high priced ($100,000+) first model. As stated in the paragraph before, the product was scarce because of the low amount of units produced and the new technology used, which increased the

perceived value of the product. Tesla’s premium branding strategy was way different than the traditional automotive industry, which is more focused on mass marketing and production. Similar to for example Apple computers, Tesla effectively used this strategy to approach their ultimate goal of selling cars to more mainstream consumers at affordable prices (Mangram, 2012), where the production of cars for mainstream consumers has since started. As this case shows, this type of strategy can be very successful but the question is whether the effect will be the same for a company that has just encountered a CSR crisis. VW’s seems to try and copy this strategy for their fully electric car fleet by starting off electrically with a premium priced car like the Phaeton.

The previously mentioned research conducted by Janssen et al. (2012) gives an implication for further research about luxury products and CSR fit effects for familiar brands. They give a particular example of such a further research as “luxury products sold by a brand

previously affected by ethical scandals probably suffers lower perceived fit with CSR, if any”

(p. 53). The current research could test this effect as the Phaeton could represent the luxury product, where the CSR credibility can be compared with the more mainstream product. This can also show if the Tesla strategy with their premium pricing also could be effective for a company that has just suffered a CSR crisis. Cars in general can be seen as luxury products, but as the research conducted by Gartman (2004), history already shows that mere ownership of a car does not distinguish class hierarchy. The type of car conveyed status as the simple, functional car is distinguished from the more luxury makes. This makes it likely that the Phaeton will be seen as a more luxury product than the Polo car. This leads to the last hypothesis of this research:

H4: Responding to a CSR scandal by introducing a fully mainstream model instead of a high-end model will lead to a lower perceived credibility when the respondent will perceive the

car as a more luxury product.

2.9 Conceptual model

Method

3.1 General

The current research on the effectiveness of the response in the Dieselgate example to a CSR crisis on the perceived credibility of the response is a quantitative experimental

research. Next to that there are three different variables included, which act as a moderator on the main effect of the study. These variables are the perceived CSR motive, perception of a luxury product and the sustainability orientation of the respondent.

The research design for this experiment was a vignette method survey. By using a vignette survey the quantitative data can be quickly gathered and analyzed, which will give the researcher an advantage to effectively process the results. By using vignettes it is possible to manipulate the independent variables, which makes it an effective way to understand the effect of the manipulations on the dependent variable (Saunders & Lewis, 2011; Jasso, 2006). A hypothetical situation is described where the respondent’s opinion is asked. The values of the independent variable (car response) were manipulated, which have been used to examine the causal relationship. The research uses a between groups design, where the respondents were randomly assigned to the two different conditions used, which make it a true experiment (as cited in Walker, 2005). The survey was conducted via the mostly used online survey tool

qualtrics.com. Social media was primarily used to gather respondents, as this research has

used a convenience sample. Furthermore, the snowballing method has been used to contact as much people as possible (Hox, 2005). The goal was to obtain 150+ respondents, which ultimately has been reached. As the majority of potential respondents were going to be Dutch and considering the difficulty to understand CSR related topics in English, the survey has been conducted in Dutch. Multiple Dutch native speakers with an excellent level of English have carefully translated the scales from English to Dutch in order to guarantee a lower respondent dropout rate.

The survey started with a general welcoming message that introduces the survey as part of a master thesis that focuses on the effect of several responses to a CSR crisis. There was a notion of informed consent included, where the information gathered in the survey and the consent of participating in the research will remain confidential, which should lead to a smaller social desirability of answers (Saunders & Lewis, 2011). The survey started with a general explanation of CSR. This explanation included a CSR definition as well as the framework of the survey itself. Afterwards, a message about the Dieselgate scandal at Volkswagen was shown from an existing newspaper article that neutrally communicated the news. This article gave the respondent a brief overview of the scandal and its implications. Subsequently, there were two different strategies (the stimulus) that Volkswagen could apply in order to restore their own credibility. One of the conditions involved the strategy that is actually used by Volkswagen at the time this study was written, by introducing the premium Phaeton class as a fully electric vehicle. The other (hypothetical) condition involved

introducing a fully electric VW Polo line, which is one of the most popular mainstream vehicles of the brand. The actual response by Volkswagen was copied from an online article that actually exists (Nu.nl, 2015). The manipulation condition with the Polo car was the same article, but replacing the Phaeton components with the Polo car. In both of the articles small adjustments were made to communicate the minimum price and popularity of both vehicles. Both the general information and the two manipulation articles have been added to the

appendices of this study. Both of the strategies described in the manipulations were used with a purpose to restore the faith consumers have in VW. After equally dividing the respondents in one of the two conditions, they were asked questions in order to measure the variables explained further on in this chapter. The two manipulations were randomly distributed between the respondents, as methodically outlined by Jasso (2006).

3.2 Pre-tests

There were two pre-tests executed for the current research. In the first pretest (n=30) a few conclusions could be made. The first and very positive conclusion was that the

dependent variable, the newly created credibility scale, had a Crohnbach’s Alfa of 0.85, which can be seen as a good level of internal consistency. This scale was adjusted on just one item, which was an unjust formulated sentence. The most surprising conclusions came from the moderator variable scales that are extracted from existing research. The CSR motive scale had a very low Crohnbach’s Alfa, which means that the scale had to be adjusted. Also, the sustainability orientation scale had a Crohnbach’s Alfa of less than 0.7, which means that adjustments for this scale also needed to be made.

Additional feedback from the respondents regarding the first pre-test have been taken into consideration and multiple adjustments have been made, such as making the reading text shorter, changing the CSR motive from a seven-point to a five-point scale, using a different scale for sustainability orientation of the respondent (which is combined from existing scales), general formulation adjustments and translation perfection.

In the second pre-test the results had improved, which led to finally conducting the official survey. A few language and formulation adjustments were made, but not a lot of changes were needed after the second pre-test. The Crohnbach’s Alfa of the sustainability orientation scale had risen to a good internal consistency and the CSR motive scale had an adequate internal consistency.

3.3 Variables

In this section all the used variables will be outlined and discussed.

Strategic response

The strategic response by VW was the independent categorical variable in this research, whereas the respondent was placed in one of the two experimental groups (fully electric high end Phaeton versus the mainstream Polo car line).

Sustainability orientation

Sustainability orientation played an important role for the current research as a moderator. As there are not a broad variety of scales measuring sustainability orientation in the current existing literature, the best scale to use for this specific research was formulated by Maignan (2001), which consists of 5 items (i.e., I consider the ethical reputation of

businesses when I shop). In addition, one item of the scale by Mohr, Webb and Harris (2001)

was added (i.e., I think that companies carry responsibility for the society). The total scale of six items is measured on a seven-point Likert scale, ranging from absolutely disagree (1) to absolutely agree (7).

Perceived luxury product

The perception of a luxury product was a moderating variable, whereas this is measured as the product being a necessity or a luxury product. This was done on a one-item seven-point semantic differential scale of 1 (pure means of transportation) to 7 (luxury product), based on the research by Hauck and Stanforth (2007).

CSR motive

The perceived CSR motive was also used as a moderating variable in the current research, whereas the hypothesis was that the CSR motive moderates the effect of strategic response to a CSR crisis on the perceived credibility of the company. Becker-Olsen et al. (2006) have developed a scale that was used for this study. After the respondents read the crisis article and the stimulus material, they were asked to rate the perceived motivation of CSR involvement on a five-point Likert scale with three items (1= community interested /5=

self-interested,1=customer-focused/5=firm-focused and 1= socially motivated/5=profit motivated). The higher the score on this scale, the more the motive was perceived as

self-serving instead of public self-serving.

Perceived credibility

The dependent variable in this study is the perceived credibility of the response according to the information given in the different manipulation conditions. As the existing literature mostly uses credibility scales that focus on the company or brand, a new credibility scale was formulated that measures the credibility of the response. The new scale consists of five items (i.e., Volkswagen’s response is a suitable solution for the problems created). Additionally, two items from the scale used by Coombs and Holladay (1996) were used (i.e.,

I think that the message in VW’s response is clear). Also one item from Turker (2009) was

used (i.e., VW is committed to take social responsible actions). The final scale of eight items was measured on a seven-point Likert scale, ranging from absolutely disagree (1) to

Manipulation check

The survey also consisted of a manipulation check, where after the stimulus material there was a question asked which car response option was used according to the article. Respondents who did not correctly answer this question in compliance with the manipulation group they were in, were not included in the research.

Control variables

To control for other factors that can influence the relationships that were researched, several control variables were added. The widely used control variables age, gender and education were used, as well as the knowledge of Dieselgate’s existence and the ownership of a VW car. As assumed that most people are aware of Dieselgate’s existence, their a-priori perceptions of the scandal or the brand could have affected the outcomes of the study. Also being the owner of a VW car could have led to diminishing effects.

3.4 Sample

A total of one hundred and sixty five respondents (N=165) completed the survey. The convenience sample and the gathering process have already been described in the general method section. The respondents that did not complete the survey at once were given a week to continue finishing the survey before their data got removed. In total there were ninety-six unfinished surveys gathered (N=96). A critical factor in the current research is the

manipulation check, whereas remembering with which car VW responded to the scandal is of key importance to analyze the results. The survey consisted of one item where was asked which car VW plans to produce fully electric, to check whether the respondent read the manipulation text in a correct matter. All the respondents that were in the Phaeton manipulation text that have answered Polo to this question have been taken out of the

analytical part. The same goes for the other way around, all the respondents that were in the Polo manipulation that answered the question with Phaeton were also taken out of the analytical part of this thesis. By doing this it can be concluded that the respondents had the right kind of response in mind when filling in the survey, which increases the validity of the research. After having conducted the manipulation check, a total of one hundred and fifty one respondents were included for the actual analysis (N=151). This means that fourteen

respondents (N=14) did not pass the manipulation check and were not included in the

analysis of the results. The sample size largely exceeds the minimum of 100, as suggested by Hair, Anderson, Tatham and William (1998).

From the total included respondents (N=151), the average age was 27 years old (SD=7.47) where the ages ranged from 20 up to 60 years old. A total of 57% of the

respondents were male (N=86) and 43% female (N=65). Out of the respondents, the largest group of 69.5% studied academic education/university (N=105), whereas higher professional education was the second largest group (N=37). Most of the respondents knew about the Dieselgate scandal in advance (94.7%, N=143), where only eight respondents did not. Finally, 6,6% of the respondents owned a Volkswagen car themselves (N=10).

Results

The data gathered from the survey was used for the analytical part. First the general data characteristics will be discussed, like the process of missing values, reliability of the scales and the correlation matrix. Subsequently, the results from the regression analysis and the hypothesis testing will be outlined.

4.1 Statistical analysis

First step in the process was checking for missing values. After analyzing the results it can be concluded that there were no missing values gathered. This is probably due to forcing the respondents to answer each question. The survey was designed in such a way that

counter-indicative items were already accounted for in the answer possibilities.

Reliability analysis

A principal axis factoring analysis (PAF) was executed on all the items used in the survey. The factorability of all the three variables was examined, which consisted of 17 items. All the 17 items correlated at least .4 with another item, which suggests reasonable factorability. The Kaiser-Meyer-Olkin (KMO) measure verified the sampling adequacy for the analysis, KMO = .8. Bartlett’s Test of Sphericity (BTS) ² (136) = 1135.062, p < .001, which indicates that the correlations between items are large enough for PAF. Both tests indicate that a factor analysis is suitable for this research (Hair et al., 1998; Tabachnick & Fidel, 2007). Given these indications, the factor analysis was executed with these 17 items.

The initial eigenvalues show that the first factor explained 27% of the variance. The second factor explained 19% and the third factor 9%. The fourth factor had an eigenvalue of just over one, but will not be included in this research. The three factor solution, which explained 56% of the variance, was preferred because of the theoretical foundation found, the

leveling off that occurred with the eigenvalues on the scree plot after three factors in

accordance with Kaiser (1974), the insufficient value of factor loadings and the difficulty to interpret the fourth factor. When conducting the PAF with the 17 items, an Oblimin rotation was used. All the items used had principal loadings over 0.4.

To examine the consistency of measures used in this research, reliability checks were conducted for all the scales used (i.e. CSR motive, sustainability orientation and response credibility). Cronbach’s Alpha was measured for all of the scales, which gives an indication of the internal consistency. If Cronbach’s Alpha is > 0.8, it indicates that there is a high level of internal consistency (George & Mallery, 2003). Cronbach’s Alpha’s above 0.6 are seen as adequately consistent (George & Mallery, 2003; Murphy & Davidshofer, 1988). In some situations removing an item in the scale leads to a better consistency, which was also checked in the analysis.

No. of items

M (SD) Skewness Kurtosis Alpha Response credibility 8 3.59 (1.12) -.15 -.49 .86 Sustainability orientation 6 5.48 (.89) -.75 .64 .82 CSR motive 3 3.92 (.72) -.26 -.32 .67

Table 1: Descriptive statistics for the three factors.

As shown in table 1, the scales Response Credibility (N=8) and Sustainability Orientation (N=6) both had a high reliability (>.80). CSR motive (N=3) can be seen as an adequate reliable scale (>.60). The corrected item-total correlations indicate that all the items

Deleting one of the items for the different scales did not result in a higher reliability. The final moderator, luxury perception, was a one-item scale and was not included in the factor analysis.

In general, these analyses resulted in these three factors, which were also in accordance with previous theoretical research. Therefore, they have also proven to be moderately to very internally consistent. The generated data was well suited for statistical analyses.