The influence of tax planning activities on corporate marketing outcomes

(1)

The influence of tax planning activities on

corporate marketing outcomes

Master Dissertation - Dual award Master of Science

Advanced international business management and marketing

Submission date:

Personal details:

5 December 2016

Stijn Hollink

Supervisors:

Sint Lucassstraat 8

mr. drs. H.A. Ritsema

9718LR GRONINGEN

dr. S. Bhattacharya

Student numbers:

Contact details:

S2363208

s.j.j.hollink@student.rug.nl

B150687986

s.hollink2@newcastle.ac.uk

(2)

(3)

Abstract

The aim of this research is to identify the effect of corporate tax strategies and tax planning on corporate marketing outcomes. The interdisciplinary approach used to label aggressive corporate tax strategies as negative CSR helped this study to draw on existing bases of knowledge. Due to the increased media attention and information leakages, the consequences of corporate tax strategies are bigger than ever. Since most studies focussed on direct financial consequences of tax planning, this study focusses on consumer reaction to tax planning.

Using an experiment (N=170), the effects of corporate tax strategies on various elements of corporate marketing outcomes were analysed. The analysed elements are; Corporate Brand Credibility, Corporate Reputation and Consumer satisfaction. The results suggest that compared to a positive, an aggressive corporate tax strategy can negatively influence all three elements of corporate marketing outcomes and vice versa. This relation is influenced by personal attitudes towards tax and personal tax moral.

Give the increased customer interest in CSR and in corporate tax strategies, combined with the increasing level of media attention for tax planning, this research’s findings advocate a change of corporate tax practices. Corporations should not only focus on financial gains when designing a tax strategy but should also address the broader consequences of aggressive or positive tax strategy. This could be achieved by setting up a more multi-disciplinary tax department.

(4)

Acknowledgements

This dissertation project is the final part of my dual award master at the University of Groningen and the Newcastle University Business School. Before continuing, I would like to thank some people for their contribution to this dissertation.

(5)

This research addresses the relationship between corporate tax planning strategies and related marketing outcomes. In order to do so, it examines the consequences of corporate tax planning on; corporate brand credibility; corporate reputation; and consumer satisfaction. These concepts are chosen because they form the core of consumer based perceptions, and therefore are important determinants of corporate marketing outcomes (Hur et al., 2014). This introduction will first address the research background. After highlighting the background, the managerial relevance will be addressed. Subsequently, the theoretical relevance will be explained. Doing so, this introduction will also introduce the research gap. The introduction will then continue with stating the research question. The chapter is finalised with discussing the structure of the remainder of this study.

1.2 Background of tax planning

Based on the decreasing governmental power (Christensen and Murphy, 2004) around the globe and the fact that the capabilities and wealth of (multi-national) corporations are clearly on display (Donaldson and Walsh, 2015), consumers expect corporations to engage in CSR activities that go way beyond Friedman’s shareholder's theory (Friedman, 1970). There is increasing attention for the way in which multinational corporations contribute to modern day societies. One of the ways in which corporations try to mitigate this increased attention is by emphasising their CSR activities in their annual reports. The payment of their “fair share of taxes” and assuring “tax transparency” is a commonly used reasoning to convince customers of their good intents and can be found in annual reports of large corporations (EY, 2013; Unilver, 2016; Vodafone, 2016). However, these good intents aside, corporations increasingly engage in tax planning activities. While using sophisticated tax structures, multinational corporations can reduce their tax liabilities drastically. As a consequence, items regarding their aggressive tax planning activities or corporate tax strategies often dominate the international financial news on multinational corporations (Johnson, 2015). The more and more problematic trade-off between financial and social responsibilities connected to corporate tax are fruitful ground for debate but also for academic research.

1.3 Managerial Relevance

(9)

media coverage, combined with the increasing chances of getting caught may imply that corporation’s product market outcomes might be adversely affected by their tax planning activities (Graham et al., 2014).

Lately, the discussion on tax evasion and the usage of tax havens has also gained traction with international policy makers (Riedel et al., 2015). The fact that the European Commission, led by the EU’s commissioner for competition Margrethe Vestager, is cracking down on tax avoidance and employs different initiatives to battle harmful tax practices can be seen as an example of this increased attention (OECD, 2000; OECD, 2004; Oliver, 2015). The case of Starbucks which was affected severely on its commercial side in the UK by tax activists (Jopson, 2012) and was recently fined for receiving illegal governmental support due to favourable tax rulings in The Netherlands (Lennihan, 2015) is a showcase of both these pressures on a corporations CTS. Taken together, these developments urge corporations to rethink their tax strategies and to acknowledge that tax planning consequences are not restricted to financial ratios alone.

1.4 Theoretical relevance

After multiple studies the influence of CSR activities on corporate marketing outcomes is clear. These studies, for instance, emphasise the importance of being good corporate citizens (Matten and Crane, 2005). However, these studies mostly discard the influence of CTSs. The number of studies on the relation between CTSs and consumer behaviour and associated corporate marketing outcomes is limited, and the research setting is solely limited to the United States. Therefore, this study which examines this relation can be seen as a valuable addition to theory. Understanding the influence of CTSs on corporate marketing outcomes is important for a number of reasons. First, taxes are a major part of corporations’ liabilities and therefore materially affect the profitability of corporations. Corporate tax rates fluctuate between 12,5 and 34,43% across Europe (OECD, 2016). Second, there is no academic consensus on the influence of CTS. Some studies consider it as negative CSR (Erle, 2008; Friese et al., 2008; Lanis and Richardson, 2012), while others disagree with this negative classification (Gassner, 2001; Freedman, 2004). Finally, the increased freedom of multinational corporations on the global market due to decreased governmental power gives corporations more freedom to plan their tax strategies (Christensen and Murphy, 2004; Scherer and Palazzo, 2011).

(10)

1.5 Research question

The aim of this study is to explore the possible influence CTS can have on corporate marketing outcomes according to existing literature and to examine to what extend this influence can be found in practice. Therefore, the main research question of this study is as following; To what extent do tax planning activities influence corporate marketing outcomes? 1.6 Structure of the paper

(11)

2 Literature review 2.1 Introduction

The goal of this literature review is twofold. First, the descriptive part will further introduce the concepts of CTSs and CSR. The descriptive part will also discuss the labelling of CTSs as CSR. Thereafter, the concept of corporate marketing outcomes will be elaborated on and divided into relevant elements. From paragraph 2.7 onwards, the second part of this chapter unfolds. This part will focus on discussing the state of, and reviewing, the current research on the elements that together form corporate marketing outcomes. Based on this review, hypotheses for each of the three concepts will be developed individually. Possible moderators and the related hypotheses will be addressed at the end of this literature review.

2.2 Corporate Tax Strategies

(12)

Not included in study

Included in study

Illegal

Legal

Corporate tax Strategies

Tax Evasion

Tax Avoidance

Tax Planning

Declaring less profits

Use conflicting tax

treaties Structuring group entitities

Overstating deductions Use transfer pricing that

cause tax base erosion mind when managing a Keep fiscal aspects in

firm

Conceal interest in

offshore entitities Profit shifting into low tax countries Assessing possible societal recations

Figure 2.1 - Clearifying types of tax activities This study will define tax avoidance as any activity that reduces a corporation’s cash effective tax rate over a longer period of time Dyreng et al. (2008) and includes all transactions from investing in a municipal bond to engaging in tax shelters (Hanlon and Heitzman, 2010). In order to classify different types of tax planning, this study uses the typology of (Hardeck and Hertl, 2014) They distinguish two types of so-called Corporate Tax Strategies (CTSs) (Hardeck and Hertl, 2014). CTSs incorporate the preliminary strategic decision making as well as the actual tax avoidance activities (Hardeck and Hertl, 2014). These tax corporate tax strategies can be labelled aggressive or responsible. When using aggressive CTSs, corporations interpret tax laws aggressively and exploit loopholes (OECD, 2008; Hardeck and Hertl, 2014). Therefore, aggressive CTSs do not comply with the original intention of the legislator (OECD, 2009). Responsible CTSs, on the other hand, intend to not only comply with the letter of the law but also with the intention of the legislator (Hardeck and Hertl, 2014). These responsible CTSs also refrain from exploiting loopholes to diminish tax expenditures. (OECD, 2008).

(13)

engage in aggressive CTSs are much more divergent. Hanlon and Slemrod (2009) find that the industry firms operate in, influences the extent to which they engage in aggressive CTSs. They also find that poorly governed firms are less likely to engage in aggressive CTSs (Hanlon and Slemrod, 2009). Graham et al. (2014) find that reputational concerns along with accounting incentives are the primary motives for corporations not to engage in aggressive CTSs. Other factors that influence the extent to which corporations engage in aggressive CTSs can be found in; influence of individual executives (Dyreng et al., 2010);, product market power (Kubick et al., 2015); aggressiveness of business strategy (Higgins et al., 2015); consumer concentration (Huang et al.); and the extent to which corporations hold large amounts of consumer-based brand equity (CBBE) (Austin and Wilson, 2015). Although the outcomes of these studies are somewhat mixed, it is justifiable to say that the incentives for companies to engage, or not to engage, in aggressive CTSs are clear and well researched. What is unclear and received almost no attention in previous research is one important consequences of an aggressive CTSs. Namely, the reaction of the consumer to CTSs and therefore the way CTSs influence corporate marketing outcomes. To be able to build this part of the research on a mature field of study it is worthwhile to see if a link between CTSs and CSR can be established. This link will be addressed in the next paragraph.

2.3 Corporate social responsibility and Corporate Tax Strategies

The field of Corporate Social Responsibility (CSR) is a very mature field of study. There is rich and useful literature on the influence of CSR on certain consumer behaviour and associated corporate marketing outcomes. For this study to build on this literature, CTSs has to be linked to CSR. This link will be clarified, and the end of this section after the different CSR definitions are discussed, and relevant literature is addressed. CSR can be defined in multiple ways, from ‘‘how business takes account of its social and environmental impacts in the way it operates, maximizing the benefits and minimizing the downsides’’ (UKGovernment, 2004p.3) to ‘‘the continuing commitment by business to behave ethically and contribute to economic development while improving the quality of life of the workforce and their families as well as of the local community and society at large’’ (Holme and Watss, 2006 p.10). One of the field’s most cited definitions comes from Carroll; “the social responsibility of business encompasses the economic, legal, ethical, and discretionary expectations that society has of organisations at a given point in time’’ (Carroll, 1979p. 500). Because Carroll’s framework, that goes along with his definition offer the best opportunities to link CSR to CTSs his definition of CSR is most suited for this study. Although his reference is rather old, it’s still commonly used in studies to date (Sheehy, 2015)

(14)

certain dichotomy in the literature on this link between CSR and CTS. The majority of authors consider aggressive CTSs as negative CSR (OECD, 2004; Erle, 2008; OECD, 2009; Preuss, 2010; Jenkins and Newell, 2013; Hardeck and Hertl, 2014). However, there are some authors in the field who dispute this (Gassner, 2001; Freedman, 2004; Hanlon and Slemrod, 2009). Supporters of labelling aggressive CTSs as negative CSR refer to decreasing power of governments to prevent the usage of aggressive CTSs (Hardeck and Hertl, 2014). Subsequently, they address the implications the decreased revenues have on their public services (Scherer and Palazzo, 2011; Hardeck and Hertl, 2014). Besides this, supporters find evidence that corporations who engage in negative tax practices are more likely to engage in tax avoidance too (Chun et al., 2013). The work of (Christensen and Murphy, 2004) is also in line with the labelling of aggressive CTS as negative CSR. They address a change of opinion of the general public towards tax avoidance. According to them, it is, therefore, logical that taxation will become an issue of increased importance (Christensen and Murphy, 2004). Opponents of labelling aggressive CTSs as negative CSR stress the fact that consumers benefit from corporations focussing on reducing tax liability because it ensures competitiveness and increases the chance of long-term success (Freedman, 2004). Based on the literature there is also a possibility that the qualification consumers give to a CTSs depend on their characteristics. Consumers with a negative tax attitude could be more likely to have sympathy for corporations that use aggressive CTSs (Freedman, 2004). On the other hand, customers might be averse towards responsible CTSs because they fear that corporations shift the related costs to them (Gassner, 2001; Freedman, 2004). Despite the theoretical endeavours in the field, there are limited empirical studies to the relationship between CSR and CTSs. An empirical study on the relationship between CSR and CTSs worth mentioning is the study of (Lanis and Richardson, 2012). They find that the level of CSR disclosure is a strong indicator of the extent to which corporations engage in aggressive CTSs (Lanis and Richardson, 2012). The most important empirical work on the relationship between CSR and CTS is the study of Hardeck and Hertl (2014). This study will be addressed in the remainder of this section.

(15)

(Schwartz and Carroll, 2003). Finally, the ethical domain ensures that corporations meet the expectations of members of society which are not codified in order to protect stakeholder’s moral rights (Schwartz and Carroll, 2003). The first two domains are required by society according to (Schwartz and Carroll, 2003) whereas the latter one is expected by society.

The study will now address the three domains and their relation to CTSs step-by-step. Doing so it will show that an aggressive CTS does not meet all three domains as introduced by (Schwartz and Carroll, 2003). Although aggressive CTSs matches with the economic and partly with the legal domain of the CSR pyramid by aiding the maximization of profit and obeying the law, aggressive CTSs falls short when it comes to the ethical domain (Hardeck and Hertl, 2014). It matches only partly with the legal domain because the behaviour of corporations engaging in aggressive CTSs should be qualified as opportunistic compliance (Schwartz and Carroll, 2003). This is due to the fact that the corporations exploit loopholes in legislation and take advantage of the lack of governmental power on the global market (Scherer and Palazzo, 2011; Hardeck and Hertl, 2014). Due to the inappropriate use of power differences, large corporations are able to use aggressive CTSs to shift the tax burden to smaller corporations or consumers. Therefore aggressive CTSs does not meet the requirements of the ethical domain as introduced by Schwartz and Carroll (2003). Because aggressive CTSs fail to meet the ethical and part of the legal domain they can be labelled as a negative CSR activity (Hardeck and Hertl, 2014). Responsible CTSs do not violate the ethical and legal domain, and can therefore be labbeld as positive CSR activities (Hardeck and Hertl, 2014).

This theoretical endeavour to label aggressive CTS as negative CSR is also tested empirically. In two separate studies Hardeck and Hertl (2014) find corporations that engage in aggressive CTSs are negatively evaluated on their CSR activities by consumers. For positive CTSs they find this to be the other way around (Hardeck and Hertl, 2014). Based on the studies mentioned above the link between CTS and CSR is established.

2.4 Consumer Reactions to Corporate Tax Strategies

As mentioned above, the literature on consumer reactions to corporate tax strategies is limited. Because this study conducts an experiment that is based on the engagement of consumers the study has to find some theoretical foundation for this reaction to justify this approach. Therefore, it will first address the studies that have found evidence for these reactions to aggressive CTSs. After that, it will address relevant CSR-research to provide this study with the additional theoretical foundation. The choice for CSR-research is based on the reasoning mentioned earlier that aggressive CTSs can be labelled as negative CSR. The rich CSR provides enough literature on the reaction of consumers to negative CSR; this study will now address the key finding of these two areas of literature.

(16)

and Slemrod (2009). In their research, they evaluate stock prices of corporations caught in tax shelter involvement. Compared to other industries they find a severe consumer/taxpayer backlash for the retail industry that deals directly with consumers (Hanlon and Slemrod, 2009). This indicates that consumers are indeed sensitive to the extent to which corporations engage in aggressive CTSs (Hanlon and Slemrod, 2009). In light of the research design of this study, their findings are of particular relevance because it indicates that evidence for consumer reactions can be found in studies that do not engage with consumers directly. The second relevant study that indicates that CTSs matters to consumers is the study by Austin and Wilson (2015). They also use a research design which does not require them to engage with consumers. They evaluate the amount of consumer-based brand equity (CBBE) corporations hold and the extent to which these corporations engage in tax planning (Austin and Wilson, 2015). They find that corporations that hold valuable CBBE refrain from using aggressive CTSs. They do so due to prevent consumer backlash. This is another clear indication that consumers indeed respond to variations of CTSs.

The first CSR research that needs to be addressed in this case is the study by Mohr and Webb (2005). In their United Stated based experiment to the effects of CSR on consumer responses, they find that the way consumers evaluate corporations is affected positively trough high levels of CSR. They also found that a lower price cannot compensate a low level of social responsibility. Next to these finding Mohr and Webb (2005) also find that the negative effect of low levels of CSR is greater than the positive effect of high levels of CSR. In this case, high levels of CSR match with positive CSR and low levels of CSR match with negative CSR.

A second CSR study relating CSR and consumer response was performed by Abdeen et al. (2016). The purpose of their study was to investigate the relation between CSR beliefs and purchase behaviour (Abdeen et al., 2016). Their New Zeeland-based quantitative research finds that philanthropic, ethical and legal CSR beliefs can be associated with support intentions held by consumers. Only ethical CSR beliefs have the power to influence purchasing behaviour (Abdeen et al., 2016). In light of this study, their research is useful because it indicates that CSR activities deployed by corporations indeed influences consumer behaviour.

2.5 Corporate marketing outcomes

(17)

corporate identity, is powerful but necessarily limited. Besides moving from individual concepts to an institutional-wide focus, the power/balance in marketing has also shifted. This shift was first described by Greyser (1997) who distinguished three power relations; (1) the manipulative model, (2) the service model and (3) the transactional model (Greyser, 1997). These three relations differ in the allocation of power. In the first, the marketer holds most of the power enabling him for instance to create consumer desires. In the second and the last the power has shifted to consumers. Marketers, in this case, have to service the consumer (Greyser, 1997). This model was complemented later by Balmer and Greyser (2006). They introduced the fourth relation namely the corporate model or exceptional stage (Balmer and Greyser, 2006). According to them the consumers and the other stakeholders hold power in this model. The marketer has to work with them and acts as a stakeholder servant. This corporate marketing role implies that corporations should not only focus on profit but should also be concerned with meeting societal needs to ensure stakeholder support (Balmer and Greyser, 2006). Next to that stakeholders also expect corporations to carefully balance current societal needs with those of future generations (Balmer and Greyser, 2006). This extended model provides a fruitful foundation for this study’s allegation that corporations should pursue more goals than profit maximisation.

(18)

2.6 Corporate Tax Strategies and corporate marketing outcomes

After addressing the research done to date on the relationship between CTSs and CSR and the concept of corporate marketing it’s time to address the influence CTSs can have on corporate marketing outcomes. The importance of engaging in positive CSR activities can be explained using the resource-based perspective (RBP). RBP describes the relation between a corporation’s characteristics and its performance (Hur et al., 2014). A corporation’s resources determine how a corporation can transform inputs into outputs and services (Hur et al., 2014). A sustainable competitive advantage is thus created by manipulating and controlling the resources of the corporation (Hur et al., 2014). In light of the sustainable competitive advantage intangible resources such as culture and corporate reputation are of specific importance because they are rare and hard to imitate and or substitute (Hur et al., 2014). According to the RBP, CSR activities can influence these intangible resources and therefore the business benefits of a corporation (McWilliams et al., 2006). CSR affects a consumer’s attitude towards a corporation (Pivato et al., 2008). CSR influences consumer satisfaction through higher levels of “corporation – consumer” identification in case of corporations that engage in positive CSR activities (Luo and Bhattacharya, 2006). These higher levels of identification can enhance a consumer’s desire to support a corporation (Dutton et al., 1994). This implies that the more a corporation engages in positive CSR activities, the greater the consumers’ motivation is to support a corporation and to identify themselves with the corporation (Lichtenstein et al., 2004; Hur et al., 2014). This increased identification and support will lead consumers to react to a corporation’s positive characteristics and react increasingly favourable towards a corporation. Therefore, it’s likely that engaging in positive CSR activities will boost corporate marketing outcomes (Hur et al., 2014). The three concepts related to corporate marketing outcomes that will be addressed in this study are; Corporate brand credibility; corporate reputation; and the more consumer based concept of customer satisfaction. After addressing these related concepts individual hypotheses will be developed for each construct. These hypotheses will describe the expected relation between the concept and CTSs.

2.7 Hypothesis development

This section will address the influence of the different levels of CTSS on the three dependent variables. The section builds on the previous chapter which described and discussed the related concepts. Besides this, the previous section enables this section to label aggressive CTS as negative CSR. At the end of each paragraph, the related hypotheses are stated.

2.7.1 The effect of CTSs on Corporate Brand Credibility

(19)

this is related to the extent to which consumer experience a brand as trustworthy and value it’s expertise (Erdem et al., 2002). The brand can also be used to transfer a symbolic meaning to consumers which help them with communicating their values and beliefs (Chaplin and Roedder John, 2005). Compared to other marketing elements, the enduring influence of a brand itself is distinctive because it incorporates a corporation’s past and present marketing strategy (Klein and Leffler, 1981).

Next to their experience with a product, consumers also care about the influence of their consumptions on other stakeholders and the community as a whole. Therefore, it’s likely that they identify themselves easier with corporate CSR activities (Bhattacharya C. and Sen, 2003). The studies of Dutton et al. (1994) and Schneider et al. (1995) both found that consumers find corporations more trustworthy and attractive when a corporation’s set of beliefs matches their own. Therefore, CSR activities that address social issues that are relevant to a consumer can make a corporation favourable with a consumer (Pivato et al., 2008). Relevant CSR activities also proof to be an indication of greater management competency (McWilliams and Siegel, 2001). Because of these two reasons CSR can increase the trustworthiness and expertise of a corporation and can, therefore, increase corporate brand credibility (Hur et al., 2014).

With the link between CSR and CTSs, established in section 2.3 in mind the study can now turn to the influence of CTS on corporate brand credibility. Similar to CSR, responsible CTSs might be a good match with a consumer’s personal beliefs. Responsible CTSs is also likely to be qualified as greater management competency by consumers. Therefore, this study argues that responsible CTSs positively influence corporate brand credibility. Consequently, this study hypothesises that aggressive CTSs negatively influences corporate brand credibility. Therefore, the first hypothesis of this study is stated as follows:

H1a: Aggressive CTSs have a negative influence on corporate brand credibility. H1b: Positive CTSs have a positive influence on corporate brand credibility. 2.7.2 The effect of CTSs on Corporate Reputation

A second important determinant of corporate marketing outcomes is corporate reputation (Hur et al., 2014). Corporate reputation is a valuable intangible asset because it signals a corporation’s main characteristics and can, therefore, be a source of sustainable competitive advantages (Branco and Rodrigues, 2006; Melo and Garrido-Morgado, 2012). Corporate reputation is of specific importance when consumers lack specific product information because they will project the information they do have on the product (Schnietz and Epstein, 2005). Corporate reputation also helps to protect a corporation from negative consumer perceptions (Lange et al., 2011).

(20)

(Melo and Garrido-Morgado, 2012). Because of this CSR is seen as, and justified as, the most effective way to improve a corporation’s reputation (Jones, 2005; Porter and Kramer, 2006). One of the studies on this subject worth mentioning is the study of Lai et al. (2010) which shows that corporate reputation is positively related to the consumer’s perception of CSR engagement. Next to this, research of Fombrun (2005) indicates that corporate CSR engagement enhances corporate reputation extrinsically.

Based on the CSR research the study can turn to the influence of CTSs on corporate reputation. As argued in section 2.3 aggressive CTSs can be labelled as negative CSR. Therefore, this study argues that aggressive CTSs are likely to negatively influences corporate reputation. On the other hand, this study hypothesises that responsible CTSs could influence corporate reputation positively.

H2a: Aggressive CTSs have a negative influence on corporate reputation. H2b: Positive CTSs have a positive influence on corporate reputation. 2.7.3 The effect of CTSs on consumer satisfaction

A third important determinant of corporate marketing outcomes is consumer satisfaction. This more consumer based concept is chosen because of its fit with the other two concepts. Besides that, it ensures that this study also captures some direct consumer measures related to CTSs. Consumer satisfaction refers to the extent to which a consumer has positive emotional evaluations of, and is pleased with a supplier (Flint et al., 2011). Although there are other definitions of consumer satisfaction, the above is best suited for this research because it is consistent with more than 50 years of empirical research on consumer satisfaction (Oliver, 1980).

The positive emotional evaluations that are the core of consumer satisfaction enabled previous studies to link consumer satisfaction to CSR. As a reaction to perceived positive CSR consumers are likely to favourably evaluate their attitudes towards a firm (Aaker, 1996; Sen and Bhattacharya, 2001; Mandhachitara and Poolthong, 2011). The link between CSR and consumer satisfaction was first tested empirically by Luo and Bhattacharya (2006). Their results show that a firm’s CSR activities increase consumer satisfaction. These empirical results offer enough foundation for this study to connect positive CSR activities with higher levels of consumer satisfaction (Luo and Bhattacharya, 2006).

(21)

impact of positive CTSs on consumer satisfaction. Hence, the following hypotheses are put forward.

H3a: Aggressive CTSs have a negative influence on consumer satisfaction H3b: Positive CTSs have a positive influence on consumer satisfaction. 2.8 Possible moderators and hypotheses development

Based on existing literature it is likely that the described relations will have moderators. The first important research to address here is the study conducted by Sen and Bhattacharya (2001). They find that the response of consumers to CSR activities deployed by corporations is influenced by the congruence between the perceived character of a corporation and their character (Sen and Bhattacharya, 2001). Based on the above-described linkage between CSR and CTSs it is, therefore, likely to expect a congruence between consumers and corporations depends on personal tax mentality. A consumer’s tax mentality is based on some important building blocks. Hardeck and Hertl (2014) find that among these building blocks, personal tax moral and personal attitude toward taxation have a particularly high impact. This study will first introduce the possible moderating function of personal attitude towards tax avoidance and will then introduce personal tax moral and it’s possible moderating function.

2.8.1 Personal attitude towards tax planning

(22)

H4a: A negative personal attitude towards tax planning will strengthen the influence of CTSs on corporate brand credibility

H4b: A negative personal attitude towards tax planning will strengthen the influence of CTSs on corporate reputation

H4c: A negative personal attitude towards tax planning will strengthen the influence of CTSs on consumer satisfaction

2.8.2 Tax morale

As mentioned a consumer’s tax morale could also be an important moderator. The difference between personal attitude towards tax planning and a consumer’s tax morale, in this case, lies in the legality of the tax strategy. Besides matching with a consumer's beliefs, a consumer’s attitude towards the illegal evasion of tax can also influence the relation between a corporation’s CTS and its corporate marketing outcomes. Schmölders (2006) defines tax morale as the attitude taxpayers have towards the idea that others have an obligation to meet their tax liabilities. However, in light of the argumentation earlier in this study, the definition of Hardeck and Hertl (2014) is better suited. They define tax moral as the consumer’s attitude towards legal practices aimed at reducing tax liabilities (Hardeck and Hertl, 2014). Based on the labelling of aggressive CTSs as negative CSR it is likely that consumers with high tax morale, implying that in their opinion corporations have a strong obligation to fulfil their tax liabilities, will penalise corporations that use aggressive CTSs. Based on this line of reasoning the following hypotheses is stated:

H5a: A high tax morale will strengthen the influence of CTSs on corporate brand credibility H5b: A high tax morale will strengthen the influence of CTSs on corporate reputation H5c: A high tax morale will strengthen the influence of CTSs on consumer satisfaction 2.9 Conceptual Framework

Based on the concepts as introduced above and the described moderators a possible conceptual framework is drawn in figure 2.1.

(23)

3 Methodology 3.1 Introduction

To answer the research question and test the hypotheses, a suitable research design has to be developed and described. The methodology part of this study will address this. It will start with an introduction of the research strategy and the related research paradigm. After this introduction, the methodology will be explained in greater depth. Finally, it will address the constraints of the chosen methodology and ethical issues this research encountered.

To answer the relevant research questions and to examine the relationships mentioned in the conceptual framework, this study will conduct an experiment. Using fictitiousness news articles about corporations that deploy CTSs this study will examine the reaction of consumers on CTSs deployed by corporations. Based on the literature review from the previous chapter, the fields of corporate tax strategies, CSR and corporate marketing outcomes can be labelled as mature and well explored. This labelling justifies the quantitative approach uses in this study to examine to what extend the possible relations mentioned in the literature review can be found in practice (Blumberg et al., 2014).

3.2 Research Paradigm

(24)

study to draw conclusions and explain principles based on factual data. The experiment as a technique to collect data also aligns well with the chosen research paradigm (Blumberg et al., 2014). Now the reasoning behind the research is clear the methodology itself will be addressed. 3.3 Research design

Because corporations hardly ever disclose tax planning strategies, it’s complicated to collect data that enables this study to examine the relation between CTSs and corporate marketing outcomes directly from these corporations. This is solved by organising an experiment. The benefit of an experiment is the possibility to provide consumers with the required information.

This study uses a two times one between subject design. This means that the study varies the level of CTSs aggressiveness (positive or aggressive) in only one industry. The between subjects design implies that participant can only be part of one of the treatment groups, either the aggressive or the positive (Mulligan, 2002). The participants will be assigned randomly to one of the two conditions.

3.3.1. Survey Structure

At the start of the study, participants were served with one of the two stimuli (aggressive or positive). First, the stimuli perception was checked. After that, questions about the dependent variable and the two moderators were asked. Finally, the survey contained questions that measured demographical factors and several questions to gauge the quality of participation.

3.3.2. Stimulus industry and brand selection

To overcome concerns regarding prior knowledge and therefore treads to the internal validity, this study (Brown and Dacin, 1997) uses a fictitious corporation in the news articles used in the experiment. The fictitious corporation created for this research is called “Biashara” and the corresponding product brand “Ladha Ice Cream”. This kind of food is widely consumed throughout Western Europe (Pendrous, 2013) and therefore of particular relevance for the participants.

(25)

point of reference for other studies (Holzmüller, 2005). Therefore, the choice for the FMCG sector improves the possible generalizability of the results (Holzmüller, 2005). Within the FMCG sector, this study will take place in the groceries industry because this is the most important industry to consumers (AxelSpringer, 2009).

3.3.3. Stimulus description

Now the focal industry and the fictitious corporation used during this study are specified the fictitious articles can be discussed in greater detail. These fictitious articles are used as the study’s stimuli. In line with the choices of other experimental studies, this study uses the cover page of a well-known newspaper for the fictitious articles. This kind of stimulus is preferred because tax-related information is almost always reported on by journalists due to tax secrecy (Hardeck and Hertl, 2014). Second, one of the roles of the media is addressing corporate ethical issues (Dyck et al., 2010). This study uses the front page of the Financial Times because this helps to create reliable and objective stimuli compared to more popular newspapers. Second, the Financial times is the most popular financial newspaper in Western Europe. Finally, a business paper is most appropriate to report on CTSs. Both articles first introduced; the corporation; the specific brand; it’s sector and industry and the European headquarters. One article reported on an aggressive CTSs deployed by Biashara while the other reported on a positive CTSs used by Biashara. Both the stimuli for the aggressive and positive CTSs can be found in Appendix A.1 and A.2. The questionnaire can be found in Appendix A.3.

The positive CTSs article commented on the fact that Biashara was endorsed for their responsible CTS by an NGO. After that, it explained the goal of Biashara’s CTS. After that, the article addressed that, opposed to its competitors Biashara refrained from tax planning. The article on positive CTSs ended with outlining the definition of tax avoidance, the fact that it is legal and with the social consequences of tax planning such as the shift of the tax burden to SME and consumers.

The article on aggressive CTSs was structured in a similar way. It first addresses the fact that Biashara was caught using aggressive CTSs. After that, the aggressive article defined such strategy and described its objective and the fact that the used strategy is legal. After that, the article provided some examples of aggressive tax strategies that could be used (e.g. arbitrage opportunities, transfer pricing). The social consequences of tax planning as described in the article on positive CTSs were addressed in this article too.

3.3.4 Questionnaire and items

(26)

Variable Type

Variable

Independent Corporate Tax strategy (CTS)

Dependent Corporate marketing outcomes - Corporate brand credibility (CBC) Dependent Corporate marketing outcomes - Corporate Reputation (REP) Dependent Corporate marketing outcomes - Consumer Satisfaction (SAT) Moderator Personal attitude towards tax (ATT)

Moderator Personal tax moral (MOR)

Table 3.1 Description of variables

Corporate tax strategy (CTS) differed between the questionnaires. One group received an article about an aggressive CTS and the other group on a positive CTS. The dependent variables measure the consequence of variance in tax strategy. Given the challenges mentioned before regarding the qualification of aggressive CTSs as negative CSR the questionnaire first had to test if CTSs change the perception of a corporation’s CSR policy held by participants. By analysing the differences, this study evaluated if CTSs affected the participant’s beliefs towards the corporation’s CSR activities. To so this study used items from Wagner et al. (2009) to test this. These results and other manipulation checks will be addressed in Chapter 4 “findings”.

The study’s dependent variables are; Corporate brand credibility (CBC), Corporate Reputation (REP) and consumer satisfaction (SAT). The items to test the dependent variables are also extracted from previous research. Corporate brand credibility was measured using three items from Erdem et al. (2002), Corporate reputation was measured using three items from Newburry (2010) and three items from Flint et al. (2011) were used to measure consumer satisfaction. These items were also used in combination in previous research (Hur et al., 2014)

The items to test the moderating variables are also based on previous research. Four items from Lewis (1979) are used to measure attitude towards tax, two items to measure tax morale from Körner and Strotmann (2006) are incorporated. All constructs and related items are incorporated in Table 3.2. The items to measure CSR and the moderating variables were also used together in previous research by (Hardeck and Hertl, 2014)

The questionnaire also addressed control variables such as; gender, age, level of education and main daily occupation. These were collected to enable the study to describe the sample. The survey ended with three question that checked the quality of participation.

(27)

Construct Items

CSR

(Wagner, 2009)

In my opinion Biashara is a socially responsible company

In my opinion Biashara is concerned with improving the wellbeing of society

In my opinion Biashara follows high ethical standards

Personal attitude towards tax avoidance

(Lewis, 1979)

People who earn more and who pay more in taxes are more justified in finding loopholes to reduce their tax payments/bills

Avoiding taxes by exploiting legal loopholes is unfair because only the well-off can afford to employ the accountants needed to find them

We should say good luck to people who avoid paying taxes by finding legal loopholes

To avoid paying taxes by finding legal loopholes is unethical

Personal Tax Moral

(Körner and Strotman, 2006)

Illegal Tax evasion is in no case ethical

People who engage in tax evasion should be considered as unethical

Corporate Brand Credibility (Erdem, 2002)

Biashara will deliver what it promises Biashara’s product claims will be believable Biashara has a name you can trust

Corporate Reputation (Newburry, 2010)

Biashara is a company I have a good feeling about Biashara is a company that I admire and respect Biashara has a good overall reputation

Consumer Satisfaction (Flint, 2011)

I would describe my feelings towards Biashara as delighted I would describe my feelings towards Biashara as happy I would describe my feelings towards Biashara as pleased

(28)

agree, strongly agree) due to several reasons. First, the odd scale offers participants the choice to give a neutral answer. In this case option 3, “Neither agree nor disagree”. Second, the five-point Likert scale is preferred because it provides an easier interpretation of answer possibilities for participants compared to Likert scales with 7 points (Havlíková, 2016).

3.3.5 Survey development

To test the survey, a pilot test was conducted. The pilot test had N=8 respondents. These were almost all friend and family. The participants were asked questions about the readability of the survey and their understanding of the questions. Some questions were commented on as difficult to read or almost identical to other. Therefore, the wording of these questions was adapted and some question where deleted from the questionnaire. Another problem that emerged from the pilot test was the difficulty some participants had with understanding the English questionnaire. Possibly due to the technical subject of the questionnaire even some more advanced English readers experienced these difficulties.

Because the study aimed at a significant amount of Dutch respondents translating the survey was necessary. Back-translation or reverse translation was done to translate the questionnaire. This form of translating is commonly used in international business research because it checks extensively for the correctness and robustness of the translation (Douglas and Craig, 2007; Chidlow et al., 2014). Two bilinguals did the translation itself. Afterwards, some small adaptions were made to the wording of some questions to ensure readability and prevent too much similarity between questions.

After the pilot-test, the development continued with a pre-test of the questionnaire. The participants of the pre-test with N=19 were again mostly friends and family. The pre-test was done to ensure reliability and validity of all the measured items (Sarantakos, 2012). The pre-test did not indicate problems with the validity and reliability and of the questions. The fact that some of the questions were used together in previous studies supports this finding.

The participants in the pilot- and pre-test did not participate in the final study. This was done because their participation could be biased since they have seen the questions before and are to some extent familiar with the reasoning behind these questions. The final survey is presented in Appendix 1.

3.4 Data Collection

(29)

consumed an ice cream product before. Qualtrics was utilized for the design and online distribution of the survey. The web-based survey increased the speed and efficiency of the data collection (De Bruijne and Wijnant, 2014). It also helps to reach large numbers of participants in geographically dispersed areas with limited resources (Greszki et al., 2015). The questionnaire was spread predominantly by social media such as Facebook and LinkedIn.

3.5 Data analysis

The raw data was first inspected visually in Qualtrics. This did not uncover problems or unexplainable variances in response. After that, the data was exported. This data was cleared for uncompleted responses and some responses were deleted due to quality issues. These quality issues are addressed in the “findings” chapter. Thereafter, sample characteristics were computed and descriptive statistic where provided. The actual analysis was performed using IBM SPSSS version 24. The following relevant tests were conducted; ANOVA, regression and the computation of Cohen’s d. These are appropriate because the study tries to identify if differences between the dependent variables are caused by the independent variable and what the size of these effects are. Cohen’s d was computed to identify the size of these effects. A more extensive explanation of why these test in particular where used can be found in paragraph 4.4. 3.6 Ethical considerations

(30)

4 Findings 4.1 Introduction

Based on the design described in chapter 3 the collected data is analysed. This chapter will describe this analysis. This chapter will first address the preliminary analyses of the data. Furthermore, the construct validity and reliability are checked and reported. This chapter will then continue with the manipulation check to see if the manipulation succeeded. Finally, this chapter will test the hypothesis as stated in chapter 2.

4.2 Preliminary analysis

To test the stated hypotheses, the study first executed some preliminary analyses. First, the data was checked for missing or abnormal values. Incomplete entries were removed from the dataset. Second, before testing this study addresses some characteristics of the data set. Finally, the constructs are checked for reliability and validity. This was done by checking, Cronbach’s alpha, Factor Loadings and Average Variance Extracted.

4.2.1. Data cleaning

The survey was online for two weeks (13 days). After these days, the data set contained N=251 respondents. To clean the data set listwise exclusion was used. Although this means that this study has to assume that data is missing completely at random (MCAR) in this case listwise is preferred because of powerful interpretation benefits (Peugh and Craig, 2004). Besides, a preliminary analysis showed that the difference between variables for missing data was small ranging from N=70 to N=77 respondents. Due to survey settings, it was impossible to skip questions. Therefore the number of possible incomplete responses is limited to N=77. This implies that the benefits of pairwise exclusion are small and listwise exclusion is preferred (Peugh and Craig, 2004). After removing these incomplete and aberrant responses, a valid data set remained with N= 174 (69,7%) respondents. Both pairwise and listwise results can be found in Appendix B.1.

(31)

Scenario type

Respondents

Aggressive CTS Scenario N = 86

Positive CTS Scenario N = 84

Table 4.1. Respondents for each scenario

Respondents were assigned to a scenario using Qualtrics’ randomizer tool. The tool evenly presents one of the scenarios to a respondent. The negligible difference in participants confirms this random evenly distribution between scenarios.

4.2.2 Sample characteristics

As described in 4.2.1 the final data set had N=170 respondents. The survey asked questions about demographic characteristic. These demographics are used to characterise the sample. An extensive overview can be found in Table 4.2. The table additionally shows the characteristics for both of the individual scenarios. This shows no considerable differences. This section will continue with a short description of the most notable aspects of the demographics.

The table shows that most participants are between 17 and 27 (53,5%). The table also shows that most participants are from The Netherlands (88,2%). Almost all other participants come from other European countries (except for one from Brazil). 56,5% of the respondents are male, while 43,5% are females. Based on statistics from Eurostat, 49% of the EU population are male (Eurostat, 2015). Therefore, males are slightly overrepresented in the sample. According to the same statistics, females account for 51% of the EU population (Eurostat, 2015). Females are therefore slightly underrepresented in the sample. Additionally, 53,5% of the participants hold a bachelor degree. Most of the participants had a job (55,9%), while 40,0% of the sample consisted of students.

4.2.3 Descriptive results

(32)

Demographics Complete Sample Aggressive CTSs Positive CTSs N % N % N % Sample 170 100 86 50,6 84 49,4 Age - 18 - 26 years 91 53,5 43 50,0 48 57,1 - 27 - 36 years 25 68,2 13 65,1 12 71,4 - 37 - 46 years 21 80,6 13 80,2 8 81,0 - 47 - 56 years 17 90,6 9 90,7 8 90,5 - 57 - 66 years 13 98,2 7 98,8 6 97,6

- 67 years and older 3 100,0 1 100,0 2 100,0

Gender - Male 96 56,5 47 54,7 49 58,3 - Female 74 100,0 39 100,0 35 100,0 Country of origin - Austria 1 0,6 0 0,0 1 1,2 - Brasil 1 1,2 1 1,2 0 1,2 - France 1 1,8 0 1,2 1 2,4 - Germany 6 5,3 2 3,5 4 7,1 - Italy 1 5,9 1 4,7 0 7,1 - Spain 2 7,1 1 5,8 1 8,3 - The Netherlands 150 95,3 78 96,5 72 94,0 - Turkey 1 95,9 1 97,7 0 94,0 - United Kingdom 2 97,1 0 97,7 2 96,4 - Missing 5 100,0 2 100,0 3 100,0

Main daily activity

- Work 95 55,9 49 57,0 46 54,8

- Study 68 95,9 34 96,5 34 95,2

- Retired 5 98,8 1 97,7 4 100,0

- Stay at home mum/dad 0 98,8 0 97,7 0 100,0

- Unemployed 2 100,0 2 100,0 0 100,0

Highest level of education

- Less than High School 0 0,0 0 0,0 0 0,0

- High School 8 4,7 3 3,5 5 6,0

- Some college, no degree 27 20,6 14 19,8 13 21,4

- Associate's degree 0 20,6 0 19,8 0 21,4

- Bachelor Degree 91 74,1 51 79,1 40 69,0

- Ph.D. 1 74,7 1 80,2 0 69,0

(33)

With these descriptive results, the study gains some valuable insight. The skewness value deals with the symmetry of the responses’ distribution (Huizingh, 2014). The kurtosis value in turn deals with the peakedness of the responses’ distribution (Huizingh, 2014). Taken together, these tools assess whether the gathered data is normally distributed. Normally distributed implies that the greatest frequencies of responses are detected in the middle and responses with a lower frequency are found towards the extremes (Huizingh, 2014). If the data is perfectly normal both skewness and kurtosis values are 1. According to Huizingh (2014) skewness and kurtosis values between -1.96 and 1.96 can be considered as normally distribute. The data in Appendix B.2 is consistent with this consideration. Most items are normally distributed with kurtosis and skewness values between -1,856 and 1,349. However, item “MOR1 – Illegal tax evasion is in no case ethical” has kurtosis value above 1,96. In the positive CTS condition, the value is 2,458. In the aggressive CTS condition, the kurtosis value of MOR1 is 3,238 which is also above 1.96. However, these two values are not utterly high therefore and limited to only one item. Therefore, no problems due to normality are expected further testing of the data.

A preliminary look at the descriptive results also shows that there are indeed differences between the mean values of the variables in the positive CTS condition and these values in the aggressive CTSs condition. Looking at the variables, the mean in the positive condition ranges from 3,24 to 3,91 while these values range from 1,81 to 2,66 in the negative condition. When analysing the moderators, this picture is different. Moderator mean values are relatively similar. The values of the positive condition range from 3,28 to 4,31, while values associated with the negative condition range from 3,33 to 4,45. Although the differences are clear, no implications can be connected to these differences in this stage of the study.

4.2.4 Reliability analysis

(34)

Constructs Number

of items Cronbach's alpha

Corporate social responsibility (CSR) 3 0,94

Corporate brand credibility (CBC) 3 0,80

Corporate reputation (REP) 3 0,91

Consumer Satisfaction (SAT) 3 0,94

Personal attitude towards tax (ATT) 4 0,61

Personal tax moral (MOR) 2 0,60

Table 4.3 Cronbach's Alpha

Because factor loadings represent the extent to which an item explains a construct, it can also be used as an indicator of the reliability of these constructs (Kolenikov, 2009). To prove this reliability, the factor loading of each item should at least 0,7 (Kolenikov, 2009). Due to the size of the tables, the results of this test can be found in Appendix B.3. In the aggressive scenario, the values of all variables pass the threshold with values between 0,70 and 0,91. The values of the moderators range from 0,56 to 0,80. Two values are lower than 0,7 (0,56 and 0,60). In the positive scenario, the values of all variables pass the threshold with values between 0,82 and 0,92. The values of the moderators range from 0,57 to 0,90. Two values are lower than 0,7 (0,58 and 0,57). Although two values are below 0,7 the factor loading does not indicate serious problems with the reliability, due to the small deviations and the fact that the other variables are well above 0,7.

4.2.5 Validity analysis

Validity was checked for convergent validity as well as discriminant validity. Convergent validity measures the correlation between different measures that are based on the same construct (Moutinho and Huarng, 2015). This can be checked using average variance extracted, hereafter AVE. This AVE should be at least 0,5 (Moutinho and Huarng, 2015). In both scenarios (positive and aggressive) all values except one meet this requirement. The exact results can be found in Table 4.4.

(35)

Variables AVE 1 2 3 4 5 6 1 CSR_A 0,71 0,84 2 CBC_A 0,56 0.47 0,75 3 REP_A 0,73 0.73 0.53 0,86 4 SAT_A 0,83 0.68 0.54 0.81 0,91 5 ATT_A 0,43 -0.16 -0.15 -0.27 -0,31 0,65 6 MOR_A 0,62 -0.14 -0.07 -0.15 -0,20 0.33 0,79

Table 4.4a - Validity aggressive condition

Variables AVE 1 2 3 4 5 6 1 CSR_P 0,82 0,91 2 CBC_P 0,74 0.77 0,86 3 REP_P 0,78 0,63 0.62 0,88 4 SAT_P 0,80 0.63 0.54 0.84 0,89 5 ATT_P 0,50 0.29 0.28 0.38 0.38 0,71 6 MOR_P 0,80 0.33 0.37 0.26 0.32 0.42 0,90

Italics = no significant correlation

Table 4.4b - Validity positive condition

4.3 Manipulation check

To check if participants perceived the test stimuli as intended a manipulation check was performed. A one-way ANOVA is suitable to compare the mean scores of both groups. The ANOVA was conducted for the variable CSR because the articles were intended to influence the CSR perceptions of participants. There was a statically significant effect of the condition on CSR, F (1,170) = 220,269, p < 0,00. Based on this evidence the study concludes that participants perceived the articles substantively different (positive mean; 3,835, aggressive mean; 1,80). The results of this ANOVA can be found in Appendix B.4. These results support the labelling of aggressive CTS as negative CSR as presented in the literature review.

4.4 Hypotheses testing

(36)

2013). Thereafter, the size of these effects is determined through Cohen’s D (Field, 2013). The Cohen’s D is the most commonly reported statistic to describe the size of an effect (Field, 2013).

4.4.1. ANOVA related assumptions

To ensure the validity of the results the study first checks the relevant assumptions related to an ANOVA test.

The first two assumptions deal with the variables involved. First, the dependent variables need to be measured on a continuous scale. In this case all dependent variables CBC, REP and SAT are measured as an interval. Therefore, assumption one is met. According to the second assumption, the independent variable has to consist of two or more categorical groups. Besides this, these groups have to be independent (Field, 2013). The independent variable is indeed categorical (positive or aggressive), and because participants could only be in one of the groups, the variable is also independent. Therefore, the second assumption is also met.

The first more complex assumption has to do with the fact that an ANOVA is based on a normal distribution. Therefore, the data has to be relatively normal distributed. As addressed in the descriptive characteristics, the data is normally distributed because the skewness and kurtosis values are both with the relevant limits. These values can be found in Appendix B.2. Another commonly used way to test normal distribution with ANOVA is the Shapiro-Wilk test of normality. In order to be normally distributed the Shapiro-Wilk test has to report insignificant values. The results of the Shapiro-Wilk test are reported in Appendix C.1. The values are significant. Therefore, the assumption of normality is only partly satisfied. However, this does not inhibit us from using ANOVA because ANOVA is robust for the valuation of normality when samples are equal, which is the case here (Härdle and Simar, 2007). Based on the group size of N=84 and N=86 being large enough and the previous tests regarding normality of the data the literature Field (2013) ensures that the ANOVA can deal with deviations from normality.

Another important assumption deals with outliers. ANOVA is more sensitive to outliers than other well know analysis (Field, 2013). Again the test in the descriptive, which focussed around skewness and kurtosis did not raise serious problems due to outliers. Therefore, this requirement is also met.

(37)

4.4.2 ANOVA analysis

The ANOVAs were performed to evaluate the effect of corporate tax strategies on the dependent variables. Because the independent variable only has two conditions pairwise significance is not an issue. The size of the effects is determined through computing Cohen’s d (Härdle and Simar, 2007). For Cohen’s d values above 0,8 the impact of an effect is considered to be large (Härdle and Simar, 2007). The results will be addressed for each dependent variable separately. For presentation reasons, the results are summarised in table 4.5 and table 4.6. First, the results show that corporate tax planning strategies had a highly significant effect on corporate brand credibility F (1,168) = 64.75, P < 0.001. The aggressive CTS condition had significantly lower corporate brand credibility (M = 2,52 SD = 0.67) than the positive condition (M = 3.47 SD = 0.84). This effect had a large impact on corporate brand credibility (d 1.052 > 0.8). Taken together, these results indicate that corporate brand credibility is positively affected by positive CTSs compared to negative CTSs. Therefore, both H1a and H1b can be accepted.

Secondly, the results reveal that corporate tax planning strategies had a highly significant effect on corporate reputation F (1,168) = 150.58, P < 0,001. Corporate reputation was significantly lower in the aggressive CTS condition (M = 2.15 SD = 0.85) than in the positive CTS condition (M = 3.67 SD = 0.77). This effect had a large impact on corporate brand credibility (d 1.371 > 0.8). Based on the results corporate reputation is positively affected by positive CTSs compared to negative CTSs. Therefore, both H2a and H2b can be accepted.

Finally, the results indicate that corporate tax planning strategies had a highly significant effect on consumer satisfaction F (1,168) = 136.103, P < 0.001. Consumer satisfaction was significantly lower in the aggressive CTS condition (M = 2.04 SD = 0.80) than in the positive CTS condition (M = 3.48 SD = 0.80). This effect had a large impact on consumer satisfaction (d 1.334 > 0.8). These results confirm that consumer satisfaction is higher for positive CTSs compared to negative CTSs. This offers enough support to accept H3a and H3b.

Appendix D.1 shows the results for these ANOVAs and Cohen’s d. Relevant values are highlighted.

Variable Aggressive CTSs Positive CTSs Cohen's d

(1) (2)

M SD M SD (1) vs. (2)

Corporate Brand Credibility 2,52 _0,7 _3,47 _0,83 _1,05

Corporate Reputation 2,15 _0,84 _3,67 _0,77 _1,37

Consumer Satisfaction 2,04 _0,8 _3,47 _0,8 _1,334

The influence of tax planning activities on corporate marketing outcomes