A
BSTRACTCompanies engaging in cause-related marketing (CM) often increase product prices to (partially) cover the costs; however, they do not communicate such price changes explicitly in their campaigns. Yet, knowledge is lacking on what consumers actually expect who is paying for the donation in implicit-stated CM campaigns and how this perception impacts consumer attitude and rewarding or non-rewarding behavioural intentions toward the firm. Against this background, the current study had three objectives that address important research gaps in the literature. First, we examined how consumers actually perceive who is paying for the donation in CM campaigns. Second, we investigated under what circumstances consumers attribute internal or external motives to the firm and perceive higher levels of moral hypocrisy in terms of a moderating effect of advertisement execution style (i.e., campaign focuses on charity or on product). Third, we investigated the process of how the interplay between who pays and the moderating variable advertisement execution style leads to attitude toward the campaign, potential rewarding behaviour and negative opinioned communication intentions. In terms of the first objective, we found that consumers expect the company to pay for the donation in actual CM campaigns. In terms of our second goal, this study found little evidence for the moderating effect of advertisement execution style. Finally, in terms of our third objective, we found perceived corporate motives and perceived moral hypocrisy to play a critical role in determining consumer attitude and behavioural intentions, but the proposed moderated mediation sequence was not supported. More specifically, our findings indicate that people who think the consumer pays for the donation will attribute less internal attributive motives to the firm, perceive higher levels of moral hypocrisy, and consequently have a less positive attitude toward the campaign than people who think the company pays for the donation. Although we did not find the moderated mediation sequence due to the fact that advertisement execution style did not influence consumers’ responses as expected, the present study, however, makes several noteworthy contributions to CM, consumer purchase decision-making, and societal marketing practice.
P
REFACE1.
I
NTRODUCTIONIn recent years, consumers increasingly expect organizations to act responsibly toward the wider society and the natural environment in which they operate. Responding to society’s changing expectations of business, a growing number of companies have adopted Corporate Social Responsibility (CSR) policies. Although no universally accepted definition of CSR exists (Van Marrewijk, 2003), two features are often used to differentiate CSR activities from other deeds: (1) they partly or entirely benefit society and/or general interests, and (2) they are not obligated by law (Arvidsson, 2010).
Research shows that consumers not only take CSR initiatives into account when purchasing products, but also when evaluating the company (Aaker, Vohs and Mogilner, 2010; Brown and Dacin, 1997; Sen and Bhattacharya, 2001). For instance, Aaker et al. (2010) suggest judgements of warmth (i.e., the intent), and competence (i.e., the ability) to play important roles in consumers’ perceptions of organizations and brands. When firms are perceived as both warm and competent, consumers become more willing to buy brands from these organizations. Additionally, a more recent study revealed that companies who only act competent are evaluated less positively than companies who are both moral and competent or only moral. This means a firm must not only focus its attention on increasing its bottom line, but also on being a good corporate citizen. Or, differently stated, companies without morality are a no go (Tauber, Leliveld and Fennis, working paper).
these social initiatives as sincere actions and thus may or may not reward the firm (Barone, Miyazaki and Taylor, 2000; Ellen, Mohr and Webb, 2000; Sen and Bhattacharya, 2001). Various studies found consumers’ perceptions of the underlying firm’s motives for engaging in CM to play a critical role in explaining consumers’ responses to CM (Becker-Olsen, Karen and Hill, 2005; Barone et al., 2000; Barone et al., 2007; Ellen et al., 2006; Folse et al., 2010). According to the attribution theory, consumers may assign two primary types of motives to a firm’s initiative: intrinsic motives (i.e., philanthropic) or extrinsic motives (i.e., self-interested or profit-driven). Perceptions of insincerity occur when the primary motivation of the company appears to be profit-driven rather than of philanthropic nature (Bowie, 1991). Research suggests that consumers will punish firms that are perceived as profit-driven and thus insincere in their social involvement (Sen and Bhattacharya, 2001; Simmons and Becker-Olsen, 2004), whereas they will reward firms that are rejecting their basic self-interested nature (Ellen et al., 2000).
how these perceptions impact consumers’ attitudes toward the campaign and consumers’ rewarding or non-rewarding behavioural intentions toward the firm.
2.
T
HEORETICAL FRAMEWORK2.1. CAUSE-RELATED MARKETING
The first CM campaign dates back to 1983, when American Express launched a campaign whereby it contributed one cent for every card transaction and $1 for every new card issued to the renovation of Ellis Island and the Statue of Liberty. The campaign became a huge success: the use of the card increased by 28% compared with the same period in the previous year and the company raised 1.7 million dollars for the restoration project (Wall, 1984). From that moment, CM is one of the fastest growing marketing communication tools due to its win-win-win situation for all three parties involved: the for-profit company, the NPO and the consumers (Endacott, 2004).do-gooding is not without critics. It seems that CM efforts often go hand in hand with public cynicism and suspicion, arising from: the small amount of money actually delivered to charities (Endacott, 2004), perceived exploitation of the NPO (Varadarajan and Menon, 1988), and consumers’ misunderstanding about the details of CM campaigns (Pracejus et al., 2003).
2.2. DIRECT EFFECTS OF CONSUMERS’ PERCEPTIONS OF WHO PAYS ON THEIR RESPONSES
As already mentioned in the introduction, CM is often associated with price increases to cover some of the costs of the campaign (Krishna and Rajan, 2009). Yet, while CM has received a fair amount of research attention, knowledge regarding consumers’ perceptions and reactions to price increase in a CM campaign is rather limited. Barone etal. (2000) found that price trade-offs in CM can have important impact on product
implicit-stated CM campaigns. Moreover, we analyse how such perception impacts consumers’ beliefs, attitudes and intentions.
Attitudes are a key construct in the explanation of consumer behaviour in general, and socially responsible behaviour in particular (Gawronski, 2007). The fact that attitudes are predispositions to consumer responses points to their relation to actual consumer behaviour and their direct impact on purchase decisions (Assael, 1999). However, consumer attitudes do not always predict buying behaviour (Stanton, Etzel and Walker, 2004). Attitudes are multidimensional: they consist of three components, of which the intentional component is most directly related to behaviour. Therefore, in the current research we not only take into account attitude’s cognitive and affective elements, but also the intentional element in form of rewarding and non-rewarding behavioural intentions towards the firm.
In case of CM initiatives, having a positive attitude toward the initiative should correspond to a positive evaluation of the campaign, leading eventually to an increased purchasing intent. However, we suggest that price increases in a CM campaign could easily make consumers distrust the underlying firm’s motives to engage in CM. Therefore, we predict that people who think the consumer pays for the donation (a) have a less positive attitude, (b) are less likely to engage in rewarding behaviour, and (c) are more likely to engage in negative opinioned communication, compared to people who think the company pays for the donation (hypothesis 1).
2.3. THE MEDIATING EFFECT OF PERCEIVED CORPORATE MOTIVES AND MORAL
HYPOCRISYVarious studies have confirmed consumers’ perceptions of a company’s motivation for engaging in CM to be one of the most critical factors in explaining consumers’ responses to CM (Becker-Olsen, Karen and Hill, 2005; Barone et al., 2000; Barone et al., 2007; Ellen
et al., 2006; Folse et al., 2010). For instance, Becker-Olsen et al. (2005) found that
perceived corporate motivation is likely to influence consumers’ attitudes toward companies and their social efforts. Similarly, Barone et al. (2000) demonstrated that the influence of a CM campaign on product choice depends on consumers’ perceptions of the firm’s motivation to start such a campaign. Former findings are consistent with results of a more recent study by Folse et al. (2010), which indicates that the relationship between size of the monetary donation and consumers’ intention to participate in CM campaigns can be explained by the inferences consumers make about firm motives. Inferences about firm motives also appear to be an important factor influencing sincerity and credibility (Gao, 2009); two fundamental factors in the outcomes of CSR activities (Barone et al., 2007; Hillenbrand et al., 2013; Hur et al., 2014). So, even though a CM initiative may seem to be an altruistic action, consumers’ perceptions of a firm’s motives for the activities drive their evaluations of the campaign and influence beliefs, attitudes and intentions. In other words, it seems more important for consumers to know why a company is doing it than what it is doing (Gilbert and Malone, 1995).
In comparison to other CSR initiatives, CM is especially vulnerable to higher suspicion about the motives of firms engaging in it (Barone et al., 2000). This vulnerability stems from CM’s core philosophy, namely, the direct relationship between the sales of a company’s product and its support of a social cause (Strahilevitz and Myers, 1998). Consumers might question whether a company’s support of the cause is designed to benefit the cause or the organization (Webb and Mohr, 1998). For example, Reebok's support of the Amnesty International "Human Rights Now!" tour was viewed by some as an indication that the sportswear giant cares about social issues, but by others as only an attempt to enhance its product sales (Elsbach and Sutton 1992).
egoistic and self-serving (e.g., to increase sales or improve the company’s reputation). Consumers responded most positively to CM programs that are intrinsically motivated (Ellen et al., 2000). Yet, another study revealed that perceived honesty about the motives may be equally important as the motives themselves (Forehand and Grier, 2003). They state that scepticism is not simply driven by a firm being profit motivated, but rather by a discrepancy between stated objectives and firm actions (e.g., if the objectives are stated purely social, while the firm’s actions appear to be self-serving). Due to the important role this discrepancy plays in CM effectiveness, we not only take into account the perceived corporate motives, but also perceived moral hypocrisy when predicting consumer responses in the current study.
Moral hypocrisy is a form of corporate deception that occurs when companies have the desire to appear moral while, if possible, avoid the cost of actually being so (Batson, Kobrynowicz, Dinnerstein, Kampf and Wilson, 1997). A classic example of corporate deception and moral hypocrisy is the one of tobacco firm Philip Morris, who, in 1999, contributed $60 million to charities, while, in that same year, spent $108 million on an advertising campaign to publicize their benevolence. Although several charities benefited from their donation, the tobacco giant spent so much more on advertising that their “moral behaviour” was roundly parodied in the press (Batson, Collins, and Powell, 2006; Stoll, 2002).
perceive higher levels of moral hypocrisy, compared to people who think the company pays for the donation (hypothesis 2). Moreover, we predict (a) perceived internal attributive firm motives, (b) perceived external attributive firm motives, (c) and perceived moral hypocrisy to mediate to relationship between who pays and attitude toward the campaign (hypothesis 3), rewarding behavioural intentions (hypothesis 4), and negative opinioned communication intentions (hypothesis 5).
2.4. THE MODERATING ROLE OF ADVERTISEMENT EXECUTION STYLE
Many studies have demonstrated the influence various visual aspects can have on a campaign’s effectiveness. For instance, the use of a picture in an ad leads to greater recall of the ad (Lutz and Lutz, 1977) and a more positive attitude toward the ad, the brand and the product category (Kelly, Slater and Karan, 2002). In particular, concrete pictures of a product are more effective in stimulating vivid visual imagery processing, which in turn positively influences attitudes toward the ad and brand than either a less concrete pictures or no picture at all (Babin and Burns, 1997). However, specific research on the influence of visual appeals in CM campaigns is rather limited.
Interestingly, in case of CM, two partners are involved in the offer, creating a unique situation with respect to who should be depicted in the ad: the brand/product or the social cause. The majority of CM campaigns show the brand as the focal visual, while the cause plays a more secondary role. For instance, an advertisement of the Dutch fashion brand Goosecraft depicts a picture of a girl in a red jacket and contains the CM message only in small print at the bottom (see Appendix A). On the other hand, ads can also be constructed in a manner that highlights the cause, with the product taking a lesser prominent role (Lafferty and Edmondson 2009; Chang 2011). Good examples of cause-focused advertisements is the alliance between Benetton and the World Food Program and the advertisement of a detergent, “Dawn saves wildlife’, in which baby animals play the main role in order to create an emotional appeal (see Appendix B).
perceived differently and, more specifically, have an interesting and prevalent effect on consumers’ perceptions because causes are likely to enhance emotional reactions (Kim
et al., 1998). Moreover, focussing on the cause in the campaign may enhance the brand’s
image, since companies involved with CM are generally perceived by consumers as being generous, altruistic and responsible (Adkins, 2010). Thus, the type of picture used in the campaign and in particular whether the ad is cause-focused or product-oriented, appears to have a significant influence on how consumers perceive the CM initiative and the firm.
As we have stated in the previous section, we not only suggest that consumers’ perceptions of who pays directly influence their attitudes and behavioural intentions, but also that perceived corporate motives and perceived moral hypocrisy could mediate, and thus explain these consumer responses. However, scepticism is not simply driven by a firm being profit motivated, but rather by a discrepancy between stated objectives and firm actions (Forehand and Grier, 2003). To overcome consumer scepticism, the CM message needs to be communicated in such a way that it boosts the credibility of the message (Dawkins, 2004; Du and Vieira, 2012).
Based on the aforementioned section, we suggest that the influence consumers’ perception of who pays has on perceived corporate motives and perceived moral hypocrisy will depend on advertisement execution style (i.e., moderation effect). We suggest that when a company uses a cause-focused ad, while, at the same time, increases the price to cover some of the costs of the campaign consumers may perceive a discrepancy that triggers scepticism; on the one hand the firm portrays itself in the ad as being social and altruistic, while, on the other hand, it increases the price, making consumers pay for the donation in the campaign. Therefore, we predict that people who
motives to the firm when the advertisement is cause-focused than when it is product-oriented (hypothesis 7), however, in this case we do not expect to find a difference in perceived moral hypocrisy.
2.5. MODERATED MEDIATION
In previous sections we have described all the potential direct effects, simple mediation effects and moderation effects separately. Now we will make assumptions that combine the effects in a moderation mediation model, also known as conditional indirect effects. Moderated mediation models attempt to explain both how and when a given effect occurs (Frone, 1999). Moderated mediation occurs when the strength of an indirect effect (i.e., a mediation effect) depends on the level of some variable (i.e., the level of a moderator). With regard to our study, we predict that the strength of perceived corporate motives and perceived moral hypocrisy depends upon the company’s advertisement execution style.
Based on the aforementioned sections, we expect that participants’ perceptions of who pays is related to consumer attitude toward the campaign through indirect effects, such that the effect of who pays on consumer attitude toward the campaign will be moderated by advertisement execution style and mediated by perceived internal attributive firm motives, perceived external attributive firm motives, and moral hypocrisy (i.e., moderated mediation) (hypothesis 8).
motives, perceived external attributive firm motives, and moral hypocrisy (i.e., moderated mediation) (hypothesis 10).
2.6. OVERVIEW
3.
M
ETHODOLOGY3.1. DESIGN AND PARTICIPANTS
A 3 (who pays for the donation: consumer pays versus company pays versus ambiguous) x 2 (CM campaign’s execution style: cause-focused versus product-oriented) between-subjects factorial design was used to investigate consumers’ perceptions and responses to a CM campaign. A total of 312 subjects participated in the study. Thirteen participants did not pass the attention check and thus were eliminated from the analyses, leaving a final sample of 299 (176 males and 123 females; Mage = 35.52, SD = .63) on which to run the analyses. Further descriptions of the sample can be found in table 1. The participants were randomly assigned to one of the six conditions. The survey took place via Mechanical Turk (Mturk) and participants were paid $0.75 to complete the 5-minute survey. Table 1. Descriptive statistics of the sampleCharacteristic Frequencies Percentage %
3.2. PROCEDURE
An online questionnaire including a scenario was set out on MTurk, using USA citizens only. Participants were assigned randomly to one of the six experimental conditions and invited to participate in a study on cause-related marketing. Participants learned that cooperation in the survey was voluntary and their answers would be collected completely anonymous. After they agreed to the terms, they were instructed to carefully read a short description of a campaign for the fictitious toilet paper brand Quilton. A fictitious brand name was used to prevent triggering brand associations and emotions that might bias participants’ responses. The participants were then asked to take a thorough look at the advertisement, which was the first manipulation and could have two different execution styles: (1) cause-focused, containing photos of the social cause and a headline saying: “The most caring toilet tissues”, (2) product-oriented, containing photos of the product and a headline saying: “Ultra soft & strong toilet tissues”. Moreover, the ad proclaimed that 5% of the sales price would be donated to support the World Vision, an American charity that is dedicated to helping disadvantaged children go to school (see Appendix B for examples of the ad stimuli). The advertisement was followed by the second manipulation wherein respondents learned who pays for the donation, which was effectuated by a short note that could contain three different descriptions: (1) the product price increased, meaning that the consumer pays for the donation, (2) the product price did not increase, meaning that the company pays for the donation, (3) no note was provided (i.e., the ambiguous condition). After that, the respondents were requested to fill in the questionnaire, which also contained two manipulation checks and an attention check. Participants were then debriefed and thanked for participation. The questionnaire used in this study can be found in Appendix C.
3.3. MEASURES
product to family and friends and the likelihood of purchasing it), and (5) non-rewarding behavioural intentions (i.e., negative opinioned communication intentions). Although, due to random assignment to the conditions, we do not expect trait variables to influence our DV’s, we will check for them to make sure they really don’t. So to account for demographic variables that could potentially impact perceptions of moral hypocrisy, we included gender and age as control variables. Moreover, we included also cause familiarity, cause importance and consumer scepticism to CM in general as control variables, as they could confound consumer attitude toward the campaign and behavioural intentions toward the firm. All variables were tested on validity and reliability. To check the construct validity, factor analysis was conducted. Factors were included based on the content interpretation, the KMO (>.6), the eigenvalue (>1.0) and the scree plot. Next to factor analysis reliability was tested by the measurement of Cronbach’s alpha (α).
3.3.1. CONSUMER ATTITUDE TOWARD THE CAMPAIGN
3.3.2. PERCEIVED CORPORATE MOTIVES
Perceived corporate motives were measured on a six-item seven-point scale with bipolar adjectives. Participants were asked to what extent they evaluated the company Quilton to be not self-interested (1) - very self-interested (7), not firm focused (1) - very firm focused (7), not profit motivated (1) - very profit motivated (7), not community interested (1) - very community interested (7), not customer focused (1) - very customer focused (7), and not socially motivated (1) - very socially motivated (7). Factor analysis resulted in a KMO value of .83, which is significant (p < .001). Varimax rotation confirmed the two dimensions with the first three items grouped as extrinsic perceived motives (α = .81) and the latter three as intrinsic perceived motives (α = .92). The reliability of both factors was found to be very good, therefore subsequent analyses were conducted using composite scale means calculated on the basis of the item means. 3.3.3. PERCEIVED MORAL HYPOCRISY
Perceived moral hypocrisy was conceptualized as the result of the ethical attribution of misalignment between self-interestedness and others-interestedness (i.e., altruism) in the CM initiative. The concept was measured by three items consisting of statements anchored by a seven-point Likert scale, running from “strongly agree” to “strongly disagree”: (1) “Quilton acts hypocritically”, (2) “Quilton’s behaviour is not in line with what they preach”, (3) “Quilton pretends to be something that it is not” (based on Wagner et al., (2009)). The items were reverse coded and combined into a perceived moral hypocrisy scale (α = .98), ranging from 1 to 7 with higher scores indicating a higher perception of moral hypocrisy. Factor analysis resulted in a KMO value of .78, which is significant (p < .001)
3.3.4. POTENTIAL REWARDING BEHAVIOUR
reversed the scale and combined these questions into a potential rewarding behaviour scale (ranging from 1 to 7) with higher scored indicating a higher possibility of rewarding behaviour (α = .95). Factor analysis showed a KMO value of .90, which was significant (p < .001).
3.3.5. NEGATIVE OPINIONED COMMUNICATION INTENTIONS
Similarly to potential rewarding behaviour, negative opinioned communication intentions were measured with six questions, e.g., “I would criticize without any hesitation how the company puts its business first, rather than the charity” and “I would correct someone who overestimates Quilton’s philantrophic giving during any conversation about it”, based on Kim and Rhee’s megaphoning scale (2011), and Moon’s public communication behaviours scales (2011). The questions were measured on a seven-point Likert scale ranging from “very likely” to “very unlikely”. We reversed the scale and combined these questions into a negative opinioned communication intentions scale (ranging from 1 to 7) with higher scored indicating higher intentions to negative opinioned communication (α = .94). Factor analysis resulted in a KMO value of .89, which was significant (p < .001).
3.4. CONTROL VARIABLES
To control for other variables that could influence the effect of who pays and advertisement on the five variables stated in the previous sections, three control variables were included: cause familiarity, cause importance, and consumer scepticism to CM in general.
3.4.1. CAUSE FAMILIARITY
3.4.2. CAUSE IMPORTANCE
Cause importance was measured with three items, again anchored by a seven-point scale with bipolar adjectives proposed by Lafferty and Edmondson (2009): unimportant (1) - very important (7), means nothing to me (1) - means a lot to me (7), very insignificant (1) - very significant (7). Again reliability was found to be excellent (α = .96). Factor analysis resulted in a KMO value of .75, which was significant (p < .001). Therefore items were combined into one cause importance seven-point scale, with higher scores indicating a higher sense of cause importance.
3.4.3. CONSUMER SCEPTICISM TO CM IN GENERAL
Consumer scepticism to CM in general, was measured using four questions that were adapted from Mohr et al. (1998), e.g., “Most statements made by a company in advertising or package labels about supporting a cause are true”. The items were measured using a seven-point Likert scale, ranging from ‘strongly agree’ to ‘strongly disagree’. The last three items were reverse coded, so that a higher score means a higher sense of scepticism. Because reliability of the scale was found to be very good (α = .88), subsequent analyses were conducted using composite scale means. Factor analysis showed a KMO value of .82, which is significant (p < .001).
3.5. MANIPULATION CHECKS
4.
RESULTS
This section displays the results of the analyses to test the hypotheses and answer the main research question of our study. First, a chi-square test was performed to evaluate the two manipulations in the study. Second, the influence of the control variables on the dependent variables was checked. Third, Two-Way ANCOVA’s were conducted to test the direct effects and to answer the main research question of this study. In addition, several exploratory analyses were performed to find further evidence for the answer to who consumers’ perceive pays for the donation in CM campaigns. Fourth, moderated mediation analyses of Hayes (2012) were conducted to test the conditional indirect effects. Finally, an overview of all the confirmed and rejected hypotheses is given.
4.1. MANIPULATION CHECKS
A chi-square test was performed to evaluate to what extent the two manipulations, as intended by this research, were well received by the participants. The first manipulation in the study was checked by the question: “Who paid for the donation?”. Results indicate a strong association between the price increase (consumer pays), no price increase (company pays) and ambiguous condition, and participants’ perception of these conditions, χ2 (4) = 199.50, p < .001. However, a closer look to the manipulation check reveals that some respondents did not interpret or remembered the manipulation correctly. In the price increase (consumer pays) condition, seven respondents (7.1%) responded incorrectly to the question by answering, “the company Quilton paid for the donation”, whereas in the no price increase (company pays) condition, four respondents (4.0%) answered incorrectly to the question by stating, “the consumer pays for the donation”. Surprisingly, in the ambiguous condition, where nothing was explicitly stated about the price, only a few respondents answered the question with the expected “I don’t know” answer (12.8%). Specifically, 61.6% of the participants answered: “the company pays”, whereas 25.6% indicated: “the consumer pays”. We will come back to this finding later in the discussion section.
what kind of pictures did you see?”. Results indicate a strong association between the cause-focused or product-oriented advertisement and participant’s perception of these advertisement execution styles, χ2 (2) = 167.61, p < .001. However, not all participants perceived or remembered the advertisement execution style correctly. Within the product-oriented ad condition, 13 respondents (8.6%) responded incorrectly to the question by answering “the cause”, while six respondents could not remember to what the advertisement visually focused on. Within the cause-focused ad condition, 23 respondents (15.5%) responded incorrectly to the question by answering “the product”, while two respondents could not remember to what the advertisement visually focused on.
A total of 244 (81.6%) participants responded well on both manipulation checks. However, exclusion of participants who incorrectly answered the two questions did not yield different results on the dependent variables and were therefore included in the rest of the analyses.
4.2. CONTROL VARIABLES
To control for other variables that could influence the effect of who pays and advertisement on the dependent variables, five control variables were included: gender, age, cause familiarity, cause importance, and consumer scepticism to CM in general. 4.2.1. GENDER
.03, p = .980. Men (M = 3.03, SD = 1.55) do not significantly differ from women (M = 3.38,
SD = 1.93) on their perceptions of moral hypocrisy, t (225.5) = -1.64, p = .101. Men (M =
3.88, SD = 1.45) do not significantly differ from women (M = 3.78, SD = 1.58) on their rewarding intentions to the firm, t (297) = .56, p = .575. Finally, men (M = 2.60, SD = 1.44) do not significantly differ from women (M = 2.84, SD = 1.52) on negative opinioned communication intentions, t (297) = -1.37, p = .171. We can conclude that gender does not affect the outcomes of our dependent variables. Therefore, gender will not be taken into account in further analysis.
4.2.2. AGE
4.2.3. CAUSE FAMILIARITY
Simple linear regression analyses were conducted to analyse whether respondents’ familiarity with the cause predicted their responses on the six DV’s. Results suggest that cause familiarity did not significantly predict scores on attitude toward the campaign, β = .050, t (297) = 1.02, p = .311. Multiple R squared, indicates that approximately 0.3% of the variation in scores on attitude toward the campaign was predicted by cause familiarity, an insignificant effect (F (1, 297) = 1.03, p = .311). The same insignificant pattern holds for internal attributive firm motives (β= -.073, t (297) = -1.779, p = .076; R2 = .003, F (1, 297) = .81, p = .370), external attributive firm motives (β= .042, t (297) = .898, p = .370; R2 = .011, F (1, 297) = 3.16, p = .076), perceived moral hypocrisy (β= -.015, t (297) = -.292, p = .770; R2 = .000, F (1, 297) = .09, p = .770), and negative opinioned communication intentions (β= .047, t (297) = 1.069, p = .286; R2 = .004, F (1, 297) = 1.14, p = .286). However, we did find a significant effect for cause familiarity on rewarding behavioural intentions to the firm, β= .163, t (297) = 3.689, p < .001. Multiple R squared indicates that approximately 4.4% of the variation in scores on rewarding behavioural intentions to the firm was predicted by cause familiarity, a significant effect (F (1, 297) = 13.68, p < .001). These results suggest that respondents’ familiarity with the cause positively influenced their rewarding behavioural intentions. In other words, the more familiar respondents are with the charity World Vision, the more likely they engage in firm rewarding behaviour. Therefore, cause familiarity will be taken into account as a covariate when analysing rewarding behavioural intentions in further analyses.
4.2.4. CAUSE IMPORTANCE
words, the more important the charity World Vision is to the respondents, the more positive their attitude toward the campaign. Therefore, cause importance will be taken into account as a covariate when analysing attitude toward the campaign in further analyses.
Second, results suggest that cause importance significantly predict scores on internal attributive firm motives, β= .258, t (297) = 4.844, p < .001. Multiple R squared, indicates that approximately 7.3% of the variation in scores on internal attributive firm motives was predicted by cause importance, a significant effect (F (1, 297) = 23.466, p < .001). These results suggest that cause importance positively influenced internal attributive firm motives. In other words, the more important the charity World Vision is to the respondents, the more internal attributions they ascribe to the firm’s motives (i.e., the more they think the firm is socially-motivated). Therefore, cause importance will be taken into account as a covariate when analysing internal attributive firm motives in further analyses.
Third, results suggest that cause importance significantly predict scores on external attributive firm motives, β= -.186, t (297) = -3.855, p < .001. Multiple R squared, indicates that approximately 4.8% of the variation in scores on external attributive firm motives was predicted by cause importance, a significant effect (F (1, 297) = 14.858, p < .001). These results suggest that cause importance negatively influenced external attributive firm motives. In other words, the more important the charity World Vision is to the respondents, the less external attributions they ascribe to the firm’s motives (i.e., the less they think the firm is profit-motivated). Therefore, cause importance will be taken into account as a covariate when analysing external attributive firm motives in further analyses.
In other words, the more important the charity World Vision is to the respondents, the less they perceive the firm as acting moral hypocrite. Therefore, cause importance will be taken into account as a covariate when analysing perceived moral hypocrisy in further analyses.
Fifth, results suggest that cause importance significantly predict scores on rewarding behavioural intentions, β= .320, t (297) = 6.308, p < .001. Multiple R squared, indicates that approximately 11.8% of the variation in scores on rewarding behavioural intentions was predicted by cause importance, a significant effect (F (1, 297) = 39.787, p < .001). These results suggest that cause importance positively influenced rewarding behavioural intentions. In other words, the more important the charity World Vision is to the respondents, the more likely they engage in rewarding behaviour toward the firm. Therefore, cause importance will be taken into account as a covariate when analysing rewarding behavioural intentions in further analyses.
Finally, however, no significant effect was found for cause importance on negative opinioned communication intentions (β= -.098, t (297) = -1.867, p = .063; R2 = .012, F (1, 297) = 3.49, p = .063). Therefore, cause importance will not be taken into account as a covariate when analysing negative opinioned communication intentions in further analyses.
4.2.5. CONSUMER SCEPTICISM TO CM IN GENERAL
campaign. Therefore, consumer scepticism to CM in general will be taken into account as a covariate when analysing attitude toward the campaign in further analyses.
Second, results suggest that consumer scepticism to CM in general significantly predicts scores on internal attributive firm motives, β= -.635, t (297) = -9.105, p < .001. Multiple
R squared, indicates that approximately 21.8% of the variation in scores on internal
attributive firm motives was predicted by consumer scepticism to CM in general, a significant effect (F (1, 297) = 82.893, p < .001). These results suggest that consumer scepticism to CM in general negatively influenced internal attributive firm motives. In other words, the more sceptical respondents are to CM in general, the less internal attributions they ascribe to the firm’s motives (i.e., the less they think the firm is socially-motivated). Therefore, consumer scepticism to CM in general will be taken into account as a covariate when analysing internal attributive firm motives in further analyses.
sceptical respondents are to CM in general, the more they perceive the firm as acting moral hypocrite. Therefore, consumer scepticism to CM in general will be taken into account as a covariate when analysing perceived moral hypocrisy in further analyses. Fifth, results suggest that consumer scepticism to CM in general significantly predicts scores on rewarding behavioural intentions, β= -.583, t (297) = -8.448, p < .001. Multiple R squared, indicates that approximately 19.4% of the variation in scores on rewarding behavioural intentions was predicted by consumer scepticism to CM in general, a significant effect (F (1, 297) = 71.376, p < .001). These results suggest that consumer scepticism to CM in general negatively influenced rewarding behavioural intentions. In other words, the more sceptical respondents are to CM in general, the less likely they engage in rewarding behaviour toward the firm. Therefore, consumer scepticism to CM in general will be taken into account as a covariate when analysing rewarding behavioural intentions in further analyses. Finally, results suggest that consumer scepticism to CM in general significantly predicts scores on negative opinioned communication intentions, β= .539, t (297) = 7.841, p < .001. Multiple R squared, indicates that approximately 17.2% of the variation in scores on negative opinioned communication intentions was predicted by consumer scepticism to CM in general, a significant effect (F (1, 297) = 61.483, p < .001). These results suggest that consumer scepticism to CM in general positively influenced negative opinioned communication intentions. In other words, the more sceptical respondents are to CM in general, the more likely they engage in negative opinioned communication toward the firm. Therefore, consumer scepticism to CM in general will be taken into account as a covariate when analysing negative opinioned communication intentions in further analyses.
toward the campaign, (2) perceived internal attributive firm motives, (3) perceived external attributive firm motives, (4) perceived moral hypocrisy, (5) rewarding behavioural intentions, and (6) negative opinioned communication intentions, while controlling for the covariates based on the outcomes stated in section 4.2. Besides exploring the direct effects, these analyses also help answering the main research question of this study; how consumers perceive who pays for the donation in implicit-stated CM campaigns. The outcomes on the DV’s of the experimental condition of who pays (consumer pays or company pays) that are most similar to the outcomes of the comparison group, i.e., the ambiguous condition, demonstrate how consumers actually perceive who pays for the donation in CM campaigns. In addition, several exploratory ANOVA’s were conducted to find further support for the outcome.
Second, by means of the PROCESS macro Hayes (2012), the indirect effects were studied of our two IV’s via perceived internal and external attributive firm motives, and perceived moral hypocrisy on consumer attitude toward the campaign, rewarding behavioural intentions and negative opinioned communication intentions. This tool is an advanced regression-based approach that combines mediation and moderation analysis and “focuses on the estimation and interpretation of the conditional nature (the moderation component) of the indirect and/or direct effects (the mediation component of X on Y in a causal system)” (Hayes, 2013, p.10). Upon analyses of these results we confirm or reject the stated hypotheses of our study.
4.3.1. TESTING THE DIRECT EFFECTS
4.3.1.1 CONSUMER ATTITUDE TOWARD THE CAMPAIGN
.001). The ambiguous group (EMM = 5.64, SE = .11) did significantly differ from the consumer pays group (p < .001), but not from the company pays group (p = .065). These results indicate that participants who thought the company pays and those who were told nothing about the price (i.e., ambiguous) are most similar and both hold a more positive attitude toward the campaign, compared to respondents who thought the consumer pays. The main effect of advertisement execution style as well as the interaction effect were not significant (respectively F (1, 191) = 1.69, p = .195, η2 = .006 and F (2, 191) = .55, p = .547, η2 = .004). In table 2 the estimated marginal means and standard errors of the six groups are given for consumer attitude toward the campaign. The covariates, consumer scepticism to CM in general and cause importance, were significantly related to consumer attitude toward the campaign (respectively F (1, 291) = 28.38, p < .001, η2 = .089 and F (1, 291) = 21.78, p < .001, η2 = .070).
4.3.1.2 INTERNAL ATTRIBUTIVE FIRM MOTIVES
attributive firm motives (respectively F (1, 191) = 28.43, p < .001, η2 = .089 and F (1, 191) = 11.03, p < .001, η2 = .037).
4.3.1.3. EXTERNAL ATTRIBUTIVE FIRM MOTIVES
price (i.e., ambiguous), and lowest among respondents who thought the company pays. No significant main effect for advertisement execution was found, F (1, 190) = .42, p = .518, η2 = .001. In table 2 the estimated marginal means and standard errors of the six groups are given for external attributive firm motives. The covariates, age, consumer scepticism to CM in general and cause importance, were all significantly related to external attributive firm motives (respectively F (1, 190) = 18.79, p < .001, η2 = .061; F (1, 190) = 16.19, p < .001, η2 = .053 and F (1, 191) = 6.33, p < .05, η2 = .021). Figure 1. Interaction effect of who pays and advertisement execution style on perceived external attributive firm motives. 4.3.1.4 PERCEIVED MORAL HYPOCRISY A Two-Way ANCOVA was conducted to explore the direct effects of our two IV’s on perceived moral hypocrisy, while controlling for consumer scepticism to CM in general and cause importance. A significant main effect of who pays on perceived moral hypocrisy was found, F (2, 291) = 114.80, p < .001, η2 = .441. In accordance with hypothesis 2C, respondents who thought the consumer pays for the donation perceived significantly higher levels of moral hypocrisy (EMM = 4.66, SE = .12) than respondents who thought the company pays for the donation (EMM = 2.22, SE = .12) (p < .001). The 3,5 4 4,5 5 5,5 6 Company
pays Consumer Pays Ambiguous
ambiguous group (EMM = 2.65, SE = .11) perceived significantly lower levels of moral hypocrisy than the consumer pays group (p < .001), but higher levels than the company pays group (p = .014). These results indicate that consumers who think they themselves have to pay for the donation perceive the highest levels of moral hypocrisy, followed by those who were told nothing about the price. People who think the company pays perceive the lowest levels of moral hypocrisy. The main effect of advertisement execution style as well as the interaction effect were not significant (respectively F (1, 291) = 1.98, p = .160, η2 = .007 and F (2, 291) = 1.71, p = .182, η2 = .012). In table 2 the estimated marginal means and standard errors of the six groups are given for perceived moral hypocrisy. The covariate consumer scepticism to CM in general was significantly related to perceived moral hypocrisy, F (1, 291) = 33.22, p < .001, η2 = .102. However, the covariate cause importance was not significantly related, F (1, 291) = 3.23, p = .073, η2 = .011). 4.3.1.5. REWARDING BEHAVIOURAL INTENTIONS
A Two-Way ANCOVA was conducted to explore the direct effects of our two IV’s on rewarding behavioural intentions, while controlling for consumer scepticism to CM in general, cause familiarity and cause importance. A significant main effect of who pays on rewarding behavioural intentions was found, F (2, 290) = 34.86, p < .001, η2 = .194. In accordance with hypothesis 1B, respondents who thought the consumer pays for the donation were significantly less likely to engage in rewarding behaviour (EMM = 2.99, SE = .12) than respondents who thought the company pays for the donation (EMM = 4.39,
SE = .12) (p < .001). The ambiguous group (EMM = 4.13, SE = .12) did significantly differ
related to rewarding behavioural intentions (respectively F (1, 290) = 23.50, p < .001, η2 = .075; F (1, 290) = 8.92, p < .01, η2 = .030 and F (1, 290) = 17.73, p < .001, η2 = .058). 4.3.1.6. NEGATIVE OPINIONED COMMUNICATION INTENTIONS
Table 2.
Group means 2x3 design for each of the six dependent variables
Dependent
Variable Who pays Advertisement execution style Estimated Marginal Means (Standard Error) Attitude toward the campaign Consumer pays Cause-focused Product-oriented 3.92 (.16) 3.87 (.17) Company pays Cause-focused Product-oriented 6.16 (.16) 5.78 (.16) Ambiguous Cause-focused Product-focused 5.69 (.16) 5.60 (.16) Internal attributive
firm motives Consumer pays Cause-focused Product-oriented 3.83 (.17) 3.65 (.17) Company pays Cause-focused Product-oriented 5.54 (.17) 5.13 (.17) Ambiguous Cause-focused Product-oriented 5.13 (.17) 5.05 (.17) External attributive firm motives Consumer pays Cause-focused Product-oriented 5.63 (.17) 5.34 (.17) Company pays Cause-focused Product-oriented 3.98 (.17) 4.53 (.17) Ambiguous Cause-focused Product-oriented 4.79 (.16) 4.79 (.16) Perceived moral
hypocrisy Consumer pays Cause-focused
Product-oriented
Note. Covariates appearing in the model are evaluated at the following values: Consumer scepticism to CM in general = 3.63; Cause familiarity = 2.50; Cause importance = 4.09; Age = 35.51.
4.3.2. CONSUMERS’ PERCEPTION OF WHO PAYS IN IMPLICIT-STATED CM CAMPAIGNS
words, participants who were told nothing about price changes have similar response patterns as participants who were told the price did not increased thus the company pays for the donation. Based on previous ANCOVA-analyses we would conclude that consumers expect the company to pay for the donation in actual CM campaigns. However, in order to find more support this assumption, several exploratory analyses are performed. This time we take into account how respondents actually perceived who paid, by using the manipulation check question, “who paid for the donation?” as IV. 4.3.2.1. EXPLORATORY ANALYSIS CONSUMER ATTITUDE TOWARD THE CAMPAIGN
A One-Way between subjects ANOVA was conducted to compare the effect of who paid on attitude toward the campaign, in “the company paid”, “the consumer paid”, and “I don’t know” answer. There was a significant effect of who paid on attitude toward the campaign for the three answers, F (2, 296) = 62.30, p < .001. Because we want to know how consumers perceive who pays in the ambiguous condition (i.e., when they answered “I don’t know”), we compare respondents who answered the manipulation question with “I don’t know” with respondents who answered “the company paid” and “the consumer paid”. Post hoc comparisons using the Tukey HSD test indicated that the mean score for the “I don’t know” answer (M = 5.63, SD = 1.12) did not significantly differ from “the company paid” answer (M = 5.93, SD = 1.06) (p = .631). However, the “I don’t know” answer was significantly different from “the consumer paid” answer (M = 4.08, SD = 1.76) (p < .001). Taken together, these results suggest that respondents who were told nothing about a price change, expect the company to pay for the donation in a CM campaign. 4.3.2.2. EXPLORATORY ANALYSIS PERCEIVED INTERNAL ATTRIBUTIVE FIRM MOTIVES A One-Way between subjects ANOVA was conducted to compare the effect of who paid on perceived internal attributive firm motives, in “the company paid”, “the consumer paid”, and “I don’t know” answer. There was a significant effect of who paid on perceived internal attributive firm motives for the three answers, F (2, 296) = 51.35, p < .001. Post hoc comparisons indicated that the mean score for the “I don’t know” answer (M = 4.96,
SD = 1.04) did not significantly differ from “the company paid” answer (M = 5.40, SD =
“the consumer paid” answer (M = 3.78, SD = 1.65) (p < .001). Taken together, these results suggest that respondents who were told nothing about a price change, expect the company to pay for the donation in a CM campaign.
4.3.2.3. EXPLORATORY ANALYSIS PERCEIVED EXTERNAL ATTRIBUTIVE FIRM MOTIVES
A One-Way between subjects ANOVA was conducted to compare the effect of who paid on perceived external attributive firm motives, in “the company paid”, “the consumer paid”, and “I don’t know” answer. There was a significant effect of who paid on perceived external attributive firm motives for the three answers, F (2, 296) = 26.99, p < .001. However, post hoc comparisons indicated that the mean score for the “I don’t know” answer (M = 4.89, SD = 1.13) was not significantly different from both “the company paid” answer (M = 4.36, SD = 1.31) (p = .193) and “the consumer paid” answer (M = 5.49,
SD = 1.25) (p = .145). Taken together, these results do not clearly indicate how
1.01) did not significantly differ from “the company paid” answer (M = 4.46, SD = 1.26) (p = .925). However, the “I don’t know” answer was significantly different from “the consumer paid” answer (M = 2.93, SD = 1.41) (p < .001). Taken together, these results suggest that respondents who were told nothing about a price change, expect the company to pay for the donation in a CM campaign. 4.3.2.6. EXPLORATORY ANALYSIS NEGATIVE OPINIONED COMMUNICATION INTENTIONS A One-Way between subjects ANOVA was conducted to compare the effect of who paid on negative opinioned communication intentions, in “the company paid”, “the consumer paid”, and “I don’t know” answer. There was a significant effect of who paid on negative opinioned communication intentions for the three answers, F (2, 296) = 25.73, p < .001. However, post hoc comparisons indicated that the mean score for the “I don’t know” answer (M = 2.72, SD = 1.47) was not significantly different from both “the company paid” answer (M = 2.19, SD = 1.15) (p = .193) and “the consumer paid” answer (M = 3.37,
SD = 1.60) (p = .145). Taken together, these results do not clearly indicate how
consumers perceive who pays in implicit-stated CM campaigns.
Again these results show that participants not knowing who paid for the donation appear to be more similar to participants indicating the company paid than to participants saying the consumer paid. In other words, participants who were told nothing about price changes have similar response patterns as participants who were told the price did not increased thus the company pays for the donation. In conclusion, both the ANCOVA-analyses and exploratory ANOVA-analyses indicate that consumers expect the company to pay for the donation in actual CM campaigns.
4.3.3. TESTING THE CONDITIONAL INDIRECT EFFECTS - MODERATED MEDIATION ANALYSIS
significantly negative, whereas a positive lower and upper bound can be interpreted as a positive mediation effect.
In this analysis only the two experimental groups for the independent variable who pays are used: the consumer pays versus the company pays. Advertisement execution (cause-focused versus product-oriented) was set as the moderator. Internal attributive firm motives, external attributive firm motives and perceived moral hypocrisy, were set as mediators. Due to the fact that the two independent variables are categorical, we dummy coded the categorical IV’s. Both IV’s have 2 categories; therefore 1 dummy variable was created for each IV. The three covariates, consumer scepticism to CM in general, cause familiarity and cause importance, were retained in the analysis.
4.3.3.1. THE DIRECT EFFECT OF WHO PAYS ON THE DV’S: ATTITUDE TOWARD THE CAMPAIGN, REWARDING BEHAVIOURAL INTENTIONS AND NEGATIVE OPINIONED COMMUNICATION INTENTIONS In accordance to previous results of the ANCOVA analyses and consistent with hypothesis 1A, respondents who think the consumer pays for the donation have a less positive attitude toward the campaign than respondents who think the company pays for the donation (B = -.5.55, p < .01). The fact that we found a direct effect even when taking into account the three mediators, suggests a partial mediation. While previous results of the ANCOVA analyses found evidence for hypothesis 1B, here we did not find a direct effect of who pays on rewarding behavioural intentions (B = -.282, p = .205). The fact that we found no direct effect when taking into account the three mediators, suggests a full mediation effect. While previous results of the ANCOVA analyses found evidence for hypothesis 1C, here we did not find a direct effect of who pays on negative opinioned communication intentions (B = 0.21, p = .931). The fact that we found no direct effect when taking into account the three mediators, suggests a full mediation effect.
4.3.3.2. THE DIRECT EFFECT OF WHO PAYS ON THE MEDIATIORS: PERCEIVED INTERNAL AND EXTERNAL ATTRIBUTIVE FIRM MOTIVES, AND PERCEIVED MORAL HYPOCRISY
internal attributions to the firm’s motives than respondents who think the company pays for the donation (B = -2.00, p < .001). Hypothesis 2A is again confirmed.
In accordance to previous results of the ANCOVA analyses and consistent with hypothesis 2B, respondents who think the consumer pays for the donation ascribe more external attributions to the firm’s motives than respondents who think the company pays for the donation (B = 1.78, p < .001). Hypothesis 2B is again confirmed.
In accordance to previous results of the ANCOVA analyses and consistent with hypothesis 2C, respondents who think the consumer pays for the donation perceive higher levels of moral hypocrisy than respondents who think the company pays for the donation (B = 2.68, p < .001). Hypothesis 2C is again confirmed.
4.3.3.3. CONSUMER ATTITUDE TOWARD THE CAMPAIGN
significant as evidenced by a 95% bias-corrected bootstrap confidence interval that does not include zero [-1.29; -.33]. Hypothesis 3C is confirmed.
Hypothesis 6A predicted that advertisement execution style would moderate the relationship between who pays and perceived internal attributive firm motives, such that people who think the consumer pays for the donation in a CM campaign ascribe less internal attributive motives to the firm when the advertisement is cause-focused than when it is product-oriented. However, consistent with previous results from the ANCOVA analyses, no such interaction effect on perceived internal attributive motives is found (B = .34, p = .41, 95% CI [-.47; 1.15]). Hypothesis 6A needs to be rejected.
Hypothesis 6B predicted that advertisement execution style would moderate the relationship between who pays and perceived external attributive firm motives, such that people who think the consumer pays for the donation in a CM campaign ascribe more external attributive motives to the firm when the advertisement is cause-focused than when it product-oriented. Consistent with previous results from the ANCOVA analyses, a significant interaction effect on perceived external attributive motives is found (B = -.84, p < .05, 95% CI [-1.53; -.15]). Although we were not able to interpret the interaction effects more precisely, figure 1 indicates that, indeed, people who think the consumer pays for the donation in a CM campaign ascribe more external attributive motives to the firm when the advertisement is cause-focused than when it product-oriented. Hypothesis 6B is confirmed.
Hypothesis 6C predicted that advertisement execution style would moderate the relationship between who pays and perceived moral hypocrisy, such that people who think the consumer pays for the donation in a CM campaign perceive higher levels of moral hypocrisy when the advertisement is cause-focused than when it is product-oriented. However, consistent with previous results from the ANOVA analysis, no such interaction effect on perceived moral hypocrisy is found (B = -.07, p = .84, 95% CI [-.82; .67]). Hypothesis 6C needs to be rejected.
that people who think the company pays for the donation in a CM campaign ascribe more internal attributive motives to the firm when the advertisement is cause-focused than when it is product-oriented. However, consistent with previous results from the ANCOVA analyses, no such interaction effect on perceived internal attributive motives is found (B = .34, p = .41, 95% CI [-.47; 1.15]). Hypothesis 7A needs to be rejected.
Hypothesis 7B predicted that advertisement execution style would moderate the relationship between who pays and perceived external attributive firm motives, such that people who think the company pays for the donation in a CM campaign ascribe less external attributive motives to the firm when the advertisement is cause-focused than when it is product-oriented. Consistent with previous results from the ANCOVA analyses, a significant interaction effect on perceived external attributive motives is found (B = -.84, p < .05, 95% CI [-1.53; -.15]). Although we were not able to interpret the interaction effects more precisely, figure 1 indicates that, indeed, people who think the company pays for the donation in a CM campaign ascribe less external attributive motives to the firm when the advertisement is cause-focused than when the advertisement in product-oriented. Hypothesis 7B is confirmed.
Hypothesis 8 predicted that perceptions of who pays is related to consumer attitude toward the campaign through indirect effects such that the effect of who pays on consumer attitude toward the campaign will be moderated by advertisement execution style and mediated by perceived internal attributive firm motives, perceived external attributive firm motives, and perceived moral hypocrisy. Preacher et al. (2007) note that evidence of moderated mediation exists if the estimates of the indirect effects transmitted through the mediator variable are significant across levels of the moderator variable.
cause-focused advertisement, 95% CI [-1.62; -.73] and product-oriented advertisement, 95% CI [-1.48; -.58]. For perceived external attributive firm motives the confidence interval did contain zero for cause-focused advertisement, 95% CI [-.29; .23] and product-oriented advertisement, 95% CI [-1.18; -.14]. The confidence interval for perceived moral hypocrisy did not contain zero for cause-focused advertisement, 95% CI [-1.60; -.42] and product-oriented advertisement, 95% CI [-1.66; -.40]. The overall index of moderated mediation is not significant for non of the mediators, as the confidence intervals do contain zero (perceived internal attributive motives, 95% CI [-.23; .59], perceived external attributive motives, 95% CI [-.11; .13], perceived moral hypocrisy, 95% CI [-.25; .27]).
4.3.3.4. REWARDING BEHAVIOURAL INTENTIONS
Hypothesis 4A predicted that perceived internal attributive firm motives mediate the relationship between who pays and rewarding behavioural intentions. Results of PROCESS’s simple mediation analyses also revealed support for this assertion. The indirect effect of who pays on rewarding behavioural intentions through perceived internal attributive motives is significant as evidenced by a 95% bias-corrected bootstrap confidence interval that does not include zero [.23; .55]. Hypothesis 4A is confirmed.
Hypothesis 4B predicted that perceived external attributive firm motives mediate the relationship between who pays and rewarding behavioural intentions. Results of PROCESS’s simple mediation analyses also revealed support for this assertion. The indirect effect of who pays on rewarding behavioural intentions through perceived external attributive motives is significant as evidenced by a 95% bias-corrected bootstrap confidence interval that does not include zero [-.39; -.09]. Hypothesis 4B is confirmed.
Hypothesis 4C predicted that perceived moral hypocrisy mediate the relationship between who pays and rewarding behavioural intentions. Results of PROCESS’s simple mediation analyses did not reveal support for this assertion. The indirect effect of who pays on rewarding behavioural intentions through perceived moral hypocrisy is not significant as evidenced by a 95% bias-corrected bootstrap confidence interval that does include zero [-.25; .05]. Hypothesis 4C needs to be rejected.
Figure 3. Full mediation model of effect who pays and advertisement execution style on
rewarding behavioural intentions, mediated by perceived internal attributive motives and perceived moral hypocrisy.
4.3.3.5. NEGATIVE OPINIONED COMMUNICATION INTENTIONS
Hypothesis 5A predicted that perceived internal attributive firm motives mediate the relationship between who pays and negative opinioned communication intentions. Results of PROCESS’s simple mediation analyses also revealed support for this assertion. The indirect effect of who pays on negative opinioned communication intentions through perceived internal attributive motives is significant as evidenced by a 95% bias-corrected bootstrap confidence interval that does not include zero [-.35; -.06]. Hypothesis 5A is confirmed.
Hypothesis 5C predicted that perceived moral hypocrisy mediate the relationship between who pays and negative opinioned communication intentions. Results of PROCESS’s simple mediation analyses did not reveal support for this assertion. The indirect effect of who pays on negative opinioned communication intentions through perceived moral hypocrisy is significant as evidenced by a 95% bias-corrected bootstrap confidence interval that does not include zero [.41; .73]. Hypothesis 5C is confirmed. Hypothesis 10 predicted that perceptions of who pays is related to negative opinioned communication intentions through indirect effects such that the effect of who pays on negative opinioned communication intentions will be moderated by advertisement execution style and mediated by perceived internal attributive firm motives, perceived external attributive firm motives, and perceived moral hypocrisy. Preacher et al. (2007) note that evidence of moderated mediation exists if the estimates of the indirect effects transmitted through the mediator variable are significant across levels of the moderator variable.
In conclusion, no significant moderated mediation effect is found. Thus, hypothesis 10 needs to be rejected. However, we do have found simple mediation effects whereby perceived internal attributive firm motives and perceived moral hypocrisy mediate the effect of who pays on negative opinioned communication intentions. When including the
mediators, there is no significant direct effect of who pays on negative opinioned communication intentions (B = .03, p = .90, 95% CI [-.44; .50]). The final model can be found in Figure 4. Finally, an overview of all the statistical tests conducted and the supported or not supported hypotheses can be found in table 3.
Figure 4. Full mediation model of effect who pays on negative opinioned communication
5.
DISCUSSION AND CONCLUSIONCompanies initiate CM campaigns to contribute to society and provide additional benefits to consumers. These campaigns are often associated with price increases to which consumers may react negatively (Krishna and Rajan, 2009). However, companies do not communicate such price changes explicitly in their campaigns. Yet, knowledge is lacking on how consumers perceive who is paying for the donation in implicit-stated CM campaigns. Against this background, the current study had three objectives that address important research gaps in the literature. First, we examined how consumers actually perceive who is paying for the donation in CM campaigns. Insights on this perception might enrich literature, which tries to explain consumers’ responses and especially shed light on why and when consumers perceive a company as acting moral hypocrite. Second, we investigated under what circumstances consumers perceive higher levels of moral hypocrisy in terms of a moderating effect of advertisement execution style. Third, we investigated the process of how the interplay between who pays and the moderating variable advertisement execution style leads to attitude toward the campaign, potential rewarding behaviour and negative opinioned communication intentions.
be thought of as buying a cause-related product while the consumer pays for the donation. Gneezy et al. 2012 found people engaging in costly prosocial behaviour to interpret that behaviour as a signal of their prosocial identity and therefore are more likely to behave prosocially in subsequent behaviour. On the other hand, costless prosocial behaviour does not affect an individual’s self-perception, resulting in reduction of subsequent prosocial behaviour. Our finding that consumers who buy a CM product expect the company to pay for the donation can thus be seen as prosocial behaviour that comes at no cost; consumers help donating to a charity without actually paying for it. Moreover, our study suggests consumers’ expectation that the company pays to positively influence their attitudes and rewarding behavioural intentions. This increase in purchase intentions might of course be beneficial for both the firm and social cause. However, based on the work of Gneezy et al. 2012, this costless prosocial behaviour might hurt society on the long term, as it leads to a reduction of subsequent prosocial behaviour. In conclusion, the fact that consumers expect the company to pay might be less beneficial than it looks at first sight.
