When Should a Brand Cut Ties With a Scandalized Endorser?

Abstract

It is important for brands to appropriately respond when their athlete endorsers are involved in a scandal. The present study examines how consumer evaluations of endorsed brands are influenced by a brand’s response to an endorser scandal. A 2 (brand response strategy type: maintenance vs. termination) × 2 (endorser scandal type: competence-relevant vs. competence-irrelevant) between-subjects experiment is conducted. Specifically, the authors focus on the perceived appropriateness of the response strategy and consumer attitudes toward an endorsed brand. Subjects were recruited from Amazon Mechanical Turk (N = 111). Consumers perceive that terminating an endorsement contract with a scandalized endorser is more appropriate than maintaining the relationship. This is particularly true when celebrity endorsers are involved in competence-relevant scandals. A further analysis provided support for the idea that perceived “appropriateness” mediates the relationship between competence-relevant scandal and consumer attitude toward an endorsed brand. A termination strategy was considered appropriate when scandalized endorsers engaged in competence-relevant scandals, which eventually lead to more favorable consumer evaluations toward an endorsed brand.

Keywords

celebrity endorsement scandals response strategy advertising crisis communication

Companies consider celebrity endorsement to be an effective marketing strategy. Roger Federer and LeBron James both earned approximately US$60 million in 2017 solely from their endorsements (Badenhausen, 2017b). Nike enjoys a favorable return on investment from the endorsement contract with James, given that his signature shoe brand sold more than any other brand endorsed by any other basketball player (Badenhausen, 2017a). Yet this marketing communication strategy is also recognized to be a double-edged sword. Celebrity endorsers are often involved in misbehaviors that receive keen media attention (Rowe, 2017), and this potentially hurts endorsed brands (J. S. Lee, Kwak, & Moore, 2015; Sato, Ko, Chang, & Kay, 2019). The collective shareholder values of Tiger Woods’s endorsed brands decreased approximately US$5–US$12 billion when his sexual infidelity scandal was exposed to the public (Knittle & Stango, 2013).

Endorsed brands have only limited options to consider—either maintaining or terminating a relationship with a scandalized athlete endorser (Agyemang, 2011). Executives tend to respond to crises in a manner that they consider “suited” to their brands rather than following a predictable set of rules. For example, the luxury watch brand, TAG Heuer, announced the decision to break its endorsement contract with Maria Sharapova when her performance-enhancing drugs (PEDs) scandal broke out (Doerr, 2016). In contrast, Nike decided to maintain its endorsement relationship with Sharapova (Thomas, 2016) during this time. Earlier, Nike had made a similar decision to maintain its relationship with Tiger Woods when his sexual infidelity was exposed to the public, yet had decided to immediately cut ties with Manny Pacquiao, the professional boxer, when he made comments that were insensitive to homosexuals (Perez, 2016). These examples highlight inconsistent brand responses toward misbehaving endorsers.

From a business standpoint, consumers are the definitive target audience for endorsed brands. Appropriate communications generate favorable brand responses from target audiences (Cheshin, Amit, & van Kleef, 2018). Communication executives need to determine the appropriate response strategy from the customer perspective to maintain favorable attitudes toward their brands. Perceived “appropriateness” refers to the extent to which consumers perceive whether a focal communication does not violate relational or situationally accepted norms (Nicotera, 1995; Seele & Lock, 2015). For example, most people think that spitting on the ground is not an appropriate behavior because it violates generally accepted standards. However, such perceptions of appropriateness can vary (Weber, Kopelman, & Messick, 2004). Sport fans accept that baseball players commonly spit on the field during games. Analogously, consumers perceive that the appropriateness of a brand’s response towards a scandalized athlete can also depend on the situation.

In this respect, scholars have recognized the importance of transgression types (J. S. Lee et al., 2015). The key consideration is whether focal transgressions are relevant (or irrelevant) to the competency of an endorsers. Using the above examples, the PEDs incident of Maria Sharapova was “competence-relevant” because of the direct relationship between the incident and her athletic performance. The discriminatory remark issue of Manny Pacquiao was “competence-irrelevant” as it had nothing to do with his athletic performance. Applying the competence relevancy (or irrelevancy) dichotomy, prior research indicates that the competence relevancy of a transgression influences the interpretation (J. S. Lee et al., 2015) and the evaluation of a wrongdoers (Sato, Ko, Park, & Tao, 2015). These studies suggest that consumers are more critical to competence-relevant issues due their immediate congruencies to sport performance. However, there have been no studies as to how these factors interact with decisions (i.e., termination or maintenance) to influence fan responses or how these factors affect endorsed brands.

The current study was conducted to investigate the effects of brand response strategy type (maintenance vs. termination) and scandal type (competence-relevant vs. competence-irrelevant) on consumer judgement about the response strategies and endorsed brands. We focus on how brand response strategies and scandal types interact to influence (1) consumers’ perceived appropriateness of response strategies and (2) fan attitudes toward an endorsed brand in the context of athlete transgressions. We also assessed a moderated mediation model that incorporates the above factors to better explain the underlining causal relationships.

Theoretical Background and Hypotheses Development

Competence Relevancy of Athlete Endorser Transgressions

Existing research has examined various types of transgressions to understand the underlined mechanism regarding their impact (Brown, Murphy, & Maxwell, 2018; Coombs, 2007; Kim, Kim, & Cameron, 2009). The concept of “competence-relevance” has been utilized to classify athlete transgressions (C. Lee, Bang, & Lee, 2013; J. S. Lee et al., 2015; Sato et al., 2015). Competence-relevant scandals are related to the domain of sports performance, while competence-irrelevant scandals emphasize endorsers’ moral behavior in their everyday life. Doping issues, for example, are competence-relevant because these have the potential to boost or alter athletic performance, whereas extramarital infidelities are considered competence-irrelevant as they are unrelated to the on-field athletic performance.

The legitimacy of using competence relevancy is illustrated by several frameworks. Dawar and Lei (2009) utilized a brand association framework to argue that the severity of brand scandals depends on the relevance of the scandal to the brand’s key benefit. In the current research context, the athletic competencies of athlete endorsers are a key brand element since they establish and maintain their performance status based on the excellence in sport (Arai, Ko, & Ross, 2014). Sato and his colleagues (2015) confirmed that perceived competence relevancy, which they called “performance relatedness” in their study, leads to more negative evaluations toward the wrongdoing endorsers. Prior case studies also indicate differences in people’s response to competence-relevant and competence-irrelevant scandals. Although a majority of sport spectators show no tolerance toward doping (Becker & Scheufele, 2008), the case of Michael Phelps is an exception. His “nonperformance-related” drug use (i.e., recreational use of marijuana) did not diminish fan support for Phelps. In fact, Kellogg’s stock value dropped when they ended their endorsement contract with Phelps after the incident (J. Walsh & McAllister-Spooner, 2011).

Perceived Appropriateness of Brand Response Strategies: Maintain or Terminate?

The goal of a crisis response strategy is to minimize the damage by appropriately communicating with stakeholders (Coombs, 2007; Kim et al., 2009). Perceived appropriateness is defined as the extent to which individuals perceive whether a focal communication does not violate relational or situationally accepted norms (Nicotera, 1995; Seele & Lock, 2015). Response strategies that are perceived appropriate by consumers are effective in restoring images (Benoit & Drew, 1997). In athlete transgression contexts, marketing and communication managers from an endorsed brand need to communicate with consumers promptly to explain their decisions, either maintaining or terminating the contract with misbehaving endorsers (Knittel & Stango, 2013). Hence, brands need to understand which response strategy, either to terminate or to maintain, is perceived as more appropriate in athlete transgression settings.

Various brands have made inconsistent decisions as to whether to maintain or to terminate their endorsement ties with transgressed athlete endorsers. For example, Nike cut ties with Lance Armstrong while the relationship with Maria Sharapova was maintained (although both cases were related to PED). Miller and Laczniak (2011) summarize decisions made by brands with transgressed endorsers (including Kobe Bryant, Michael Phelps, and Tiger Woods) and showed that companies tended to respond in accordance with their endorsement agreements. Specifically, when Tiger Woods’s scandal emerged, AT&T, Accenture, and Gillette implemented the termination strategy, while Nike and EA Sports decided to maintain the relationship with Woods. These examples clearly indicate that response strategies are executed without a set standard.

Cutting Endorser Ties to Maintain Relationships With Consumers

The current research proposes a conceptual model to explain the formation of consumer judgments toward endorsed brands in athlete scandal settings based on classic theories of cognitive information processing. Balance theory posits that individuals seek to maintain balanced mental states when they consider relationships among multiple attitude objects (Heider, 1958; Solomon, 2014). The relationship involves three attitude objects: (1) a person, (2) her perception of an attitude object (e.g., endorsed brand), and (3) a celebrity endorser (Solomon, 2014). Individuals strive to maintain a harmonious balance (or consistency) between related objects by altering their attitudes toward each object. A balanced relationship is maintained by making only one of the three relationships positive or by making all relationships positive (Heider, 1958). For example, a celebrity in an endorsement contract creates attitude imbalance when a preexisting positive perception toward a particular brand is confronted by a negative perception due to a scandalized endorser. According to balance theory, when consumers are faced with this situation, they seek to reduce the tension by changing their perception of the brand and/or the celebrity endorser in order to restore attitude balance.

In the context of the current study (see Figure 1), the baseline expectation is that typical sport fans evaluate a scandalized endorser negatively. This has been supported elsewhere (Brown, 2016; J. S. Lee & Kwak, 2016; Sato et al., 2019). The negative relationship between sport fans and scandalized endorsers results in an imbalance in the relationship. To correct this imbalance, previous research (McCracken, 1989) suggests that consumers will evaluate an endorsed brand negatively. Branding executives may preempt this consumer reaction by cutting ties with the scandalized athlete. MINI USA’s decision to pull its commercial of Abby Wambach, a U.S. soccer player, after she was arrested for driving under the influence (Associated Press, 2016) is an example of a termination strategy. A termination strategy appears more effective than a maintenance strategy in mitigating consumers’ critical evaluation. Hence, the authors present the first hypothesis followed by the second:

Hypothesis 1: A termination response strategy will be perceived as more appropriate than a maintenance response strategy.

Hypothesis 2: Endorsed brands will be evaluated more favorably when they execute the termination response strategy than the maintenance response strategy.

Figure 1.

Conceptual model.

Differential Effects of Termination Response Strategy

The idea that an endorsed brand should always cut the relationship with misbehaving endorsers can be supported by balance theory. However, termination strategies can generate different consumer responses, depending on the situation, because perceived appropriateness of communication is closely associated with expectations (Cheshin et al., 2018). It follows that the effect of the termination strategy can be moderated by scandal type. According to the accessibility-diagnosticity model (Feldman & Lynch, 1988), consumer judgment is likely to reflect the content of information that is relevant to a judgment task. For example, when consumers make judgments about a celebrity athlete endorser, they may focus on the relevant diagnostic information (e.g., athletic performances of that athlete) more than on some less relevant information (e.g., the athlete’s favorite movies). Consumers are likely to engage in such information processing because they are inclined to avoid effortful information processing tasks (Fiske & Taylor, 1980; Sherman, Lee, Bessenoff, & Frost, 1998). Based on this theoretical perspective, consumer judgments are influenced by scandal information that is relevant to the focal endorser (e.g., a scandal related to athletic performance of the focal athlete endorser). On the other hand, less relevant information (e.g., a scandal related to a person’s private life) about the endorser can be overlooked or viewed as less important. Therefore, the following hypotheses follow:

Hypothesis 3: The termination response strategy will be perceived as more appropriate when athletes are involved in competence-relevant scandals rather than competence-irrelevant scandals.

Hypothesis 4: Endorsed brands will be evaluated more favorably when they execute the termination response strategy against athletes who are involved in competence-relevant scandals rather than competence-irrelevant scandals.

Method

Stimuli Development

The authors used experimental stimuli involving a fictitious athlete (named Phil Franklin) and sports drink brand (called “Equinox”) adopted from previous celebrity endorsement research (Sato, Ko, Kaplanidou, & Connaughton, 2016). Fictitious athletes and brands are used in order to control for confounding variables such as any prior experience, likeability, and familiarity with an athlete (Money, Shimp, & Sakano, 2006). Consistent with prior research, a PED incident was used to represent a competence-relevant scandal, and an extramarital affair was used to depict a competence-irrelevant scandal (C. Lee et al., 2013; J. S. Lee et al., 2015). Four news stories were written incorporating scandal types and response strategies.

Participants and Procedures

One hundred twenty-seven participants were recruited using Amazon Mechanical Turk (39.6% female; mean age = 32.41, SD = 9.81). Amazon Mechanical Turk is a useful online platform that can recruit participants with diverse backgrounds (Buhrmester, Kwang, & Gosling, 2011), and participants’ demographics are more similar to the population in the United States compared to other online panels (Ipeirotis, 2010). In fact, Amazon Mechanical Turk has been utilized commonly in the sport management studies (Billings, Broussard, Xu, & Xu, 2019; Brown et al., 2018). The authors employed a 2 (response type: maintenance vs. termination) × 2 (scandal type: competence-relevant vs. competence-irrelevant) between-subject experimental design. Participants were randomly assigned to one of four conditions. The authors did not find it necessary to include a control condition (i.e., brand neither maintains nor terminates the endorsement agreement) since endorsed brands decide to simply either maintain or terminate the relationship with scandalized endorsers.

With regard to scandal manipulations, participants in both competence-relevant and competence-irrelevant scandal conditions received identical information except for scandal type. The stimuli described a brief introduction to Phil Franklin and his accomplishment as an athlete in both conditions. In the competence-relevant scandal condition, information about Phil Franklin’s use of a PED was inserted in the news feed. Similarly, participants in the competence-irrelevant condition received information about his extramarital affair. These scenarios were presented as online news within an ESPN webpage. The scandal stimuli employed in the study are presented in Figure 2.

Figure 2.

Scandal stimuli.

Participants were exposed to the first stimuli and responded to filler questions that were unrelated to the study. Then, participants were exposed to another online news feed. Participants in both maintenance and termination conditions received almost identical information (other than the endorsed brand’s decision) regarding the endorsement agreement with Phil Franklin. Participants then received a description of the sports drink brand maintaining or terminating the endorser. In order to increase the realism of this study, the news feed appeared as if it were on an actual social network site (see Figure 3). After the exposure to the stimuli, participants answered questions regarding focal study variables, control variables, and manipulation check questions.

Figure 3.

Brand response strategy stimuli.

The authors included several screening questions that asked participants about the contents of the scenario. The median absolute deviation technique (Leys, Ley, Klein, Bernard, & Licata, 2013) was also used to detect outliers. Overall, two outliers and 14 participants who either failed to answer the contents of the scenario or did not complete the experiment were excluded. As a result, 111 useful responses were analyzed. The number of participants in each condition is as follows: Competence-Relevant Scandal × Maintenance (n = 31), Competence-Relevant Scandal × Termination (n = 26), Competence-Irrelevant Scandal × Maintenance (n = 29), and Competence-Irrelevant Scandal × Termination (n = 25).

With regard to the sample size, the G*Power software was used to calculate the minimum sample size required for this study (Faul, Buchner, & Lang, 2009). The G*Power software computation based on the present study’s model that has a maximum of five predictors (two independent variables and three covariates) for two dependent variables (perceived appropriateness and brand attitude) with the medium effect size (.20) and power as .95 indicated that the minimum sample size required for this study was 66. In addition, the sample size in this study was further supported as more than 20 subjects were assigned to each experimental condition (Hair, Anderson, Tatham, & Black, 2009). Descriptive statistics are shown in Table 1.

Table 1.

Descriptive Statistics for Each Condition.

Scandal Type	Competence-Relevant		Competence-Irrelevant
Response strategy type	Maintenance	Termination	Maintenance	Termination
N	31	26	29	25
Perceived appropriateness	4.87	8.12	5.28	6.76
Brand attitude	5.22	7.73	4.67	6.61

Measures

Perceived appropriateness of endorsed brands’ response strategies was measured using 1 item adopted from the previous literature (Benoit & Drew, 1997; Rice, 1993). It was measured by using a semantic differential scale ranging from 1 = inappropriate to 10 = appropriate. The previous literature supports using single-item scales; these are as effective as multi-item scales in terms of capturing the focal constructs (Kwon & Trail, 2005). It is also especially useful when data are collected online (Jordan & Turner, 2008). In fact, many experimental studies in the sport management realm that focus on athlete endorsers and/or scandals have used single-item scales (Carlson & Donavan, 2013; P. Walsh & Williams, 2017).

Brand attitude was measured using items adopted from Dawar and Pillutla (2000). Brand attitude was measured with three semantic differential items (i.e., 1 = bad to 10 = good, 1 = unfavorable to 10 = favorable, and 1 = negative to 10 = positive). The scale demonstrated adequate reliability (α = .98). These items were averaged to create an overall index of brand attitude.

In addition, the authors measured intentionality of acts, controllability of situations, and scandal severity as covariates. The selection of these control variables was guided by studies that demonstrate that consumer perception and evaluation can be influenced by the intentionality of perpetrators (Struthers, Eaton, Santelli, Uchiyama, & Shirvani, 2008), the controllability of the situation (Claeys, Cauberghe, & Vyncke, 2010), and the seriousness of the incident (Zhu & Chang, 2013).

Participants were asked to answer each control variable using 1 item of a 7-point Likert-type scale. Actual items used were “the incident was intentionally created by the athlete (intentionality of acts: 1 = strongly disagree; 7 = strongly agree),” “how likely do you think the athlete could have prevented this incident from occurring (controllability of situations: 1 = not at all; 7 = very much likely),” and “in my opinion, the incident caused serious problems (scandal severity: 1 = strongly disagree; 7 = strongly agree).”

Lastly, participants were asked to answer manipulation check questions. Participants rated whether a scandal was relevant to the endorser’s professional attribute (i.e., athletic performance). The actual item read “this case directly influenced the athlete’s performance” adopted from prior research (Sato et al., 2015). It was measured on a 7-point Likert-type scale ranging from 1 = strongly disagree to 7 = strongly agree. Brand response strategy manipulation was assessed by a 7-point bipolar item (terminated to maintained), asking participants to answer the endorsed brand’s decision regarding the relationship with the endorser.

Data Analysis

We employed a multivariate analysis of covariance (MANCOVA) to assess the proposed hypotheses since the present study has two dependent variables. A series of one-tailed tests were conducted as the proposed hypotheses were one-directional predictions based on the theoretical background. In this case, two-tailed tests may increase the probability of Type II error (Hair et al., 2009). We also used PROCESS Model 7 based on 2,000 bootstrap samples as an optional analysis to further understand the underlying psychological processes. Data analyses were executed by using IBM SPSS 23.0.

Results

Manipulation checks confirmed that competence relevancy of scandals and brand response strategies were successfully manipulated. Participants in the competence-relevant condition reported significantly higher levels of competence relevancy (M = 5.68, SD = 1.09) than those in the competence-irrelevant condition (M = 2.37, SD = 1.42, p < .01). With regard to brand response strategy type, participants in the maintain condition perceived that the relationship between the endorsed brand and the endorser was maintained (M = 5.97, SD = 1.26) rather than the terminate condition (M = 2.22, SD = 1.36, p < .01). These findings indicated that both scandal- and brand response–type manipulations were successful.

For the hypotheses testing, we first ran an MANCOVA. Specifically, the effects of scandals and response strategies on perceived appropriateness and attitude toward endorsed brands were assessed, controlling for intentionality, locus of control, and severity. As a result, none of the covariates emerged significant: intentionality of act, Wilks’s λ = .98, F(2, 103) = .89, p = .41, locus of control, Wilks’s λ = .99, F(2, 103) = .52, p = .60, and scandal severity, Wilks’s λ = .98, F(2, 103) = 1.18, p = .31. Hence, the authors excluded the covariates from further analyses to avoid a Type I error (Hair et al., 2009). The results demonstrated the marginally significant main effect of scandal, Wilks’s λ = .96, F(2, 106) = 2.32, p = .051 (one-tailed), and the significant main effect of response strategies, Wilks’s λ = .75, F(2, 106) = 17.29, p < .001 (one-tailed). This main effect verified that our Hypotheses 1 and 2 were supported (for perceived appropriateness, M _maintenence = 5.08 vs. M _termination = 7.44; for attitude toward brands, M _maintenence = 4.95 vs. M _termination = 7.18).

More importantly, the interaction between scandal type and response strategy type was marginally significant, Wilks’s λ = .97, F(2, 106) = 2.34, p = .050 (one-tailed). A following univariate analysis revealed the significant main effect of response type, F(1, 107) = 27.81, p < .001 (one-tailed), η² = .21, on perceived appropriateness. Although the main effect of scandal type, F(1, 107) = 1.12, p = .143 (one-tailed), η² = .01, was not significant, the interaction between response type and scandal type was significant, F(1, 107) = 3.85, p = .026 (one-tailed), η² = .04. The results of simple effect test revealed that perceived appropriateness in the competence-relevant scandal condition (M = 4.87, SD = 2.31; bar ①) and the competence-irrelevant scandal condition were not significantly different, M = 5.28, SD = 2.90; bar ③; F(1, 107) = .44, p = .254 (one-tailed), when the endorsed brand maintained the relationship with scandalized endorser. On the other hand, when the endorsed brand terminated the relationship with the scandalized athlete, participants perceived that the response was more appropriate when the athlete was involved in the competence-relevant scandal (M = 8.12, SD = 1.80; bar ②) than the competence-irrelevant scandal, M = 6.76, SD = 2.20; bar ④; F(1, 107) = 4.23, p = .021 (one-tailed). Therefore, our Hypothesis 3 was upheld (Figure 4).

Figure 4.

Simple effect test on perceived appropriateness of response strategy.

Regarding Hypothesis 4, another univariate analysis for brand attitude revealed that the main effects of scandal, F(1, 107) = 4.37, p = .019 (one-tailed), η² = .04, as well as response type, F(1, 107) = 31.34, p < .001 (one-tailed), η² = .23, were significant on brand attitude. Participants in the competence-relevant scandal condition (M = 6.36, SD = 2.43) reported more favorable attitude toward endorsed brand than the competence-irrelevant condition (M = 5.57, SD = 2.28). In addition, participants evaluated endorsed brands more favorably when the termination strategy (M = 7.18, SD = 1.64), rather than the maintenance strategy (M = 4.95, SD = 2.44), was implemented. Although these results further supported the second hypothesis that endorsed brands terminating the relationship with scandalized endorsers are more favorably evaluated than those maintaining the relationship, the interaction effect between response type and scandal type was not significant, F(1, 107) = 0.51, p = .238 (one-tailed), η² = .01, rejecting our Hypothesis 4 (Table 2).

Table 2.

F Values for Dependent Variables in MANCOVA and ANOVA.

		Univariate ANOVA
	MANCOVA	Perceived Appropriateness	Brand Attitude
Independent variables
Scandal type (A)	2.32**	1.12	4.37**
Response type (B)	17.29***	27.81***	31.34***
A × B	2.34*	3.85**	0.51
Covariates		—	—
Intentionality	0.89	—	—
Locus of control	0.52	—	—
Scandal severity	1.18	—	—

Note. ANOVA = analysis of variance; MANCOVA = multivariate analysis of covariance.

*p < .10. **p < .05. ***p < .01.

As an optional analysis, we employed PROCESS Macro Model 7 (Hayes, 2009) to further understand the processes in which consumers develop attitude toward endorsed brands. Specifically, we assessed the following model: Response × Scandal → Perceived Appropriateness → Brand Attitude. Response type was coded as 0 = maintenance and 1 = termination. Scandal type was coded as 0 = competence-irrelevant and 1 = competence-relevant. A bootstrap analysis using 2,000 samples with a 95% confidence interval (CI) revealed that perceived appropriateness did not significantly mediate the relationship when brands execute maintenance strategy (indirect estimate = −.28, SE = .47, CI [−1.17, 0.72]). To the contrary, the results showed a significant indirect effect for the termination condition (indirect estimate = .92, SE = .38, CI [0.19, 1.71]). Specifically, perceived appropriateness significantly mediated the relationship between scandals and brand attitude when brands implement termination strategy. This means that endorsed brands should terminate the endorsement contract with misbehaved endorsers who engaged in competence-relevant scandals. Consumers perceived that as an appropriate strategy, which eventually minimizes the damage on attitude toward the endorsed brands.

Discussion and Implication

The present study adds to the growing literature of crisis communication in celebrity endorsement by examining consumer judgments of scandal response strategies and endorsed brands. The authors conducted an experiment focusing on celebrity scandal types (i.e., competence-relevant vs. competence-irrelevant) and brand response strategy type (i.e., maintain vs. terminate) to examine consumer perception of appropriateness of the brand response strategy and attitude toward focal endorsed brands. The findings of this study, and suggested decisions for endorsed brands, are summarized in Figure 5 and Table 3.

Figure 5.

Possible decision-making model for endorsed brands. Note. Brand managers terminate athlete sponsors when a competency-related scandal emerges. However, when the scandal is not relevant to athlete competency, brand managers face the choice of either sponsor termination or further messaging to influence consumer interpretation that preserves the positive effects of sponsors.

Table 3.

Summary and Evaluations of Decision Based on the Findings.

Athlete Endorser Scandal Type	Possible Endorsed Brands’ Decision	Actual Industry Examples	Evaluation Based on the Findings
Competence-relevant	Maintain	Nike’s response to Maria Sharapova’s PED	×
	Terminate	Nike’s response to Lance Armstrong’s PEDAdidas’ response to Tyson Gay’s PED	◎
Competence-irrelevant	Maintain	Nike’s response to Michael Vick’s dogfightingSpeedo’s response to Michael Phelps’ marijuana case	×
	Terminate	Nike’s response to Tiger Woods’ extramarital affairDick’s response to Ray Rice’s assaultSpeedo’s response to Ryan Lochte’s storytelling	◯

Note. ◎ = a highly suggested decision; ◯ = a suggested decision; × = an unrecommended decision; PED = performance-enhancing drug.

The results indicate that consumers perceive the termination strategy as more appropriate. Termination, rather than a maintenance strategy, reflects more positively on the brand regardless of the type of scandal. The same pattern of the results was obtained with regard to consumer attitude toward the endorsed brand. Specifically, consumers show more favorable attitude toward the endorsed brand that implemented the termination strategy rather than the maintenance strategy. These findings are consistent with the basic tenet of the balance theory (Heider, 1958). It is unlikely that consumers will want a brand to continue its endorsement of an athlete who is involved in negative incidents.

Consistent with the results of the current study, Chang (2018) has also suggested that cutting ties with scandalized athlete endorsers can be a viable means to prevent a negative image transfer. This is an argument based on a learning perspective toward athlete endorsers. When the appropriateness of the response strategy and consumer attitudes toward an endorsed brand are the primary concerns, endorsement contracts with the scandalized endorsers should be terminated. This finding runs counter to the strategies implemented by brands such as Nike that often maintain endorsement contracts with tainted athlete endorsers. In contrast, athletic supply retail chains such as Dick’s and Modell’s Sporting Goods made a different decision. They quickly reacted to terminate endorsement deals with National Football League (NFL) running back Ray Rice in 2014 after a video surfaced showing that he assaulted his girlfriend (Castillo, 2014).

The authors also examined whether the effects of the termination strategy on consumer judgment vary based on the type of scandal. Consistent with our third hypothesis, results indicated that consumers perceive that a termination strategy is even more appropriate when the scandal is related to the focal endorser’s expert domain. This finding can be interpreted based on an accessibility-diagnosticity model (Feldman & Lynch, 1988). When a focal endorser’s expertise overlaps with a scandal type (i.e., athlete and PED), the issue can become more accessible to consumers because the category (i.e., sports performance) is utilized in processing the scandal information (Fiske & Taylor, 1980). Information about a competent-relevant scandal such as use of PEDs, relative to a competent-irrelevant scandal such as an extra marital affair, would be processed more easily due to this being a shared category. Therefore, the scandal will be perceived as more diagnostic and can significantly influence a subsequent judgment task (Feldman & Lynch, 1988).

Finally, consumer attitudes toward an endorsed brand that implemented a termination strategy was not found to differ based on the type of scandal. Therefore, our fourth hypothesis was rejected. The authors speculate that consumers show the same level of brand attitude in competence-relevant and competence-irrelevant conditions due to the limits of transfer effects. In the current study, consumers were exposed to the brand information only once. From a learning perspective (Collins & Loftus, 1975), however, repeated exposures to the endorser–brand pairing strengthen the association in memory. Thus, transfer effects from the response appropriateness judgment to a brand evaluation might have required more repeated exposures (Chang & Ko, 2018). In reality, consumers are exposed to scandal information over and over again through various media. The negative impact of competence-relevant scandals may thus be magnified to critically influence endorsed brands. However, our optional moderated mediation analysis at least supported the indirect effect through perceived appropriateness. Specifically, consumers’ attitudes toward a brand that terminates an athlete endorser involved in a competence-relevant scandal was indeed more favorable. This suggests that consumers agree that terminating endorsers in a competence-relevant situation is appropriate.

Overall, the results demonstrate that the competent-relevant scandal was more directly processed and had a stronger impact on perceived appropriateness of the endorsed brand’s response strategy. Previous studies in athlete endorsement also indicate that competence-relevant scandals have stronger negative impact (J. S. Lee & Kwak, 2016; Sato et al., 2015). However, those studies have largely overlooked how endorsement decisions after the transgressions can influence subsequent judgments about endorsed brands. The current study indicates that a termination strategy is desirable, especially when athletes are involved in competence-relevant scandals. Based on these findings, the decisions made by the sponsors of Lance Armstrong were appropriate as they quickly abandoned him after it became apparent that his performance was influenced by illegal PEDs. Nevertheless, the decision made by Nike for the PED scandal of Maria Sharapova was not in accordance with these findings. This point deserves some consideration. The cases of Armstrong and Sharapova were different in terms of severity (i.e., Armstrong repeatedly used banned substances). However, if consumer judgment and attitudes toward the brand are the primary and only concerns, Nike’s decision to retain Sharapova may bring negative outcomes to the brand.

Limitations, Future Research Directions, and Conclusions

There are several limitations to acknowledge. The authors utilized a fictitious scandalized endorser to conduct the current study. Although it is certainly beneficial in controlling for external variables that could influence the outcomes of this study, the lack of realism may raise a concern for the generalizability of the study. In the future studies, using real endorsers could provide improved research opportunities. Consumers appear to have significant psychological connections with human brands (Arai et al., 2014; Chien, Kelly, & Weeks, 2016). If real endorsers have strong psychological connections with consumers (e.g., identification; Carlson & Donavan, 2013; Chien et al., 2016), it may encourage consumers to adopt defensive behaviors. Therefore, terminating the endorsement contract with the scandalized athlete may have an adverse impact on an endorsed brand.

The researchers focused on the crisis response strategies of sports product brands in this study. Response strategies could be influenced by various factors including the investments made to support endorsed brands and endorsers’ responses (e.g., apology) after the incidents, and these could affect attitudes toward endorsed brands. These factors could be included to more holistically and realistically test the effects of endorser scandals on the endorsed brands. Future research should be conducted to examine whether the current research findings can be replicated in endorser–brand incongruent settings (e.g., Maria Sharapova endorsing Walmart). Also, the study assessed the manipulation of response strategies with continuous scale items. These could be measured with a dichotomous scale so that screening participants who failed to answer the manipulation question could be removed from the further analyses; this might have provided a larger effect size. Finally, the impact of endorser scandals can fade as time goes by. If a brand does not terminate its relationship with an endorsed brand (as in the Sharapova example), there is a need to study the outcomes. Future studies can also employ longitudinal design to assess how consumers’ information processing can change their evaluation toward endorsed brands over time.

When athlete endorsers are involved in scandals, endorsed brands should implement appropriate marketing communication strategies. Based on the findings, the authors believe that endorsed brands should terminate the relationship with scandalized athletes regardless of the scandal types. Consumers perceive termination strategies to be appropriate, and this perception will contribute to maintaining favorable attitudes toward endorsed brands. The current study is particularly valuable as it focuses on communication strategies of stakeholders (i.e., endorsed brands) rather than wrongdoers. This is important from the perspective of communication professionals but has been largely overlooked. Therefore, the findings substantially contribute to the endorser scandal and crisis communication literatures.

Footnotes

Authors’ Note

Shintaro Sato is now affiliated with Waseda University, Higashifushimi Campus, Nishitokyo, Tokyo, Japan.

Acknowledgments

The authors thank Dr. Devon Johnson for insightful suggestions. We also thank the editor and anonymous reviewers for constructive feedback.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Shintaro Sato

References

Agyemang

(2011). Athlete brand revitalisation after a transgression. Journal of Sponsorship, 4, 137–144.

Arai

Y. J.

Ross

(2014). Branding athletes: Exploration and conceptualization of athlete brand image. Sport Management Review, 17, 97–106.

Associated Press. (2016, April 4). Abby Wambach sponsor MINI USA withdrawing ads that feature retired soccer star. NY Daily News. Retrieved from http://www.nydailynews.com/sports/soccer/abby-wambach-sponsor-mini-usa-withdrawing-ads-article-1.2587840

Badenhausen

(2017a, February 17). The 2017 NBA all-stars: Players who earn the most money from endorsements. Forbes. Retrieved from https://www.forbes.com/sites/kurtbadenhausen/2017/02/19/the-nba-all-stars-2017-players-that-earn-the-most-money-from-endorsements/#3dac08e578f6

Badenhausen

(2017b, June 7). Ronaldo, LeBron top the world’s highest-paid athletes of 2017. Forbes. Retrieved from https://www.forbes.com/sites/kurtbadenhausen/2017/06/07/ronaldo-lebron-top-the-worlds-highest-paid-athletes-of-2017/#27e23d0a4a1c

Becker

A. B.

Scheufele

D. A.

(2008). Public perceptions of steroid use in sport: Contextualizing communication efforts. International Journal of Sport Communication, 1, 444–457.

Benoit

W. L.

Drew

(1997). Appropriateness and effectiveness of image repair strategies. Communication Reports, 10, 153–163.

Billings

A. C.

Broussard

R. M.

(2019). Untangling international sport social media use: Contrasting US and Chinese uses and gratifications across four platforms. Communication & Sport, 7, 630–652. doi:10.1177/2167479518790014

Brown

K. A.

(2016). Is apology the best policy? An experimental examination of the effectiveness of image repair strategies during criminal and noncriminal athlete transgressions. Communication & Sport, 4, 23–42.

10.

Brown

K. A.

Murphy

Maxwell

L. C.

(2018). Tried in the court of public opinion: Effects of involvement in criminal transgressions on athlete image. Communication & Sport, 6, 283–307.

11.

Buhrmester

Kwang

Gosling

S. D.

(2011). Amazon’s Mechanical Turk: A new source of inexpensive, yet high-quality, data? Perspectives on Psychological Science, 6, 3–5.

12.

Carlson

Donavan

(2013). Human brands in sport: Athlete brand personality and identification. Journal of Sport Management, 27, 193–206.

13.

Castillo

(2014, September 9). Ray Rice’s last sponsor, Nike, has dropped him: No brand is standing by disgraced NFL player. Adweek. Retrieved from http://www.adweek.com/news/advertising-branding/ray-rices-last-sponsor-nike-has-dropped-him-160005

14.

Chang

(2018). When infamy becomes fame: The positive side of negative athlete publicity. Journal of Sport Management, 32, 401–411. doi:10.1123/jsm.2017-0082

15.

Chang

Y. J.

(2018). The effects of association strength on attention and product evaluation: Reconsidering endorsement effectiveness. European Journal of Marketing, 52, 1257–1279.

16.

Cheshin

Amit

van Kleef

G. A.

(2018). The interpersonal effects of emotion intensity in customer service: Perceived appropriateness and authenticity of attendants’ emotional displays shape customer trust and satisfaction. Organizational Behavior and Human Decision Processes, 144, 97–111.

17.

Chien

P. M.

Kelly

S. J.

Weeks

C. S.

(2016). Sport scandal and sponsorship decisions: Team identification matters. Journal of Sport Management, 30, 490–505.

18.

Claeys

A. S.

Cauberghe

Vyncke

(2010). Restoring reputations in times of crisis: An experimental study of the situational crisis communication theory and the moderating effects of locus of control. Public Relations Review, 36, 256–262.

19.

Collins

A. M.

Loftus

E. F.

(1975). A spreading-activation theory of semantic processing. Psychological Review, 82, 407–428.

20.

Coombs

W. T.

(2007). Protecting organization reputations during a crisis: The development and application of situational crisis communication theory. Corporate Reputation Review, 10, 163–176.

21.

Dawar

Lei

(2009). Brand crises: The roles of brand familiarity and crisis relevance in determining the impact on brand evaluations. Journal of Business Research, 62, 509–516.

22.

Dawar

Pillutla

M. M.

(2000). Impact of product-harm crises on brand equity: The moderating role of consumer expectations. Journal of Marketing Research, 37, 215–226.

23.

Doerr

(2016, March 11). TAG Heuer drops tennis star Maria Sharapova in the wake of failed drug test. Forbes. Retrieved from https://www.forbes.com/sites/elizabethdoerr/2016/03/11/tag-heuer-drops-tennis-star-maria-sharapova-in-the-wake-of-failed-drug-test/#362984ba2df4

24.

Faul

Buchner

Lang

A. G.

(2009). Statistical power analyses using G*Power 3.1: Tests for correlation and regression analyses. Behavior Research Methods, 41, 1149–1160.

25.

Feldman

J. M.

Lynch

J. G.

(1988). Self-generated validity and other effects of measurement on belief, attitude, intention, and behavior. Journal of Applied Psychology, 73, 421–435.

26.

Fiske

S. T.

Taylor

S. E.

(1980). Social cognition. New York, NY: Random House.

27.

Hair

J. F.

Anderson

R. F.

Tatham

R. L.

Black

W. C.

(2009). Multivariate data analysis (6th ed.). Upper Saddle River, NJ: Prentice Hall.

28.

Hayes

A. F.

(2009). Beyond Baron and Kenny: Statistical mediation analysis in the new millennium. Communication Monographs, 76, 408–420.

29.

Heider

(1958). The psychology of interpersonal relations. New York, NY: John Wiley.

30.

Ipeirotis

P. G.

(2010). Demographics of Mechanical Turk (NYU Working Paper No. CEDER-10-01). Retrieved from SSRN: http://ssrn.com/abstract=1585030

31.

Jordan

J. S.

Turner

B. A.

(2008). The feasibility of single-item measures for organizational justice. Measurement in Physical Education and Exercise Science, 12, 237–257.

32.

Kim

H. J.

Cameron

G. T.

(2009). Making nice may not matter: The interplay of crisis type, response type and crisis issue on perceived organizational responsibility. Public Relations Review, 35, 86–88.

33.

Knittel

C. R.

Stango

(2013). Celebrity endorsements, firm value, and reputation risk: Evidence from the Tiger Woods scandal. Management Science, 60, 21–37.

34.

Kwon

Trail

(2005). The feasibility of single-item measures in sport loyalty research. Sport Management Review, 8, 69–88.

35.

Lee

Bang

Lee

(2013). Regaining fans’ trust after negative incidents: Fit between responses and nature of incidents. Sport Marketing Quarterly, 22, 235–245.

36.

Lee

J. S.

Kwak

D. H.

(2016). Consumers’ responses to public figures’ transgression: Moral reasoning strategies and implications for endorsed brands. Journal of Business Ethics, 137, 101–113.

37.

Lee

J. S.

Kwak

D. H.

Moore

(2015). Athletes’ transgressions and sponsor evaluations: A focus on consumers’ moral reasoning strategies. Journal of Sport Management, 29, 672–687.

38.

Leys

Ley

Klein

Bernard

Licata

(2013). Detecting outliers: Do not use standard deviation around the mean, use absolute deviation around the median. Journal of Experimental Social Psychology, 49, 764–766.

39.

McCracken

(1989). Who is the celebrity endorser? Cultural foundations of the endorsement process. Journal of Consumer Research, 16, 310–321.

40.

Miller

F. M.

Laczniak

G. R.

(2011). The ethics of celebrity-athlete endorsement: What happens when a star steps out of bounds? Journal of Advertising Research, 51, 499–510.

41.

Money

R. B.

Shimp

T. A.

Sakano

(2006). Celebrity endorsements in Japan and the United States: Is negative information all that harmful? Journal of Advertising Research, 46, 113–123.

42.

Nicotera

A. M.

(1995). Conflict and organizations: Communicative processes. Albany: State University of New York Press.

43.

Perez

A. J.

(2016, February 17). Nike drops Manny Pacquiao after derogatory comments toward gays. USA Today. Retrieved from http://www.usatoday.com/story/sports/boxing/2016/02/17/manny-pacquiao-nike-drops-anti-gay-comments/80511952/

44.

Rice

R. E.

(1993). Media appropriateness. Human Communication Research, 19, 451–484.

45.

Rowe

(2017). Sports journalism and the FIFA scandal: Personalization, co-optation, and investigation. Communication & Sport, 5, 515–533. doi:10.1177/2167479516642206

46.

Sato

Y. J.

Chang

Kay

(2019). How does the negative impact of an athlete’s reputational crisis spill over to endorsed and competing brands? The moderating effects of consumer knowledge. Communication & Sport, 7, 385–409. doi:10.1177/2167479518783461

47.

Sato

Y. J.

Kaplanidou

Connaughton

(2016). Consumer evaluation of athlete endorsers: Comparative evaluative judgment approach. Journal of Sport Management, 22, 553–565.

48.

Sato

Y. J.

Park

Tao

(2015). Athlete reputational crisis and consumer evaluation. European Sport Management Quarterly, 15, 434–453.

49.

Seele

Lock

(2015). Instrumental and/or deliberative? A typology of CSR communication tools. Journal of Business Ethics, 131, 401–414.

50.

Sherman

J. W.

Lee

A. Y.

Bessenoff

G. R.

Frost

L. A.

(1998). Stereotype efficiency reconsidered: Encoding flexibility under cognitive load. Journal of Personality and Social Psychology, 75, 589–606.

51.

Solomon

M. R.

(2014). Consumer behavior: Buying, having, and being (11th ed.). Englewood Cliffs, NJ: Prentice Hall.

52.

Struthers

C. W.

Eaton

Santelli

A. G.

Uchiyama

Shirvani

(2008). The effects of attributions of intent and apology on forgiveness: When saying sorry may not help the story. Journal of Experimental Social Psychology, 44, 983–992.

53.

Thomas

(2016, September 9). HEAD join Nike in staunch support for banned Maria Sharapova. Sky Sports. Retrieved from http://www.skysports.com/tennis/news/12110/10308223/head-join-nike-in-staunch-support-for-banned-maria-sharapova

54.

Walsh

McAllister-Spooner

S. M.

(2011). Analysis of the image repair discourse in the Michael Phelps controversy. Public Relations Review, 37, 157–162.

55.

Walsh

Williams

(2017). To extend or not extend a human brand: An analysis of perceived fit and attitudes toward athlete brand extensions. Journal of Sport Management, 31, 44–60.

56.

Weber

J. M.

Kopelman

Messick

D. M.

(2004). A conceptual review of decision making in social dilemmas: Applying a logic of appropriateness. Personality and Social Psychology Review, 8, 281–307.

57.

Zhu

D. H.

Chang

Y. P.

(2013). Negative publicity effect of the business founder’s unethical behavior on corporate image: Evidence from China. Journal of Business Ethics, 117, 111–121.