“The Dynamic Pursuit of Consumers’ Social Identity Goals”
Carlos J. Torelli
Carlson School of Management, University of Minnesota, 3-150 CSOM, 321 - 19th Avenue South, Minneapolis, MN 55455. Phone: (612) 625-2903, E-mail: email@example.com
Papers to be presented in the session: (* indicates the author who agreed to present if the proposal is accepted). None of the papers have been published yet. 1. American = Men? Gender and Cultural Dynamics in the Marketing of Male-Symbolic Brands to Women
Carlos J. Torelli*
Carlson School of Management, University of Minnesota, 3-150 CSOM, 321 - 19th Avenue South, Minneapolis, MN 55455. Phone: (612) 625-2903, E-mail: firstname.lastname@example.org
Nanyang Business School, Nanyang Technological University, Singapore
Hean Tat Keh
Department of Marketing, Guanghua School of Management, Peking University
Carlson School of Management, University of Minnesota
Nanyang Business School, Nanyang Technological University, Singapore
Content areas: Branding, Self-Identity, Social Identity
Methods area: Experimental Design & Analysis
“The Dynamic Pursuit of Consumers’ Social Identity Goals”
Session Chair: Carlos J. Torelli, University of Minnesota
Research on the role of brands for signaling social identity goals has gained momentum in the last few years (e.g., Berger and Heath 2007; White and Dahl 2007). This growth is timely as we see in the marketplace that boundaries defining group-symbolic brands are increasingly blurred (e.g., through increased cross-gender and cross-ethnic marketing and faster product adoption-abandonment) and consumers’ social roles are increasingly fluid (e.g., through digital social networking). This symposium attempts to advance our knowledge about the social identity function of brands by: (1) emphasizing that social identity goals are fluid in nature and can change due to situational factors and/or life-changing events, (2) demonstrating that brands can symbolize multiple social identities and that people can flexibly use symbolic brand meanings for fulfilling their changing identity goals, and (3) acknowledging that interactions with brands and their symbolized meanings can impact social identity goals and subsequent behaviors. Each paper in this symposium substantiates one or more of these propositions. Together, the symposium portrays a dynamic view of consumer-brand interactions—brands can flexibly serve both as inputs and outputs in identity-processes, and consumers react to the symbolism in brands to meet their changing social identity goals.
The first paper by Torelli, Chiu, Keh, and Amaral demonstrates that certain brands can symbolize multiple social identities and the specific symbolism changes according to specific personal and social concerns. Their results show that although both men-and women perceive male- (vs. female-) symbolic brands to be more symbolic of American culture, women highly identified with their gender prefer female-symbolic brands to be American symbols. But when the American identity is salient, women prefer to use male-symbolic brands to symbolize American culture. These effects disappear when female consumers are made to feel “less American” and view these brands as symbols of a dissociative referent.
The second paper by Cheng, Ng and Hung investigates how consumers use brands to manage changing identities. They show that a life event (e.g., graduation) that causes an identity to expire (e.g., student) and a new identity to emerge (e.g., worker) can result in mixed feelings. This in turn leads consumers who were used to buying brands that symbolize the expiring identity (e.g., student brands) to like these brands less, and to like brands that symbolize the new identity more—i.e., a tendency to “move on.” However, for consumers who don’t consider identity issues when making brand choices (i.e., brands are less identity-relevant) mixed emotions about identity transitions have null effects on brand preferences.
The third paper by El Hazzouri, Carvalho and Main shows that being seen as possessing positive characteristics of a dissociative reference group can be threatening to the self. To avoid this threat, individuals would refrain from showing such positive characteristics, even by making oneself appears as incompetent. The authors found that the threat of being associated with a feminine stereotype impaired male participants’ performance in a gift-wrapping task, as a way of signaling that they do not have the artistic characteristic stereotypically linked to women.
The last paper by Kreuzbauer and Chiu shows that even in brief encounters with brands, the symbolic meanings associated with brands can change social identity goals and impact subsequent social behaviors. They found that the casual act of being handed a branded product associated with distinctiveness (connectedness) meanings can temporarily satisfy distinctiveness (connectedness) needs and inhibit its expression in a subsequent social interaction (i.e., less [vs. more] willingness to work alone in an unrelated task).
This symposium discusses a very important topic that would inform consumer researchers about the psychological processes underlying the use of brands in fulfilling social identity goals, and help branding professionals make better decisions. We anticipate this session will attract those interested in consumer-brand interactions at both individual and social group levels, from both social-cognitive and motivational paradigms. In combination, the papers will elicit a lively discussion about broad-level questions related to the fluidity in consumers’ pursuit of social identity goals through consumption, the multiple social meanings of brands, and the way by which consumption can affect the social identity goals that are salient in a given context.
1. American = Men? Gender and Cultural Dynamics in the Marketing of Male-Symbolic Brands to Women
Carlos J. Torelli, University of Minnesota; Chi-yue Chiu, Nanyang Technological University, Singapore; Hean Tat Keh, Peking University; Nelson Amaral, University of Minnesota
Results show that both men-and women perceive only male-symbolic brands to be symbols of American culture as well. Despite this, women highly identified with their gender prefer female-symbolic brands as American symbols. However, women favor male-symbolic brands, via associations with the American identity, when this common in-group identity is salient. (50 words)
2. Brand Preferences During Identity Transitions
Shirley Y. Y. Cheng, University of Illinois at Urbana-Champaign; Sharon Ng, Nanyang Technological University, Singapore; Iris W. Hung, National University of Singapore
This research studies changes in brand preferences emerging from an identity transition. Results show that mixed feelings about graduation can lead students who normally prefer student-symbolizing brands to like these brands less. We investigated the process explaining why students are more ready to “move-on” in terms of their brand preferences. (50 words)
3. I am Too Much of a Man to Be Artistic: Signaling Incompetence to Reduce the Threat of Being Associated with a Dissociative Group
Mohammed El Hazzouri, Sergio W. Carvalho, Kelley Main, University of Manitoba
This research shows that the threat of being associated with a feminine stereotype impaired male participants’ performance in a gift-wrapping task, as a way of signaling that they do not have the artistic characteristic stereotypically linked to women. This effect disappears when men are cued with their ‘inartistic’ nature. (49 words)
4. Substituting Brands with Social Behavior to Satisfy Identity Needs
Robert Kreuzbauer and Chi-yue Chiu, Nanyang Technological University, Singapore
This research demonstrates that consumers’ pursuit of social-identity goals is a dynamic process continuously shaped by their product interactions. We show that the mere act of being handed a brand can satisfy a salient distinctiveness-(connectedness) need, and results in a reduced likelihood of expressing this need in a subsequent social-task. (50 words)
American = Men? Gender and Cultural Dynamics in the Marketing of Male-Symbolic Brands to Women
Carlos J. Torelli, University of Minnesota; Chi-yue Chiu, Nanyang Technological University; Hean Tat Keh, Peking University; Nelson Amaral, University of Minnesota (820 words)
Marketers are sensitive to gender differences in consumer preferences, as reflected by the development of brands that target a specific gender (e.g., Victoria’s Secret for women). Some of these brands resonate strongly within the gender group they set out to target and become symbols of the group (or gender-symbolic brands) (Keller 1993). Yet, increasingly we find brands that try to leverage their equity by broadening their appeal to consumers of the opposite gender (e.g., Jockey women’s underwear). Surprisingly, consumer research has little to say about the factors that contribute to a brand’s success in bridging the gender divide. If any, recent research suggests unfavorable consumer responses to product-brand mismatch along the gender dimension of brand personality (Grohmann 2009), a finding that seems to recommend against marketing across the gender line.
The present research distinguishes between knowledge of consensual assumptions regarding the cultural symbolism of brands and personal agreement with these assumptions. Specifically, we propose that in the U.S., both men and women have high agreement on which brands can represent the two genders. More intriguingly, we argue that due to some socio-historical biases against women in the U.S., both men and women perceive a widespread consensus in American society that only male-symbolic brands are symbols of American culture. Analogous to previous research on “American = White” (Devos and Banaji 2005), we term this collective perception the “American = men” fallacy. We propose that how people act on and react to the “American = men” fallacy depends on their gender and extent of identification with their gender. Men, regardless of how much they identify with their own gender, would accept this perception as valid, and personally believe that only male-symbolic brands are symbols of American culture. Women, particularly those who identify strongly with their gender, personally disagree with the consensual perception and feel that female-symbolic brands should also be accepted as symbols of American culture.
We further suggest that women may like a male-symbolic brand via its association with the American identity. When this common in-group identity is salient, female consumers would tend to like male-symbolic brands that are symbols of America. However, when female consumers are made to feel “less American,” they would tend to dislike male-symbolic brands that symbolize American culture. Because of the weak cultural associations of female-symbolic brands with the common American identity, cultural identification would have minimal effects on male consumers’ liking for female-symbolic brands.
Study 1 tested the collectively perceived “American = men” fallacy. To test this hypothesis, we asked three different samples of participants from the two genders to rate the extent to which a large and varied group of brands (from pretests) symbolize American culture, men, and women (using the cultural symbolism scale, or CS, Torelli, Keh, and Chiu 2010). We found that brands that were rated more highly on male symbolism were also rated more highly on American culture symbolism (r across all brands = .58, p < .001). In contrast, the correlation between American culture symbolism and female symbolism was significantly lower (z = 3.22, p < .005) and not significantly different from zero, r = -.13, p > .1.
Study 2 investigated gender differences in consumers’ personal agreement with the perceived consensus regarding the cultural symbolism of gender-symbolic brands. Participants from both genders rated three brands (male-symbolic, female-symbolic, or gender-neutral) on the extent to which they symbolized American culture from the perspective of an average American (perceived consensus), as well as from their own personal perspective (after a filler task). They also indicated their gender identification. Results showed that both men and women perceive a consensus in American society that only male-symbolic brands are symbols of American culture. However, men agreed with the consensual view regardless of their level of gender identification. In contrast, women with relatively low identification with their gender personally agreed with the cultural consensus that only male-symbolic brands are symbols of American culture, whereas women with relatively strong identification with their gender (although aware of the cultural consensus) personally felt that female-symbolic brands are also symbols of American culture.
In Study 3, we recruited two samples to investigate female consumers’ evaluation of male-symbolic brands as a function of whether the “women = American” identity link is salient. We manipulated the “women = American” link in two ways. First, in sample 1, we increased the salience of a positive American identity to enhance women’s identification with the common in-group identity. Next, in sample 2, we directly manipulated the salience of the link by highlighting women’s contributions to American society versus the history of gender discrimination in the U.S. We measured brand evaluation and willingness to promote the brand. Results indicated that women evaluated a male-symbolic brand more favorably either when a positive American identity was salient or when women’s contributions to America were highlighted. In contrast, they evaluated a male-symbolic brand less favorably when they were reminded of the exclusion of women from the common American identity.
Brand Preferences During Identity Transitions
Shirley Y. Y. Cheng, University of Illinois at Urbana-Champaign; Sharon Ng, Nanyang Technological University; Iris W. Hung, National University of Singapore (887 words)
Brands are intimately related to self-identity—consumers like brands that match their active self-concepts (Aaker 1999) and use brands to tell others who they are (Escalas and Bettman 2005). Prior research has extensively studied brand preference when consumers have stable identities. However, how do identity transitions influence brand preference? This question remains largely unexplored despite its theoretical and practical relevance.
In face of an identity transition, consumers anticipate both the end of a current identity and the start of a future identity. Consumers may feel happy finishing an important life stage (e.g., finishing school). At the same time, they may feel upset about starting an uncertain future identity (e.g., starting work) and thus experience mixed emotions. This research seeks to understand brand preference in response to mixed emotions emerging from identity transitions.
We propose that some consumers are more likely than others to use brands for managing identity transitions. Those who consider their identity in their brand choice (i.e., enactors) would have an instrumental view of brand and identity construction. To them, identity-symbolizing brands can help in constructing their identity. So we hypothesize that when they feel mixed emotions about a transition, they would show a tendency to “move on”—their liking for brands that symbolize the expiring identity decreases but liking for brands that symbolize the new identity increases. However, for those who don’t tend to consider identity issues when making brand choices (i.e., non-enactors), mixed emotions about identity transitions would have negligible effects on brand preferences. We tested this hypothesis and its underlying mechanism with undergraduate students, using graduation as the focal identity transition. In all four studies, we measured how much participants normally consider their student identity when they choose brands (i.e., levels of enactment) and use the rating as a continuous variable in the analyses.
In study 1, undergraduates first wrote an essay about how graduation makes them feel and rated the extent of mixed emotion they feel about graduation. Then, in an ostensibly unrelated study, the participants rated their liking for a student-brand (i.e., a brand that symbolizes student identity). As predicted, enactors’ level of mixed emotion about graduation negatively correlated with liking of the student-brand. Whereas for non-enactors, mixed emotion did not correlate with their liking of the student-brand.
Study 2 provided evidence on the causal effects of feeling mixed emotions on enactors’ tendency to “move on” in their brand preference. We manipulated participants’ feeling toward graduation using a print ad on a photo-taking service for graduating students. The ad either portrayed graduation as a joyful or sentimental event; results confirmed that the sentimental ad elicited higher level of mixed emotion. Then, participants rated their reactions toward the student-brand relative to a work-brand (i.e., a brand that symbolizes work identity). As predicted, the mixed emotion manipulation decreased enactors’ purchase intention, liking, and self-brand connection of the student-brand relative to the work-brand. The emotion manipulation did not affect responses of non-enactors.
We argued that enactors and non-enactors differ because enactors often use brand choice for constructing their identities. Thus, when dealing with the transition, enactors would find student-brands (vs. work-brands) less (vs. more) appealing. Inducing enactors to focus on preserving the expiring identity during identity transition (rather than constructing a new identity) should attenuate the effects of mixed emotion on brand preferences. We tested this hypothesis in study 3. Specifically, we manipulated the focus of identity management in two between-subject conditions. In the identity-preserving condition, participants read a description of a social networking platform that could help immigrants connect to people in their home country (“Find your old friends from your home country! Savoring the old you!”). In the control condition, participants read that the social networking platform could help immigrants connect to people in the new country (“Make new friends in your new home! Creating the new you!”). All participants were asked to imagine being an immigrant when reading the message. Then, in two ostensibly separate studies, participants rated mixed emotions about graduation and responses to the work-brand. Results in the control condition replicated the effects found in previous studies: Enactors who felt higher (vs. lower) level of mixed emotions preferred the work brand more (vs. less). More importantly, results in the identity-preserving condition supported our hypothesis. Priming the participants to focus on preserving an old identity attenuated enactors’ tendency to “move on” in their brand preference—feeling mixed emotions did not increase their preference for the work-brand. This finding provides evidence on the proposed mechanism of enactors’ “moving on” effect.
Lastly, we conducted a field study to examine whether enactors were indeed more likely to “move on” in their brand preference when they faced the real transition. We surveyed graduating seniors after they completed all their examinations. Results showed that for non-enactors and enactors who had not found a job, the more (vs. less) they miss being a student, the more (vs. less) they like a student-brand. However, for enactors who have already found a job, how much they miss student life no longer predicted liking for the student-brand. We reasoned that having a job at the time of the survey marked the actual transition from student identity to work identity. For enactors, having a job means that they have a new identity to enact, thus they moved on and their brand preferences dissociated from their feelings about student life.
I am Too Much of a Man to Be Artistic: Signaling Incompetence to Reduce the Threat of Being Associated with a Dissociative Group
Mohammed El Hazzouri, Sergio W. Carvalho, Kelley Main, University of Manitoba (883 words)
It is widely accepted that consumers often make consumption decisions in ways to protect their self-views. Previous research has shown that consumers are reluctant to use products that send negative cues (Banister and Hogg 2004) and avoid selecting products that are associated with dissociative reference groups, especially when the product is more symbolic in nature (White and Dahl 2006). Such actions are presumably driven by a desire to avoid the negative associations of the dissociative referent. However, in many real-life situations, people may not easily stay away from these negative associations. For instance, if a guy is just asked to wrap a gift for a co-worker, would he worry that doing so very neatly might be perceived by others as possessing the ‘artistic’ trait often associated with women? This research argues that, under certain conditions, such situations can be self-threatening and heighten the need to distance the self from undesirable associations, even by making oneself appears as incompetent.
Research on stereotype threat (e.g., Steele, Spencer, and Aronson 2002) suggests that people can feel a threat to the self when they become aware that their poor performance on a task can confirm the negative reputation associated with the groups they belong to. The presence of such threat seems to impair performance by introducing an additional pressure in order to disconfirm the negative stereotype (e.g., Aronson et al. 1999). We turn this relationship upside-down by extending the effects to the stereotype threat from the positive reputation associated with a dissociative reference group. We predict that when people’s performance on a task can associate them with a dissociative reference group, they will also feel a threat to the self that leads to poorer performance. However, in this case, performing poorly is not driven by added pressure to disconfirm a negative stereotype, but instead by the desire to confirm that one does not fit the positive stereotype. Poor performance would signal that one does not possess the characteristic associated with the dissociative referent, and hence help dissociating from it.
Study 1 tested this prediction using a 2 (stereotype threat: present vs. absent) x 2 (gender: male vs. female) between subjects design. Participants were first primed with a positive stereotype of women: they are more artistic than men. In a subsequent, unrelated task, stereotype threat was manipulated by informing half of the participants that a gift-wrapping task they were about to do was designed to test the wrapping materials (stereotype threat absent) or to test their artistic ability (stereotype threat present). As expected, results demonstrated that stereotype threat significantly reduced performance for male but not for female participants. Further, results revealed a reduction in perceived competence, and performance satisfaction for men and not for women. No gender differences emerged when the stereotype threat was absent.
Study 2 was conducted to rule out the alternative interpretation that men’s poorer performance may be driven by self-handicapping strategies (e.g., Steele and Aronson 1995) and not by the heightened dissociative needs argued here. Stone (2002) showed that when Caucasian athletes were threatened by the stereotype that Caucasians are incompetent in sports, they self-handicapped by practicing less before a sports-related task. Study 2 was similar to study 1, but the wrapping exercise was presented as a practice phase in which participants would have up to 8 minutes to practice before performing the main and more complicated gift-wrapping task (which did not occur). Results showed that male participants in the threat present condition took significantly more time to practice than those in the threat absent condition, presumably because doing so would signal their incompetence in terms of gift-wrapping (i.e., lack of artistic abilities). No effect of threat was found among female participants.
Study 3 was designed to more directly assess the notion that a desire to signal a lack of artistic abilities underlies men’s poorer performance in the gift-wrapping task. We reasoned that men cued with their ‘inartistic’ nature should not exhibit these heightened needs, which in turn should mute the effects. We tested this proposition in a field study using a 2 (stereotype threat: present vs. absent) x 2 (inartistic cue: present vs. absent) between subjects design. Male participants were primed with the stereotype that females are more artistic than males, and were then told that gift-wrapping is either indicative or not of artistic ability. Men's inartistic nature was cued by having the experimenter replace a ribbon with an already made bow stating that males are not artistic enough to use a ribbon/no explanation was provided (present vs. absent). Study 1’s results were replicated in the inartistic cue absent condition. In contrast, when cueing with men’s inartistic nature, the stereotype threat present condition positively impacted participants’ performance, perceptions of competence, and performance satisfaction. Interestingly, when inability was signaled, performance, perceived competence, and performance satisfaction reached the same level as when the stereotype threat was absent.
Our findings show that the prospect of being perceived as possessing the traits of a dissociative referent can be self-threatening, which in turn can trigger subsequent behavior aimed at confirming that one does not fit the stereotype of the undesirable group. To the extent that the barriers that separate gender and ethnic-based products become blurred (i.e., cross-marketing), the self-threat effects described here are likely to be more common. Our findings provide a framework to understand how consumers cope with such threats and adjust their behaviors to dissociate from undesirable social identities.
Substituting Brands with Social Behavior to Satisfy Identity Needs
Robert Kreuzbauer and Chi-yue Chiu, Nanyang Technological University (812 words)
People use brands to communicate important aspects of self-definition (Escalas and Bettman 2005). For example, a marketing professor might choose to wear for class a suit from a fashion label in order to signal distinctiveness from his students. Similarly, a college student might wear a T-shirt with the college’s logo to signal connectedness with her institution. In both examples, brand usage provides the means to satisfy identity needs (e.g., distinctiveness and connectedness respectively). This research departs from prior work that explains how consumers use brands to achieve an optimal connectedness-distinctiveness balance (Berger & Heath 2007) and demonstrates that the mere act of being handed a brand can satisfy a salient distinctiveness (connectedness) need, and results in a reduced likelihood of expressing this need in a subsequent social task. This shows that consumers flexibly use brands and social behavior to satisfy social identity needs.
According to optimal distinctiveness theory (ODT, Brewer, 1991), individuals desire to attain an optimal balance of connectedness (belongingness) and distinctiveness within and between social groups and situations. These basic social needs can be flexibly satisfied through alternative means of attainment (e.g., by buying a product or by engaging in social activities, Kruglanski et al. 2002). Distinctiveness and connectedness needs are in constant opposition with each other. In most experimental conditions (cf. Brewer, 2003), situations that tilt self-categorization processes in one identity direction arouse the opposing need until the optimal point is reached again. This research uncovers an alternative mechanism for achieving optimal distinctiveness. We show that the balance between these two needs can also be achieved by satisfying the need for distinctiveness (or connectedness) through a brief encounter with a branded product, which consequently brings both needs back to an optimal level. We further show the flexible manner by which consumers can satisfy distinctiveness (or connectedness) needs. The mere act of being handed a low-involvement product (e.g., a branded tea-bag) associated with distinctiveness (connectedness) made people less interested to work alone (in a group). In other words, consumers can interchangeably use consumption and other social behaviors in order to restore the balance between distinctiveness and connectedness. Four experimental studies demonstrate the hypothesized effects.
The first two studies were set up to test these basic predictions. In study 1, participants were shown an advertisement about a tea-bag brand. The advertisement conveyed the notion that using the tea-bag brand would make a person feel distinctive. After seeing the advertisement, half of the participants were handed the same tea-bag shown in the advertisement (distinctiveness condition), whereas the other half were handed a different tea bag (control condition). We reasoned that receiving the “distinctive” tea-bag would help participants in the ‘distinctiveness’ condition to satisfy their need for distinctiveness more than those in the control condition. To assess this, after receiving the tea-bag, participants were presented with an ostensibly unrelated writing task in which they rated their willingness to work alone or in groups. Results showed that participants in the ‘distinctiveness’ condition were less willing to work alone than participants in the control condition were. Study 2 replicated the findings for a tea-bag brand associated with connectedness (instead of distinctiveness). In this case, participants who satisfied their need for connectedness when handed the tea-bag brand were less inclined to work in groups in the subsequent writing task.
The last 2 studies were designed to rule out an alternative interpretation of the findings based on possible ‘disappointment’ on the part of control participants handed a different (or ‘wrong’ bag), and to prime the target social need in a more controlled manner. Study 3 followed a procedure similar to that used in study 1. However, we first primed all participants with distinctiveness by presenting them with a set of related words and images. We also added two conditions in which participants read a neutral advertisement about the tea-bag brand. The design was then a 2 distinctiveness of tea-bag message (distinctiveness or neutral) X 2 tea-bag received (same in the ad or different). As expected, only participants handed the ‘distinctive’ (and same) tea-bag showed a decreased willingness to work alone compared to their counterparts handed a different bag, and also compared to those in the two ‘neutral message’ conditions that were handed the same or a different tea-bag. Study 4 replicated the findings with the same design, but using this time a tea-bag associated with connectedness. That is, participants who received the “connected” tea-bag brand were less willing to work in a group than participants in the other three conditions. Results from these two studies persisted after controlling for gender, mood and individual differences in the need for distinctiveness and belongingness.
In summary, our findings suggest that consumers’ pursuit of identity goals is a dynamic process continuously shaped by their interactions with products. We show that people’s interaction with a low-involvement product, such as a tea-bag, can trigger social identity needs that affect their subsequent social behaviors.