Jennifer Edson Escalas James R. Bettman*
*Jennifer Edson Escalas is an Associate Professor of Management at the Owen Graduate School of Management, Vanderbilt University, Nashville, TN, 37203, (615) 322-3493, fax (615) 3437177, e-mail: firstname.lastname@example.org. James R. Bettman is the Burlington Industries Professor at the Fuqua School of Business, Duke University, Durham NC 27708-0120, (919) 660-7851, fax (919) 681-6245, e-mail: email@example.com.
We propose that consumers appropriate brand symbolism that comes from celebrity endorsement to construct and communicate their self-concepts. Study 1 finds that celebrity endorsement enhances self-brand connections when consumers aspire to be like the celebrity, but harms them when consumers do not; this effect is more pronounced when the brand image is congruent with the celebrity’s image. This effect is further moderated by the degree to which a brand communicates something about the user, with more symbolic brands having stronger effects than less symbolic brands. Study 2 finds that the effect of celebrity endorsement on selfbrand connections is augmented when consumers’ self-esteem is threatened. Consumers selfenhance by building connections to favorable celebrity images or distancing themselves from unfavorable celebrity images.
About 20% of U.S. ads feature celebrities (Solomon 2009), and the percent of ads using celebrities in other countries, such as Japan, is thought to be even higher. Traditional explanations of celebrity endorsement persuasion effects are based on the source effects literature and find that 1) celebrity endorsement increases the attention paid to an ad (Buttle, Raymond, and Danziger 2000); 2) celebrities are generally attractive, which helps persuasion when consumers are worried about social acceptance and others’ opinions (DeBono and Harnish 1988) or when the product is attractiveness-related (Kahle and Homer 1985, Kamins 1990); 3) celebrities may be credible sources if they have expertise in a particular area, such as an athelete endorsing shoes (Ratneshwar and Chiaken 1991) or a beautiful model endorsing make-up (Baker and Churchill 1983); and 4) celebrities are often well-liked, possibly leading to identification and consumer persuasion in an attempt to seek some type of relationship with the celebrity (Belch and Belch 2007). In traditional dual process models (e.g. ELM; Petty, Cacioppo, and Schumann 1983), celebrities are most often considered a peripheral cue: they are important in persuasion only when consumers are not involved in the product category or in processing the ad. However, celebrities may provide central information when an aspect of the celebrity matches the product (as with beauty products and attractiveness; Kahle and Homer1985). Also, as affective peripheral cues, celebrity endorsements may lead media weight to have an impact on sales in mature categories (MacInnis, Rao, and Weiss 2002). However, our approach differs from more traditional explanations of celebrity endorsement effects on persuasion, focusing on the cultural meanings associated with celebrities. Our framework is based on the idea that people engage in consumption behavior in part to construct their self-concepts and to create their personal identities (Richins 1994; McCracken 1989; Belk 1988). We examine celebrity endorsement based upon McCracken’s (1989)
perspective: as consumers construct their self-concept by using brands, they appropriate the symbolic meanings of brands derived, in part, from celebrity endorsement. According to McCracken’s (1986) theory of meaning movement, symbolic properties of the celebrity first become associated with the brands the celebrity endorses. Next, these symbolic meanings are transferred from the celebrity to consumers as they select brands with meanings congruent with their...